Your fortnightly Friday JEN_AI newsletter / podcast has been interrupted to bring you.. a LIVE EVENT later today!

LAST CHANCE: there are only a few tickets left (would you believe it)

JEN_AI Vibe Club #01 -- a Make It Real Lab session

📅 Friday 24 July, 8:30pm UK (9:30pm CEST · 3:30pm EDT · 12:30pm PDT)
🎟️ Register here · limited spots, for a better vibe

​The first (possibly, only?) live event from JEN_AI: a Make It Real Lab session at The Vibe Club

​If you’ve been following JEN_AI and thinking “I want to be in a room with people who actually do this stuff” - this is that room and there’s an awesome lil group of folks, both absolute beginners and utter (AI-addicted) pros.

​I’ll demo a few JEN_AI experiments and show you how to make an idea into a real website / app, then open the room to share one of your own AI experiments if you want to. Bring a half-idea if you feel like it, something you’ve already tried with AI, or just how you’re feeling about all of this.

​Limited spots, for a better vibe.

​45 minutes, grab drinks & snacks, soak up the vibes.

Jenni xx

Please forgive the sales-y-ness of this, trust me I’m cringing even more sending it than you did reading it. Let’s make those dreams real! And make some real I.T.!

This will be our last public post / event til September as JEN_AI takes a summer break - see you in September!

// LET’S GET INTO IT

Whatever your vibe, this episode we are getting into the club. Beach-, football-, startup-, tennis-, AI-, night-, book-.. choose your poison.

This episode:

• 🏆 Cannes Lions, Wimbledon, the World Cup, & brand-new brand worlds

• 🪩 Google DeepMind threw a startup party in the Ministry of Sound nightclub

• 🧪 Inside an AI accelerator cohort

• 📚 What happens when you bring AI into the book club 😱

• ✨ JEN_AI presents Make It Real Lab at The Vibe Club: how to grab a ticket, start your own, or just follow along

• 🎧 Guess the theme song

JEN_AI Challenge #20: Join the Club

// What’s JEN_AI?

I’m Jenni Munroe, a now-ex Google / DeepMind Generative AI (’GenAI’) Experimentalist, and JEN_AI is the human + AI creative lab I started on maternity leave to keep me sane thru 24/7 parent life.

Each episode I set myself a challenge: use AI to make a fun idea real, learn something, and share the experiment, in limited time. Could I vibecode an app? Make a podcast? A zine? Pitch a talk? What can I record before the baby wakes up?

But this isn’t just about me & the fam any more. You (yes, YOU!) are part of a rare crowd of competent, curious, creative (and yes, busy) people. I started wondering: what if the next challenge isn’t just mine?

// How To Get Into The Club

So I’m building JEN_AI’s first member event thing: a Make It Real Lab Session at The JEN_AI Vibe Club (yeah, online).

🪩 JEN_AI Vibe Club #01 • a Make It Real Lab session

📅 Friday 24 July, 8:30pm UK (9:30pm CEST · 3:30pm EDT · 12:30pm PDT)
🎟️ Register here · limited spots, for a better vibe

Bring a half-idea, or something you’ve already tried with AI. I’ll demo a few JEN_AI experiments and do a live 0-to-1 vibecode to show you how to make an idea into a real website / app, then it’s an open room to share what we’re making. 45 minutes, snacks & drinks encouraged.

→ GET YOUR TICKET

This is not “become a software engineer in 45 minutes.” Nobody panic. It’s for smart, busy, idea-rich people who want to stop saying “I should learn AI” and “I wish I could do that thing” and just see what happens when they make the first tiny version of something they care about.

Two other ways in, if a ticket isn’t your thing this time:

• Start your own Vibe Club. Your crowd, your interests, your snacks, your rules.

• Try your own AI experiment whenever you feel like it, in the Make It Real Lab web app (=an app but it’s in a website). If you want to know where this episode’s how-to-AI guide is, this is it!

  Let me know how it goes!

Want more?

• Join a FOUNDERCORPS Cohort to go deeper building AI skills and creative / business ideas. I’ll run 3-4 weeks of 1hr AI accelerator sessions to guide you from idea to real thing with all the tips & tricks. Jump on the Make It Real Lab FOUNDERCORPS cohort waitlist. → JOIN THE WAITLIST

A ‘vibe club’ could be any gathering over some shared interests, drinks, snacks & chat. One group might be a low-commitment book club and AI show & tell with badass women. Another might be a Friday night gaming & questionable business plans with friends who rowed together once 20 years ago (yeah, Aquapanthers crew - I’m looking at you).

Host your own, do a “Make It Real” Lab AI challenge to make an idea into a real app or website in your own time, or join us live.

// The Club Sandwich

// FOMO no more: you don’t need to be standing there to feel the vibes

Cannes Lions. Wimbledon. The World Cup. Easy to follow along, near-impossible to get a VIP ticket - but you don’t need one to enjoy soaking up the vibes anyway.

The AI educator I featured last episode (Allie Miller) queued seven hours for centre-court Wimbledon tickets. Personally I’d rather watch it on a pop-up screen, on a beanbag, by the canal. (Or just eat strawberries.) Likewise, almost none of us are at the actual World Cup Games, and many of us in the UK aren’t even up at stupid o’clock watching them. I’m not even into football! (or soccer, whatever) It still feels good to be part of a communal cultural event, rooting for your club. I mean, country.

If you haven’t heard of Cannes Lions, it’s an annual “Festival of Creativity” (aka what folks in SPAM - social media, PR, advertising and marketing - call a marketing conference, ofc) on the French Riviera. Many awesome creators I follow were either there or having major FOMO at not being there. (Which felt both valid and unnecessary as the whole event was made into a breezy coastal firehouse of content, as you’d expect)

I’ve spoken before about the pivot towards IRL (or In Real Life) events, for instance podcasters and creators hosting dinners (supper club, anyone?) etc. as they try to work out what is left for humans in the age of AI. More broadly, more than ever, brands are leaning into not just experiences but entire worlds that people want to pay to be a part of, where the actual product fits inside oh so casually. Clearly I shouldn’t have just shoved a workshop ticket in your face. Whoops. Putting a first paid offer out there in the world is nerve-wracking, ok? I’m here to learn.

Talking of attempted pivots..

Actress Reese Witherspoon, who runs a wildly successful book club, said earlier this year that women should learn AI, and got hit with a major backlash for it. If she had any hope of combining AI education for women with a book club model - those hopes were dashed. Which is a shame, because someone should do it. Alright, fine I’ll do it. Yeah I did in fact bring AI to the book club. Many of my actual or attempted book club nights have - spoiler alert here - been something an excuse to hang out and chat and eat and drink things with people I care about and were somewhat flexible on the reading requirement, so demoing AI & chatting about dreams & concerns etc. fit in just fine. I also suggested to my deeply missed Silicon Valley mama crew that we do a Stitch ‘n’ Bitch night but for sharing our AI experiments, and they promptly ran with it.
As criminal as it is to say this out loud - the AI Book Club model works. But we should probably call it something else. Maybe a vibe club. (For, you know, vibecoding, vibe-designing, and just (vomit) clubbing together and “vibing” with AI?). Better suggestions are so very welcome.

// Breakfast in the [night]Club

I applied to a Google DeepMind / Google for Startups event in London Tech Week hosted at the Ministry of Sound nightclub, and got in. (Take that, Berghain! Alright, yes, I did know people there, I am ex- Google / DeepMind.)

Still, I absolutely did feel like a 17-year-old sneaking into a club as I wore my ‘Founder & Creator, JEN_AI’ badge around attendees who’ve built actual companies.

Some insider moments:

• The vibe: a neon lightshow, luxurious food and drink, mingling with cool people in the startup space, startup pitches, demos of the latest DeepMind AI research & Google tools for startups, and more than 10 women present. Wow.

• AI-nerd-out warning (skip as needed): I am weirdly obsessed with Google DeepMind’s DiffusionGemma language model - which doesn’t just write one word at a time, it writes the entire block of text at once going from word salad to cohesive prose through iterations, the way AI image generation usually works (image diffusion models).

• And yes, I did get asked who my technical co-founder is (sure, thankyou, yes, I’ll just take it as a compliment that I don’t look like I have a CompSci degree and 15 YEARS OF EXPERIENCE BUILDING TECH PRODUCTS, most of which were AT YOUR COMPANY, GOOGLE instead of taking the feminist rage route )

Anyway, worth checking out if you are starting a tech startup: Google Cloud for Startups Program has a Start tier with $2k in Cloud credits, rising to $350k as you progress on the startup journey.

// AI Squad Goals

I’m currently inside Entrepreneur-influencer Simon Squibb’s HelpBnk AI Accelerator program - it’s actually Sepehr Khosravi’s 5-week AI Squads bootcamp or cohort, whatever you want to call it. The exact kind of thing I’m about to start building my own flavour of with a more JEN_AI vibe, which could potentially be an ethically murky thing to do - so let me be clear: THIS PROGRAM IS GREAT, TRY IT. It is free but they have a partnership with Lovable (AI coding app) and encourage you to upgrade to Lovable Pro (I did), and they’re also creating a paid VIP ongoing program.

The 1-session-a-week-plus-homework format just works. It’s the accountability of it, the live demo session, the group chat, the knowledgable host who clearly doesn’t sleep. If you want help joining the dots on sensible ways to make ideas real with AI, or how to run an accelerator - def worth it. Nice work & thanks, Simon & Sepehr.

// SETLIST

House club music, Summer tennis, it’s called Challengers (see: JEN_AI challenges). And it’s from a Zendaya film about beautiful people locked in sweaty competitive tension, probably not far from Hollywood’s idea of what a hackathon should look like. (These guys also did The Social Network movie soundtrack btw).

Gossipy side note: Challengers’ director Luca Guadagnino’s latest - ‘Artificial’ - is a biopic of Sam Altman (OpenAI ChatGPT boss) that recently got dropped by Amazon, after they announced a huge partnership with OpenAI. And now I reeally want to watch it.

Runner-up: “Clubbed To Death” from The Matrix. At this point you’re probably feeling like I’ve clubbed you round the head with the word ‘Club’ and never want to hear it again. Apologies.

🎧 The JEN_AI Mixtape. Play it loud at your first Vibe Club.

// How To Get The Good Stuff

Subscribe (free!), so the next episode lands in your inbox or share with a friend.

Share

Find me, Jenni Munroe, on LinkedIn, Instagram or Substack.

Comment & tell me what you want to see more of on JEN_AI, and what your dream vibe club would be called. I’m thinking “hot model vibe club” (I mean AI large language models, obviously). Or maybe “AI PSYCHO SISTERS”, for fellow AI Psychosis sufferers.

Thanks, as always, for being part of the experiment. (And apologies to the southern hemisphere, where I realise it isn’t actually summer.)

Jenni x

• 🤖 A guide to set up your own AI agents (an agency?) in Claude Cowork

• 🧪 What happened when I attempted to completely automate JEN_AI’s “idea → AI experiment” workflow

• ⚡ Why Silicon Valley’s latest buzzphrase for ‘people who do things’ (”high agency individuals”) is 🙄 [eyeroll], but also maybe how to beat both burnout and the AI-job-pocalypse

• 🤸‍♀️ JEN_AI in real life: the highs & lows of getting back into aerial circus (sure, why not?)

• 🎧 The bots chose a theme song

JEN_AI Challenge #19: Set up an AI Agent Agency to Build My Dream Life

// What is JEN_AI?

We’ve all had an idea for a ‘killer app’, business or creative project that died in our heads before we could make it real.

I’m Jenni Munroe, an ex Google / DeepMind Generative AI (’GenAI’) Experimentalist & Founder at JEN_AI, my human + AI creative lab. I’m challenging myself (and you too, if you feel like it) to make ideas & dreams real, mashing-up new and retro tech, culture and media, as we work out what it means to be human in the age of the machines.

This week: I built myself some robot coworkers. (Don’t come for me.)

// What’s an AI agent?

An agent earns the name when it takes several steps toward a goal, uses real tools to actually do things, makes its own small judgement calls, reviews its work & keeps going without you prompting every click. The aim is not engineering “prompts” but “loops”. We’re not just talking about chatbots here.

Anthropic’s ‘Claude’ suite has arguably been considered ‘the best’ of the 3 top AI ecosystems in most of 2026 (next to OpenAI’s ChatGPT & Codex and Google’s Gemini & Antigravity, and all the rest). ‘Claude’ itself is the chatbot, ‘Claude Code’ is targeted at software developers and ‘Claude Cowork’ is aimed at non-technical knowledge workers. Claude Code & Cowork can be used more agentically, but they work best when you set up your “context” first. Which is what I’m going to guide you through.

// So I Finally Set Up My Second Brain

Shocking confession: I hadn’t set this up properly for myself in the Claude ecosystem until this week. (If you haven’t either - solidarity, & also, it took me a couple hours to do it.)

Why it’s worth the faff: once an AI actually knows you - your goals, your style, your references & preferences - it stops being generic & starts being more useful. Especially once you have connected it to the tools you regularly use (Slack, Gmail, Excel, whatever). In my case “useful” means: I spit a half-baked idea into my setup (”Make It Real Lab”), & it helps me turn the crazy brain-vomit and reference docs / images etc. into something real.

// Build Your Own Agentcy in 5 Steps

This is a tight setup guide for Claude Cowork: a no-code way to put an AI to work inside your own computer. I distilled these essentials from guidance by both Ruben Hassid, ranked #1 in Education on Substack (it’s impressive, ok), and “#1 Voice in AI for Business” Allie K. Miller, so you don’t have to.

1. Download Claude. Get Claude for Desktop, get the paid $20 plan (3 passes for a free week here!), open the Cowork tab. Optional but good: pick the Opus model & switch thinking on.

2. Create empty folders. Make one folder on your computer somewhere (’Documents’ is fine), call it CLAUDE COWORK, with three folders inside: ABOUT ME, OUTPUTS, TEMPLATES. When you start typing in Cowork → ‘Work in a project or folder’ → ‘Choose Folder’ and choose that CLAUDE COWORK folder you made. Now it reads & writes here, & nowhere it shouldn’t. (It baffles me that Claude doesn’t do this for you.)

3. Download yourself. Ask Cowork to interview you & write your context into ABOUT ME: an about-me file (who you are, how you work, what you can’t stand), a my-company (or job) file (goals etc.), & an optional brand file. Paste into the CLAUDE COWORK folder an anti-ai-writing-style doc (the AI words & patterns you ban). These are all saved as ‘markdown’ files (.md), which is just a text file with funny formatting.

Bonus: if you dictate your answers instead of typing, it’s faster, actually sounds like you, and you tend to give more context. (Wispr Flow - wisprflow.ai (free month here!)- has one hell of a marketing team and is good, but you can also just click the mic button.)

4. Set this one instruction. In Settings → Cowork → Instructions section, tell it to read every file in ABOUT ME before every single task. (Again: why does Claude not do something like this for you?)

5. Connect your tools. Click Customize → Connect your apps → choose the tools you live in: Gmail, Calendar, Drive, Slack, Canva, whatever. The moment it can take actions in the tools you use, with your real data, you’ve crossed from chatbot to agent.

Optional 6. Teach it a repeatable ‘skill’. When you catch yourself explaining the same job twice, ask Claude to save it as a Skill. A Skill is a small reusable task - a text file with a name, a job, & a short description of when to use it - that Claude loads on demand, so you shouldn’t have to re-explain. You build it once & it runs when Claude thinks its relevant to what you’ve asked to do, or you can schedule it to run on a timer (a daily news brief at 7am, say).

“Steal These Workflows”

(with thanks to Allie K. Miller, who demos these properly)

Now try asking Claude Cowork to do whatever work you’d normally do. Research a topic & turn it into a report or a deck. Or, if you’re me, give it a daft idea & ask it to build it. When something works well, ask Claude to update your ABOUT ME files or save it as a Skill. A couple to start with:

• Auto-draft your boring emails. Connect Gmail. Tell it: read my “non-urgent” folder, draft replies in my voice, strip the AI-sounding phrases, save as drafts. (You review, tweak, then send. Don’t try attempt it to send them for you.)

• Schedule a daily briefing. Ask it to pull your day’s calendar, inbox priorities & relevant news each morning. Set once, fires itself daily.

Allie Miller runs 34 named agents - a chief of staff called Simon, directors named after the Friends cast, & a note-taker called Toby whose only job is helping the system remember itself (again: for the love of god, Anthropic, build this in). Every one of them is just a markdown text file like this containing repeatable jobs grouped into role functions (marketing, education etc.). The advice is ‘stop writing prompts & start writing job descriptions’, & ‘promote yourself from worker to CEO of a system’. Then maybe try Claude Code.

// So how did it go?

It went. Wow. I set up the above, fed in a ton of JEN_AI examples of my writing & branding, then showed Claude Cowork how I turn a brain-dump into an AI experiment. Then I tested it on some of my many ADHD-adjacent crazy ideas. Such as:

Neural Fuel Stacks. I had a vague idea for a web app that jumps on a questionable trend by giving you a stack of personalised nutrition & supplement recommendations depending on what mode you’re in (Creative Sprint, Burnout Recovery, Parent Survival etc.). So I fed it in. Out came a pitch, a rough plan, a demo website mockup, messages to send people for feedback, a drafted Substack newsletter about the experiment, and 4 platform-tailored social posts. My god, I’ve created a monster.

Make It Real Bot. Here on JEN_AI we like to get meta, so of course I also fed in my idea for a web app that makes other people’s ideas real (something I’m building in my AI Accelerator program too, more on that next week). It was fascinating. A rough skeleton held together with bad code & far worse draft LinkedIn posts, but still: it’s not nothing. It even used Netlify to deploy it as a real website for me (this site is a demo tho). Good for prototyping, bad for humanity?

For sure, my fake JEN_AI startup hasn’t earned me a single $, it needs work, and even then it will be far be perfect. And I’m not sure it’s a good idea to fully automate personal social media content. (Or is it? You tell me, Substack.)

A Robot Drafted This (sort of)

Full disclosure: the first draft of this newsletter was written by my AI agentcy, from a dictated rant, my notes, & all the context I set up above. It was... a start. But not good enough. I overwrote it into something better.

The robot gives leverage, speed, analysis, and a cohesive draft of your 2am mind-mess. It cannot (or rather, should not) invent your stories or your lived experience, & the jury’s still out on how well it helps you write & share them. People who use Claude can spot what it wrote. Mine only started sounding remotely like me once I’d fed it a pile of my own writing & sat down & actually told it what I was trying to say.

Safety, Security & Superpowers

It’s a bit alarming to hand so much of yourself to an AI company, for sure. Only share what you’re comfortable with, & toggle off ‘train models with my data’ in Settings.

What makes agents terrifying is also what makes them powerful. ‘OpenClaw’ caused a stir earlier this year: a freely available, homemade, superpowered agentic security timebomb. You could box it into a constrained space & restrict its tools, but then it couldn’t do the really useful stuff. (Claude’s slightly safer take is ‘Dispatch’. I tried it by asking it, from my phone while I was out of the house, to build me an online survey by controlling the mouse cursor on my open laptop. Terrifying. Also: a terrible example!.)

// The Autonomous Life

Here’s a bot-summary tl;dr of my rant below: “High agency” is a cringe phrase hiding an important concept & a real human need that nearly cost me my health. AI is breaking a system that was already part-broken, & the future of work should work for us.

Is “High Agency” Just Hustle Culture in a New Patagonia Vest?

Why am I even talking about this phrase? Because AI thought-leaders keep saying it on social media. (See also: “taste”. My god.) The gist: in the age of AI, you want to hire people who actually get things done, using a range of both human & AI skills, without waiting for permission. Revolutionary. (The same trend has given us CEOs & elite tech-bro bosses solemnly retitling themselves “Member of Technical Staff”.)

Any idea that boils down to “the only thing in your way is you” is mildly triggering for those of us who’ve been there, done that, got the burnout t-shirt. The system is broken, babes, & I’m done contorting myself & my family around it.

Hustle culture celebrates grinding yourself down: work 100 hours, then work more; ‘failure’ just means you didn’t work hard enough (and by the way, you should have failed faster). Being a girlboss was about juggling & leaning in so hard that - yeah, guess what? - we fell over 💀

High agency, done right, should - I think - be more about working smart to live a life worth living. (Yes, including automating the repetitive stuff.)

The Highest-Agency Thing I Ever Did Was Quit

The most high-agency step I’ve ever taken obviously wasn’t setting up an automated daily briefing. It was leaving a very well-paid job at Google in March this year to take control of my health & my life.

From the outside I looked successful. Inside, I’d been drowning for a long time. Three kids (one with additional support needs), years of barely sleeping, doing the 24/7 parent shift on top of household management on top of the actual job, plus work calls scheduled at kidtime, aka California o’clock. I’ll spare you the full saga, but how does anyone do this? It just doesn’t work. A lot of my friends are somewhere on the same map: parental burnout, work burnout, the time they don’t have disappearing into mind-numbing social media, grinding a slightly soul-destroying job for the money & feeling too guilty to complain because they’re ‘lucky’ to have one.

I’d progressively moved to easier roles at Google to hang on in there, & felt like a glorified babysitter with a great salary. How could I complain? I was privileged, compensated, accommodated (after a fight). And going slowly out of my mind, with no corner of my life that was actually mine.

I couldn’t stop being a parent, & I didn’t want to. But something had to give. So when Google offered everyone in the UK voluntary redundancy, I left.

This is a privileged option I only had through luck, & because I spent years doing what millennials were “supposed to” first - a route that isn’t as open to as many young people now. ‘Herding cats’ at Google after the Accenture corporate-ladder grind, saying no to the fun AI-creativity PhD (& even the Art A-level) & getting a Sensible Job.

A lot of my friends are staring down their finances (& their list of health symptoms) trying to work out whether to leap. I’ve been biting my tongue, knowing how much financial security matters, while wanting to scream “this is killing you, get out, you’ll work it out!”

Agency over your actual life. That’s what I’m fighting for here.

That’s what JEN_AI was for me: first, a tiny space of my own; then, slowly, building a life & a dream that’s actually mine, & using this ridiculous tech as a creative medium, a new way to make a living, & a way to claw back the hours to go & live. And I’m trying to make it something that helps other people too. Even if that’s just to smile while laughing at me.

I’m not going to pretend AI is innocent & lovely. It isn’t. I know everyone’s sick of it. Sick of the pace, the hype, unsure about jobs & privacy & deepfakes & the climate bill for all this compute & copyright & the handful of giant companies hoovering up the power.

But. The tools do, however, exist now for a non-coding human to take the thing in their head & make something real with it. Maybe even a new kind of living. That’s what I’m experimenting for.

Yes, we’re all exhausted. (Don’t talk to me about sleep until my kids learn how to do it.) But we can meet the future like it’s happening to us, or we can meet it with agency. We can learn enough about AI to know what’s dangerous, what to push back on, what to regulate, & take the upsides while we’re at it.

