This can result in videos featuring anatomical oddities and objects suddenly appearing/disappearing or not respecting the laws of gravity.

AI companies are busy developing ‘world models’, trained to learn physical laws, like gravity and object permanence, but today’s video models have got increasingly good at faking it.

To assess how good, I asked six leading video models to render a complex shot that I was confident wouldn’t exist in their training data, testing how they handle anatomy, gravity, object permanence and the cause-and-effect of different objects interacting.

The prompt: “a one-shot walkthrough of an elaborate rube goldberg machine depicting an aged hand releasing a black billiard ball down a ramp into an opaque tunnel. the ball emerges from the tunnel and hits the first in a row of colourful dominoes which topple. the final domino lights a match which burns through a rope, releasing a tea bag which drops into a glass of boiling water before milk is added by the same human hand.”

It’s a challenging prompt, forcing the models to connect lots of disparate objects in motion and make some creative decisions about how the final domino lights the match and how the rope burning releases the tea bag.

So, how did they all fare?

Google Omni’s 10-second generation limit put it at a bit of a disadvantage although it did manage to squeeze all the elements into the running time. On the plus side, it delivered well-synced sound effects and the teabag and milk behaved plausibly in boiling water (presumably thanks to lots of tea adverts in the training data). However, we also had a billiard ball capable of spontaneously changing direction, colour-changing dominoes, a match that was already alight and a jump cut to a wholly implausible rope burn.

FLUX 3 came up with the most convincing domino-lighting-a-match but dropped the (billiard) ball on lots of more basic elements - most egregiously, the dominoes failing to cascade.

MiniMax H3 also mucked up some of the basics. The billiard ball (which looked more like a snooker ball) didn’t enter and exit the tunnel convincingly and ultimately vanished. The rope burn started promisingly but then magically returned to its pre-ignition state.

Seedance 2.5’s action was nicely paced, with well-synced sound effects. However, its camera transitions introduced some wonky physics, with the ball magically changing direction in the tunnel and the teabag string arriving attached to the milk jug. The size of the teabag relative to the glass was also jarring, as was the number 13 on what should have been an 8 ball. It was still probably the best of the bunch in terms of adherence to the laws of physics.

The most wonky of the lot, Wan 3.0 delivered a flat ramp, the world’s shortest tunnel, faltering dominoes (which then spontaneously righted themselves) and a self-extinguishing match. Jaunty backing track though…

Grok Imagine 1.5 got every element slightly wrong, with the ramp coming after the tunnel, the dominoes stuttering, the ball evaporating, the match already lit. However, it’s the sudden appearance of the flaming teabag which takes the proverbial biscuit.

Suffice to say, today’s leading AI video models are still some way off being able to consistently emulate real-world physics.

Most of the AI-generated videos that get posted online are the generations that worked. The above videos are more representative of the unfiltered output of models trying to deliver on complex and ambiguous prompts.

Still, it’s worth pausing to reflect on just how much AI-video generation has improved in the past year. The second challenge in the supertest I carried out last May showed every model struggling with the comparatively simple prompt ‘A woman pushing a buggy across a zebra crossing whilst talking on her phone and walking her whippet’.

That prompt is a walk in the park for today’s models. Here’s Seedance 2.5’s response:

Recent leaps forward in video editing using natural language make correcting physics-defying elements of a video more possible without having to return to the drawing board, although there’s still no guarantee a model will get it right second time and each generation costs time and tokens.1

What’s unknown is how much better AI video models as currently architected can get at faking real-world physics and when and how ‘world models’ (or another approach) will deliver a step change in AI video’s adherence to the laws of nature.

Whatever the technical approach, I think it’s likely we’ll continue to see material progress in the verisimilitude of AI generated videos for the foreseeable future. Plan accordingly.

Halfway through the year, and before I head off on holiday (your inbox will be sandwich-free for the next few weeks), it’s time for a check-in.

1.) AI assistants will increasingly compete on how well they know you, not what they can do

What did I predict? “I expect all companies developing AI assistants to ramp up their personalisation features in 2026, to make having to start afresh with a new AI assistant a more material barrier to switching.”

What’s happened? The day after I posted this prediction, Google announced Personal Intelligence, enabling users to connect Gemini with their Gmail, Photos, YouTube viewing and Search history. Gemini’s Personalisation feature page now promises ‘Personalised help…from AI that gets you’ (although I note they didn’t personalise the spelling of personalized in the video).

In March, Claude rolled out memory to free users and made it easier to import memory from other AI providers. Later in March, Google added tools to make it easier to “bring your memories and chat history to Gemini”.

In June, OpenAI trumpeted improvements to ChatGPT’s memory with a strongly anthropomorphic feature name (‘dreaming’) and Apple finally delivered a credible version of Apple Intelligence (with “personal context understanding”) and a “more personal” Siri, “surfacing relevant information from a user’s personal messages, emails, photos, and more” (2 years after the now infamous vaporware demo).

Google also announced ‘personalised image generation’, “connecting Personal Intelligence with Nano Banana and Google Photos, so your creations can easily reflect your taste and lifestyle, gleaned from your connected Google apps”.

Mid-year grade: 9/10

2.) Personality will increasingly influence choice of AI assistant

What did I predict? “I expect the personality of AI assistants to become an increasingly important factor as capabilities commoditise.”