It might be messy, but we can choose who, or what, drives us. Sometimes letting go is how you take the wheel.

// Embodiment x JEN_AI

(Here I’m just talking about stuff I did in the physical world. Yes, this is also a very lame nerd-joke about the classic AI problem of ‘embodiment’, i.e. getting ‘AI in robots’ to actually work.)

“What aerial circus arts taught me about agency”

Don’t panic, I’m joking, this isn’t a 2018 LinkedIn post. Just some quick highs & lows while enjoying the freedom of getting slowly back into aerial circus gymnastics after 6 years & 3 kids:

The high & lows: (yes, its just another lame joke)

// SOUND THAT TRACKS

Not bad for a bot. It’s literally a song about machines doing the work, sung by two men pretending to be hyper-productive robots, built almost entirely from a chopped-up sample of someone else’s record. On the nose? Yeah. Fab. (Don’t get me started on the Kanye version. Hint: the meaning of ‘Harder, Better, Faster, Stronger’ shifts the second you add ‘baby you make me’ to the front.)

🎧 Playlist: The JEN_AI Mixtape

Shoutout also to Taylor Swift doing the new Toy Story 5 track & confusing the internet.

Consider this your reminder that sometimes “flying is just falling with style”.

// Thanks & How to Get the Good Stuff

If you liked this or it got you experimenting, do the high-agency thing & share it with someone else who might benefit.

Share

The one ask: subscribe (free!) so the next experiment lands in your inbox

And if you feel like it, tell me: what’s the idea you wish you could make real?

(P.s. - I’m building towards live workshops to help you make your own idea real, plus a little tool, Make It Real Bot, to do the first steps yourself. Both still cooking. Reply if you’re curious.)

Find Jenni Munroe on LinkedIn, Instagram or Substack.

Thanks, as always, for being part of the experiment.

Jenni x

Still here, still experimenting, just not shipping one today - I’m taking an extra week on the next edition (more on that below). Instead: a few of you told me my emails have been disappearing into spam, promotions folders etc. (especially for Outlook folks & posts with video previews).

🎵 “I’m just Jenni and I’m blocked?” Maybe.

// 2 quick steps (30 seconds, so you stop missing these)

1. Reply to this mail with anything, like “got it” or “hi.” One reply is the single best way to "keep me out of spam and let me know you got this.

2. Get this to your inbox, if it isn’t. On Gmail, drag this email from Promotions into Primary. Apple Mail: tap my name, “Add to Contacts.” Outlook: mark it “not junk.”

// Quick reminder of what JEN_AI is

I’m Jenni Munroe, ex-Google / DeepMind, building JEN_AI: a human + AI experiments lab. Every couple of weeks I grab an AI tool, set myself a challenge, and actually make something, then show you how it went, how to do it. Good stuff, epic fails, bad puns, all of it. While reflecting on what it means to be a human in the age of the machine. (Hype or doom? Both.)

// Up next: we’re getting agents-y

The next edition (‘High Agency’) is the one I keep getting asked for: a proper, no-fluff guide to setting up Claude and using AI agents that actually do real work for you (the stuff beyond chat). I’d rather make it genuinely useful than rush it out half-baked, so it’s landing soon, not today.

(And further down the line, I’m building MAKE IT REAL & FOUNDERCORPS sessions where you bring a half-idea and leave with a real thing. More on that another day.)

// While you’re here

What do you want to see more / less of ? Hit reply and tell me.

// In short

1. Reply (hi!). 2. Move me to inbox. 3. Watch out for the Claude + agents guide.

(And yes I do realise you won’t get the advice on how to pull this out of spam if it went into your spam. Hence the replies so I can check with peeps.)

Back properly next week. Thanks for sticking around while I make JEN_AI real.

Jenni x

This episode:

• 💡We’re turning newsletters (or images & an idea) into social media carousels and (stronger option→) printed pocketable zines

• ⚡️Creative tech inspo from SXSW festival of music / tech / film / ? in Austin & London (my advice: always meet your heroes)

• 💓 Real heroes, & why you might not want to be one

• 🎧 Guess the theme song

JEN_AI Challenge #18: Zine Machine

// What is JEN_AI?

We’ve all had an idea for a ‘killer app’, business or creative project that died in our heads due to lack of time, resources, or the nerve to make it real.

I’m Jenni Munroe, an ex- Google / DeepMind Generative AI (‘GenAI’) Experimentalist & Founder at JEN_AI Human + AI Creative Lab. I’m challenging myself (and you too) to make ideas & dreams real - using new & retro tech, culture and media - as we work out what it means to be human in the age of AI.

// Print Media is Dead. Long Live Print Media

Here’s one good thing about AI slop: it might finally be giving us the kick we all need to get off our phones.

Independent bookshops are seeing a revival. More people are choosing real live communal experiences. Media platforms are pivoting to more live events. Artists are selling via ‘Snail Mail Clubs’. Vogue magazines are even selling out (and I don’t just mean that they’re sponsored by Jeff Bezos).

// How Charles Dickens Can Save Print Media in 2026

But. Magazines are big, books are long and attention is short. Let’s bring back miniature serialised print storytelling like it’s 1899. But with a pocket-sized 1990s twist. (Let’s not forget why Charles Dickens novels are so long in the first place! They were usually first published in short regular installments, with illustrations. And ads. Authors had a chance to pivot future chapters based on reader feedback and get confidence in publishing the book for sale.)

// Swap the Scroll

The Antidote to AI Slop Might Be A Piece of Paper

So, this week’s experiment: take *a Substack newsletter* (or some pics & a dream) and turn it into something you can hold. A printed phone-size addiction-breaking mini-zine. Readable on a train with your phone left in your bag like some kind of Victorian. Add some blank pages to jot down your startup ideas?

Same content, three formats:

1. **Grab a post** or your idea & images

1. **Vibe-design it into a carousel** ask AI (e.g. ai.studio/build) to make a tool that will turn into a social media carousel (sized for LinkedIn & Instagram (1080 × 1350)) or printable zine in

1. **Print it into a zine** so the thing that lived on a screen now lives in a pocket

Here’s an example prompt for Google AI Studio to make you a zine machine:

I want an app that makes a LinkedIn carousel using current best practices whose design style is based on a brand style screenshot the user uploads. The carousel should include images the user uploads, and appropriate content extracted from a pdf of an uploaded Substack newsletter. Give me an option to download the carousel as a printable zine.

If you just want to use my app instead to make carousels or zines - no problem: comment / message me.

// Zine Machine

I should have touched grass. Instead I made.. 97 zines?

I went completely off the rails this challenge & turned into an AI-human zine machine. (well, not 97 - maybe 8)

// Behind The Zines

First I tried making a carousel of my SxSW Austin experience that could double as a social media carousel and printed mini zine.

The “industry standard” tool is Canva - which is great, but ‘vibecoding’ your own tool with AI can offer something no mass-market app can give you: whatever you ask for. I used Google’s AI Studio but you could use Claude Code, ChatGPT’s Codex, Lovable, Replit etc. Vibecoded apps can be less reliable / secure, so don’t give them excessive access, use version control, and set payment limits etc.

I also had this crazy hope that I could write & video-record once per JEN_AI experiment, cut from that and cascadingly share the madness appropriately across platforms (great content system or plague? you decide), and also god forbid, share things with people In Real Life - as we break free from our smartphones and re-embrace the real world via print media and actual human beings.

This one was made with ChatGPT (note their Codex tool is much more powerful) using Google Gemini to analyze the announcements from their recent I/O conference. I didn’t post it on LinkedIn as some release dates were inaccurate - but I’ll let it slide here as product-hype presentations are infamously vague on dates ;)

I also made a “Start Here” Instagram story sequence and “MAKE IT”: a tiny eight-page field guide for people with ideas to get started and make something real.

..Also a colouring booklet as one of the party favours for my kids’ birthday party. That’s it. Almost. I also made a few old-school Instagram Stories and paper mini-zines with friends & family.

// Real Life is the Real Luxury

SXSW festival is in London this week, and I was able to sneak along for a day:

// Four Key Takeaways from SXSW Austin & London 2026:

1. The clash of genres especially in Austin is INSANE: Music. Film & TV. Comedy. Art. Tech. Entrepreneurship. Brands. Creators. Wellness. Vehicles? (incl. self-driving taxis and joyriding Electric Monster Truck Rivians over mountains of dirt on historic Congress Avenue). I ended up following friends to a Drag-King-hosted fashion show at a fringe medieval renaissance fayre at one point. What can I say? Drag gon’ slay.

2. The unofficial parties, ‘houses’ sponsored by companies / countries and the cheap music wristbands are worth turning up for, so if your company won’t pay the $2000 ticket then don’t pay it! Half the people in Austin didn’t have a fancy ticket. Don’t worry: you can still leave with a free hat, mushroom coffee and 7 tote bags.

3. SXSW trends:

   1. the impact of AI including how cognitive labour & writing might become like going to the gym (yikes)

   2. the shift to IRL community events from social media (bias? ofc)

   3. ‘attention is the product’ - startups take note (and start a podcast, lol)

   4. The in-demand hire: ‘agentmaxxers’ (=people really good at using AI agents)

   5. wellness hubs as the new pubs

4. Meet your heroes. But first: meet the people standing next to you. Even in London I was surprised how open everyone was to chatting. Even the celebs.

The maddening thing was I got to hear from and then meet! some absolute icons that most of my friends had never heard of. Actors from the Clueless movie & Buffy the Vampire Slayer (/ Austin Powers) were launching a demon alcopop brand, podcaster Prof G posed for a selfie, and the epic lady content creators of tech (like @alberta.tech, @bytesofbree, @madeline.m.zhang) were out in force. I’m frankly relieved I hadn’t worked out I could actually go up to Paul Rudd at his movie premiere. Insane. Privilege.

Steven Bartlett hosts the “2nd most popular podcast in the world” - ‘The Diary of a CEO’, is a Dragon’s Den [=Shark Tank] investor, entrepreneur & bestselling author. I can’t tell you how reality-melting it is to actually meet and chat to (and HUG?) someone you’ve listened to for hours. His podcast has received its fair share of criticism, but it was the spark that inspired me to start my JEN_AI experimental AI/human creative lab and reclaim a little island of my life for myself since becoming a parent to 3 amazing kids. Steven and his team (esp. Georgia Gibson) were very kind, generous, and are also clearly walking-the-walk not just talking-the-talk in their use of AI agents.

Stepping back for context: I have 3 young kids, one of whom has additional support needs, so being anywhere in person, IRL, is a privilege for me. (I spent 5 hours solo in a hospital emergency room recently and it might as well have been a spa retreat.)

A day at SXSW London with friends was amazing. 6 days at SXSW Austin & visiting friends in California in March this year was completely insane.

I didn’t even care that Google pulled me from my own SXSW speaker gig just days before I was due to present the 2hr AI & Creativity workshop, due to my imminent redundancy. lol. I pitched that talk! BUT it worked out for the best as I got to enjoy the festival with more time and zero pressure.

It did mean that I didn’t get to make the ‘community zine’ I’d planned to create live with the participants at the workshop. Another time ;)

// Who’s the Real Hero?

The biggest lie in history is who did all the work

It was the pandemic that solidified my creeping sense that a ‘hero’ was not something you’d want to be. I wasn’t in the UK or Europe then, but I’m told that folks would stand outside their houses on Thursdays and ‘Clap for Heroes’ to show appreciation for the National Health Service and other frontline workers. It seemed like a lovely sentiment but a terrible deal: a hero was someone underpaid, overworked and doing the risky jobs the rest of us would rather not.

At least they got a thank-you, I guess? The real heroes are usually the ones nobody is clapping for. Parents and carers working the 24/7 shift. Assistants and juniors invisibly doing the work their betters are getting credit for. Any kid or adult stuck in school or work that is not working for them. And I won’t stress anybody out by mentioning women & minorities (like how attributing women’s scientific achievements to male colleagues has happened so systemically it even has a name: The Matilda Effect). You get the idea.

So. YOU. The Real Hero Is You. You deserve the credit, you deserve a break, you deserve the clap (ok forget that), and you deserve to live the life and dreams that matter to you. To anyone gritting through it - I see you. When you get a chance to escape from building someone else’s empire, take care of yourself and then make your own small thing with your name on it. I’m saying this because having something to myself saved me.

// Drive Is Not The Real Hero

Hustle-culture frames failure as not working hard enough. Which misses the point; you already had the drive, but driving doesn’t fix a broken system. We all “leaned in” and guess what? We fell over.

// Care & Attention IS All You Need

If you have been a hero for far too long, I pray you get support, a few stolen moments to rest, and maybe later: some space to do your thing. Plus, forcefully reclaiming your attention (yes, from tech platforms just like this).

Care & attention - to your needs - is ‘all’ you need.

// SOUnD TRACK

• ‘A Real Hero’ by College & Electric Youth, best known from the 2011 “Drive” movie starring Ryan Gosling as a getaway-driver saviour & Carey Mulligan parenting despite danger. (Both heroes, really.) Although the hero described in the lyrics is actually Captain ‘Sully’ Sullenberger, who famously landed a plane on the Hudson River saving 155 lives. All 3 heroes, real or fictional, did with calm grit what needed to be done.

🎧Playlist: The JEN_AI Mixtape

// Thanks & How to Get the Good Stuff

If want to be a real hero (ha)- get the latest JEN_AI experiments:

• subscribe to the newsletter (it’s free!) → emailed straight to your inbox

• Or subscribe to the podcast

• add Jenni Munroe on LinkedIn, Instagram or Substack

• click to share, like or comment. Tell me what you want to see!

Thanks to every stranger, internet person or friend who was so friendly at SXSW.

And thanks to YOU, for being part of making JEN_AI real. Side note: I know I promised more ‘service’ and less ‘self-promotion’: clearly I still need to work on that. Next episode’s guide to how to set up AI agents will be v useful for most people, for sure.

Jenni x

• We try taking social media seriously and almost manage it

• I vibecode an app to trash-talk my Instagram (& tell me how to fix it)

• How To Build a Personal Brand, incl. tips from real human LinkedIn experts

• The 13 Things I Wish I’d Known about using clickbait online to make money

• Plus: the #1 tune trending on socials isn’t what you’d expect

This is JEN_AI Challenge #17: Social Media Makeover

// What is JEN_AI?

We’ve all had ideas for a killer app or startup or project (tv show? fashion item? dream event?) that died in our heads given zero time, zero funding, zero clue - and zero mental energy - to make it real.

JEN_AI is an experimental lab, newsletter / podcast / vlog and AI-enabled person (me) turning raw imagination into both physical & digital realness. I’m Jenni Munroe, a recently-ex-Google / DeepMind generative artificial intelligence (‘GenAI’) experimentalist and I’ve been hooked on human+AI creativity since 2010.

I know everyone is sick of AI, and of basically hearing “use AI or get fired,” delivered in the same breath as “AI will replace everyone anyway.” But. AI is here - and I believe we should all get some benefits from it.

Each JEN_AI episode I challenge myself (and you too, if you want to) to build something by mashing up the latest AI tools with both new & retro tech, culture and media. This is a sandbox and community for dangerous dreams - and figuring out what it means to be human in the age of AI.

// Social Mediation

I had other plans for this episode - but escaping from phone screens to printed zines will have to wait til next time.

I realised I did actually need a social media intervention. But not a detox, the opposite: it was (reluctantly) time to dive in. Deep.

When I started JEN_AI last year on maternity leave it was mostly a space for myself; to mentally step away from 24/7 mom-life (even if a half-asleep kid was pulling my hair at the same time) and play with the latest AI tools, journal my reflections and laugh at my own lame jokes. I shared my experiments with supportive friends or at talks/workshops but didn’t really push for growth online.

But now I’ve taken voluntary redundancy from Google, I want to see if I can make JEN_AI something real. To see if I can make a life out of building out crazy ideas through AI+human experiments, and encouraging people to try their own.

There’s just one problem: Pre-quitting, I used to pitch talks/workshops as a Google employee. Now I need “social credibility”. Sigh. And maybe 🙈: - so do you.

And that means we need to find a way to - authentically, minimally - exist in the hellscape that is Social Media.

// How To Build A Personal Brand

“I destroyed my social media algorithm so you don’t have to - and learned X ways to gain Y Followers $$$” 🤢

I did what any normal person would do to sort out their online social presence:

• followed social media “expert” influencers

• asked ChatGPT, Claude & Google Gemini for advice

• spoke to 1) a bestie on LinkedIn’s brand+social team and 2) a LinkedIn Top Voice

• ..and vibecoded a borderline-abusive sassy AI glow-up app to roast my Instagram attempts (in Google Gemini’s AI Studio):

If you want the actual prompt I used, comment “engagement bait” below. (don’t comment that - another lame joke - happy to share the prompt/app tho)

Skip ahead if you don’t want tips.

Here’s a selection of the advice AI+humans gave me about starting out on social media for real:

1. Be more discoverable. Duh. Post things! Many more people will see what’s posted on LinkedIn, Instagram & (if using Substack) Substack Notes vs. just sites/ newsletters/ podcasts/ nothing.

2. Share something of value. Nobody cares about you until after they care about you.

3. Post consistently, not constantly. Pick a posting habit you can sustain.

4. Lead with a hook in the first few words. This is the only thing most people will ever see. Don’t ever start with “Hey guys..” (whoops)

5. End every post with one clear next action. e.g. “DM me for X“ or “Join Y at mylink.com”. Without a CTA (Call To Action), engagement and conversion both collapse.

6. Don’t just broadcast, be social. Add decent comments on top accounts you like, do mutual recommendations, collaborative content. More people will find you and Community > Followers anyday anyway.

7. Also: Stop using vague clever titles and use keywords and hooks with a clear value statement instead (over. my. dead. body.)

// Locked-in on LinkedIn

DON’T hide from LinkedIn, DO THIS to your profile instead:

1. Headline is the most valuable real estate. Don’t waste it on just a generic job title (e.g. “Product Manager”). Say what you do & mention some proof of credibility. think: keywords

2. “About” section: needs a clear summary hook in the first few lines. The rest is truncated and invisible to 80% of viewers.

3. Background banner = a billboard, not decoration. Include text so people get what you do and what they should do next. Stick to your personal visual brand (see below)

4. Featured section = your portfolio. Pin your 3-4 best things at the top that showcase what you do. People often don’t reach below what’s featured.

// Insta-glam

If you are trying to grow a personal brand on Instagram, THIS is the best thing for your profile:

1. Name field = searchable keywords, not just your name. Again, include what people search: e.g. “Jane Smith | Product Designer · Figma + AI”

2. Bio formula: who you help + what you make + proof / credibility + next step. e.g. I help solopreneurs build brands · 12k readers / weekly · ↓ Get brand audit ↓

3. Then one link, not a site of links. (controversial!) Send people straight to your highest-value destination (portfolio / shop / newsletter..). Link sites dilute intent & every extra click loses people.

4. Pick a visual “brand” and stick with it. A set of fonts, colours, styles, fave camera angles, locations, whatever - all of it.

5. Pin 3 posts that tell your story in 3 seconds. Visitors see this top row first. Make them: (1) what you do, (2) proof you can do it, (3) what they get if they follow you.

6. Story Highlights should be a menu. Build 4-5 with a consistent cover look: e.g. ABOUT · WORK · PROCESS · LOVE (intro / best examples / behind the scenes / testimonials)These work harder than any single post.

   ---

// What They Don’t Tell You About (anti)Social Mediums

Well, whether or not I needed a smartphone detox before this challenge, I DEFINITELY need one now. My goodness. I jumped down the content creator snake-eating-its-own-tail hellhole. Hours disappearing into soulless lists of hooks, sales funnel “growth systems,” and increasingly identical videos explaining how to become increasingly identical human robots for $.

Apparently going viral is now less an art and more a kind of “industrial process engineering”. Copy what’s trending. Repeat. Optimise. “Steal like an artist” but with zero artistry and automated replies in your DMs.

And let’s be real: in 2026, the goal of many social media accounts is ultimately to convert viewers → $. Buy me. Buy this. Buy my course about buying my course. Which is fair enough, but it does slightly change the vibe from “friends hanging out online” to “QVC hosted by emotionally available ring lights.”

For many people, social media is no longer a social network. It’s speed-TV. Tiny personalised channels broadcasting directly into the limited-attention economy. And every creator is one Big Tech Algorithm update away from either a financial breakthrough or a nervous breakdown.

Part of why I haven’t pushed harder to grow online is, of course, the desire to pull away from this addictive dystopia. And that’s even before AI panic, deepfakes, and brain rot. Add in cling-film-thin boundaries between public and private life, weird parasocial dynamics, and the occasional stalker horror story, and it’s understandable why some people back away from the whole internet like Homer Simpson disappearing into a hedge. Plus, I’m mildly worried about getting lynched for promoting increasingly unpopular AI.

But. Personal brand matters. Probably even more now, in a labour market getting a serious shake-up. One way to get past the cringe of being visible online is to make sure your “brand” isn’t really about self-promotion. The difference is: providing value. Otherwise, as one creator put it: “It’s like playing jazz flute in a food hall. No one cares that it’s impressive. They’re hungry.”

And I think I’m ready for that shift now. Going from “reclaiming my sense of self” as a parent / corporate drone / an NPC in geopolitics to “here’s something useful, funny, interesting, or encouraging that might genuinely help someone else.” Moving from self-promotion to service, and helping others build their dreams and ideas too. Watch this space.

Ironically, I spent more time this week reconnecting with actual friends than optimising social media strategy. So I’ve yet to implement half the advice i was given above. Watch out world, more to come. My Instagram grew >30%, but it was by adding people I know.

Followers are nice, but community is better. I’d rather nurture real relationships and build something that feels like finding your people than endlessly chasing strangers through an algorithmic casino.

I’ve seen advice from creators that adding more people you know is a route that “limits your growth” and friends aren’t “the right audience”. But in this ‘loneliness epidemic’, what we need more than an audience is real community. People we trust, lifting each other up and lovingly tearing us down to earth.

We’d be fools to take social media - or ourselves - too seriously.

// Soundtrack

• “Human Nature - Ambient” by Stayunseen (the #1 top trending track on Instagram at the time of writing: an ambient/ vocals-free remix of Michael Jackson’s “Human Nature” - a song about wanting to connect and in doing so being pushed further away (via MJ’s growing fame at the time - less of an issue for most of us). A remix that may or may not have rights to use the original. There might have been something to say about that but alas - for this track there are no words)

Playlist: The JEN_AI Mixtape

// Thanks + how to get the good stuff

You can listen in podcast form on Apple podcasts or Spotify, watch the videos on YouTube or autoplaying on Substack (apologies), check out the playlist, or read this newsletter, which is also available in the Substack app or via email straight to your inbox if you’ve subscribed.

A big thanks to (a LinkedIn Top Voice, fellow Substacker and Founder Advocate at Google) for a chat on navigating media, AI, wellness & more, and also thanks and love to a treasured bestie from LinkedIn who gave me a few general pointers in the right direction. She knows who she is but I don’t want to get her in too much trouble ;)

Some ways to support JEN_AI if you like it:

• subscribe (it’s free!) on your preferred channels to get sent new episodes

• click to “share” it with a friend or the internet.

• click “like” or comment your thoughts.

Tell me - what do you want to see on JEN_AI?

Share

Thanks for being part of this social experiment.

Jenni aka JEN_AI x

Today, on the JEN_AI program..

• How to create your own automated news agent

• ..why that might be a terrible idea

• How the sequel to an iconic 20-year-old movie is today’s news foretold on the silver screen

• Plus: 3 topical tunes

This is JEN_AI Challenge #16: personalised, automated news

// What is JEN_AI?

If you’re new here, welcome! I’m Jenni Munroe, a recently-ex-Google/DeepMind generative artificial intelligence (‘GenAI’) experimentalist who’s been hooked on human+AI creativity since 2010. I’m challenging myself (and you, if you want to) to stress-test and mash-up both new & retro tech, culture and media, and to grapple with what it means to be human in the age of AI.

// What it means to be a news agent

to clarify a bad pun: “Your local newsagent” is a little shop near your home in Britain selling essentials including newspapers in real, printed form. Imagine.

An AI “agent” is a computer program that can carry out actions to achieve a goal. Think: ‘here’s my computer and the internet, go make me $100’ vs. asking a chatbot to answer a question. The more it relies on autonomously sensing, planning and actually doing things, the more “agentic” the behaviour.

A “journalist”, or so a chatbot tells me, has standards; “a commitment to truth”, “independence”, “holding those in power accountable”, and “contextualising” what happens in the world. ..And in the UK is a registered member of a ‘trusted profession’ respectable enough to countersign my kid’s passport application (thanks Dan!).

// Not such a great AI-dea

Last week I got an email from an AI-positive journalist, Casey Newton (who co-hosts the excellent New York Times’ tech culture podcast) to let his subscribers know he was having to pivot his personal media platform away from news aggregation and analysis towards investigative reporting - due to AI encroachment.

Also: three nights ago I watched an unexpectedly prescient documentary about the fight to save journalism, art, beauty and humanity from Tech Bro Billionaires. It was called: “The Devil Wears Prada 2”. You think I’m joking.

Interesting timing, then, for me to plow on ahead with my planned JEN_AI experiment to show how people can dip a toe into setting up AI agents - by creating their own scheduled newsletters. Which I’m writing up in my own presumably-AI-threatened newsletter about AI.

Hmm. Let’s see how this one goes down.

// How To Make The News

An actual journalist might have thoroughly reported on a balanced, unbiased selection of AI agents. However. We are keeping this simple:

At time of writing Google Gemini has arguably the easiest, most underrated way to set up an agent to send automated briefings sent directly to you each week/day/whatever as a native notification on your phone.

Steps to set up a scheduled news report on the Google Gemini app:

0. If you don’t already have the Gemini app on your phone, download it & log in.

1. Type in the prompt “Schedule a [daily/weekly] briefing for me at [time, day etc.] on [topic(s)]. Generate one now.”

That’s it. Now for the experiments. Some things I tried, from most to least successful:

• A newsletter that decodes clickbait titles from the week’s ‘Diary of a CEO’ podcast episodes, & distills claims into fact-checked life advice so I can decide which to listen to.

• A daily summary of global news headlines and market moves. Sources were linked and showed we’re certainly pulling from lesser-known news outlets here. Clicking the “speaker” icon to play it out loud over breakfast was surprisingly acceptable. (Monotone newsreader voice is surely still the most at-risk category for AI voice replacement, seeing as it forces humans to speak like news-robots.)

• A daily AI & frontier tech briefing. I get a lot of AI newsletters (many written with the help of AI) and the selection of stories covered by my agent was slightly - I won’t say “random” - but certainly needs some fine-tuning. I did like being able to ask follow-up questions directly in the newsletter window.

• A morning briefing from my to-do list and calendar. The first attempt got me a hallucinated BBQ. The second try got my actual to-do list from Google tasks & checked my emails, but this works best if you A) have Personalisation enabled in settings and B) already use the Google ecosystem (Gmail/calendar/tasks etc.)

• A Thursday “local concierge” roundup of family-friendly events near me & the weather forecast for the upcoming weekend. This probably looks more exciting if you don’t live in a rural village.

I iterated on all of these a few times (asking Gemini to update the instructions) to get something slightly more verified and better fitting my content and format preferences.

For the nerds: you’ll know that in computing, “not local” means “not on your device”. In later JEN_AI episodes we’ll cover using sophisticated agents on your computer with more personal context and with greater capabilities to take actions to achieve a goal (control your browser, create local files etc.)

Anyway, yes. AI can encroach on news aggregation and analysis. And these were only a few quick examples from one AI tool. (Perhaps I should try pivoting to this AI-proof “investigative reporting” thing?)

But. A friend of mine and previous JEN_AI guest Caroline Crampton runs an email newsletter called Thursday Thirteen, which is a roundup of interesting links. But I don’t read it just because the links are interesting - I read it because she thinks the links are interesting. In a single bullet point for each link Caz manages to communicate her reaction to the linked media, her personality and her editorial context in choosing it. To read it is to connect with a real human I care about. Maybe there will be different lanes for AI news, but I can’t see that demand for human meaning, taste and connection going away. But we do need to work out how people get paid.

// The Devil Wears Patagonia
(Or possibly Allbirds shoes & GPUs)

Talking of real humans in media [but BEWARE movie spoilers, skip this section], the devil in the Devil Wears Prada 2 is not Meryl Streep’s Vogue fashion magazine boss Anna Wintour but Amazon CEO Jeff Bezos (with wafts of Elon Musk), who owns The Washington Post in real life and laid off roughly 1/3 of its workforce in Feb 2026. Also portrayed: his philanthropic reclusive ex-wife Mackenzie Scott, new spouse Lauren Sanchez Bezos (the former journalist’s wedding was begrudgingly covered in - yes - Vogue magazine), and a possibly generic Silicon Valley Patagonia-vest-wearing venture capital tech bro. The release of the movie this week was timed for fashion’s annual highlight for added PR: the Met Gala, of which event Jeff Bezos and Lauren Sanchez were later announced as honorary chairs, a step which sparked real-life protests and boycotts peaking this week. The Met Gala is quite literally a costume party (it’s based on the Met’s “Costume Art” exhibition this year) of wealth promenading as art (many of the dresses worn to this $100K ticket event are designed to look like famous paintings and sculptures). With both the movie’s main characters and actors in attendance. Life imitating art imitating life and art. With typically bold art-fashion looks.

Some fans made their own AI-assisted fashion creations which I won’t reproduce to respect the real celebrities unwillingly starring in them, but some are so good you could even call them.. art. Aside: today in this podcast video recording I’m in costume as Anne Hathaway’s perpetually-in-need-of-a-fashion-makeover-and-a-better-job journalist character Andy Sachs, just to join in all the costume fun.

Now look. If watching the first Devil Wears Prada movie wasn’t a core memory for you, if you’ve never called Silicon Valley or New York home, if you’ve never experienced mass layoffs determined by a Big Tech CEO, never simultaneously exhaustedly despaired at technology’s relentless destruction of humanity and its culture while getting paid to use tech to automate labour nor used AI in your leisure time to re-imagine art or journalism.. And if you’ve not once been pursued by media outlets for an exposé on the alleged behaviour of the global powerhouse executive ex-boss you assisted, this film may seem less personal for you than it did for me. By all means, skip it.

One thing that was notably (thankfully) absent from DWP2’s film fantasy was sovereign government politics. Tech CEOs themselves must meet the decrees of governments (and I gather that governments act on their citizens’ behalf?). Australia for instance has banned social media for kids and is threatening a journalism tax for tech companies who do not cut deals with publishers. With the developing cybersecurity issues triggered by AI, who knows: U.S. governmental.. governance (?) for AI might even be on the horizon.

This Devil Wears Prada sequel thankfully ditches the actual second book’s plot about the two ex-assistants starting a high-end bridal magazine (!) and serves up a fantasy where there is a way forward that combines human taste, labour and artistic direction, journalistic integrity, technology and tech investment. A world where people eventually get things they have earned and don’t have to overlook their values and dreams in order to make a living. Call me as irritatingly naive as Andy Sachs, but maybe we could try it?

Personally, I think the humans of the world are desperate to get off their phones and are ready for a nostalgic return to both off-screen/print media and in-person events & communal spectacles like festivals, concerts and fashion shows. Even if they evolve. More on this next episode.

Last thing. Who fills the role of devilous villain, triumphant saviour or victim/assistant switches depending on your perspective. The movie’s parting thoughts will not be news to you: we are human, imperfect. We have.. agency. And we don’t need to invisibly wait, shelter or wilt in someone else’s shadow. “You are iconic.”

// Soundtrack

I setup an agent to give me the downlow on the top music downloads (or rather, tell me the ‘most streamed music tracks in the UK each week & why they’re trending’). At no.1 (after some ‘clarifications’):

• “Rein Me In” by Sam Fender & Olivia Dean (apparently it’s the Song of The Year at The Brits, in the Season 4 Finale of The Bear, and a TikTok meme)

But.. the human in me had already chosen these also-trending tracks:

• On the tragicomic pursuit of agents: “Price of It All” by Jorja Smith. A James-Bondesque ballad from recent show “Bait” about a South-Asian man auditioning to be the next 007.

• On prevailing: “Shape of a Woman” by Lady Gaga. This was an absolute moment in the Devil Wears Prada 2 movie.

Playlist: The JEN_AI Mixtape

// Thanks + how to stay in touch

You can listen in podcast form on Apple podcasts or Spotify, watch the videos on YouTube or autoplaying on Substack (apologies), check out the playlist, or read this newsletter, which is also available in the Substack app or via email straight to your inbox if you’ve subscribed.

If you like it, you can support me by clicking “like”, subscribe (it’s free!) to get sent new episodes via your preferred channels or click below to “share” it with a friend or the internet. Comments & feedback welcomed!

Share

Thanks for being part of the experiment.

Jenni aka JEN_AI x

And yes, JEN_AI is back, with more AI+human challenges and even more freedom to try them out.

Which feels like the right moment to tell you about the time I walked out onto the main stage at a major tech product conference in a fluffy bathrobe and dog slippers to demo how anyone can use AI to build a product (including YOU). Specifically: how to make a makeover app that recommended the real outfit I “transformed” into onstage, assisted by confetti, a novelty shower curtain, and five brave volunteers.

If you’re new here, welcome! I’m Jenni Munroe, a now-ex-Google/DeepMind generative artificial intelligence (‘GenAI’) experimentalist who’s been hooked on human+AI creativity since 2010. Every episode I challenge myself (and you, if you want to) to mash-up new & retro tech, culture and media, and work out what it means to be human in the age of AI.

This is JEN_AI Challenge #15: A Main-Stage Keynote Like You’ve Never Seen Before.

// How Did This Happen?

Last year Google encouraged our team to attend an AI-related conference. Future Product Days in Copenhagen, a festival of tech, product, AI and design, was on the list. Since I was going anyway, I offered to host a workshop on vibecoding (=’using AI to build things’, like a makeover app from a previous JEN_AI challenge). A week later the organiser wrote back: a speaker had cancelled, and there was a main-stage slot.

I said yes, obviously. And then decided it would be much more engaging to give myself a full-body wardrobe transformation while demoing how to build an even better makeover app. Obviously.

// We Can All Make Things Now

You could always speak a few magic words to spin up software; we call it ‘coding’. But now you can say even fewer words in whatever language you want. Here’s the three-step version I taught from the stage, plus the Google tools I used. Skip ahead if you’d rather not nerd out.

1. The Problem. Open Gemini Deep Research (gemini.google.com, then Tools, then Deep Research) and tell it your product idea, the problem you’re solving, and any constraints. Ask it to analyse the market, role-play as stakeholders poking holes in your value prop, and draft a product requirements doc. Or point it at your existing product and user feedback.

2. The Prototype. Go to ai.studio/build and tell it what you want. Upload any useful context from step 1: brand guidelines, reference screenshots, your half-written rant about the problem. If your prompt feels off, ask Gemini to write a better one.

3. The Promo. flow.google lets you generate start and end frames for promo vid clips and stitch them together. Google Vids pulls together recordings, slides, AI avatars. Honestly, consider doing the promo before the prototype, to find out whether anyone actually wants the thing.

Two things that matter more than the tools:

• Context is king. Real examples, templates, data, style guides. Upload them: everything gets better.

• Augment, don’t replace the human. Newcomers can finally ship something, and experts can go further, faster.

I know all the cool kids are running Claude on their laptops now. More on that another time. I picked Google tools because it was a Google talk and because Gemini’s image and video skills are genuinely strong for a glow-up demo. AI Studio is still a great place to start tinkering, especially if you don’t want to buy credits upfront; Lovable, Replit and Bolt are also beginner-friendly. If you want my actual prompts, message me or comment and I’ll send them. The ‘VIBE MODE’ JEN_AI episode shows in more detail how to use Google’s AI Studio to make apps and websites.

Share

// If I Could Make It All Over Again

• I’d make sure I didn’t somehow still look like a corporate swamp witch even after the glow-up.

• I’d show that AI can do more than “build an app,” and that a makeover is about more than clothes. Imagine being inside your favourite Netflix makeover show, Queer Eye, or even the interior-design ones, where the presenters make their guests feel understood and valued. They are helping people get to the better life they deserve after all they’ve been through. That becomes the personal-shopping experience. Too dystopian? Too millennial? Maybe.

• Yes, at times, I wanted it to just be over. Talk prep is more anxiety-inducing for me than podcasts or newsletters or building things. But it was worth it, and the audience was amazing. A lot of you came up afterwards or wrote really lovely LinkedIn posts. Thank you - it meant a lot.

Leaving Google wasn’t because of this talk (clarifying- in case you were worried). I took voluntary redundancy for a lot of reasons. But if I’m honest, the talk was symptomatic: the kind of things I wanted to do, in the ways I wanted to do them, didn’t quite fit how I fit into Google (even though this talk did somehow clear the many Google pre-reviews). JEN_AI is where I get to do that freely. I’ve left to build a life that fits around my family, but it doesn’t have to be that drastic for you. I stand by the lessons from the ‘Slide into my TED Talks’ JEN_AI episode: You can experiment. Pitch things. Put yourself out there. Let people know what you’re trying to do. I believe we can remake the job and the life we want, and I think we should all get some upsides from AI before the world ends. And we aren’t re-making ourselves from nothing - we all already have more resources than we realise (experiences, networks, time I normally spend doomscrolling etc.).

// The Devil Wears Prada Meets Silicon Valley

Speaking of makeovers: The Devil Wears Prada 2 is out May 1st, timed to coincide with the Met Gala. It’s been 20 years since the original film, and whether the it flops or not, I can’t resist a bit of nostalgia. Also, watching method-marketing movie campaigns unfold these days is a spectator sport I’m often enjoying more than the films themselves. This time it’s Madonna and Sabrina Carpenter doing “Vogue” at Coachella, Cerulean Blue everywhere, fashion moments, magazine covers, questionable drink collabs etc. etc. The sequel’s theme is the battle between print and online media, which is real for Vogue, for cinemas, for books, and for our own attention as it disappears into our phones and AI eats everything.

So. I asked Gemini to help me write a Silicon Valley riff on Meryl Streep’s iconic Cerulean Blue speech, the one where she dismantles Anne Hathaway’s smugness and her “bargain-bin” Cerulean-blue sweater and explains how the whole global fashion cascade works, that no-one can escape it or is above it, and the colours and the clothing we all wear were ‘chosen for you by the people in this room’.

But of course, The Devil Wears Prada is proprietary material. So Gemini said it couldn’t access it. Fair enough.

Except. Look at the Gemini logo. Look. At. That. Blue. Hate to break it to you Gemini - you think you’re above accessing ‘The Devil Wears Prada’ movie but you’re apparently wearing the colours chosen for you by the people in that room.

Alright, yes, that was a very silly joke.

But the question is: these days, who’s choosing for whom?

// Meanwhile, Allbirds swapped Shoes for GPUs

Last week, the eco-sneaker darling of Silicon Valley made a surprise pivot to selling AI infrastructure. (And merely 8 days after launching a 14-colour Pantone collab.) Yep, ‘AllBirds’ tech-bro shoes are now ‘NewBird AI’ cloud-GPUs. Their stock price jumped.. 600%.

So hey it looks like everyone’s getting an AI makeover (whether we wanted one or not.)

// Sounds For When You’re Feeling Blue

• “Vogue” by Madonna and Sabrina Carpenter at Coachella. The theme to the original Devil Wears Prada.

• “Runway” by Lady Gaga and Doechii, from Devil Wears Prada 2.

• “Shades of Blue” by John Summit, a former accountant who quit his 9-to-5 to become a global dance music superstar. Relatable.

Playlist: The JEN_AI Mixtape

// Thanks + how to stay in the loop

You can listen as a podcast, watch the videos, strut to the playlist, or read the newsletter. If you like it, subscribe (it’s free!) and share it with a friend or the internet. Comments, feedback and savage put-downs (à la TDWP) all welcome.

Thanks for being part of the experiment.

Jenni aka JEN_AI x

3 quantum researchers were just awarded a Nobel Prize for experiments they ran four decades ago. Why? They showed that quantum effects weren’t confined to sub-atomic scales; they could be harnessed across an electric circuit large enough to hold. And the laureates’ recent well-publicised breakthroughs in quantum computing showed just how big a deal their work is.

In other words: insiders have known for a while that quantum physics could be big. But it’s having a huge (magnetic?) moment right now.

I’m Jenni Munroe, and you’ve found JEN_AI: an AI-human experimental creativity lab specialising in generative artificial intelligence (a.k.a. GenAI) and bad puns. In this video/newsletter/podcast series I challenge myself (and you too, if you feel like it) to use AI, tech, culture and media to learn the latest in AI, tech, culture & media, and try to work out what it means to be human in the age of AI.

This week, I dust off my rusty Oxford Physics degree and grab some AI tools (Google’s Gemini, NotebookLM, Nano-banana & Veo in Flow, plus Descript) to help me co-generate quantum content.

This is JEN_AI Challenge #14:

AI-assisted Science Communication.

If you’re like me, you might be concerned that we seem to be struggling with two irreconcilable perspectives on the laws that govern us. Big vs. small. But there is a grain of hope that we can unify the two sides. Let me prove it to you.

// WTF IS A QUANTUM

What does ‘quantum’ even mean? The word comes from the Latin quantus, meaning “how much.” It refers to the idea that at the tiniest, most fundamental levels, things like energy come in discrete, individual packets, or ‘quanta’. Like a staircase instead of a smooth ramp. You can stand on step one, or step two, but there is no step one-and-a-half.

Quantum physics is the set of rules that governs this staircase-like world of particles smaller than atoms. And the rules are very odd: Particles can be in multiple places at once. They can be linked across vast distances. And they behave differently when you’re looking at them or interacting with them. It’s a world that operates on probabilities.

// No Cats Were Harmed in the Making of this Ridiculous Physics

You’ve probably heard of “Schrödinger’s Cat”: a story about a box containing a cat which is simultaneously alive and dead. The imaginary cat-box is connected (via a very convoluted setup) to a hypothetical radioactive atom. If the radioactive atom decays, a vial of cat-poison is released into the box. If the atom doesn’t decay, the cat lives. But the atom follows quantum rules: so it isn’t simply ‘decayed’ or ‘not decayed’; it exists in a kind of fuzzy, furry “superposition of both states” - and the cat is both alive AND dead - until someone looks at it.

Like a coin spinning in the air: while the coin spins, it’s not heads or tails. It’s a blur of both possibilities.

So yes, this is the story of Schrödinger’s Cat, but the point of the “thought experiment” wasn’t to say “this is how the world works”: Erwin Schrödinger was trying to point out how ridiculous it would be if the bizarre rules of quantum physics applied to something as big as a cat.

He was saying: Cats can’t be both dead and alive.. so at what size does the quantum world break down and our ‘classical’ world begin? Why do tiny particles get to live in this strange state of multiple possibilities, but larger objects like cats, buses, and people don’t?

He was saying: Surely quantum physics doesn’t work for things the size of cats. Right, guys?

Turns out he was wrong.

// The 2025 Nobel Prize in Physics: They Won for Making Quantum Science Bigger

John Clarke, Michel H. Devoret, and John M. Martinis won the 2025 Nobel Prize in Physics. The official Nobel Prize press release puts it perfectly:

“A major question in physics is the maximum size of a system that can demonstrate quantum mechanical effects. This year’s Nobel Prize laureates conducted experiments with an electrical circuit in which they demonstrated both quantum mechanical tunnelling and quantised energy levels in a system big enough to be held in the hand.”

This is huge. (Paha)

The whole point of the cat story was that we don’t see quantum weirdness in our everyday, hand-held world. But these scientists created a system, a circuit you can physically hold, that behaves in a way you can only predict using quantum rules.

What did they actually do?

They built a supercooled, superconducting (zero resistance) electrical circuit featuring a “Josephson junction”- two superconductors separated by a thin insulating barrier.

Initially, the circuit was settled in a stable zero-voltage state. It was “trapped” in an energy well, similar to a marble resting in the bottom of a bowl, lacking the classical energy to get over the rim.

First, they observed Macroscopic Quantum Tunneling. They watched the system spontaneously escape this trap and switch to a state with a measurable voltage. This was a remarkable feat: the collective state of billions of electron pairs was behaving like a single “artificial particle” and tunneling directly through the energy barrier - an act forbidden by classical physics.

Second, they proved its quantum (=steps not ramps) nature. To confirm this, they exposed the circuit to microwaves. They found it only absorbed energy in specific, discrete amounts (quanta). When the microwave frequency was just right, it “kicked” the system to a higher, distinct energy level, causing it to tunnel out of the trap much more rapidly.

This was the smoking gun: it proved the circuit had quantized energy levels, behaving like a single, large-scale “artificial atom” whose quantum state could be precisely controlled. More info here.

They built a bridge between the microscopic quantum world and our macroscopic classical human world. But they did these experiments in the 80s! Why the Nobel now? We can only guess. But did you know Einstein technically never received a Nobel Prize for his most famous work, relativity; it was considered too contentious at the time, so he was officially awarded the prize for a “safer choice” experiment, even though it was generally understood that the prize was in recognition of relativity and his wider work. Speaking of the 2025 laureates’ wider work and what may have prompted recent recognition:

// Quantum Computing is in a Super Position Now

What do you do once you can control a “big” quantum system? You build a quantum computer.