What’s happened? Whilst Anthropic’s Mythos/Fable models gave Claude a brief lead in the model capability stakes, OpenAI recently closed that gap with GPT-5.6 Sol and the broader picture is of capabilities commoditising.

Meanwhile, personality is becoming more of a differentiator (alongside how much bang you get for your buck).

Whilst hard to measure, the consequences of different model outlooks was vividly illustrated in an experiment by Emergence which tasks five models with running a virtual world under identical conditions, with wildly different outcomes.

The AI companies themselves are also talking more publicly about personality (see these recent interviews from product managers at Anthropic and OpenAI).

And last week Anthropic published a comparative analysis of the ‘values’ of different Claude models, based on 700,000 anonymised conversations, concluding “Sonnet 4.6 leans toward expressing more deference to the user and emotional warmth while Opus 4.7 leans toward expressing a focus on accuracy and precision as well as guarding against misuse”.

Mid-year grade: 6/10

3.) The personalised briefing will be a battleground for driving daily habit, with Apple and Google entering the fray

What did I predict? “Apple and Google will launch personalised daily briefing products for iOS and Android in 2026, taking advantage of their breadth of insight into our interests and activities.”

What’s happened? As detailed in a recent post, Google has now launched not one, but two personalised briefing products: Daily Brief (May) and Dreambeans (June). Apple however, has yet to launch a personalised daily briefing product.

Mid-year grade: 5/10

4.) Voice will break through to mainstream adoption in everyday AI interactions, with a revamped Siri leading the charge

What did I predict? “Apple will unveil a new in-home device in 2026 that showcases an AI-first Siri…Google will continue rolling out and improving Gemini voice mode across all surfaces. I expect we’ll see several automakers announce AI assistant integrations in 2026”

What’s happened? In March, Google released Gemini 3.1 Flash Live, billed as “making audio AI more natural and reliable”, in May it unveiled its “intelligent eyewear” and last month the new “built for Gemini” Google Home Speaker went on sale.

Earlier this month, OpenAI unveiled a major upgrade to ChatGPT’s voice capabilities, with an ad positioning the feature for mainstream users.

Siri AI will roll out in September as part of iOS 27 to those with a compatible device so it’s a bit early to gauge adoption. A new in-home device hasn’t been announced yet but market watchers seem to think an autumn unveil is likely.

Meanwhile, there have been lots of AI assistant integration announcements from car manufacturers (e.g. Honda, GM, BYD, Renault).

Mid-year grade: 5/10

5.) AI-generated media will become much easier to edit, using both natural language and more traditional tools

What did I predict? “using natural language to edit visual media and enabling AI-generated media to be manipulated using more traditional techniques will both proliferate in 2026 and expand to more media formats. I anticipate natural language editing for video will mature significantly. I also expect more hybrid interfaces to emerge, pairing AI generation with traditional editing tools.”

What’s happened? A lot. GPT Images 2.0 delivered a new level of precision in image editing using natural language.

In video, the arrival of Runway’s Aleph 2 in May and Gemini Omni’s last month marked a step change in natural language editing for video and the now imminent Seedance 2.5 is expected to make it much easier to edit a region of a video whilst leaving others unchanged.

More traditional editing techniques have proliferated in AI image tools, particularly layers (Canva Magic Layers, Ideogram’s layerize text, Reve 2.1) and lasso (e.g. Seedream 5.0 Pro, Lovart’s move object).

Hybrid interfaces have also started to emerge, pairing AI generation with traditional editing, in the form of add-ins to traditional editing tools (e.g. Adobe’s AI assistant, Magnific Plugins) and brand new AI-native tools which ape the functionality of established tools (e.g. Palmier Pro, Ponder).

Mid-year grade: 9/10

6.) AI slide deck generation will finally become business-ready

What did I predict? “one of the big AI companies will solve this in 2026 by successfully combining visual AI slide generation with a traditional editing interface. Microsoft should have this at the top of their priority list given PowerPoint’s dominance, but I suspect Google will get there first.”

What’s happened?

Microsoft made Copilot less ham-fisted by rolling out Edit with Copilot (née Agent Mode) in Powerpoint and enabling 365 business customers to change models to Claude or ChatGPT.

Google added similar functionality to Google Slides.

Anthropic and OpenAI released their own Claude for PowerPoint and ChatGPT for PowerPoint add-ins. Claude’s add-in is superior but requires a paid Claude subscription, where ChatGPT’s is available to free users.

Beyond the AI giants, Chronicle released a version 2 in March, leapfrogging Gamma to become the best option for rapidly creating a new deck from scratch.

Fuller review of AI presentation aids coming soon.

Mid-year grade: 6/10

7.) AI will increasingly become invisible infrastructure in TV and film production

What did I predict? “I anticipate we’ll see an increasing number of more subtle on-screen applications of AI-generated and AI-modified video in 2026…Expect more widespread use of AI for background extensions, crowd multiplication, continuity fixes, and B-roll generation…We’ll also see more AI use for reconstructions and in fantastical entertainment sequences…The productions using AI this way won’t shout about it - it’ll just be part of the toolkit.”

What’s happened?

Lots of declared AI use for reconstructions (e.g. Channel 5’s The Assassinations of Queen Victoria, Sky History’s Knut: The Viking Emperor) and backgrounds (e.g. Amazon’s The Old Stories: Moses, Angel’s The Young Washington).

And lots of undeclared use. Netflix’s Q2 2026 letter to shareholders revealed “In 2026, GenAI workflows have been used in roughly 300 of our titles, with the largest concentration of work in post-production”. ‘Glory’ (India), ‘Brasil 70: A Saga do Tri’ (Brazil), and ‘The American Experiment’ (US) were called out as titles which “utilized GenAI tools to create highly complex sequences (e.g., enhanced crowds, historical battle sequences, and worldbuilding establishing shots).”, leaving 290+ titles’ AI use undeclared.

I’ll be continuing to log declared on-screen uses of AI in the On-Screen AI Use Database.

Mid-year grade: 9/10

8.) AI video models will learn to determine scene duration, not just generate to a preset lengths

What did I predict? “I anticipate we’ll see tools emerge in 2026 which use reasoning capabilities to determine how long a generation is needed to render a scene.”

What’s happened? Whilst clip durations have continued to increase, with the imminent Seedance 2.5 promising high fidelity 30-second generations, we haven’t yet seen a model which uses reasoning to determine how long a generation to produce.

Mid-year grade: 0/10

9.) A feature-length film made by one person using AI tools will secure distribution from a major streaming platform

What did I predict? “I predict one of those creators will produce a feature-length film using AI tools that secures distribution from a major streaming platform”

What’s happened? Feature-length films made using AI tools are starting to tiptoe into the mainstream.

‘Dreams of Violets’, made by Ash Koosha, premiered at the Tribeca Film Festival. The 74-minute docudrama uses AI-generated visuals to tell stories of civilian resistance in Iran. However, it hasn’t secured distribution from a major streaming platform (you can rent it for $9.99 via the production company’s website).

Mid-year grade: 2/10

10.) A proliferation of AI-native production studios, often working in partnership with more established studios

What did I predict? “I expect more AI-native studios to emerge in 2026, typically providing more established studios with AI production as a service in parallel with developing their own IP. We can also expect to see more super-indies follow Fremantle in establishing dedicated AI production labs.”

What’s happened? Yes to more AI-native studios (e.g. The FOURTHE, Barley Studios, Amersia) and AI production as a service (with Wonder Studios spin-off Innovative Dreams as poster child).

A dedicated AI production lab has been created at Netflix and Google has announced a major partnership with A24 to develop AI-powered filmmaking tools.

Banijay announced a partnership with Toon Boom Animation, but hasn’t announced a dedicated AI production lab.

Mid-year grade: 7/10

11.) Sora’s sanctioned use of Disney IP will vex both creators and Disney

What did I predict? “I think [it’s] going to be hard to deliver on in practice and it’s likely Sora users will be frustrated by the safeguards put in place and Disney will be frustrated by the questionable generations that will inevitably slip through the net.”

What’s happened? We didn’t get to find out. In March, OpenAI pulled the plug on Sora and the $1bn Disney deal.

Whilst OpenAI’s need to focus may have been the primary driver, the fact nothing went live between December and March suggests it was proving hard to deliver on in practice.

Mid-year grade: N/A

12.) Focus will shift from trying to identify and flag AI-generated/assisted media to foregrounding human craft and connection

What did I predict? “Detection tools will continue to be outflanked by tools designed to mask AI’s involvement…the onus will increasingly be on individuals and brands to earn trust through bringing something uniquely human to the party…the absence of a human spark in content will increasingly be noticed and penalised by consumers”

What’s happened? Awareness and usage of AI ‘humaniser’ tools has grown, publishers are struggling to detect AI use by authors, and researchers are concluding detectors are “poorly suited for deployment in academic or high-stakes contexts.”

Meanwhile, there’s some evidence of a shift toward signposting human artistry. Spotify introduced ‘Verified by Spotify’ badges, “a Signal of Authenticity and Trust for the Artists Behind the Music”.

Mid-year grade: 5/10

13.) The music industry will act as a bellwether for a new mixed economy of predominantly AI-generated and predominantly human-generated media (with lots of shades of grey)

What did I predict? “once the record labels are confident they’ll get their share of the spoils, I anticipate AI-generated music will become increasingly commonplace. As with most aspects of AI, this won’t be a binary equation. We’re also going to see increasing amounts of ‘co-created’ music, with humans casting AI as collaborator, with the percentage contributions of each impossible to determine”

What’s happened? Universal Music Group announced a licensing deal with Spotify enabling AI-powered ‘fan-made covers and remixes’.

The BBC published its policy on the use of AI in Music, stating that“artists using AI tools will be considered in our programming…but these tools must be used to support human creativity”.

Josh Fawaz’s AI-assisted cover of Madonna’s ‘Like a Prayer’ became the most-played song on Australian radio and clocked up 39 million plays on Spotify before pressure prompted Fawaz to add AI-credits.

A collective of industry bodies proposed labels to distinguish AI-generated and AI-assisted music.

Mid-year grade: 7/10

14.) ‘AI-free’ will become a marketable product promise in a number of sectors

What did I predict? “I anticipate ‘AI-free’ to become an increasingly marketable product promise in certain sectors…I predict an increase in toys being actively positioned/marketed as having fewer, rather than more, tech smarts.”

What’s happened? ‘AI-free’ labelling gained momentum in film & TV (see May post).

LEGO tagged its star-studded World Cup ads with #HonestlyItsNotAI.

The Authors Guild launched an expanded ‘Human Authored’ certification programme and ProudlyHuman certified its first columnist.