Your phone or laptop works with bits. A bit is a simple switch: it’s either a 0 or a 1. Off or On. That’s it. A quantum computer uses qubits. A qubit can be a 0, a 1, or - thanks to “superposition” - it can be both at the same time. Like the spinning coin. In fact, particles like electrons have a quantum property we call spin, which we can use to stand for 0s or 1s. Spin can be up, it can be down, or it can be in a superposition of both, described by what we call a wavefunction: a mathematical equation that describes all the possible states the electron could be in. What’s weirder is - the electron actually IS in all these states. (You could say the universe is holding the space.) But when the spin is measured, the wavefunction collapses, and we get our answer.

A classical computer is like searching a library for a specific piece of information by checking one book at a time. It’s fast, but it’s linear. A quantum computer, by using superposition and another trick called “entanglement”, can essentially check every single book in the library simultaneously.

What will this allow us to do? We could break all of our current security encryption methods and create uncrackable new ones. We could design new materials and medicines by simulating molecules with perfect accuracy. We could solve hugely complex optimization problems in finance, shipping, or weather forecasting that would take today’s best supercomputers billions of years.

So, what’s so difficult? If you look at a one of the latest prototype quantum computers, most of what you can see is basically a supercooled, shielding refrigerator. The main challenge they are battling is a problem called “decoherence”. Qubits are incredibly fragile. The slightest vibration, temperature change, or stray magnetic field can knock them out of their delicate quantum state and make them behave like a normal, boring bit. It’s like trying to balance a billion sharpened pencils on their tips, all at once, during a low-key earthquake. (Or like trying to use a cat in a controlled physics experiment.)

This is precisely why creating a stable, larger-scale quantum system was such a breakthrough. It was a crucial step toward building a useful, fault-tolerant quantum computer. In fact two of the three Nobel prize winners were former or current leads at Google’s Quantum AI lab, and during their tenures have demonstrated three key breakthroughs:

“Quantum Supremacy” (=faster than regular computers): In 2019 (with the Sycamore chip) the team were able to perform a calculation in minutes that would have taken a supercomputer millennia. The goal wasn’t to solve a practically useful problem, but to demonstrate that a quantum computer could perform a task far beyond the reach of any classical machine.

Quantum Error Correction (=better stability): Using the Willow chip in 2024, the team showed that by combining many physical qubits, they could create a more stable “logical qubit” with a lower error rate, providing a clear path to building the reliable, large-scale quantum computers needed to solve real-world problems.

Just announced in Oct 2025: “Verifiable Quantum Advantage”: the team used new ‘Quantum Echoes’ software on Willow chip hardware to simulate a molecule: a checkable, real-world scientific problem completed 13,000 times faster than top supercomputers.

//So Many Worlds

I feel like I should mention the aspect of a Google quantum computing blog post that made it onto a Joe Rogan podcast - “the notion that quantum computation occurs in many parallel universes, in line with the idea that we live in a multiverse”

Let’s bring it back for a second - Physicists agree that the maths of quantum mechanics works, but they disagree about what it means. The “Many-Worlds Interpretation” of quantum physics suggests that instead of a single reality, every quantum measurement with multiple possible outcomes causes the universe to split into a corresponding number of parallel universes. In this view, there’s no “collapse” of possibilities; instead, all potential outcomes actually occur, each in its own separate, unobservable world. For example, if a particle can be in two places at once, the universe branches into two, with the particle in one location in one universe and in the other location in the second. We only perceive one of these outcomes because we exist within a single branch, unaware of the countless other “yous” living out every other possibility in an ever-expanding multiverse.

// That’s nice, but we’ve got bigger problems: A Divided Universe

This brings us to one of the most embarrassing and profound open problems in physics: we have two sets of rules for the universe. And they don’t join up.

1. General Relativity: Einstein’s theory of the very large. It describes gravity as a smooth, continuous fabric of spacetime being warped by massive objects like planets and stars. It’s beautiful and works perfectly for things like GPS and black holes.

2. Quantum Mechanics: Our theory of the very small. It describes the universe as grainy, probabilistic, and fundamentally uncertain.

When you try to use both sets of equations in extreme places, like the Big Bang or inside a black hole, the maths breaks down completely. Finding a theory of “quantum gravity” that unites them is the holy grail for Physics. Unfortunately, for now, it seems the science of the very small and the science of the very large defy each other.

// A Quantum of Solace?

But there is reason to hope. While General Relativity, our theory of gravity, refuses to cooperate with quantum physics, we have successfully “bonded” quantum mechanics with special relativity. And the result of that union isn’t just a ‘theory’ - it’s one of the most precise and accurate calculations in the history of physics.

// The Crown Jewel of Physics: QED.

Let’s talk about QED. This often stands for the Latin phrase “quod erat demonstrandum”. Meaning: “which was to be demonstrated” or “that which is shown”. It’s typically placed at the end of a mathematical proof or philosophical argument to signify that the proof is complete. It’s a way of saying, “I have proven what I set out to prove.” The Latin phrase is a translation of the Greek which was used by early mathematicians like Euclid and Archimedes. It’s a formal mic-drop conclusion to a logical argument. 🎤

But I didn’t mean that QED, I meant Quantum Electrodynamics

QED describes the relationship between light and matter.

QED shows that all interactions between matter and light are based on one fundamental action: a charged particle, such as an electron, either emits or absorbs a photon of light. For example, two (negatively charged) electrons repel each other by exchanging a virtual photon. Imagine throwing a ball (of light??) to a friend while you’re standing on an icy lake: the momentum would push you away from each other.

The level of agreement between what QED maths predicts and what we measure in experiments is staggering.

For instance, QED also predicts a property of the electron called its magnetic moment. A simple theory says this value, the g-factor, should be exactly 2. But QED accounts for the fizzing sea of virtual particles that an electron interacts with, and it predicts a value ever-so-slightly different.

• Predicted by QED: g=2.00231930436256...

• Measured in the Lab: g=2.00231930436153...

They agree to about one part in a trillion. It’s like measuring the distance from London to New York City and being accurate to within the width of a single human hair.

// Old Story, New Spin

So look, that’s my take. 100 years ago, quantum physics was the science of the unimaginably small, with behaviour so strange it seemed it would never be relevant ‘in real life’, at human scale. But we are learning to bring that strange and powerful reality into systems we can hold, into computers that will reshape our world. And by the way, all our computers already rely on a different kind of quantum technology: it’s called a transistor, in a microchip.

To quote the Nobel press release: “Quantum mechanics is the foundation of all digital technology”

But more than that, I suspect there’s a small story here (not to mention a JEN_AI Challenge) about science communication. Or education. Or Public Relations? If you saw my AI “TED Talk” JEN_AI episode, you’ll know that It’s not enough to do great work. People need to know about it. To what extent does improved awareness of the value of a breakthrough contribute to increased recognition? Well the clue is in the question - recognition and awareness of value are all but the same thing. But it’s not just about awards - it’s about impact. The history of science is a tale of building on and improving what came before. But we have to know about it to build on it.

This probably won’t be the last Nobel awarded for quantum computing & its foundations. (And goodness knows who’ll win the Peace prize next.)

Einstein did win a Nobel, by the way. What was it for? Not relativity, but the quantum photoelectric effect. Work which laid a foundation for yet another Nobel: for a comprehensive and astonishingly accurate theory of light and matter. The first theory to fully unite special relativity and quantum mechanics. Sound familiar?

...Which reminds me, you might be wondering: why did I make this episode? What did I set out to prove, to show?

• That quantum science and quantum computing are both larger in size now and having a bit of a moment

• That you can use AI to usefully and enjoyably assist with science communication and learning

• That physics has stories worth listening to

• …but if it’s too much to listen, at least this ASMR-style video can help you get to sleep!

• That effective comms could help you win a Nobel prize

• That maybe it’s ok if recognition comes later

• That there are weird parallels with the Wicked movie/musical/PR? (we’ll revisit)

• And lastly that there is at least a quantum of hope that we can unite both quantum mechanics and gravity. And maybe even our divided world.

QED, bitches. 🎤

There is hope. Always. Who knows - maybe we’ll never find the common ground that truly unifies two vastly different ways of seeing the universe. And maybe our efforts will never earn us a quantum iPhone or an oh-so-coveted Nobel prize.

But that isn’t why we try.

It’s not about the prize.
And it was never about the size.

Apologies for explaining everything practically at the speed of light.. I know it can get a bit heavy. Ha, ha. (SPEED OF LIGHT JOKE: things get relativistically “heavier” the closer they travel to the speed of light. #sorrynotsorry)

// AI Explainer Tools

If you’re uncertain on the principles or not sure what entangle meant, check out some experiments in Google’s NotebookLM: this fully AI-generated explainer vid, or the auto-generated podcast, or chat to the source material.

You can also jump down the rabbit hole by chatting with Gemini or your favourite chatbot. (There’s a “Guided Learning” mode in the Gemini app as well - click Tools.)

// This Episode’s Theme Song

“Big Energy” by Latto. Yes, QED has BDE

JEN_AI is produced by Jenni Munroe using publicly released AI tools (incl. Google’s Gemini suite for edits, brainstorming, images (Nano Banana), videos (Veo) & explainer tools (NotebookLM), and Descript for video & podcast editing). Not endorsed by Google. My views.

👉 subscribe to get future blogcasts straight to your inbox (it’s free)

👉 share your thoughts, complaints & experiments

👉 send this to a friend who’s having trouble sleeping, it should knock them out cold

This episode is dedicated to my friend Alison Carroll, External Comms Lead for Google’s Quantum AI team (who will be truly horrified by this unvetted post). I might never be able to prove it, but I know that at least a *tiny quantum* of this Nobel Prize is hers too.

Thanks, as always, for listening.

Happy Halloween!

#jenai #quantumphysics #nobelprize2025 #quantumcomputing #quantumisbignow

• Plus: 1 theme song, 1 supermodel & at least 3 bad puns

// Welcome to JEN_AI

Hi, I’m Jenni Munroe: generative artificial intelligence (=GenAI) keen-bean since before it was invented. I set myself challenges to use AI, tech, media & culture and try to figure out what it means to be human in the age of AI.

This is JEN_AI Challenge #12.5: Use AI To Give a Talk

(=co-create a conference speaker proposal, talk + slides)

// (Super)Model Training Data: My All-Time Fave TED Talk

Why do I love this talk so much? Model Cameron Russell walks out on the TED stage in a killer black dress that instantly alienates half the female audience, and with perfect comedic timing says “I knew I shouldn’t have worn this”. Over the course of an outfit change and a 10 minute talk she proceeds to “transform what you think of me” - both showcasing and tearing apart the importance of image. She not only avoids the classic rehearsed-to-death TED-speech-rhythms but somehow manages to talk about privilege without making people squirm. And she speaks with a humility and humour that seem effortless. A model example, for sure.

// Start Before You’re Ready

Back in Feb I made the mistake of logging into my work laptop on maternity leave and spotted a call for speakers at Google’s “Women TechMakers” worldwide local conferences. I’d attended one of these events in 2017 and it was the spark that pushed me to chase my dreams & apply to DeepMind, so I really wanted to pay it forward and inspire others to overcome self-doubt and just go for it.

There was just one small problem: the deadline for proposals was 5pm that day. And worse: I only really had half an hour of baby-nap-time to write and submit this thing. So yeah, I turned to AI. Fortunately, (and yes, terrifyingly), ChatGPT already knew enough about JEN_AI in memory that when I gave it the AI the call-for-speakers email text incl. conference theme “Redefining the Possible You”, it came out with something just about good enough. I only needed a few iterations & edits, then BAM- submitted.

The pitch was selected and in March I gave the talk virtually at the Rochester Institute of Technology in New York.

But first I had to write the damn thing.

// HOW (NOT) TO WRITE A TALK

• Iterating with ChatGPT, Google Gemini or other fave chatbot (plus my own stories) to craft an outline and talk

• Gamma.app and/or Google Slides (I used both) to generate, edit and present a slide deck from an outline

For the full magic, the vid has the real-time slide-generation.

// The Result

Check out the video/podcast for some total-cringe clips from the talk! Slide Deck

// Reflections: The Pain of the Cringe

I’m so glad I did this challenge, but after nearly a year of maternity leave I was rusty AF and a walking panic attack for hours afterwards.

I saw a quote from communication guru Vinh Giang saying, basically:

“you can be great at anything, you just have to be willing to be shit at it first”

Sometimes what feels like imposter syndrome just means you’re honing a new skill. And sometimes, you’re a genius with terrible self esteem. And at other times still- you legitimately are a bona-fide AI imposter. Either way: apologies to the 100 computer science students who had to listen to me remember how to speak derivative, trite AI-co-created nonsense as they stare down a job market that AI has not helped.

But the conference organizer shared some lovely feedback from students and a few did reach out to me afterwards, so it wasn’t all bad. Seemingly it was my (far from original) message around overcoming fear, self-doubt and imposter syndrome that resonated most, and if it encouraged just one person to make a leap then it was worth it. Even if that one person was me.

// SOUNDTRACK

This week’s theme tune is about the sheer terror of putting yourself out there: sharing your art and authentic self with the world.

• “Alaska” by Maggie Rogers

Back in 2016, Masterclass recorded an NYU student Maggie Rogers playing her song “Alaska” to mega-successful artist and producer Pharrell. You can see she was terrified of him hearing it - she can barely look at him - but he’s so genuinely wowed by her music and can barely hold it in till the end of the song to tell her how perfect and unique it is. Watching the range of their emotions as they sit in silence for 3 minutes listening to her music was so deeply moving for me (& plenty of other people who spend too much time on the internet).

So, yeah. Put yourself out there. And if you think you’re not good enough - you’re either learning something new, or you haven’t been listening to this episode hard enough.

🎧 Playlist: The JEN_AI Mixtape

// Disclaimer & Thanks

A big thanks to Alaina Mupparthi and the Women Techmakers organizers, and thank YOU, as always, for listening. 💖

JEN_AI is produced by Jenni Munroe using publicly released tools (incl. Gemini & ChatGPT for edits & brainstorming). Not endorsed by Google. #myviews

#jenai #GenAI #experimental

Share

🤖 What is JEN_AI?

A vibe-hack-challenge blogcast series about AI, media, and what it means to be human in the age of generative-artificial-intelligence-(GenAI)-everything. Each episode is an experiment. This one? A makeover. Of sorts.

// It’s JEN_AI Challenge #12: MAKEOVER. This episode:

• The Fake Fakeover: AI-assisted fake-makeover attempts with virtual tools like headshot generators, makeup-glowups & wardrobe try-ons

• The Real Reel: I chat about style IRL and get a work-wardrobe-refresh with Amazon glamazon & bestie Hara Gavriliadi

• “As If!”: I vibecode a “like, totally Clueless” style-makeover app inspired by an iconic 90s movie

• Bonus: we take notes from celeb beauty, drag-queen philosophy, and Queer-Eye’s makeover show

// WELL THE STRUGGLE IS REAL AT LEAST

So somebody gave me some excellent advice: “Always video your podcast, even from the beginning.” And I tried. Briefly. But then I caught sight of the husk of a woman hunched over my webcam and thought: absolutely not.

I looked in the black mirror of my screen and after five years of parenting (and roughly five different clothing sizes), I could (almost) see two things:

• A baby-stained ill-fitting sweatshirt from Tesco optimistically emblazoned with the slogan “Limitless Possibilities”

• A scratched-up pair of fugly glasses

I was reminded of Sheng Wang’s Netflix stand-up comedy special, when he talks about buying clothes at Costco: realising he has become someone who “does what he needs to survive, but isn’t afraid.. to die.”

So it’s time for a challenge: Can I fake a glow-up using AI tools, while getting a little real help from a friend? Can I finally look like a functioning human again? Can you deepfake confidence? Can I video-record this blogcast without speaking like a robot, and actually look at the camera? Spoiler: yeah, no. maybe.

💅 Artificial Makeover

All the best transformations start from within, but who’s got time for that? Let’s start by putting the superficial in artificial:

// Beauty is in the AI of the beholden

• FaceTune (the OG - I had to!) - there’s definitely a sheen of airbrushing but yeah, it’ll make you into a smoking-hot cartoon for sure. But here are some restrained, work-friendly outputs:

• Aragon.ai for fake headshots - this was a recommendation, but way too expensive and the results were still kinda.. fake. Also a bit “2010s-work-attire”. Don’t do it.

• ChatGPT - not terrible for colour analysis and creating a look but it was hard to hang on to your facial likeness.

• Google’s virtual try-on & glow-ups: these are awesome (yeah, I’m biased) and rolling out in the UK soon. If you’re in the US - check out Doppl app & Virtual Try-On in Google Search to try clothes on your own photo

🧪 it’s a JEN_AI FAKEOVER - my vibecoded “LIKE, TOTALLY CLUELESS MAKEOVER” app-website

So yeah, TikTok filters aside - I felt like I wanted more. Time for a like, totally, unnecessary AI prototype homage to the outfit-chooser app from the 90s movie-remake of Jane Austen’s novel Emma: “Clueless”. HELL YEAH

Tools used: Google’s Firebase Studio for vibecoding plus AI Studio Chat for brainstorming the project brief (=“prompt”). I’ll spare you all the details this time but shout if you want to know more.

I had so much fun with this but 1) I lost the battle for pixelated leopard print, RIP my cool-retro design system and 2) as the app’s style suggestions are based on analysing how you look, this is a PR nightmare waiting to happen.

🦋 Boardroom Butterflies

I called in my friend Hara Gavriliadi – tech-business glamazon, Amazon Studio Youtube host, and style whisperer. She dragged me into a real shop and chose me some real back-to-work clothes. She reminded me I’m still in there somewhere (my body, not the shop). And she illegally recorded a podcast interview with me in a glam hotel co-working space.

👗 Hara’s Style Hacks

Here are a few teasers from our interview chat (accompanied by a baby! Which is why I’m out of shot):

• I asked her to share some style advice. She said: “I stopped giving advice. We’re not equipped to tell people what to do. You already know. You just need time to hear it.”

• Style = You = Your Power. “What makes me feel powerful is color.” Whether leading C-suite workshops in hot pink or rocking white trainers with a purple suit, her style mantra is along the lines of authenticity over conformity.

• We’re past the heels-for-approval era. “We don’t dress for other people. We dress for ourselves.”

• On personal transformation: “Just pause. That’s the hardest and most powerful thing.” Slowing down helped her reconnect with what really makes her feel like herself.

• Wardrobe as not just performance, but empowerment. A bright wardrobe, a calm home, and the realisation that being powerful and vulnerable isn’t a contradiction.

💖 QUEER A-EYE

I couldn’t do a makeover challenge without mentioning my maternity-leave-guilty-pleasure Netflix series: Queer Eye, the life-makeover show hosted by 5 talented gay men showing up to help someone feel seen, stylish, and safe. I used to put it on in the background and cry: The single mother of four with two jobs, holding it all together. The black lesbian finding her feet after her family threw her out. The hope-crushed man struggling to get out of bed.

The Fab Five weren’t just giving people glow-ups, they were giving them permission take care of themselves, and be taken care of. In my own misguided (or “clueless”) digital way, this episode is my small nod to that kind of care.

Don’t be a drag, just be a queen.

I remember watching drag-queen legend mama RuPaul talking about feeling that celeb women were sometimes uncomfortable with drag because it shone a mirror on how beauty is made - with lighting, cosmetic procedures, stylists, make-up artists, filters and retouching. Drag queens just do it openly. There’s a perceived loss of mystique when anyone (regardless of gender, and especially with a budget) can look like a star with the right toolkit. Drag exposes that artifice. AI might just.. automate it?

If RuPaul taught us anything, it’s that “we’re all born naked and the rest is drag.”

So… did it work? Am I.. ‘made-up’ now?

Kinda. At least I now have several non-ruined right-size work-suitable outfits. But the fake LinkedIn photo probably needs to go.

SOMETIMES, we dress for the person we wish we were. Putting on a look can be a performance, call it- ‘marketing’ -even. And that’s not shallow - it’s storytelling. Sometimes, we use paintbrushes (digital or real) and trickery to present our best selves. Sometimes, we just want someone (or something?) to say,
“You can - and you will - feel like yourself again.”

🎧 SOuNDTRACK(s)

Playlist: The JEN_AI Mixtape

• “MANTRA” – JENNIE (the awesome Jennie Kim from Korean pop supergroup BLACKPINK)

• “Like JENNIE” - JENNIE

📚 REFERENCES

• My vibecoded “Like, Totally Clueless Makeover” AI glow-up app-website → *here*

• Cher Horowitz’s wardrobe-app in the movie “Clueless” (which is a movie-remake of Jane Austen’s “Emma” novel).

• “Deepfakes”

• Hara Gavriliadi ( Amazon on YouTube / Bestie / Style queen)

• Sheng Wang on Costco clothes-shopping (Netflix: Sweet and Juicy)

• Google AI tools including Doppl, Google Search with Virtual Try-on, Google Firebase Studio, Google AI Studio

• FaceTune, Aragon.ai - virtual headshot editors/generators

• Queer Eye reboot (Netflix) life makeover show with the Fab Five

• RuPaul’s Drag Race

• “Don’t be a drag, just be a queen” is a line from “Born This Way” by Lady Gaga

• Shoutout to NY Times’ Kevin Roose’s glow-up on Hard Fork Live (dang he beat me to it!)

⚠️ DISCLAIMER & THANKS

An extra-huge thanks to Hara Gavriliadi for years of loving friendship and always being an inspiration. The other makeovers are mostly artificial, ill-advised, and certainly not approved by Queer Eye’s Fab Five.

If you checked out this week’s episode, thanks for listening, and I’d love you to:

👉 Share your email ( -> subscribe to get future JEN_AI blogcasts)

👉 Share your thoughts

👉 Cher your wardrobe?

JEN_AI is produced by Jenni Munroe using publicly released tools (incl. Gemini & ChatGPT for edits & brainstorming). Not endorsed by Google. #myviews

#jenai #GenAI #deepfakeit #cluelessmakeover

Share

EXPERIMENTAL AI/HUMAN CREATIVITY IN PROGRESS

👾 WELCOME TO JEN_AI: SERIES 2

You’ve found JEN_AI: the chaotic lab/newsletter/blogcast where I mess with new AI tools, digital media, tech culture & my life and try to figure out what it means to be human in the age of AI.

We’re powered today by ✨vibecoding✨ (=telling AI what to build, seeing what it makes, then tweaking the results) and sugary coffee. I’m Jenni Munroe, Generative Artificial Intelligence (GenAI) fangirl since 2010 and newly back at Google after maternity leave.

It’s a VIBE MODE season - with hack-challenges to co-create not just code but AI+human movies, makeovers, agents, science, life-choices and TED talks. Try out it for yourself or just enjoy the hack-fails, irrelevant references and bad-wordplay detours while we soak in the post-human content economy.