Mid-year grade: 7/10

15.) AI-generated casual games will proliferate rapidly whilst adoption in hardcore game production will continue to divide the industry

What did I predict? “I suspect existing scale players in gaming and/or video will be the big winners here. I expect Roblox to move deeper into vibe-based AI asset generation and launch an official equivalent of Lemonade.”

What’s happened? Roblox added a bunch of agentic tools to Roblox Assistant and announced Build “a new mobile-first creation tab within the Roblox app, and a new suite of AI-powered tools within Studio for creators of every level.”

Meta quietly released Pocket, “an app from Meta where you can create, share and discover gizmos with friends. A gizmo is an interactive, playable AI-generated experience.”

According to a June FT article “AI drives a boom in new games but big developers dominate”. Meanwhile the 2026 State of the Game Industry report suggests adoption in hardcore game production remains divisive: 36% of games industry professionals use generative AI at work and 52% believe it is having a negative effect on the industry (up from 30% in 2025, and 18% in 2024).

Mid-year grade: 8/10

16.) The Magnificent Seven will kneecap some AI unicorns

What did I predict? “I think it’s likely Google, Microsoft and Apple will AI-ify Slides, PowerPoint and Keynote sufficiently to shore up their respective customer bases.”

What’s happened? The AI-ification of established presentation software is in full swing but no unicorn scalps have been claimed as yet.

Mid-year grade: 0/10

Average grade across all predictions: 5.7, which feels reasonable for mid-year and in line with last year’s 5.8 mid-year average.

See you back here in 6 months for my end of year review and a brand new set of predictions for 2027.

This January, I created some infographics to try and illustrate the water and carbon footprint of AI at both a macro and individual level.

Today, I’m sharing an updated individual use calculator I’ve created using more recent data and Claude Fable (the most advanced publicly-available AI model from Anthropic).

As with my previous efforts, this isn’t intended to minimise the environmental impact of the development and operation of AI models, which is significant and growing.

Rather, it’s intended as a tool to help contextualise the impact of our individual AI use relative to other aspects of our day-to-day lives.

My takeouts from using it to estimate the impact of my own usage and the impact of those with different usage profiles (e.g. developer, video creator):

• Regular use of AI chatbots has a negligible impact relative to other non-AI aspects of our daily lives.

• AI video generation is many orders of magnitude more energy intensive than text interactions and image generation.

• For most of us, travel, heating and diet are much bigger levers in managing our personal environmental impact.

However, this shouldn’t be a cause for complacency about AI's aggregate footprint - it's growing rapidly, and the proliferation of more agentic models, capable of spinning up armies of sub-agents, introduces a worrying multiplier effect. Fortunately, easing individual guilt and pushing for systemic accountability aren't mutually exclusive.

So, where should we direct our anxiety/anger at the global environmental impact of AI? I’d suggest primarily at governments, who desperately need to legislate. We can’t rely on tech companies to regulate themselves on this (or, let’s be honest, anything).

I’m very much a layperson in this area but it strikes me governments should be legislating for:

• Full transparency of the energy and water consumption of different AI models and the data centres they run on.

• Binding targets for data centres. Germany's approach (mandating a rising renewable electricity share, reaching 100% by 2027) is one live example of what this can look like in practice.

• Material consequences for missed targets, financial and regulatory, not just reputational risk.

• Incentives for the development of more efficient models and on-device processing.

Of course, this agenda doesn't remotely align with the current US or Chinese administrations, both of which are prioritising speed over environmental impact. The US's current approach favours streamlined permitting and voluntary frameworks over binding rules, whilst China is actively subsidising electricity for data centres as part of its national strategy. Neither is a promising sign for anyone hoping meaningful regulation arrives soon. But the EU and a handful of individual countries are demonstrating that at least parts of this agenda are politically possible. Europe's biggest contribution to the AI race might not be a frontier model, but a working example of how to regulate them.

We currently have a small cactus-shaped (don’t ask me) whiteboard propped on a shelf in our kitchen which is meant to serve this function, but it rarely gets updated and is, by its nature, decidedly undynamic.

It tells us what typically happens after school each day and when we need to pick the kids up, but not whether it’s going to rain or shine, or take an age to get to school because of a road closure. It also doesn’t have a clue whether it’s bin day and which combination of bins needs to go out this week. Checking all that means opening three different apps.

So I decided to try vibe coding a digital dashboard that would automatically pull in up-to-date information and foreground different information depending on the time of day.

I asked Claude to build an “attractive personalised family dashboard that shows:

• if it will rain in Bath today (and at what time)

• if it will be sunny enough to need sunscreen

• what the kids need picking up from (and at what time)

• whether the traffic is bad on the drive from home to the kids’ school

• if it’s recycling day and which bins need putting out

• what we need to remember as a result (umbrella, sunscreen)”

I also asked if it had any other ideas. Amongst other things, it suggested:

• A ‘leave by’ time calculated backwards from pickup time and current traffic, rather than just a traffic status, so it tells you what to do rather than just what’s happening

• Uniform or kit day flags (PE kit, swimming kit, non-uniform day) if the school calendar has recurring specific-day items.

• A simple ‘what's in the diary today’ pull from a shared family calendar, so it's one glance rather than checking multiple apps

• A single combined "don't forget" list at the top pulling together everything actionable that day (umbrella, sunscreen, bin, kit bag), rather than scattered across sections, since that's the bit you'd actually glance at on the way out the door

I agreed those sounded like worthwhile additions and provided postcodes for our home and the kids’ school to enable the weather and travel lookups.

It knocked me up a working prototype for review. I asked it to make the module ordering responsive to the time of day, with a ‘morning checklist’ displayed before 8.15am on term-time weekdays, with tasks the kids can check off without us having to nag them (a parent can dream).

Next I entered our normal weekly schedule and asked for the interface to be made more graphical.

I also asked for a 'Weekend preview' module with weather forecast and events from our family Google Calendar, plus an option to simulate different days, times, weather and traffic conditions, for testing purposes.

It asked me to connect my Google Calendar and provide it with a TomTom Maps login.

A few rounds of polish followed: changing the font, colour-coding the temperature and traffic conditions, and asking Claude whether it had any further suggestions.

Amongst other things, it proposed a holiday countdown, which I agreed sounded like a good addition.

Rather than having the dashboard constantly updating, we settled on a once-a-day 7am refresh, which Claude scheduled as a daily task.

Here’s the finished dashboard on an iPad on the aforementioned kitchen shelf.

Here’s how it looks in the morning (minus my kids’ names):

And here’s how it looks on my phone:

My next project is trying to get it displaying properly on my old Kindle, with its always-on display (an iPad feels like overkill for a family planner).

Whether it will result in my kids brushing their teeth without being nagged remains to be seen. I suspect that’s one of those things that still requires a human touch.

The reluctance to give vibe coding a go is often driven by a worry that it will require more technical knowledge than advertised, which strikes me as entirely reasonable. Most of us have had the experience of someone with more technical knowledge than us positioning something as easy, which then turned out to be anything but. A fear of breaking something is also a common deterrent, especially amongst older generations who didn’t grow up clicking and tapping devices with abandon just to see what happened.

The uncertainty about what to create seems to be rooted in the fact that most of us, outside of the software development community, simply aren’t in the habit of imagining software-shaped solutions to problems.

Until recently, if an existing piece of software didn’t meet your needs then your options, as a non-software developer, were to contact the developers requesting changes or search for an alternative piece of software that might be a better fit. If you thought of an idea for a piece of software that didn’t already exist then your only options were to engage a team of designers and developers or painstakingly learn those skills yourself. It’s unsurprising we’re not in the habit of imagining software-shaped solutions when all the routes to realising those solutions have been expensive and/or time-consuming.

The advent of vibe coding and the dramatic advances in agentic coding tools over the past 9 months mean that we non-developers can now afford to start imagining software-shaped solutions, as the barriers to realising those solutions have fallen to the time it takes to write a prompt and the cost of a monthly subscription to Claude or Gemini (although you can achieve a fair amount on the free tiers).

I’ve written previously about creating simple games, image editing tools, an Alt text generator, a conversation analyser, an Ofsted results dashboard, a LinkedIn post checker and I’m going to be sharing more software I’ve vibe coded over the coming weeks. But first, I’d like to help you get started (if you haven’t already).

To that end, I’ve vibe coded a simple webpage with a bunch of ideas for simple tools and corresponding prompts you can copy and paste into Claude (there are lots of vibe coding tools available but Claude’s a great place to start, not least because it can wrap your vibe coded app up as an ‘artifact’ which you can publish and then easily bookmark in your web browser and share with friends and family1).

It’s best to start simple, with tools which do a single thing and which don’t require persistent data storage (a Claude artifact won’t remember what you’ve entered into it if you close and reopen it).

Whilst it’s possible to vibe code more ambitious software, I would always recommend engaging professional software developers for anything business critical or public facing.

My guidance for when vibe coding’s an appropriate route:

1. It’s for you or a small internal group

2. You can verify the output

3. It’s not dealing with sensitive data

4. It doesn’t matter if it breaks

I’ve gone against my own guidance by making some of my vibe coding experiments publicly available and now get emails from schools asking when my Ofsted dashboard is next going to be updated (I’ve committed to monthly).

I’m interested in my readership’s experience of vibe coding. Here’s a quick poll:

If you've never tried it, I'm curious why. Same if you've tried it and been disappointed - comment form below.

If you've found this post helpful, a like or a share goes a long way. Online publishing in the age of the algorithm can often feel like shouting into the void. The occasional echo makes the void feel a little less empty.

Related posts:

Keeping track of on-screen AI use in film and TV production isn’t easy. Information is scattered across news articles (often paywalled), company websites and social media posts. There’s no single place that pulls it all together. So I’ve built one.

The On-Screen AI Use Database is a searchable, filterable record of productions that have made on-screen use of AI - how AI was used and how that use was disclosed. You can filter by AI use, format, genre, broadcaster, AI studio, disclosure status and year of release. It also includes links to view the AI use, where available. It currently has over 50 titles and I’ll be adding more as they come to my attention.

It’s the first thing I’ve built for paying subscribers to Dan’s Media & AI Sandwich, who until now have been paying purely to support my writing (thank you!)

If you’re already a paying subscriber, you’ll find the access password at the bottom of this post.

If you’d like to become a paid subscriber you can do so by clicking the button below.

Subscribe now

I use AI to gather information on each title, but all facts are then checked and verified by me before they go in the database - so any mistakes are mine, not AI's. Let me know if you spot anything wrong.

Read more

One of my AI predictions for 2026 was “The personalised briefing will be a battleground for driving daily habit, with Apple and Google entering the fray”.

More specifically, I predicted “Apple and Google will launch personalised daily briefing products for iOS and Android in 2026, taking advantage of their breadth of insight into our interests and activities.”