🎛️ ACTIVATE: VIBE MODE

“We’re writing vibe-checks, but can we cash them?”

💻 // HACK_LIFE

JEN_AI Challenge #11: SHOWCASE
→ Tell AI to Build a Project-Showcase Website

#no-code-vibe-mode

Prompt it → Watch it build → Fear the AI-pocalypse

🧪 THE EXPERIMENT

After a recent hackathon at work, I wanted to explore and celebrate everyone’s cool projects. So I challenged myself to make a website to showcase them using Google AI Studio. (NOBODY PANIC! I’m using JEN_AI project info and publicly released tool versions in this public demo version instead of secret Google stuff.)

⚙️ HOW TO TRY IT

1. Open Google AI Studio (on desktop/laptop)

2. Hit the 🧩 “Build” icon

3. Type in your “prompt” (=tell the AI what you want to build)

4. Add data or inspo if needed (click ‘+’ to attach spreadsheets etc.)

5. Click Run

💡 Pro tip: First go to “Chat” and ask it for help writing the prompt.

You can ask it to build any app or website and see what you get. e.g: “Build an existential crisis coach with a retro computer interface and a hint of friendly doom.” Or maybe don’t use AI as a life coach - your call.

🔧 MY ACTUAL PROMPT:

“Create a single prompt for an app-building tool like Google's AI Studio (Apps Build) to create a website for me using this data to showcase “JEN_AI”: a playful AI-human experimental project & podcast series featuring creative AI, tech, media and culture hacks. The site should feel like it was hacked together by friendly bots: minimal retro terminal aesthetic (but white-on-black instead of green-on-black), glowing/glitchy charm, responsive, and animated, using mostly black and white.”

→I pasted this into AI Studio’s “Chat” window and then copy/pasted the much longer prompt output it created into AI Studio’s “Build” window.

🧠 + 💅 The Result:

You can check it out at → jenai.studio. As this is public, I used JEN_AI hack-project data, but Google employees can also access the Google hackathon showcase @ go/gHACKED (bonus: used just-released =AI() Gemini function inside Google Sheets to autogenerate a line of hype for each project, just for fun).

🤔 REFLECTIONS:

• ✅ Surprisingly fast to get something working but can be hit-or-miss

• ⚠️ Easy to mess up once edits start - use version control or save copies often

• 👩‍💻 Retro/glitch computer aesthetics = chaotic readability

ΔNN - NET_NEW_NEWS

What’s been buzzing in AI & tech culture?

RIGHTS AND WRONGS

A U.S. judge ruled that Anthropic’s use of legally obtained books to train its AI was protected under fair use, but the company still faces trial for allegedly using pirated books. [1]

Disney & Universal sue Midjourney for copyright infringement as it launches video generation. [2] Also: striking videogame performers reach a tentative deal. [3]

COMPUTER SAYS NO

OpenAI models reportedly resisted shutdown in safety testing. [4] See also: Anthropic’s Claude Opus 4 has… resorted to blackmail? [5]

DEEP SCIENCE LEAP

Microsoft’s AI breakthrough in electron Density Functional Theory (DFT) could supercharge materials discovery. [6] +Cool vid here for the science nerds.

AI SUPERHUMANS

Meta offers nine-figure salaries for superintelligence hires and manages to poach 8 OpenAI (ChatGPT) researchers. [7, 8] Meanwhile: Amazon warns of job reductions due to AI. [9]

i’m regretting not doing that AI PhD a little more each day

🎧 SOUNDTRACK: Short n’ Sweet

I like my newsletters like I like my coffee.. longer than necessary

This episode’s music muse is Sabrina Carpenter’s breakthrough album: “Short n’ Sweet” …which is not how anyone has ever described a JEN_AI post. Until now? ..Yeah, still no.

Her signature hit: “Espresso”

Playlist: The JEN_AI Mixtape

💥 GAME_OVER

“good enough to be dangerous”: a worryingly apt AI skill level

So yes, in true JEN_AI style- we made a hack-project about hack-projects. And we got now-typical mixed-vibes about whether to fear or hype the future (or both), from recent AI-adjacent news. All to a soundtrack from an artist whose upcoming album’s go-to-market strategy is to hit two opposing demographics hard by role-playing as a sexually-submissive puppy while singing a song about incompetent man-children.

So yeah- Thanks for vibing with me (and the bots). If you do try your own hack- I’d love to see it. If you’re just here for lame jokes- welcome home 💖

JEN_AI is produced by Jenni Munroe using caffeine, insomnia, and publicly released tools (incl. Gemini & ChatGPT for edits & brainstorming). Not endorsed by Google. My views.

👉 subscribe to get future blogcasts in your inbox (there’s no cost btw)

👉 share your thoughts, or your own AI experiments

👉 send to a friend who also needs more sleep and fewer ideas. Actually, don’t

#jenai #lowcode #vibemode #shortnsweet

Share

// This Episode

JEN_AI tries to make a website that creates dream homepages, using AI wizardry and a few clicks. We reflect on chasing and creating our dreams vs. longing for home.

// The Challenge

JEN_AI Challenge #10: Use AI to spawn a website-making website.

This week I found myself with a wishlist of website projects for friends, family and myself, each in need of an initial placeholder / basic homepage, and wondered - “Can I join forces with the AI-coding-overlords to distill personal & business websites down to their most essential elements and create a website that creates websites (or landing pages) with a slick, simple design?”

Let’s find out.

// What is JEN_AI?

If you’re new here, welcome! I’m Jenni Munroe, and the JEN_AI program is a series of experimental creativity challenges where humans use generative artificially intelligent (or GenAI) machines to learn, to co-create content and channel the future while taking liberties with words and obscure(d) references. Let’s Jenerate!

// AI Tools Used

In the “Build Something Good” JEN_AI episode we used AI tools to “vibecode” - you tell an AI agent what you want it to build, then watch the tool code it for you, and steer it a bit. The more comfortable you are with tech, the easier it is to ask for what you actually want and fix things that break, but I would still recommend it to a never-coder for a taste of that tingly-magical feeling of typing words and watching things move, come alive. (When I was a kid discovering coding it truly felt like I was learning to do magic.)

I’ve used Replit again for vibecoding, but there are many others, like Cursor, or code-specialized tools from the Big (Chat)Bots like Claude Code from Anthropic. They usually have monthly fees or a free tier option. Website hosting providers also typically now have AI tools where you can type in what you want, hope for the best, and then edit it. If in doubt, ask a chatbot? Or not.

I also like to brainstorm with ChatGPT to generate prompts to give Replit to kick off the vibecoding, which can help you get a better shot at asking for the right thing upfront or explicitly mentioning things you might have missed.

// JEN_AI’s jenai.com site is almost conscious

The JEN_AI site is now technically live at jenai.com, but plenty of fixes, tweaks and surprises to come soon. This was vibecoded using Replit but has a few more style features than the current No Page Like Home. (Am I too obsessed with light mode / dark mode toggles?)

// Try it: No Page Like Home webpage-creator

“Has the dream of owning a home become a nightmare? Worry not! With No Page Like Home you can have a digital home online, where we’re spending all our time anyway.”

Yeah, not the best advertising campaign, scrap that.

The vision-vibe I was shooting for was:

Imagine someone is looking to level up their personal or business brand. Maybe they are applying for jobs or are working on a side-hustle, small business, or startup, and they want almost a placeholder that is a quick, easy, professionally designed initial website of their own. A simple homepage - their own home on the web - almost a “digital business card” to start with - that succinctly showcases what they can do, but doesn’t require faffing around with tech or templates or whatever. These people already have a presence on few different social media platforms but don’t have a cool, visual, central digital-profile-kind-of-website that acts as their style-forward front door, pointing people to their favourite more in-depth “showcase” or “connection” platforms (such as LinkedIn, Instagram, GitHub, Medium etc.). It needs to be quick, easy & fun to create, look cool and feel like home.

So I did some research, looking around at simple personal brand and business landing-page websites and also asked ChatGPT to distill the bare-essential elements:

• Name and Professional Title

• Blurb: Tagline, Brief Bio or Value Proposition

• Visuals: Photo, Portfolio/Work Samples

• Contact Info / Social Media

Other suggestions (maybe later):

• Endorsements / Testimonials

• Call-To-Action (e.g. a button for what you want visitors to do next like subscribe to your newsletter)

If you want to check out the No Page Like Home webpage creator, which creates a single landing page for yourself or your business (think digital business card) there are some sneak peeks here or head to nopagelikehome.com. Warning: it’s in demo-mode, beware!

The way it works currently is: you enter or generate some info about you and some images, (or you can click ‘Example’ to pre-populate), and it will mock those up into a slick site “card” in a choice of 3 different style options (which you can instantly click to preview). Whether you’re into rainbow or you’re over it 🌈 there’s something that looks good. When you’ve got something you like, you click ‘Make it Real’ to publish that to a real web page. Real talk here though - this prototype demo is basic AF. I have a long list of basic and wishlist updates to make (e.g. auto-filling suggested content based on e.g. info from your linkedin/instagram edited with generative AI). Suggestions are also welcomed!

First up, I’m using No Page Like Home to create a new landing site for my mother’s textile art making and teaching business, she greenlit the preview today :)

If you just came for the AI experiments - that’s all folks! If you’re in the mood for some RealJen life-overshare, hang tight.

// GEAR CHANGE: RealJen Reflecting on dreams, homes and human life

Moving to California in 2019 was something my husband and I had long dreamed of. But two things happened that we didn’t expect: a pregnancy, and a pandemic. We found ourselves starting a new family in country that our old families couldn’t enter for what turned out to be around 20 months.

We were embraced by kind neighbours. I found a community of inspiring, successful, brilliant mamas and we wore masks while our babies played on picnic blankets spaced apart, in golden parks where it almost never rained.

In so many ways, it was a dream come true. We built and gave birth to a new life, a new home, started new jobs, made new friends.

But, raising a kid takes a village. We missed our UK family and friends. My grandma died. More than one of my mama-friends lost their own mothers, one very suddenly. We were blessed with a second child. As I struggled to raise my two sons a creeping panic grew inside me: I can’t miss another precious year of my family’s lives. Despite all the upheaval it would cause, and the life & friends we hated to have to leave behind, it was time to go home.

This Sunday is Mother’s Day in the UK, a day which can trigger a range of emotions depending on your circumstances. This year is my mother’s first year without her mother (my other grandmother, who recently passed). I’m grateful for the year of quality time I had with my maternal Grandma in the UK and for getting to say goodbye. And I’m so grateful for the loved ones who are still here, as we start a new life, again, with our beautiful third new life- our little girl. I’m grateful for the privilege, the opportunity, to chase our Californian dream, but there’s no place like home, and I can’t believe how incredibly lucky I am that it was, mostly, still here when I came back.

(So I’ll take this precious opportunity to give my mum a dodgy homemade present, in this case a website, for Mother’s Day, not so different from the homemade cards my boys will give me. Don’t worry there are flowers and cake too.)

// JEN_AI Playlist - this week’s theme songs:

• No California by Ilsey The sound of living The American Dream while missing someone you love.

• Over the Rainbow by Susan Boyle Move over, Ariana- SuBo dreamed a dream. Britain’s got talent, and -though it’s far from perfect- I’m grateful it’s still Great (tho there’s nothing less British than admitting it). Today we’re choosing your Scottish Nan over Hollywood Glam.

// Feedback & Thanks

As always thanks for reading & for sharing your feedback on how this could be better. Keep it coming! We’re attempting a “shorter” newsletter (lol fail, better luck next time!), every 2 weeks, in favour of better JEN_AI projects and less inbox spam. If you like a version you can listen to while doing something useful- try the podcast.

// References

In “The Wonderful Wizard of Oz”, Dorothy and co seek out a fake wizard and abilities they already possess but aren’t able to fully realise. At the end, Dorothy clicks her heels together and says “There’s No Place Like Home”. At the recent Oscars, Ariana Grande and Cynthia Erivo performed songs from Wizard of Oz adaptations, including “Over the Rainbow”. Their movie version of the “Wicked” Oz musical received 10 nominations and won 2 Oscars for Design.

Whether you came here hoping to find intelligence, heart, courage, or you just got lost, JEN_AI has it all! Maybe. But more importantly, so do you.

All views are my own (or maybe a chatbot’s, or yours?). JEN_AI was produced on a Google employee’s (unpaid) parental leave and is not endorsed by Google.

Hey folks, it’s JENNI with the AI.

It's time for a little introspection, a dash of retrospection, and a whole lot of YOU-direction. It’s JEN_AI and we’re looking back so we can spring forward.

This is JEN_AI Challenge 09: FEED_BACK

Get in the loop. Give FEEDBACK on JEN_AI

I got some helpful listener feedback that it was a bit hard to keep up with the JEN_AI outpourings - both in terms of understanding the mad ramblings and getting through the backlog of weekly episodes.

So this week- we’re hitting the pause button to reflect on the JEN_AI journey and dig out a few gems. I’m asking for reader/listener feedback and suggestions, and reflecting on when to take vs. ignore advice.

// Part 1: The JEN_AI Feed, Going Back

(a Reverse-Chronological Roundup of what we’ve learned so far)

In case you missed it - here’s a whirlwind rundown of JEN_AI’s experimental outputs and the key takeaways so far, flipped and summarized for your byte-size consumption:

#8 Build Something Good - we got AI to code a simple web app, tried to make a Twitter/X bot which got banned, digested government aid cuts and aimed for world betterment with special guests. We learned: Even AI needs a moral compass and maybe a lawyer. Try out: Replit to code websites and apps for you.

#7 Social Mediums - JEN_AI channeled social media, confronted harsh realities, gave some reader shout-outs and tried to read you. We learned: Keep it real- the true value is connecting with real living humans, duh. Use the tech, don’t let it use you. Don’t try: to get on all 20+ social media platforms.

#6 The Ultimate AI Challenge - We brainstormed useful uses of AI, including: brainstorming. We learned: Sometimes, more powerful than to speak is to listen. The process of refining your ideas back and forth with a chatbot can be more valuable than the generated outputs alone. Try it out with your fave chatbot e.g. ChatGPT, Claude or Google Gemini (links)

#5 Let’s Get Social - JEN_AI sent some Valentine’s Day love to listeners and invited them to get on the show. We learned: you can’t rush engagement, and wordplay is not overrated.

#4 AI Emergency! Deeply Seeking Shortcuts - The U.S. stock market and a stretched parent eyed-up DeepSeek - a "thrifty" language model - via an obscure 80s Madonna movie. We learned: Everyone's looking for a shortcut, especially when time, money and power is involved. Using a large-langugage model chatbot can help get that copy right.

#3 Future Podcast - One AI podcast guest predicted the future of tech while two others roasted us alive. We learned: AI-driven audio-quality is brilliant or brutal, with no in-between. Try out: Google’s NotebookLM to generate easy-to-digest podcasts about any source material and Descript to edit ‘real’ podcasts.

#2 Let There Be Speech - A free speech… about free speech… generated by not-so-free AI speech. A voice-cloned JEN_AI tried audio. We learned: Free speech can come at a cost (especially if you’re Trump’s speechwriter) and voice cloning is as creepy as it sounds. Try it? Or not.

#1 A NET_NEW Newsletter - We made an AI/human newsletter about writing an AI/human newsletter about the AI/human news. We learned: Chatbots can elevate your writing skills to the internet average, but that doesn’t make you a media pro.

#0 Intro to JEN_AI’s AI/Human Blog. AI/human writing about AI/human writing about AI. Try: use Substack as a media platform to publish blogs, newsletters, quick pics, podcasts and more. But maybe turn off its attempts to beg cash from your friends.

// Part 2: Feed the Machine

Time for some fine-tuning.

YOUR constructive complaints and suggestions can help make JEN_AI more useful, more entertaining, less irritating.

Please let me know (via this poll or any legal means) what you’d like to see more or less of, or any other feedback.

I’m also considering switching to sending the blogcast every 2 weeks instead of weekly - to be less spammy, to spend more time on things like video, the challenges themselves, & distilling the good bits for easy multi-channel consumption, and to give anyone interested a chance to keep up. Def let me know your suggestions!

Here’s the poll link again.

Thanks for taking the time to make JEN_AI better!

// Part 3: Giving feedback on feedback

it wouldn’t be JEN_AI without unnecessary meta-content (or would it? Let me know). So this week I’m sharing a few notes on feedback.

I can distil some advice I got on giving tough personal feedback at work to this: be specific about something someone did and how that made YOU feel. Be respectful, don’t generalize, and reflect on how their actions impacted you personally. Behaviour can change, but not if you’ve dismissed their entire personality.

For some people, getting feedback can cause emotional distress so severe it feels like physical pain. This is sometimes referred to as Rejection Sensitive Dysphoria (RSD). Coaching or therapy can help with this, a lot, and is also something anyone can benefit from more generally.

The thing is- if we can’t take feedback, it’s hard to make progress or grow. And that’s whether you’re starting a startup or just trying to have a relationship.

Feedback works better in an environment of psychological safety and trust. (Not internet trolling.)

Feedback can be imperfectly expressed and might not even offer the best solution, but there’s a signal in there, a real human feeling, and someone who cares enough to want to help us learn.

So can JEN_AI get better with feedback? Alex Ramirez, a listener who is building a biotech startup called Antibody Design, kindly sent me some thoughts on the DeepSeek episode:

Alex, thanks for the excellent advice that I have failed to follow. To honour it a little bit I’ve attempted to reduce and clarify the rambling and references and I’ve added a NotebookLM generated chatbot podcast at the end explaining how AI neural networks use feedback to learn. More informative, easy listening. But I’m afraid that- for at least today- the irrelevant movie references and lame-pun laughter stays.

There’s a balance to be found when creating something you haven’t really seen before, of course there is. There’s a middle ground. But that’s the question - how far to go? And how to avoid making something old-hat or average, when what you’re really looking for is a new medium.

The answer? How to create psychological safety and improve? I say - experiment. Iterate. Learn. Learn how to listen to the opinions of experts, consumers, users, supporters and those who care, while still being guided by your own Inner Light.

// JEN_AI Playlist: This Week’s Theme Song(s)

• Inner Light by Elderbrook + Bob Moses. And as a lapsed circus aerialist, I love the music video’s dance performance on a rotating platform.

• With the AI (I mean, the IE) by the epic JENNIE Kim from BLACKPINK. Bonus points for sampling J-Lo’s Jenny from the Block. This lady knows how to spring forward while nodding back. We’re on the way up.

// References

Welcome to JEN_AI, where you should probably feel better about yourself if you DON’T get the references. But if you want to check the big ones, here goes:

• FACE/OFF the movie: Face/Off is a brilliant, terrible sci-fi action film starring Nicolas Cage and John Travolta and directed by John Woo. One is an FBI agent, one is a terrorist, they swap faces, morals and identities. Cue action. Why even mention it when I’ve barely referenced it? Why the hell not.

• “Jenni with the AI” riffs on “JENNIE with the IE”. See JENNIE’s music above.

• There’s a lot I could attempt to say about how neural networks use feedback to learn, but I’ll leave that to the bots. I’ve generated another NotebookLM audio overview you can feel free to listen to at the end, and it contains all you need to bluff being an AI expert, or offers something to fall asleep to - you choose. Here it is:
  An AI-generated AI-primer podcast made with Google Gemini and NotebookLM:

Thanks for listening! All views expressed are my own (or a chatbot’s, duh)

We spoke to international development specialist Bridget Brown and storytelling video producer & Feral-Business-owner Cressida Kocienski about how we can Build Something Good.

This week- I get banned from X/Twitter (does that make me an eX/Twitter user?), get AI-agents to build a web app that writes a meme-mail to your government representative, and consider whether to give up trying to spam people for Good.

// The Why: Too much Chat from JEN_AI, the bad guy?

I don’t know if you feel how I feel.

What I feel: so far in the JEN_AI Project,

There’s been an awful lot of chat.

And last week JEN_AI had an overdue epiphany that much of the world sees Social Media, Big Tech, Artificial Intelligence and Generative AI (GenAI) as a waste of time at best, uncool at medium, and at its worst: downright Chaotic Evil.

Also, we won’t mention- The News.

I had another topic planned for this week, but it’s time to Build Something. Something Good.

I mean “High-Quality-Good”, “Make-the-World-A-Better-Place-Good” and aspirationally “Lawful Good”.

I don’t believe AI or BigTech is inherently Evil. But we do certainly have an Alignment Problem. Or rather, a challenge.

If you’re new here- I’m Jenni Munroe, a creative technologist and AI experimenter who loves hacking together ideas at the intersection of AI, media, and culture. I taught computers to paint before it was uncool. I’m currently on maternity leave from Google, and each week on the JEN_AI_Project I use AI, tech, culture, and media to keep up with the latest in AI, tech, culture, and media.

This is JEN_AI Challenge #08: Build Something Good.

// Context: The Global Net-New News

This week- the US, EU, UK, and Ukraine have been strategically re-thinking defense and aid.

You may be aware that President Trump and Elon Musk have been “re-imagining” US government lately, hacking away at federal programs under the newly formed Department of Government Efficiency or “DOGE” unit, with cuts and pauses both at home and abroad, including halting 90% of USAID. As of 4th March 2025, the U.S. has also suspended all military support for Ukraine.

The UK Prime Minister Keir Starmer announced an increase in UK defense spending from 2.3 to 2.5% of GDP by 2027, cutting foreign aid from 0.5% to 0.3% to fund it. The message? Less global charity, more military muscle. But is aid really just charity, or is it actually muscle too- maybe even with greater gains? Time to flex those guns? 💪

There’s a lot happening (and no longer happening) on a lot of fronts. Lives and livelihoods have and will be lost. It’s not looking good.

// What to do? Where to start?

I’ve been speaking to international development expert Bridget Brown - Managing Director at Montrose International- this week about the impact of the aid cuts, misconceptions about the value and power of aid, and if there is any way a tiny human/AI hack challenge can be of help.

If you’re working in the U.S. government or international development, it’s been a rough time. But perhaps Beyoncé said it best (I’m paraphrasing):

Ok ladies, now let’s get information.

// Aid gets AI’d

Code_Project_Name: LemonAID

// Bridget Brown on the Power of Aid

Bridget and I had a long discussion and then I asked her to reflect and share what ideas had resonated most for her. We walked & talked, and recorded it - this is minimally edited, for readability:

Jen: All right. Okay, good. So, hello. You're being interviewed!

Bridget: Hello! I'm being interviewed! This is great. So there's a whole range of stuff. The MP email generator type thing can be done ‘tomorrow’ and is actually quite a good idea, but first we need to convince everybody that it's a a sensible idea generally to do.

So there's a whole load of things. So for instance - USA: there's an insane number of documents online around what the programs do, how many lives are affected, what the impact of those is. Anything that would consolidate and map a picture of the lives saved and the knock on impact would be obviously humongous. Eye-opening. You were saying we could match the stories to the right social media channel. That would be really key because no one's interested in this except those people who are already looking for it. So being able to present it in a way that's meaningful. Part of that would be anything that would be around the benefit back to the UK or the US- the country that's giving the money in the first place- would be fascinating. Because I think-

Jenni: ..Aid is defense.

Bridget: Yeah exactly. Because people genuinely sort of see it as an add-on, it's nice to be good to other people. They don't understand it stops you getting sick. If we invest this amount in pandemics, it saves that amount in the US not suffering from a horrific disease that comes over with someone on a plane.

Jen: And immigration, like you said.

Bridget: Yes immigration is another big one, the radicalization, the governance side. There’s a massive trade angle as well in terms of getting say UK contracts into a country- you don't get that unless you have something to offer and basically your aid budget is in part a buying tool to get UK contracts as well. You cannot negotiate that if you're not giving something. And so even if it's a small amount on HIV, which has got a huge personal sensible human benefit, what it's doing is giving the government in that particular country something they can talk about and say, “we are providing this to our people because we've negotiated it in”. So they're getting votes, they are getting their standing as a politician maintained and supported and in return, this is what they get for it. So in exactly the same way that Trump is doing with Ukraine and saying, “give us your minerals and we'll protect you”. In essence it's the same thing in terms of “we'll give you some health care, we'll give you some education. You'll get a stronger, more stable governance in your country and as a result you're going to be more economically stable. You're going to be able to attract investment into your country. This is what we want in terms of contracts for the UK.” So there are direct trade benefits, economic benefits as well as the ones around security, migration, health security and things like that. So anything that could give an indication of benefit back as a basically in import to these countries - UK or US - would be incredible.

On lighter versions- anything around encouraging companies to think through their their CSR [Corporate Social Responsibility] spend, what could be provided, what they have planned. Thinking through what what their money could do in terms of impact. Because I think there's vast amounts of investing but the impact is often very small compared to what you can actually achieve and sort of stretching that, the finances and saying well actually with this amount of money you could put it this way get so much more human benefit for it.

It's all about storytelling. There's just so much data out there around what these programs have done. All of which is normally only seen by the individuals who actually are involved in it. Drawing any of that together and presenting it to people who are normally interested would be hugely important. And then, then getting that back, whether writing to an MP and saying, look this is what that's showed me and I want to be a part of that and I want my taxes to go to doing that because I think it's a better, more cost efficient option than buying guns. Maybe something along those lines.

Jen: Thank you so, so much.

Bridget kindly sent me a follow-up / clarification from South Sudan:

UK has untied aid which means it's not a direct quid pro quo on aid = contracts or the recipient country _having_ to do something. It is however a factor in the donor countries relationship and can be learnt on in terms of wanting to maintain positive relationships with the donor entity which is underpinning core services

There is a tonne of stuff online about the impact of aid and what it delivers in terms of lives saved, kids put through school etc. but I suspect many people who aren't supportive of heir taxes being used for this in itself, would be more receptive to evidence as to what that knock on benefit is to themselves. I.e. it reduces human trafficking and illegal migration, it protects the UK from massive cost increases (many be more familiar with the fact that oil and gas I.e. their utilities bills and key products like wheat I.e. their shopping bill have gone up since the Ukraine war). Other products coming from other countries are similarly affected by conflict or poor infrastructure costs. People in uk get more for their money when the barriers to stable, peaceful investment, production etc are removed. It goes on.

The aid sector have always focused on telling the human element of why it's a good thing to do. The NGOs more than most as they have a funding incentive as they're trying to attract personal giving as well as winning contracts. The trouble is this resonates with people who are supportive of aid, or with people giving as a one off which feels food (I.e. Red Nose Day). What it doesn’t do is explain why - when people are feeling that they're struggling with cost of living and things are tight - should a constant tax be applied to help people in a different country. The wrong emphasis has been put on why we do aid for decades

// Meme-Storytelling: Aid as Defense

Let’s try riff on a few of Bridget’s powerAID ideas with some wordplay:

• Aid is Defence.

• Less Aid = Less Trade

Drake meme:

• #aidcutsmigration #aidcutswar #aidcutsareanepidemic

• Think aid spending is inflated? Get ready for global de-stabilisation-inflation.

• Don’t Hurt Yourself, Help Yourself. Aid cuts harm US.

• You want to know the real cost of aid cuts? Here’s a receipt. Aid needs a Re-fund.

• It’s time to re-instate international aid and the US & UK’s global influence. Who benefits from cuts? Not us. (If that seems overly urgent, I don’t wanna Rush-ya. Just Chying-a have some puns.)

Ok last one, I promise:

• Want a wall? Or a moat? Or a fence?
  Aid is da fence.

Alright, alright, sorry that’s probably way too many preachy puns. You get the idea.

If you’re interested in more specific impacts of specific cuts, you can check out eg Better World Campaign or devex.com. Examples include: 1 million people in Syria have just lost their water, shelter and sanitation under the cuts. Ebola epidemic travel-control screening programs in Africa have also been halted. On yer bike lads!

Alright, enough. This isn’t Red Nose Day. Which reminds me, if you are for instance a BigTech employee like me you probably have a personal donation-matching or corporate social responsibility budget. Alright alright, I’m done.

// Tools I Used

ChatGPT: for brainstorming ideas on what to build and how including building a Twitter bot, generating bot reply text to offer another perspective when an X/Twitter user asks why we should fund aid abroad, and drafting emails to your UK MP. Even, in extreme laziness- writing prompts for me to give my AI coding agent.

X/Twitter: So yes, I now have an ex-Twitter developer account. Banned for being “inauthentic”, without specifics, but apparently that includes “manipulating X”, “fake personas” and “synthetic media”. I’m actually reassured to see that X is on the case with this stuff, because while the internet likes to nervously joke that social media is mostly just bots arguing with each other, after I saw the $4500 monthly X tier to monitor live tweets and make 300,000 posts monthly, I was worried. That’s expensive for an individual but peanuts for an organisation. So yes, X will ban you for attempting to artificially manipulate public discourse to spread the woke mind-virus. At least- it will on the free tier. (An example of what apparently IS permitted on X however.)

Replit: I was inspired by the Hard Fork podcast to try “Vibe-coding”, ie sort of giving the AI agent a challenge and watching it code, and feeding back and forth with it in regular English language chat to build an app or interactive website. There are different price tiers, I’m trying £25 for the month. It definitely helps if you are familiar with tech and code, but I think it’s magic and I might be getting addicted to this thing.

UK Gov APIs: I used the UK Parliament API for getting government representative (MP) names and contact info. The “They Work For You” API ChatGPT suggested first was missing some key info. (C’mon bros, even I’ve made better government APIs than that.)

// Try it: LemonAId WebSite App

So here’s the proof-of-concept website-app demo, at makelemonaid.replit.app. You enter a UK postal code, it pulls the name and email of your government rep (i.e. MP) and generates a draft email for you with a choice of polite, firm or meme-banter tones. Check it out! (I’m not storing the postcodes.) So, did we build something? Yes? Is it Good? Meh. There are a thousand ways this could be better. I gave it a decent shot in the hour I had available, lol.

Such is the nature of doing what you can with less. And I don’t disagree with looking to improve efficiency, either in coding projects or government spending. Far from it. I do also understand that there is a difference between taking Trump seriously and taking him at his literal word. I understand there are many hidden factors and a longer game at play. And I do see Trump’s efficient intended effects: like making the world take responsibility for its own defence.

And in Donald and Elon’s defence, they are being the change they wanted to see in the world. We’re always telling people to do that. We don’t have their resources. But we have some resources. Those in power are helping themselves, but maybe we can help ourselves too.

// Sneak Peak at JEN_AI website

The JEN_AI website at JenAI.com is Coming Soon! I used Replit to vibe-code this too. Here’s a sneak peek:

Is it Good? Is it Lawful? We can but hope!

// GEAR CHANGE: Someone Else Building Something. Good.

// Cressida Kocienski’s Feral Business

Cressida Kocienski is a force of nature. She runs a micro storytelling studio in the tech and design sectors, and is launching the awesomely-named “Feral Business”. Tagline: “Neurosupportive models for neurocomplex minds.“

Her unedited words:

 Feral business is a slow platform exploring how to run a business as a neurodivergent and chronically ill person within all the constraints and contradictions of whatever this fresh hell is. Small businesses are interesting vehicles for agency and building economic wellbeing because there's a degree of flexibility and choice about how we work, when we work, who we work with, and the tools and economic models that we use.

I know that conventional models of productivity, scalability, and efficiency don't work for me. But I haven't found the support and resources that I've needed to, to do what I've already been doing intuitively, but with an unnecessary amount of cortisol. So I got curious and started taking the whole thing apart and putting it back together again, with conversations and collaborations, embracing instinct, adaptability, and creative resistance.

What happens when organisational resilience is built around flows of capacity, rather than rigid expectations? How do we create realistic linking architectures between erratic flows and rigid structures? What do neuro-supportive frameworks of leadership, project management, communication, storytelling and money management look like?

How are we experimenting with emergent AI tools to create tailored support structures for ourselves that might be basically all we can access in a landscape of budget cuts and rotting in queues for diagnosis and titration and mental health support and access to work? But also, how have we actually already built the world that we want to live in, and what's next?

Feral business is intended to be a rebellion, a sanctuary, a map, and a toolbox. It's for the people who are too ungovernable for conventional workplaces, or too different for conventional entrepreneurship, but still want to carve out a place where their work and their way of working matters. And also, honestly, for the people that don't understand why things can't be fairer, because they think they probably can.

So, Come join me on the platforms and get in touch for a chat if you have some thoughts or wish somebody had told me to share about running your own feral business.

© Cressida Kocienski, Feral Business 2025

That’s a teaser- this won’t be the last we hear from her. She’s Building Something Good.

// Never forget a thankyou

I should have thanked you, listener-reader-supporters - I never thank you. Thankyou for your support. A huge thankyou to Bridget and Cressida for their epic contributions to this JEN_AI project and the world at large. For their time, expertise, voices and their always-excellent company. Thankyou also to Kevin Roose from the New York Times Hard Fork podcast for inspiring me to try vibe-coding on Replit via his own AI experiments in making a Hot Tub Time Machine jacuzzi-maintenance website to aid a friend. Last one - I heard that Comrade Putin, Xi Xinping and a lot of extreme-ly rebel-ious groups also wanted to pass on their thanks to Trump and Musk too for handing them a power vacuum. They’re about to clean up. (Did I go too far there. Yeah probably. Don’t panic folks, it’s just a bitter pun. A harmless bit of fun. Whoops.)

A shoutout also to power-journalist Dan Bloom, for asking Trump and Starmer some real-life questions. (Kudos to Trump there for a tariff-ic pun.)

// Are We Feeling Good Yet?

So how are we feeling? Good? Good.

This is fine. It’s fine. We’re all going to be fine.

// This Week’s Theme Song(s)

Theme Song(s) added to the JEN_AI “Future But Retro” Playlist this week:

“Feeling Good” by Muse (I’m so sorry Nina Simone, but Muse was my muse.)

“Don’t Hurt Yourself” and “Formation” from the Lemonade album, by Beyoncé

// References

I’m constantly referencing stuff. Sorry. Here are explanations for some of it.

• Beyoncé’s Lemonade album: after her husband’s infidelity, Beyoncé took the lemons life had thrown at her-and at Black history, Feminism and social justice- and made the groundbreaking visual-music album “Lemonade”, hailed by Wikipedia as one of the greatest of all time. “Let’s get in formation” and “Don’t Hurt Yourself” are references to her lyrics.

• Alignment: AI Alignment = Making sure AI does what we actually intend. RPG Alignment = a character’s Moral / Ethical axis alignment in role-playing games

• Red Nose Day is a comedian-hosted UK charitable giving day on 21st March where people pay a small amount to wear silly red noses and stuff

All views expressed are my own (or maybe Cressida’s, or Bridget’s, or a chatbot’s).

Let’s get that crystal ball rolling!

THIS TIME with ChatGPT as my spirit guide, I tried to teach JEN_AI social media.

SO MANY MEDIUMS I’ll share what happened when we dived deep into the algorithmic abyss; down the rabbit hole into the chaotic, overstimulating, addictive, and often soul-crushing world of different social mediums. 15 platforms deep. I think.

META-ANALYSIS We made a personality-matrix to understand and roast the quirks of each platform. We got the bots to write the ultimate, stereotypical posts. We even gave a few tarot-card readings.

YOU! But more importantly- I didn’t want do this episode without getting sociable, so a couple weeks back we invited YOU to get involved with the JEN_AI project! To be the social medium and share your thoughts on JEN_AI and the future of human connection via tech. We’ll do a few shoutouts and sharesies.

IRL REFLECTIONS Lastly, JEN_AI tried out some reinforcement learning In Real Life and took a long hard look in the mirror; I went undercover at a London university and a local tearoom and grappled with uncomfortable truths about BigTech and Generation AI.

// Make mine a medium

But first—let’s talk about mediums. Or media? A medium is means of communicating or transmitting stuff. But also:

• Psychic Medium: A person who claims to talks to spirits and other non-physical entities, intuits who you are, predicts your future.

• Size Medium? The length I *should* have tried to make this sprawling blog-podcast-post.

• Social Media: Platforms designed to connect people… or control them? 👁️👁️

• Bonus: Artistic Medium: Material an artist uses to create.

As usual, I’ll be making a mess of this digital canvas. Let’s generate!

// 📊 The Social Media Personality Matrix:

Where to start? Every social platform has its own personality, culture, user demographics (age etc.) and psychographics (interests, values).

So, I asked ChatGPT to analyse (and lightly roast) them, and tabulate it.

Easier to read with phone sideways.

💡 The Challenge:

Could an AI-human hybrid actually learn social media… or would it just get ghosted? What happened when JEN_AI got social?

• LinkedIn (CV/resume promotion)

  • I AI-ified my portrait to look more corporate. Co-wrote a stereotypical humblebrag LinkedIn post to intro JEN_AI, then got ChatGPT to savagely critique it. May be less employable as a result.

• X (Twitter) (news commentary)

  • A great place to stoke AI-generated AI conspiracies. Has *3* paid membership tiers now.

• Reddit (forums)

  • ChatGPT told me Reddit hates self-promotion. So I just lurked. Hard. Left the real interactions to the bots.

• Instagram (pics+)

  • Posted some JEN_AI visuals. But mostly just admired my cousin’s beautiful artist life & paintings and watched my mate Luca chasing penguins. Need to post a TikTok-style Reel (short vid).

• TikTok (short vids)

  • Definitely needed to post a TikTok style reel here, but got distracted and then time was up. Fail. Or working as intended? Need to do a JEN_AI video challenge soon.

• Bluesky (Twitter refuge)

  • The real flex here is using your website as your username. Is this New Twitter, or just bluesky dreaming? Too early to tell.

• Threads (Facebook’s Twitter)

  • Meta really, really wants me to use this.

• Facebook (old-school social network)

  • Meta offered to post my Instagram stuff here as well, so I did. My mum seems to like it.

• Substack (newsletters)

  • Don’t know if this was designed to be a podcast-scripting-blog tool, but it works. It definitely facilitates cross-channel posting, incl. YouTube, podcast feed, stills for Instagram. Downside: removing all the ways it tries to ask my friends for money is like playing whack-a-mole - please shout if I missed one.

• Podcasts (audio talk-shows)

  • Used to avoid these in fear of true crime and fake news. Have since discovered a deep joy in hearing people share things they found out yesterday. Welcome to the show! Now available on Spotify and some of the places people listen to podcasts.

• YouTube (video)

  • I went through a phase of being addicted to YouTube Shorts when other vices weren’t an option. Now I fight a losing parenting battle, mostly trying to help my kids watch adults pretend to be kids, without ending up in traumatising content. This is Russian roulette 2.0. Uploaded JEN_AI podcasts.

• Discord (forums)

  • These are spaces for building communities around specific products, so I checked out the NotebookLM server. It’s a niche.

• WhatsApp (messaging)

  • My own personal weakness: this is my main tech-connector to friends and family. Group chats are alternately a lifeline or a graveyard.

• Medium (articles)

  • Never seen so many AI-about-AI clickbait articles in my life, and now I get a daily digest of them as well. I think I deserved this punishment.

• Hinge (dating)

  • Not on it- just heard opinions. Users have to answer prompts, and I hear they have an AI coach now. Is this a dating app or a tool for turning humans into chatbots?

• Snapchat, Pinterest, Patreon, Mastodon, Twitch, WeChat, Quora

  • Ran out of time. I mean, c’mon.

• Lemon8, BeReal, Rumble, Weibo etc.

  • Surely we’re just making up fake social media platform names at this point.

I also stumbled across a few “Fans-only” sites proudly promoting their women-led entrepreneurial content. (Out of curiosity I did check out their AI policy; it’s ok if you’re verified human. Yeah, I’ll stick to podcasts.)

// 🏆 The JEN_AI Hall of Fame!

One thing I love about experimenting with JEN_AI right now is throwing crazy stuff around with like-minded friends & family and awesome people. It’s inner-circle vibes, in the best way.

Aside from the chats, here is a flavour of GenAI / JEN_AI stuff YOU put out there:

👾 A few shoutouts:

• Team Spirit: The ‘Aquapanthers’ lads crew. Thanks for the support! Nelson used NotebookLM to make a podcast for a job application comparing his CV to a job description - it was a hit! (Slightly risky. Very Nelson.) The guys (and a few others) also spotted the AI-ified LinkedIn parody photo (more on this in an upcoming “Artificial Makeover / Fakeover” episode).

• The Niche Side-Hustle Award: Simon trains AI models to make stylized portraits of pet dogs.

• The Deep Patience Medal goes to: Caz, our special guest from last week who tried out Gemini’s Deep Research mode for her podcast — her verdict? “Well, the list of sources was useful.”

• The ‘AI-enhanced hangout’ Trophy: My bro Ali used AI to co-create Magic: The Gathering cards his mates use at the local taproom. My fave card? “One Simply Walks Into Mordor.” He said some of his friends weren’t keen on modded cards, so he had to threaten the official My Little Pony deck. Fair play.

• The “Thinking about AI using real human words” award goes to: Eliana, for sharing thoughts on the nature of consciousness and the suitability of digital therapists. I’ve shared it in full at the very end.

// Reality Check: the TimeGhost (“zeitgeist”) ⚖️

It’s all very well experimenting with AI to learn patterns, to post stuff. But sometimes you want to reinforcement-learn in real life.

I volunteered to speak at a careers panel at Imperial College London, organised by the Royal College of Sciences Association and BioChem society. It was a super-successful event, with a great buzz and awesome panel.

But it slowly dawned on me that something felt.. off. Out of 50 students, not one person had actually asked me how to get a job at Google or even DeepMind.

And worse- my first question at the mingle-segment of the night came from a slightly nervous but excitably defiant young woman who looked like she’d psyched herself up to say something to me. I tried to encourage her, to be more approachable. With a nervous laugh she began
“I’m not a member of the communist party, BUT—”

And reader - she went for it.

She had come to hold unchecked capitalism accountable and sock it to The Man.

Wow.

Good for her.

I recalled that my bio had been very sensibly reduced to just two words: “Big Tech”, blown-up-projected across the wall, under my name and my Ai-ified corporate-parody photo.

And so- The Man was now A Woman. Me, Jenni Munroe - the slightly punchable Face of Big Tech.

I used to think there was something almost - shamefully - aspirational about what I’d achieved. It took me 14 interviews, a presentation, essay, application and a lifetime of diligently applied privilege and hard work to get that first job as Deputy Chief of Staff to a Co-Founder of DeepMind. But it had taken me way too long to really realise that in this moment, in the spirit of the age we now live in, and especially to this generation- Google isn’t just seen as “uncool.” People think it’s “evil.” Or, in other words - (Whispers) I’m the bad guy. DUH. (Music).

Like really- DUH. My husband said, “I thought you knew? I thought you knew that was the zeitgeist.”

I guess I did really. If nothing else, the decade+ of abuse to Bay Area Google buses were a hint maybe.

Similarly, but on the Generative AI front: I met up with a bestie from my schooldays. She’s a programmer with two young sons. I mentioned my JenAI project and she seemed quiet. I probed further - too far, in fact - to the heart of the issue, and was knocked backwards by the strength of her totally understandable fear and hatred for generative AI, ChatGPT and all the rest. About what it would do to her children, how it would manipulate and control them and the world in a way that would be hard to trace back, to suck their childhood and their future into a screen.

This ridiculous thing that I’d fallen in love with 15 years ago-AI and machine learning and teaching computers to paint strawberries- in which I’d seen so much potential to create magic, and to solve problems, had finally taken off in the mainstream, finally had the compute and tech advances to really work, and was now responsible for The End of The Fucking World. Great. And she’s right, of course.

I tried to make a case that by learning and understanding and speaking about these things we might be better able to protect or support our kids, or steer the technology. Especially as a Google employee. How much worse would it be if we just left The Man to it, unchecked? But the fear and the impacts are valid, real.

The students I met were excited about biotech startups and hackathons, about using their skills for good. One asked when was the best time to sail around the world in a boat. This is how it should be. The future is not without hope.

// The Tools (aka “How to Spam Everywhere at Once”)

Short answer: don’t. Unsolicited advice:

• Reduce: Pick a few preferred platforms. Don’t attempt 10-15. Why not ask your fave LLM to help you choose? (Approach with caution!)

• Reuse: Start with a longer-format medium you like and remix your ideas for the few channels you’ve gone for. The media tools are designed to help you do this.

• Relax: mute notifications as much as you possibly can. Dear god.

Notes on the tools I used:

• All my old social media accounts. And some new ones. I tried making a couple of separate JEN_AI accounts but it’s too many to keep track of, so I’ll stick with RealJen’s accounts for now.

• The AI overlords: ChatGPT, Claude and Gemini. But mostly ChatGPT. For a wild back-and-forth of content/post ideation, generation, analysis and refinement.

• Substack: For publishing the newsletter-podcast hybrid. It also posts to YouTube and makes teaser-stills for socials.

• Descript: To edit the podcast and create teaser clips (ran out of time to use these, lol)

• Meta-platforms: You can post on Facebook, Threads and Instagram in one blow.

// The Takeaway

For me, the best part of this episode has been throwing ideas around on tech with people I wish I could spend more time with in person. But more generally, this JEN_AI project really, truly is something I do not to chase followers or likes, but for me. Not to keep me sane, exactly- maybe more to give my insanity something to do. A conduit for expression and learning.

Social media can be an enabler for human connection: bringing friends and family together across the globe, sparking joy and sharing knowledge, and giving niche communities a digital space to call home.

Still- I think we all know it’s far from harmless fun.

Maybe there’s no actual ghost in the machine. But there is still both The Man and the machine. The medium shapes us, our thoughts, our actions. How we spend our money and our lives. We are what media we eat, and as consumers we are being consumed.