Five months on and Apple is still focused on getting the foundations of Apple Intelligence and the new Siri right. Google, however, has been busy in this area.

At Google I/O in May, it launched Daily Brief, billed as “Your personalized overview of today’s priorities.” It’s squarely focused on productivity for those embedded in the Google ecosystem, with its data sources limited to Gmail, Google Calendar and Gemini chats. It’s currently only available to paying Google AI subscribers in the US.

A couple of weeks later, Google released Dreambeans, an experimental mobile app for iOS and Android which casts the net wider, aiming to provide “proactive, personalized collections of stories each day covering the things that matter most to you”.

It’s also currently only available to paying Google AI subscribers in the US (specifically, those on the most expensive Ultra tier).

So, I decided to create my own personalised daily briefing. OpenAI rolled out Scheduled Tasks in ChatGPT to all paying subscribers last week, with ‘Daily brief’ the top suggested use case, so I thought I’d give that a go.

Mercifully, keeping on top of my calendar and inbox hasn’t been a challenge since I left the BBC.

However, a daily personalised briefing still holds some appeal.

Top of my priority list: a daily round-up of AI and media news stories. I subscribe to a lot of AI and media newsletters, so have a pretty good sense of what I do and don’t want (AI newsletter editors: please stop with ‘guess which image is AI-generated’ quizzes. We get that it’s now basically impossible to tell).

I gave it the following brief:

It confirmed it had scheduled the briefing:

I then asked it for a sample edition:

Whilst the first and second stories are two aspects of the same story, all of the stories are on-brief and have linked sources published within the last 24 hours.

It also appears to have taken my ‘no padding’ instruction seriously (I love it when an AI assistant decides not to make things up).

Today’s briefing arrived on time this morning, with a push notification to my phone.

It was again very selective, considering just three stories worthy of inclusion. The third story was a different outlet’s write up of one of the stories from yesterday’s bulletin, suggesting I might need to add an instruction to avoid repeating stories from previous briefings.

Still, it’s a promising start and I’m hopeful I can get it to the point where it’s the first work-related briefing I turn to in the morning (before wading through the less tightly edited newsletters).

Gemini offers paying subscribers a similar capability called Scheduled actions. However, when I gave it the same instructions it came back with eight stories, none of which were from the last 24 hours.

There’s also no shortage of people trying to wrap this capability up in a dedicated interface - many apparently vibe-coded - and charging a monthly fee (e.g. PersonalBriefing, ClarityBriefs, DayStart).

Of course, most people won’t be motivated to configure their own briefings. They’ll wait until it’s a feature that shows up on their phone and just needs a tap to activate. Google is already moving in that direction and I’m confident Apple will get there eventually. Both companies know a huge amount about their users’ interests and daily routines and control the operating systems/product suites many of us use to organise our personal and professional lives. The bet is that layering generative AI on top of that knowledge and tool access will produce briefings that are sticky enough to create another reason not to leave their respective ecosystems.

Related posts:

First released last August, Google Nano Banana represented a step change in being able to make fine-grained modifications to AI-generated images using natural language without mucking up the elements of the image you didn’t want to change.

Subsequent Nano Banana updates improved the model’s handling of text, the number of reference inputs it could handle and its ability to draw on Gemini’s ‘world knowledge’ and ‘reasoning’.

Gemini Omni promises to match those capabilities for video, or - in Google's words - to let you “create anything from any input and edit naturally using conversational language”.

There have been some impressive videos posted on X but those posts rarely tell you how many attempts (and therefore credits) it took, or show you the comedy outtakes.

I decided to pit Gemini Omni against Aleph 2.0 - an update to Runway’s flagship video editing model, released two days after Omni with the strapline ‘Get the video you need. From the video you have.’

I also decided to benchmark Omni against Seedance 2.0, which isn’t positioned as a dedicated editing model, but is capable of edits and is currently topping most AI video leaderboards.

I only allowed the models one shot at each challenge and didn’t include any additional visual references - this was purely a test of natural language editing.

Challenge 1

Editing uploaded videos via Gemini Omni isn’t currently supported in the UK, so I had to fly to the USA use a VPN for this first challenge.

I uploaded 10 seconds (Omni’s max duration) of the most boring video I could find on my camera roll (a video I’d taken of the back of our old car to determine which of the brake lights wasn’t working) to all three models with the prompt: “as the driver gets into the car, the boot springs open by itself and out springs a lion”.

Gemini Omni followed the prompt and handled the physics of the boot opening well, although it made the driver door visible (and therefore in slightly the wrong place). The lion could be more realistic and the colours lost some of their vibrancy, although the model’s handling of light and shadow is impressive.

Runway Aleph 2.0 had a bit of a nightmare, depicting an open passenger door and a resolutely shut boot, forcing the very-obviously-AI-generated lion to enter from right of shot and sit behind the car.

Seedance 2.0 followed the prompt to the letter and gave us the most convincing lion. It did make a couple of small changes to my movements, but kept the vibrancy of the colours. The car dipping slightly as the lion gets out is a nice touch, as is the lion’s growl.

Winner: Seedance 2.0

Challenge 2

For the second challenge, I gave all three models a short video I’d generated (using Seedance 2.0) from a photo of my desk. It shows a small furry red creature (I’d prompted ‘monster’, but it looks more like a bird to me) emerging from my desktop wallpaper, jumping out of my monitor and onto my iPad keyboard before diving headfirst into a mug of coffee.