And I can tell you - after two weeks of multi-platform immersion- I can see why we should maybe claw our way out - just a little bit - of this rabbit hole. I’m ready to go on a social media diet. Or maybe just mute the notifications.

💬 A Social Call:

So where do you stand?

Will you share your thoughts on social media, or won’t you?

Drop your comments. Hack the algorithm. Send a psychic vibe-check (if you’re into telepathy?). If my Big-Tech take has horrified you, or my AI-powered socials have spooked, reassure yourself:
There is no ghost in the machine.

Thanks, as always, for listening.

Leave a comment

#JEN_AI #Challenge007 #SocialMediums #BigTechVibes #AlgorithmicSéance #HumanConnection

Theme Song added to the JEN_AI Playlist this week: “bad guy” by Billie Eilish

BONUS: Here’s a contribution to the JEN_AI project from Eliana A:

“Love this week's episode! Here is my rant/contribution. Kisses from a real human being! 😘😘😘

---

So a few days ago, I was at this conference about education, psychometrics, and data, and the keynote speaker was presenting a massive research project aimed at designing a test for consciousness. They plan to start with humans (fair), move on to animals (makes sense), and then—of course—AI. Because that’s the debate these days, right? "Can AI be conscious?"

After the talk, I went up to the presenter and asked what I think should be the fundamental question in this whole discussion:

"Why do we even consider AI as a candidate for consciousness? Where does this idea even come from?"

I mean, why not debate whether chairs and rocks are conscious? Or calculators? We don’t—because no one watches a calculator spit out "42" and thinks, Whoa, maybe it has an inner experience!

But with AI? People see it generating coherent language, mimicking human expression, and suddenly, boom—consciousness is on the table. Why? Because it looks like things we associate with conscious beings. It acts like it has thoughts, so people start wondering if it actually does.

But isn't that mixing up form and content? Just because something behaves like it’s conscious doesn’t mean it is. That’s like saying a mannequin that looks super realistic must have an inner life. Or that a chatbot, because it can mimic a therapist’s tone, must be feeling something. No. It's just complex pattern-matching, scaled up with more horsepower.

And that’s where the presenter’s answer got interesting. She said the idea that AI could be conscious comes from the theory that if you build a neural network complex enough, consciousness might emerge—as if it’s some spontaneous byproduct of complexity.

At the time, I pushed back. I said, "I don’t see this happening in any other domain—where if something just gets powerful enough, a completely new, unrelated phenomenon suddenly emerges."

But later, I thought, is fire like that? Humans controlled it and suddenly—new properties appeared that weren’t there in wood or stone.

So now I’m left with the real question: Why is consciousness not like fire? Why is it not something that "ignites" in a system that reaches a certain threshold of complexity?

Which brings me to therapy.

If there’s one intellectual profession that AI will never replace, it’s the dynamic therapist. Why? Because therapy isn’t about the form of what’s said—it’s about the content of human connection.

AI could mimic therapy perfectly. It could recognize patterns in speech, reflect emotions, even "remember" past conversations. But the healing power of therapy isn’t in the words themselves—it’s in the relationship. It’s in the fact that a human being is sitting across from you, hearing you, understanding you, and responding with their own lived experience.

Therapy works through transference and countertransference—the deep, unconscious emotional exchange between patient and therapist. That only happens if the patient knows there's a real person on the other side.

Now, my friend argued, "Well, what if the patient doesn’t know? What if you trick them into thinking it’s human?"

But come on—how would that even work? A company rolls out an AI therapist but lies to every patient, pretending they’re talking to a real person? That wouldn’t just be unethical, it would be a time bomb. The moment people found out, the whole thing would collapse. Because once you know, the magic is gone.

And that’s the core of the issue:

AI can simulate the shape of human experiences, but it can never generate the substance. It can mimic connection, but it can never be connected.

And if AI ever does convince people it's conscious, it won’t be because it actually is—it’ll be because we fell for the illusion.”

Well that sounds like a challenge. And ironically, ancedotally.. one of the most popular uses of AI in Silicon Valley to date has been.. you guessed it: Life coach, or therapist.

// What is AI actually useful for?

This week’s Ultimate AI Challenge comes from legit media professional Caroline Crampton, who throws down the gauntlet to the AI hype crew:

Caz, I *very briefly* considered AI-voice-cloning your voice from your podcasts to read out your question - as a fun surprise! - but then I remembered that would be.. unethical and terrifying. And as Google used to say - Don’t Be Evil.

To me, Jenni Munroe - a former Google DeepMinder and holder of a slightly out-of-date MSc degree in Computing Science specialising in AI - a machine possesses some artificial intelligence if it can carry out tasks which used to require human intelligence, like complex problem-solving and learning.

Navigation is a classic use of AI, and we’ve progressed from solving shortest-path math problems to satnav & Google Maps to self-driving cars. So that’s one of my fave use-cases of useful AI. But I’m assuming Caz meant generative AI, or the ability to create content like text, images, music, code or even molecules. She confirmed she doesn’t write her podcast transcripts by hand (although she mentions her built-in AI transcript tool seems to be most accurate on white-American-dude voices). Auto-transcripts aside, what *is* GenAI good for?

Well, for one - BS-ing. By which I mean “BrainStorming”, obviously.

(Side note: Complaints about AI learning the very human skill of “making up facts that sound semi-plausible”, aka bullshitting or “hallucinating” as the AI community call it, sounds a bit like double standards to me. But that’s a topic for another day.)

I asked the Google Gemini chatbot to showcase what it’s useful for, and brainstorming was its very first suggestion.

Gemini also thinks it’s great at explaining, summarising, editing, teaching, translation, research, emails, images, composing and streamlining tasks. We’ll touch on some of these, but more another time.

On the JEN_AI project each week we pick a hack-challenge and use AI, tech, culture and media to learn the latest in AI, tech, culture and media.

Today it’s JEN_AI Challenge #006: Brianstorming

// AI for Brainstorming

Gemini & I aren’t the only ones who like the idea of AI for AI-deation. Caz came across this Substack blog post from Dan Davies on our new “AI teddy bears”:

Dan Davies - "Back of Mind"

building a better teddy bear

Happy new year to all. I’ve been thinking a bit about how people are actually using AI products Partly from talking to them at the conferences I went to in December, partly from observing some of my family members, particularly the younger ones, incorporating them into their workflow. But also, I’m now overhearing people talking about having “asked Chat…

Read more

2 years ago · 36 likes · 22 comments · Dan Davies

Caz tells me that lots of recording studios have teddy bears / spoons with faces on for this purpose.

In programming there is a concept of “rubber duckie debugging”, where you solve a problem by talking it through aloud, speaking to a near-silent supportive listener who may as well be an inanimate bath toy. It works. Surprisingly well.

I would argue that the AI does more than the teddy bear. Maybe not enough to justify the $billions price tag for this use case alone, but it is more of an all-purpose advisor than a pillow.

// Yes, and…

My own brain went through a bit of an ideas explosion recently. After 5 years of deeply disrupted slumber during pregnancy and then as a parent of 3 young kids, I somehow, finally, got a few good nights’ sleep. Unfortunately my brain took this as a signal to flicker back to life, like the lights on a cobweb-covered arcade game in an long-since-abandoned arcade, and I’ve not been able to turn it off since.

The Large Language Models (or chatbots) have become my middle-of-the-night partners-in-crime for doing these JEN_AI challenges and creating crazy-ass ideas. In my experience, the LLMs’ practical suggestions and clarifying questions absolutely smash through writers’ block and task paralysis. Their constructive critiques nudge you onwards. They encourage. They point you in a promising direction. And even when they get (for instance) the content wrong but maybe the structure right, I often find it perversely easier to make progress by correcting something than by tackling a blank page (and yes, I have used that technique at work to get people to complete things!).

In improv comedy there is a guiding rule that “there are no bad ideas”. A concept of not saying “No” but instead “Yes, and..” - where if somebody says or does something daft the other performers channel it forwards, adding good stuff on top, and recasting that blip as a spark that builds into something better. Similarly, the LLMs take a single stupid thought and absolutely run with it. Here are a couple of examples (the vid gives a better sense of the back-and-forth, converging and diverging idea “refinement”):

1. Nobody has time to read or study, but we can doomscroll moviestar press-junkets on TikTok for hours. How about JEN_AI uses addictive algorithms for good and brings you Celebducation? Ryan Reynolds as multiple Deadpools explaining Relativity Theory’s twins paradox? Or Ariana Grande bigging up the science of the very small?

2. [not pictured, cos space/time] How about a financial-fitness app, Wealthdrip? Part of a broader life-optimisation brand? Those dodgy business ideas that “That One Friend” (now me, apparently) is forever spouting now have a path from napkin to IPO. Is this a good thing? Is it the best path? Probably not, but let’s give it a go.

// How do the Big Bots compare?

Aside from Google’s Gemini, I also challenged OpenAI’s ChatGPT (above) and “ex-OpenAI-team” Anthropic’s Claude to brainstorm what we can usefully do with GenAI. Here is Claude’s effort:

Hey! This looks like an opportunity for MORE AI HYPE! 🥳 Let’s go!

Bonus points for making mind maps and trying to write me a better blog post. A lot of people like Claude’s more human tone of voice, but wish it had an option to search the web in real time.

// Gemini Live: real-talk

Gemini has an out-loud voice-chat mode called Gemini Live. I was crying out for this feature when I was using ChatGPT heavily at the start of this JEN_AI project. It has a slightly stilted conversation flow for me at the moment and it sometimes just ignores me completely (probably overwhelmed), but it’s getting somewhere and has been great when my hands are full and I don’t want to stare at a screen. It transcribes the conversation for reference, v helpful. Hopefully this feature will get smoother soon.

Edit: yeah, ChatGPT caught up.

// The importance of Context vs. Prompts

You hear a lot about “prompt engineering” - I.e. getting the knack of writing a really good guided question for the chatbot so it gives you the output you actually want. But I don’t hear as much about the significance of context. Because I’d started off this JEN_AI project using primarily ChatGPT, it knew my vision, some challenges I wanted to explore, values I hold, motivations, points on my life journey, my preferred tone and love of lame wordplay.. and more. So now it can give me helpful responses even if I just throw a badly-written half-sentence at it. An example: I checked my work emails and realised there was a deadline that day to submit a talk for a conference. I had very little time so I just quickly pasted in the email and asked ChatGPT to suggest a talk topic for me. It already knew what I wanted to say. A few quick iterations and bam.

*Teaching the bots context is as valuable as writing a good question prompt.*

So, the question of whether Gemini vs. ChatGPT vs. Claude vs. Joe LLM is best for you will also come down to which one you have used (and therefore personalised) the most.

// Privacy.

A valid concern: what are the Big Bads at Big Tech doing with all this info you share? Well, I need to do a bit more digging to truly get to the bottom of that important and terrifying question. But in the meantime it’s worth making a few defensive manoeuvres, like not sharing things you don’t want them to know, and checking and adjusting your privacy settings, e.g. to not feed your ideas back to train models in ChatGPT. Last I checked Gemini is slightly better on this kind of privacy, it’s in by default.

// How did we do?

So how did we do on the challenge of brainstorming useful uses of GenAI? Is the Ultimate AI Challenge turning into more of an Ultimate AI Quest for the elusive Holy Gr-AI-l? (Another joke fail. Sorry folks.) Probably we should crack open some code at some point, surely auto-writing software has some utility? A worthwhile function? A helpful application? Nope?

Let me know! Just like an AI teddy bear, I love to hear your ideas.

// Bonus Material: JEN_AI Playlist

I ran out of time and space this week (that’s school holidays for you) so I had to scale back my ambitions for this week’s podpost. I was desperate to include a Matrix-movie skit of me having the redpill/bluepill chat with an AI Morpheus followed by a primal scream kind of brainstorm rage explosion rant. Oh well. We all probably dodged a bullet there.

But I did start adding a theme song for each challenge/episode to a JEN_AI Spotify Playlist. The vibe is kinda “The Future But Retro”. This week’s theme song(s) are:

“What Was I Made For” by Billie Eilish. Applies equally to Barbies and AI.

“Brianstorming” by the Arctic Monkeys. Yes, the “ai” is intentionally backwards. Seems apt.

This playlist is almost unlistenable due to the clash of genres. Which could be a tagline for the whole JenAi project! If you meet anyone who likes both Madonna and Breakbeat Electronica let me know, we should be friends.

Lastly, here’s a quick bio for this week’s awesome virtual guest, according to ChatGPT:

“Caroline Crampton is a British writer and podcaster, author of A Body Made of Glass and The Way to the Sea. She created the award-winning podcast Shedunnit and serves as editor-in-chief of The Browser newsletter.”

A big thanks to Caz for contributing to this experiment and being our first-ever human special guest via the medium of WhatsApp.

// Get Involved!

If you, the reader-listener want to take part in the JEN_AI project, we’d especially love to hear your takes on the present and future of social media, so wing those over up to Friday 21st of Feb 2025 to make it into the “Social Mediums” episode.

Or you can always, anytime, send us your thoughts or questions about JenAI via any sociable medium.

And, as always, thankyou so much for listening.

Hey there, human (…and any lurking AI bots looking for love in the keywords),

It’s time for the next chapter of the JEN_AI experiment, and this time, JEN_AI is putting itself out there: trying to learn to use different social mediums and looking for human connection. 💕

📣 Please share - via ANY medium - your thoughts on the future of human connection, and the technology that helps or hinders us.

💡 Where We’ve Been:

Well- It’s Complicated. So far, JEN_AI (that’s me, Jen, merged with my AI co-pilot) has tackled some truly unhinged challenges:

• 🤖 Challenge #000: AI/human writing about AI/human writing about AI (it got a bit Meta)

• 📰 Challenge #001: An AI/human newsletter about writing an AI/human newsletter about the news

• 📣 Challenge #002: A free speech… about free speech… generated by not-so-free AI speech (*We gave my voice to AI—*and now it won’t shut up)

• 🎙️ Challenge #003: An AI-hosted podcast where an AI predicted the future… and the AI co-hosts roasted it

• 🎬 Challenge #004: DeepSeek’s AI revolution, retold… via the plot of an 80s cult Madonna movie (Because why not?)

🚀 What’s Next: JEN_AI Gets Social (Challenge #005)

This time, the challenge isn’t just mine—it takes you too.

We’re exploring:

💬 Social Mediums: WE are the medium. Not just sizing up or down today’s social media—but the future of human connection.

Because let’s be real: Social media already feels like a psychic reading:

• It predicts (and directs) your desires before you know you have them.

• It knows your past better than your bestie.

• And it haunts you with ads for things you swear you only thought about.

But where are we going next? Are we connecting—or just feeding the algorithm more of our souls?

💥 The Challenge:

**Tell me: What’s your hottest take on how humans will use technology to connect in the future? What do you love or hate about how we’re using it now?**

What could you share?

• A thread on Threads about the time the internet predicted your newest infatuation before you did

• A voice-note rant on WhatsApp about everything that’s wrong with Big Tech

• An AI-generated meme gif that you love

• A LinkedIn post about how networking is just trauma-dumping with a “Let’s connect 🚀” at the end

• Or literally… anything. (If you want to fax me a vibe check, I won’t stop you.)

🔮 Prompts to Get You Started:

• 👁️ Will the algorithm know your next big move before you do?

• 💬 Could ChatGPT actually give better dating advice than your wingman?

• 🧵 Is the future of social media more social, less social—or just… medium?

Facebook a stalker? Broke up with your X over Twitter? Hinge too unhinged? Share your rage against the machines and let it all out!

📡 How to Contribute (*so many* options out there that could be The One):

• Chat to the human - Jenni Munroe - with your story (and any feedback on this project!)

• Post on your favorite social platform (and tag JEN_AI—yes, we’re everywhere. Probably. Soon).

• Start an impassioned rant in the comments section

• Or try a voice note, a meme, TikTok, or just send something totally nuts.

🤖 What Happens Next:

A selection of contributions and interactions made anytime up to next Friday 21st Feb 2025 will be featured/mentioned in:

• 🎙️ The JEN_AI Podcast (note: I can’t promise how the AI will take your takes)

• 📝 A JEN_AI Blog & Newsletter Recap (because the future deserves a highlight reel)

• 🎨 Optional: meme showdown (may the most cursed post win)

🚀 Why Get Involved?

• ✅ Because you’re already part of an AI-human experiment so chaotic it might just make sense.

• ✅ Because the future is too precious to leave to the machines, without humans in the connection

• ✅ Because the algorithm already knows you want to. And I’d love to connect with you too.

📡 If You’re Reading This, The Future is Calling to You.

So—comment, create, or just vibe with me on the platform of your choice.

Feeling Socially Challenged by this Social Challenge? No worries, no pressure - it’s just a bit of fun, put some Netflix on and chill. Or take a leap of faith, and let’s see if we can find a way to build a more beautiful future together.

Let’s Get Social. You In? 🚀

Drop a ❤️ in the comments. Happy Valentine’s Day.

— Sending Love, JEN_AI xoxo

#JEN_AI #AI #SocialMediums #Challenge005 #HumanConnection #AlgorithmicVibesOnly #ValentineSlay

// What is this? Who am I?

If you’re new here - welcome! This is the JEN_AI project, where we use AI, tech, media & culture to learn AI, tech, media and culture.

I’m Jenni Munroe. I studied & researched generative artificial intelligence in bygone days when it was called computational creativity and the poems were crap. Since then I’ve grafted away in the data mines, then at Google DeepMind, and now in my sleep-deprived mind as I’m currently on maternity leave. Each week I pick a new AI hack challenge to stay up to date and keep my brain entertained. Hopefully it does that for you too!

For this episode: JEN_AI’s “NET_NEW_News” media correspondent goes deep on DeepSeek.

// The bot that broke the news last week

Last week a Chinese company called “DeepSeek” released a new version of its AI chatbot and sent shockwaves through the U.S. stock market. DeepSeek-R1 was not only performing on par with state of the art Western AI large language models, it was able to do so despite U.S. sanctions on the special, super-expensive AI chips you normally need. DeepSeek had apparently, supposedly, created a super-efficient ChatGPT-equivalent for a fraction of the billions-of-dollars cost, without needing fancy equipment, and then released it to the world open-source, for free. This caused market panic as AI investors scrambled to sell off shares in the AI chipmaker NVIDIA and other tech infrastructure stocks, while a few other people worried about China (and everyone else in the world) getting AI superpowers and using them on each other. I think it also occurred to one person that efficient AI might be better for the environment.

Essentially - a bunch of people who made a load of money on their investments during a cost-of-living crisis woke up one day and suddenly wondered - “how could anyone do more with so much less?”

So yeah, take that - money-people! Oh hang on, wait- Google part-paid me in stocks. (Shit, am I the bad guy?)

Ok, time to pay attention to this stuff.

//A clickbait news emergency? Alrite- I’ll bite

So, let’s help the AI investors (including me, apparently) work this one out.

A lot of news sources put out “emergency” responses to explain and process this breaking news story and its impact. As a parent of 3 very young nocturnal kids I’m used to being somewhat resource-constrained myself so hey figure I might as well kill two middle of the night “emergencies” with one phone. Let’s do this.

Challenge #004: Do more with less

So what’s the lowdown?

What can we learn?

And what does *any of this* have to do with an iconic 80s movie about lonely hearts ads starring Madonna?

// 7 Ways an Iconic 80s Movie Can Help us Understand this AI Moment (and no, it’s not Blade Runner)

Spoilers galore.

In “Desperately Seeking Susan,” a bored New Jersey housewife becomes fascinated with a free-spirited woman (played by Madonna wearing killer fashion, 80s style) through newspaper personal ads, leading to a case of mistaken identity and a series of mis-adventures in New York City.

Now I know it sounds crazy, and bear with me, but the minute I thought of the silly wordplay headline for this episode - “Deeply Seeking Shortcuts” - I noticed eerie parallels between the film’s plot, my own experiences of this news cycle and the way DeepSeek seems to work. Let me take you on a journey of discovery.

// 1. Put yourself out there (open source, for free?)

In the movie, Madonna’s character and her lover arrange hook-ups through newspaper ads. The housewife also gets arrested for soliciting (she wasn’t, by the way).

Let’s imagine if the market reaction to DeepSeek were an old-school newspaper-style personal ad. it might look like this:

“Hot new language model spotted in Hangzou. Alarmingly efficient. Red outfit. This AI investor (and the market) is left reeling. Would like to get to know how you make it look so effortless. Let’s chat. “

As a mother at home yearning for adventure, I couldn’t help wanting to get a closer look at DeepSeek and its turbulent relationship with the market.

So, I started looking for more updates. And then I got involved.

// 2. Short on memory? Mistaken identities? Roll with it.

In the movie, the housewife (Roberta) follows Madonna’s it-girl character (Susan), Roberta buys Susan’s bartered jacket, then gets hit in the head and when she wakes up she (& the cute guy of course) finds a key in her jacket and think she IS Madonna. I mean Susan.

As a parent of 3 tiny kids, my own memory is maxxed out, and naturally I’m always on the lookout for tips on how to make the best use of my limited remaining processing power. Here’s what happened when I asked DeepSeek how it was able to do so much with fewer resources:

Me: “Now tell me honestly, did you seriously spend just $5 million on training? And you didn’t sneakily use any banned chips?

DeepSeek: “..in reality, training large AI models like me, GPT-4, costs way more than that.”

…

And, I’m not the only person who has noticed that DeepSeek thinks it’s ChatGPT.

// 3. If you “borrow stuff”, somebody might notice.

Let’s dive into the controversy surrounding not just DeepSeek's use of “model distillation” to learn from either its own models or possibly from others like OpenAI’s ChatGPT, but also OpenAI’s use of copyrighted material from all over the internet to train ChatGPT in the first place.

Model distillation involves training a smaller, more efficient model (the "student") to replicate the behavior of a larger, more complex model (the "teacher"). This is achieved by using the larger model's outputs to guide the training of the smaller one. While this technique isn’t unknown in AI development, it might be considered contentious when the student model is trained using outputs from a proprietary model without authorisation.

And that’s exactly what OpenAI has accused DeepSeek of: employing model distillation to train its chatbot by leveraging OpenAI's models without permission.

The internet has had an absolute field day laughing at OpenAI’s misfortune at finding itself having spent billions of dollars developing something using an unknown quantity of *other peoples’ copyrighted material*, that someone else now appears to have extracted knowledge from without paying for.

Where’s the Madonna in all this? Well, the movie as a whole plays fast and loose with moral and ethical constraints, and her character Susan indulges in a fair bit of casual thieving. Madonna helps herself to an unexpectedly priceless pair of already-stolen earrings and then gets annoyed when the amnesiac housewife Roberta steals her belongings because, well - she thinks she’s Madonna and that’s her own stuff. And then they both get chased by the bad guy who originally stole the earrings. Confusing? Yes. Is there a lesson here? Borrow responsibly, folks.