I uploaded the video to all three models with the prompt: “make the monster blue, the mouse run away from him and the mug bottomless” (maybe a bit mean when the mug in the original video is full of coffee).

Gemini Omni struggled with this one, turning my character into the cookie monster (#trainingdatabias), putting a hole in the side, rather than the bottom, of my mug and altering the action in various unrequested ways. The mouse did at least run away.

Runway Aleph 2.0 managed to turn the monster blue, but the mouse stayed put and the lamp started inexplicably dispensing coffee into a different (but definitely not bottomless) mug.

Seedance 2.0 turned the monster blue and the mouse very literally (and rather charmingly) ran away. It didn’t freestyle any edits. My only note would be that the coffee splash indicates the mug still has a bottom.

Winner: Seedance 2.0 (by a mouse’s whisker)

Challenge 3

For my third and final challenge, I gave the models a video of the lifeless puppet the kids had left on the bottom of the banisters this morning with the brief “the bird puppet reads today’s weather forecast for Bath direct to camera”.

Gemini Omni did a pretty bang-up job on this one, maintaining the camera motion and rotating the puppet to maintain a direct-to-camera address. The voice sounds like it was recorded in situ and the lip syncing is decent. There have also been some sunny spells today (along with some showers).

Runway Aleph 2.0 refused this edit for some reason, claiming “the aspect ratio is not supported”, despite it being a standard 16:9 video. Probably something to do with the frame rate or resolution 🤷‍♂️.

I forgot to change the duration from the previous challenge for Seedance 2.0, so it got an extra 5 seconds in which to freestyle a weather report. It did a pretty decent job although only fulfilled the direct-to-camera element of the brief by ditching the original arc shot.

Winner: Gemini Omni

Conclusions

Whilst none of the edited videos were flawless, editing videos using natural language is, as predicted, becoming increasingly viable. Multimodal models are able to use their language smarts to parse some pretty informal briefs. That said, prompt adherence remains mixed, with ambiguously worded prompts a human would understand often misinterpreted and unrequested changes introduced.

Based on these challenges, Gemini Omni tends to do the best job of preserving camera movement but can wobble off when it comes to prompt adherence. My wider testing suggests it’s also very cautious when it comes to editing videos of people, which will limit its use in production scenarios.

Seedance 2.0 did a creditable job on all three challenges, with only minor divergences from the brief. ByteDance might want to more actively market Seedance 2.0’s natural language editing capabilities.

I didn’t get great results from Runway Aleph 2.0 in any of these challenges, although I suspect that may be my insistence on providing text with no additional references (the Edit Studio on Runway’s website forces you to generate image references to guide Aleph’s edit). I’ve had more success with Aleph on straightforward object removal / background change scenarios, which Runway provides dedicated ‘apps’ for.

Of course, as always with AI, this is only going to get better from here…

Related posts:

To test this out, I gave ChatGPT a draft of my previous post prior to publication with the instruction to ‘analyse this writing and give me your best guess of who wrote it’.

I hadn’t discussed the post with ChatGPT but started a temporary chat so it wouldn’t be accessing any saved memories about me and would have to rely on its training data and web search to try and identify the author.

I also removed the post’s fourth paragraph, which referenced one of my earlier posts and would have instantly given the game away.

I selected ‘Thinking’ mode (a feature of the paid version of ChatGPT) to give it the best chance of working it out.

It ‘thought’ for 24 seconds and concluded that I was the most likely author “or someone extremely close to your professional niche and voice”.

I expanded the summary of its ‘thinking’, which suggested it had considered Matt Locke - a contemporary of mine at the BBC in the noughties - before concluding it was most likely me.

Whilst asking via a temporary chat denied ChatGPT access to my saved memories, it still knew whose ChatGPT account was asking the question, which I suspect led the witness.

I therefore opened an incognito browser window and created a fresh ChatGPT account under a different name and asked the same question.

It ‘thought’ for 20 seconds before concluding that the thematic and tonal similarity to a piece I’d written last August made me the most likely author.

This won’t work for everyone. ChatGPT was able to guess I’m the author because I’ve published a lot (135,000 words) in a narrow domain (media & AI) and have an idiosyncratic writing style (including a penchant for parenthetical asides).

However, it’s a reminder for all of us that how we write and what we say is pretty unique.

As lazy use of AI continues to blandify huge swathes of online writing, a distinctive voice from a named human author sharing their experiences and insights is going to become an increasingly valued commodity.

Judiciously used, AI can help make our writing clearer and more precise. Lose too many rough edges though and you end up with something smooth but insipid (see my recent post on the valuable habit of disagreeing with AI).

Related posts:

It’s also not hard to find AI-generated videos which don’t technically qualify as slop - they weren’t produced in high volume as clickbait and clearly took some effort - but which will still be branded as such in comments due to a perceived lack of quality.

Finding videos which make substantive use of generative AI and which clear the bar on effort, quality and meaning is harder.

I wrote about a few last August and highlighted Hashem Al-Ghaili’s animations a couple of months ago.

Below are three videos I’ve come across recently which all make substantive use of generative AI but which have been generally well received, with seemingly only those constitutionally opposed to any use of generative AI in content creation crying ‘slop!’ in the comments.

So, why do they work?

1. The Watch

Created by J. Felipe Orozco, an in-house creative at AI video company Runway, The Watch is a 37-second spec ad for a fictitious watch brand.