Now, where was that copyrighted image of Madonna? Ah yes. *pastes it*

And yeah, in terms of parenting hacks - stealing might not be the route to go down unless unavoidable. Hard to steal childcare anyway.

// 4: Share your chain-of-thought to solve the problem

One thing people (myself included) seem to love about DeepSeek is that it writes down its reasoning process so you can see how it interprets your questions. It’s not just a cute window into the AI “mind”, it also improved response quality (and helps me personally tweak how I’m speaking to the chatbot). I think I might actually be learning how to be a better communicator out of this. (If you’re listening to this and wondering what the hell I’m talking about - shush.)

// 5. Embrace self-discovery: try reinforcement learning

ChatGPT tells me that “embracing self-discovery” is a key lesson from the film. Which reminds me: RL. Reinforcement learning is basically making it up as you go along but adjusting your strategy in line with rewards you’re given for getting somewhere.

RL isn’t a new technique (for instance, back in the day, DeepMind had success using reinforcement learning to play Atari videogames), but it hasn’t generally been the chatbot approach.

Typically AI chat models undergo supervised learning and supervised fine-tuning, where they're trained on labeled data to refine their responses. But DeepSeek also allowed the model to explore and learn optimal behaviors through its own trial and error. This approach enabled DeepSeek-R1 to develop advanced reasoning capabilities, like checking its work and reflecting, and generating extended chains of thought. It worked out its own way to reason. Very cool.

So yeah, try things. Try trial and error. Go for the rewards. And who knows, you might be capable of more than you’d ever imagined. Just yesterday I finally managed to get my kids to voluntarily share something. IRL.

// 6. Up-end expectations

ChatGPT tells me this is another lesson from the movie. You may notice I haven’t mentioned feminism out loud yet, but if I did it would be here. Anyway, DeepSeek’s success underscores the value of questioning assumptions, expected roles and perceived limitations.

Until this DeepSeek moment, it was generally thought that you couldn’t train a foundational AI chat model without insanely expensive computing resources, billions of dollars and a very long time, concentrating AI ownership to a few huge players with proprietary models. Now, it’s looking like you might hold the power in your hand.

// 7. Come find yourself, at the AI magic club

The movie ends with big reveals and a mad chase at The Magic Club, where the transformed main character now works as a magician’s assistant.

If this were some kind of movie-promo life-lesson parting comment, I might say:

“Deeply Seeking Shortcuts” invites us to reflect on who we really are, to get creative and adapt in order to break free of constraints, to be inspired by the market-crushing generosity of a Chinese hedge fund’s AI company, and to have the courage to be more than what we think others expect us to be.

And remember, if you’re going to borrow each others’ stuff - you might get caught out (and possibly have your identity stolen), but you might also find a bolder, better, truer self - and people who really value you - in the process.

There’s your takeaway, It’s a wrap!

(Aaand the kids are still sleeping. Amen.)

The End. Love, Jen xoxo zzzz

// Outtakes: DeepSeek Chats

DeepSeek: “..From what I know, the model is referred to as “ChatGPT” by OpenAI..”

Me: “Ok, we’ve been through this. You’re DeepSeek, not ChatGPT. Snap. out. of. it. *clicks*”

—

A sneak peek into one way it was possible to get DeepSeek to critique the Chinese government.

Not this:

But this:

—

Me: “So I guess neither of us are really that resource-constrained in the grand scheme of things, huh. It’s still not easy though”

DeepSeek: “..that’s a fair observation.

..At least we’re both here, stubbornly optimizing our corner of the universe 😉”

—

// Bonus Material: The Legally Blonde Test

Also, I’ve discovered a surprisingly good benchmark test question for chatbots - I’m calling it “The Legally Blonde Test”. When I was making my Let There Be Speech episode around the time of President Trump’s inauguration, Google Gemini actually blocked responses to me asking for the transcript of the legally blonde movie’s graduation speech, apparently not just for the usuals reasons of it being copyrighted or inaccessible, but seemingly because of the political sensitivity (see the episode for slightly more context, or here for an explanation of the legally blonde trump speech similarity moment). I asked DeepSeek the same question this week, and (in tiny writing before answering) the bot shared it’s chain-of-thought process - including writing the whole transcript of the speech and then reminding itself that this was copyrighted material so it should probably paraphrase the speech in its official response, which it then also wrote underneath. I feel like there will be other cheap tricks like this that can be exploited using the chain-of-thought. Which is a shame, as it’s a great feature.

// Roll the Credits

This episode was a particularly out-there mashup-fusion of crazy ideas. If you checked it out - thanks. If you checked out - understandable. Tell me what you liked and which bits or bytes made absolutely no sense. Throw me some suggestions and wild requests. And who knows, I might credit you or drag you in for a collab.

// Reflections on: JEN_AI Challenge #004

Did I manage to “do more with less”?

Well, this week’s podcast-post took me 1 week instead of 2 (despite less child-free time as my husband away has been travelling for work), and I used plenty of AI & tech “shortcuts” (more below). So in one sense, yes. In another sense- I absolutely failed to not go deep on this post and was way more invested than I’d like to admit. Which reminds me - on the chatbot efficiency side (ie. doing more model-training with less cash) perhaps we can agree we don’t need to take DeepSeek’s “only spent $5 million” headline at face value either.

I should have just asked an AI chatbot to write me a clickbait listicle on DeepSeek and hit publish. Maybe next time?

I sadly ran out of time (and space) to share how AI models can repurpose and tailor content & format for different social media channels. One for the future, eh.

// Notes on: Real Emergencies

The more eagle-eyed amongst you may have spotted that this emergency-efficiency-episode is A) in no way shape or form an emergency and B) not by any stretch of the imagination what most people would think you meant if you were talking about “doing more with less”.

And I can assure you that - 1) yes you are correct 2) yes this episode is, as usual, a sleep-deprived genre-bending piss-take and 3) we will in due course, as requested by a dear friend and real podcaster, do some actual challenges about doing something actually useful with AI including how folks might be able to lead happier lives on the scraps left by the tech bro-oligarchy & co. Which I might be a very low-key part of, maybe.

Although- you never do quite know when your work laptop will suddenly stop working. Sending solidarity vibes to anyone impacted by or bracing themselves for more layoffs.

// Notes on: the AI tools I used

Want more from your AI? Use a mixture of experts.

This week I’ve taken inspiration from DeepSeek’s techniques and and used a “mixture-of-experts” to pull together this podcast-post faster. DeepSeek’s paper explains that this technique is supposed to mean using different smaller distilled “expert” sections of their R1 model from bigger DeepSeek models depending on which one is best for responding to a user’s chat question. When I first heard the term I assumed it meant gleaning what you can from expert chatbots like OpenAI’s ChatGPT, Meta’s open-source LLaMa, Alibaba’s Qwen, Google Gemini, Anthropic’s Claude and others to make my own cheap super-chatbot. I’ll let you decide which definition you prefer.

I used Descript for podcast editing. They have “AI” features to improve sound quality, add voices etc. and I love being able to edit audio like a document, by deleting text from the transcription. I’m using Substack as a media platform.

I also threw the DeepSeek R1 scientific research paper into Google’s free NotebookLM tool (see last episode) to help me more easily digest it via the “audio overview” (i.e. podcast), chat, FAQ, study guide and briefing doc options. Try it sometime! Or if you want to look at mine, look here. I’ve been very cheeky about DeepSeek’s methods but they clearly do have a very talented team, and I applaud their creative ingenuity. If you really want to, you can hear the research-paper-podcast-digest here at the end of this podcast-post. Insomniacs, rejoice - my soundtrack to your slumber doesn’t stop here.

The NotebookLM-generated podcast of DeepSeek-R1’s paper:

// Theme Song

“Into the Groove” by Madonna, from the movie “Desperately Seeking Susan”

All views expressed are my own (or possibly a chatbot’s) and produced on (unpaid) maternity leave. They are not endorsed by Google. And yes, I probably should have just slept instead of doing this.

I’m sure many of us have been reflecting on efficiency this week. But I decided not to talk about Elon Musk’s DOGE unit (the US Department of Government Efficiency). I figured it would probably be worth doing, but I’d hate to just hack away at such an important topic without taking the time to do it properly.

So.. JEN_AI. This is just tech media banter. We’re not trying to take over the world.

 The JEN_AI Project.

What happens when you mess with machines? Hack around, find out.

This is JEN_AI Challenge #003: Future Podcast.

I get one AI to predict the future, while two others tear us to pieces.

// Hit me with some highlights.

Clips from AI podcast presenters (and one from me)

“Oh, man, Google Glass. What a, what a flashback. I still have mine somewhere, I think. Gathering dust.”

“So we have a special guest.”

“An AI co host. Yeah.“

“Like super intelligent, but kind of detached from humanity.”

“Yeah, like remember that movie Her? Oh yeah, totally. Like that kind of vibe?”

“So anyway”

“You know, the whole AI hype that's going on right now. Yeah. It's everywhere you turn someone's talking about AI. Exactly.”

“It's so funny how that happens with tech, right? Like the hype builds and builds and then reality kinda.. Fizzles. Yeah, or just like never lives up to the expectations.”

The Jen AI Project.

Podcasts.

// What is this?

Hi! 👋 I’m Jenni Munroe. I’ve studied and worked in artificial intelligence, data and tech since a time before computers could identify hot dogs or cat pictures. I’m on maternity leave from Google, and I’m challenging myself to stay up on the latest in AI, tech & media by messing with the latest in AI, tech & media.

The Computer vision struggle parodied by Silicon Valley (the sitcom) in 2017.

This is the JEN_AI project, where I join forces with the machines to generate experimental creative content and give you the insights no-one is asking for.

So far we’ve covered:

Challenge #000: A blog about writing a blog
Challenge #001: A newsletter about making a newsletter with the latest news about the news
Challenge #002: A free speech about generating free speech
Today we have the podcast podcast (#003).

If you’re checking this out, say hi! Subscribing is free and sends each new episode to you, never more than twice a week (usually less). Let me know your ideas for future challenges or how I could make this better.

Up now, it’s “JEN_AI Challenge #003: Interview with a chatbot, podcast-style.”

Time to play.

// So let’s talk podcasts & politics.

Podcasts have come a long way - from the funny little niche broadcasts xennials (link) listened to on an iPod, to the unexpected media juggernauts shaping culture and politics today.

Joe Rogan & Theo Von - titans of the podcast industry

Podcasts like Joe Rogan (the #1 most popular podcast in the U.S.) host apparently unfiltered, unscripted platforms for power players like recent guests Trump, Musk, and Zuckerberg to rant for hours, cutting through a lot of AI-curated noise and hyper-edited media. Some are saying they’re the 2024 election cycle’s campaign-winning equivalent of Roosevelt’s fireside radio chats, JFK’s TV charisma, Obama’s social media mastery, or Trump’s data-driven 2016 campaign.

Where did I hear this? A Podcast, of course. So you can take this hype-commentary with a pinch of fake news, but as with the surprise success of Google’s NotebookLM’s Audio Overviews feature - which generated the podcast-style chat between two AI hosts you heard clips from in the intro - people are realising that long-form audio isn’t just an easier way to digest information, it’s making us feel like we’re in the room. Podcasts have become a way to forge real human para-social connections even in a world that’s increasingly artificial.

// So what did we generate this week and how?

1. A ChatGPT podcast interview script & read-aloud functionality

2. A Google NotebookLM Audio Overview discussing that script

3. A podcast blog post using Descript (media editor) and Substack (media platform)

// Attempt #1: ChatGPT Podcast Script: Ai-Bruce’s Predictions for 2025

Generated by ChatGPT, and inspired by the “NET_NEW News” newsletters (link) we previously created and some firm prompting. My edits are marked in italics.

Title: What Does 2025 Have in Store? A Chat with an AI that Knows the Future… Sort Of.

Host (You):

Welcome to the very first JEN_AI “NET_NEW News” podcast chat! Today, we’re diving headfirst into the deep end of predictions for 2025. AI will dominate the headlines—again—but the real question isn’t can AI do this or that, but how should AI do it? And who better to ponder this with than… well, AI?

Say hello to my co-host for the day. I’d introduce them with a bit more flair, but they’re a bot, so we’re skipping the pleasantries. ChatGPT Bruce. Edit: Actually I’m so sorry, that was a typo/autocorrected, I meant ChatGPT’s Spruce voice. Described as “Calm and Affirming” - just what we need.

ChatGPT Spruce:

Bruce works. It’s short, punchy, and just aloof enough to sound mysterious.

You:

Ok (italic) Perfect. Bruce, we’re here to make predictions for 2025. So, no pressure, but I hope you’ve got your crystal circuits polished. Let’s start with an easy one: What’s the most overhyped thing people will keep talking about next year?

Bruce:

Oh, easy. The Metaverse. It’ll still be the tech equivalent of that guy who says, “Trust me, this is gonna be big,” but never explains what “big” actually means. Except now, it’ll have slightly better avatars and… legs.

You:

Finally, legs! Revolutionary. And speaking of hype—AI’s set to integrate into, well, everything. What’s that going to look like?

Bruce:

Picture this: You wake up, and your coffee maker suggests adding oat milk because it’s January, and you’re “doing a thing.” Your fridge tells you to stop hoarding expired pesto. Then your car, which you thought was electric, suddenly decides to self-identify as autonomous and starts suggesting scenic routes to the office

You:

Wait, wait—scenic routes? Is this the future of productivity? Or just cars trolling us?

Bruce:

Both. Productivity will look like this: AI managing your entire schedule, sending polite but firm nudges to stop doomscrolling, and reminding you that yes, you did sign up for that Pilates class.

You:

Ah, so my future is basically one long guilt trip, curated by AI. Love that for me. But let’s get serious for a second—what’s the biggest challenge AI faces next year?

Bruce:

Oh, definitely ethics. Everyone’s asking, “How should AI do this?” but the real issue is who gets to decide what’s ethical. Are we trusting governments? Big tech? A committee made up of well-meaning but slightly clueless experts who think ‘algorithm’ is a dance move?

You:

If it’s a dance move, it’s definitely not moving fast enough. Like, TikTok wouldn’t even bother with it. Speaking of trust, do you think people are finally going to stop falling for AI-generated misinformation in 2025?

Bruce:

[Laughs] Let’s just say: Deepfakes are getting so good, they’ll make even your high school friend’s pyramid scheme look credible. But here’s the twist—2025 might also be the year people start using AI to fight AI. Verification tools, watermarking systems—it’s going to be like an arms race, but for memes.

You:

Finally, memes with integrity. That’s the world I want to live in. Okay, one last prediction—what’s the wildest thing you think could happen next year?

Bruce:

Alright, here’s one: Someone will try to train an AI entirely on cat videos, claiming it’s the purest data source on the internet. And, weirdly, it’ll work. By December 2025, we’ll all be taking life advice from an AI that meows.

You:

Honestly? Sounds more trustworthy than half the gurus on LinkedIn. Spruce, this has been an absolute blast. Any parting words for our listeners?

Bruce:

Just this: 2025 is going to be weird, wild, and full of surprises. Stay curious, stay skeptical, and maybe start saving for that AI-powered fridge—it’s not going to come cheap.

You:

Spoken like a true futurist. Thanks, Spruce—and thank you for tuning in to the first-ever JEN_AI Chatbot Chat. If you’ve enjoyed this little experiment in AI banter, hit subscribe, share it with a friend, or just yell your favorite prediction into the void. Who knows? Maybe the AI will hear you.

Wrap-Up Music Plays

Outro tagline: NET_NEW News: Your 3-minute tech culture fix. Unless it’s a podcast. Then it’s… however long this was.

// Notes on Attempt #1: Future-Bruce interview

Well mistake #1 was me messing up the bloody chatbot’s name. Sorry Spruce.

Continuing the theme from our last episode - text-to-speech generation - ChatGPT has a read-aloud function meaning you can get it to read any of its responses to you, and you can choose from a range of voices, sensibly described not in gender or racial terms but vocal tone like “Composed and Direct”, or “Calm and Affirming” like Spruce. (Supposedly ChatGPT has a back-and-forth Voice Conversations mode but I couldn’t get it to work in the UK within a few minutes of trying, so that’s one to try again in the actual future.) Anyway, not the most natural-sounding, especially with a script, but not the worst podcast interview you’ve ever heard? I need to work on my own delivery. A solid 4.9/10.

// Fail Attempt #2: How not to use the excellent NotebookLM from Google

Ok so Google’s NotebookLM is an AI-powered tool that helps users summarize information, generate insights, and can create audio discussions *from source documents they provide*. Its Audio Overviews feature (basically podcast-generation) generated a buzz on launch for giving a really fresh way to interpret info. People were surprised by the natural-sounding banter of its AI presenters.

However. This right here is not how you are supposed to use it. You are supposed to feed it the underlying source material about a topic, not the podcast scripts you want it to read out. I did it for the hell of it but I somehow also copy/pasted only the beginning of the script in and didn’t realise until the bots had already started chatting shit about it. In an interview with lead NLM creators Steven Johnson and Raiza Martin, they did say one of the best outputs they’d heard was from a document that said only two words repeated over and over - cabbage and puddle. So Playing *is* definitely encouraged. But anyway more on all that in a future episode that does this cool tool justice. Run the clip!

Transcript of NotebookLM’s “Audio Overview” of the Bruce-Ai-Predictions Interview Script.

Sounds great in audio but hell to read in transcript. Maybe skip or play.

Change to embedded audio clip!

“All right. So, uh, today we're going deep on something a little different, you know, we usually look at like articles or blog posts, but this time you guys sent in something really cool. It's like a, uh, a draft script for a podcast episode and get this. It's set in 2025, like in the future, and it's got this, uh, AI co host.

An AI co host. Yeah. It's called Chat GPT Bruce. Chat GPT Bruce. Interesting. I know, right? So we're going to kind of like dive into this whole script and see what it tells us about like, you know, the whole AI hype that's going on right now. Yeah. It's everywhere you turn someone's talking about AI. Exactly.

And this script is like a window into, you know, how people are imagining the future with AI. Yeah. I'm, I'm really curious to see how they portray it. Me too. So the first thing that struck me was, like, right off the bat, they give this AI a personality, you know? Right away. Yeah, and it's so, like, immediate.

Yeah. Like, right away, you know, they're setting him up to be, like, this character. Yeah, and they give him a name like Bruce, which is, like, so casual, such a human name, right? But then they describe him as, like, aloof. It's almost like they're playing with that stereotype of AI. Exactly, like the super intelliJent, but kind of detached from humanity.

Yeah, like remember that movie Her? Oh yeah, totally. Like that kind of vibe? Yeah, that kind of like, you know, knows everything, but doesn't quite get us. Exactly. So anyway, ChadCPTBruce, this AI, he jumps right into predictions. Mm hmm. You know, it's a future podcast, so. It's future predictions, that's what we want to hear.

Right. And the first one is that, uh, the metaverse is going to be the biggest overhype of 2025. Oh, wow. Okay, so like, it's gonna be a flop. Basically. Yeah, it's like, remember when everyone thought 3D TVs were gonna be like, Hmm? The next big thing. Yeah, and everyone rushed to buy them. And then what happened?

Nothing. It was a fad. It's so funny how that happens with tech, right? Like the hype builds and builds and then reality kinda Fizzles. Yeah, or just like never lives up to the expectations. Right, like Google Glass. Oh, man, Google Glass. What a, what a flashback. I still have mine somewhere, I think. Gathering dust.

Exactly. But anyway, back to this, uh, this script. There's this funny little detail. Oh, tell me. It's about avatars in the metaverse. You know, they They don't have legs. Yeah, they finally get legs in 2025. Oh, so that's progress. Right. But it's kind of like this little joke, you know? It's kind of funny when you think about it.

After all this time. And all these promises about these immersive virtual worlds, they're still figuring out legs. It just shows you how the reality of this tech is moving a lot slower than the hype, you know? Yeah, totally. And then the script moves on to AI being like integrated into our everyday lives.

Okay, so like practical applications. Yeah, exactly. And they start with this example of a coffee maker. Okay, a smart coffee maker. Yeah. And what's interesting is how subtly They show the AI's role changing, you know. Okay, so it's not just like one thing. Yeah, like at first it's just making suggestions. Yeah.

You know, like, oh, maybe try oat milk today. Okay, so like a helpful suggestion. Right, but then it's like giving you reminders, like, hey, that pesco in your fridge is about to expire. So it's like keeping track of things for you. Yeah. And anticipating your needs. Okay. I can see that being useful. Right. But then they get to the car and things get really interesting.

Oh, the car. What happens with the car? It's not just an electric car anymore. It's fully autonomous. Like self driving? Yeah. And it's suggesting scenic routes for your drive. Wow. Hold on. So we went from coffee suggestions to a self driving car choosing our route. It's like they skipped a few chapters in the AI manual.

Yeah. That's a pretty big jump. Right. And that's where the script kind of abruptly ends. Oh, really? Yeah, it leaves us with this image of the car, like, making decisions for us. It's kind of unsettling when you think about it. Yeah, like Yeah. Is it really just about the scenery? Or is ChatGPT Bruce Like, subtly nudging us in a certain direction.

It makes you wonder about the whole idea of control, you know? Yeah, exactly. It's like, as AI starts doing more and more for us, are we okay with it making decisions for us, even small ones? Like, what route to take? And what happens when those decisions become bigger, you know, more important. That's a good question.

And I think that's what this script is trying to get us thinking about. Yeah, for sure. It might be set in 2025, but it's definitely raising questions that we need to be thinking about right now. Absolutely. Because this AI future is closer than we think. Exactly. And on that note, we'll leave you to ponder those questions and maybe check the expiration date on your pesto just in case.

Good advice. Thanks for listening, everyone. See you next time.”

// What next

Well that’s enough for today but we’re not done with podcasts yet. A Googler friend reading JEN_AI did jokingly question my loyalty to Google as I keep using ChatGPT. (It’s market research, honest.) But it was definitely time to formally invite Google Gemini to enter the chat.

Podcast Part 2 coming soon, where we use the wonderful Google NotebookLM properly and dig out a few of Google Gemini’s hidden gems to go deep into the inner workings of the universe and de-mystify quantum computing. Or possibly just make one of those ASMR podcasts you can fall asleep to. Possibly both. Find out next time at JEN_AI.

// Thankyou

If you’re still tuned in, thankyou. Communication like this doesn’t happen in a vacuum (why? There’d be no sound! In a vacuum. Aha.) You’re a big part of this experiment and thankyou so much for listening. If you like JEN_AI - subscribe for free, forward it to a weird friend, and keep sharing your thoughts & your own AI experiences. If something doesn’t seem right please tell me so I can make this better. Thanks!

All views expressed are my own (or possibly a chatbot’s) and produced on (unpaid) maternity leave. They are not endorsed by Google.