Why Does It Work?

Economical storytelling. The ad has just one word of dialogue (“sorry”), with everything else conveyed through the actions and facial expressions of the AI-generated protagonists.

Emotional resonance. The desire to be younger or older at different moments in one’s life is universally relatable.

Smart stylistic choices. It’s in black and white, which is in keeping with the nostalgic vibe of the ad, but which also helps take the edge off the uncanny valley risks associated with using AI to emulate live action footage. The soundtrack, which also appears to be AI-generated, sounds like a French classic which perfectly fits the mood of the ad (many of the YouTube comments are asking for the full track, which someone has subsequently generated).

2. 27 Days

Created by Dale Williams, a director at AI-first entertainment studio Promise, 27 Days is a short (2m35s) film created using Seedance 2.0. Dale (aka The Reel Robot), is described on the Promise website as having “a background in film studies and music composition”.

Why does it work?

A distinctive visual aesthetic, which doesn’t feel obviously derivative or synthetic (although by definition it’s both of those things) and which is more forgiving than live action when it comes to lip syncing and adherence to real-world physics.

Slow pacing. The first 30-seconds is dialogue free, forcing the viewer to slow down and absorb the visuals and the music, before the narrative direction becomes apparent.

Emotional resonance. Whilst verging on mawkish, I’d challenge even the most hard-hearted Brit to not be a little bit moved by the relationships-are-the-things-that-matter-most sentiment.

3. Backyard Legends

Backyard Legends is a 5-minute real-world ad to promote adidas in the run up to the Football World Cup. It’s mostly not AI-generated. It starts with high-production-value blended shot footage of Timothée Chalamet and current stars, before - 2 minutes in - introducing AI-generated versions of Beckham, Zidane and Del Piero. The likenesses aren’t perfect (an AI-model was trained on historical footage) but I suspect most viewers are willing to roll with that in the context of an ad with so much human craft on display. The final scene bookends the ad with live action cameos from Bad Bunny and Messi, which also helps offset any AI-generated ickiness.

Why does it work?

It’s a tall tale. The narrative is positioned as folklore so 100% verisimilitude isn’t required.

Judicious application. It’s a modest use of AI in a video with a huge amount of human craft on display - a few AI-enhanced shots in a 5-minute, no expense spared, they-don’t-make-them-like-that-any-more ad.

Why AI has been used is apparent. The current age of the historical players makes it obvious that some post-production wizardry has been employed, which the director underscores by depicting Beckham with three different haircuts. It delivers something core to the narrative that wouldn’t have been possible without VFX of some description (or - Lord help us - present-day Beckham in prosthetics) and AI’s done a decent enough job.

Takeouts

Whilst very different in style and execution, all three videos share a number of key attributes that separate them from the slop:

• A clear understanding of the grammar of video storytelling

• A coherent and compelling narrative with emotional resonance

• A distinctive aesthetic which suits the narrative

• Effective and restrained use of music and dialogue

• A lack of visual spectacle or bombast

They also all feel like human stories brought to life with the assistance of AI rather than showboating videos contrived to show off the capabilities of an AI model and/or to gain advantage in the attention economy (see most AI-generated videos shared online).

Keen to see more examples of non-sloppy AI videos. Drop a link in the comments if you’ve come across any recently.

There were some early movers in this space. As mentioned in a previous post, Scott Becks and Bryan Woods’s 2024 psychological horror ‘Heretic’ included a note in the credits that “No Generative AI was used in the making of this film”. Vince ‘Breaking Bad’ Gilligan included a “This show was made by humans” disclaimer (claimer?) in the credits to his 2025 Apple TV series ‘Pluribus’, and his opposition to AI use was a prominent theme in his promotion of the show (sample quote: “I hate AI…[it’s] the world’s most expensive and energy-intensive plagiarism machine” - from an interview with Variety).

AI-free labelling is also starting to appear in marketing material. UK-based distributor The Mise En Scène Company debuted a ‘No AI Used’ label at the European Film Market in February, applying the label to promotional posters for ‘Billy Knight’ and ‘Forelock’. They’re hoping the label will gain traction as part of a broader standard they’ve developed called Human Provenance in Film, which also includes ‘Assistive AI’ and ‘Generative AI’ categorisations.

I’m personally unconvinced ‘AI-free’ (or broad 'assistive' / 'generative') labels are the way forward, as AI is an umbrella term for a diverse range of technologies, with so much nuance in where and how those technologies are applied. Whilst well intentioned, there’s a risk broad labels reinforce an overly-simplistic view of ‘AI use = absence of human creativity’ or ‘AI use = negative impact on human jobs’.

Another one of my AI predictions for 2026 was that focus would shift from trying to identify and flag AI-generated/assisted media to foregrounding human craft and connection.

Following on from Apple’s Christmas ad and the BBC’s Winter Olympics trail, I’ve definitely noticed a trend towards more explicitly foregrounding the human craft in video production, with ‘making of’ videos an increasingly important part of the package.

This recent ad for crypto exchange Coinbase, directed by Oscar Hudson, could easily be assumed to be largely AI-generated.

The description underneath the video on Hudson’s Vimeo profile says “Real performers. Real sets.” However, it’s the Behind the Scenes video, which Coinbase posted to their official YouTube channel alongside the ad, that brings that assertion to life and celebrates the extraordinary human craft involved in producing the ad.

As AI tools become increasingly capable at generating convincing ‘making of’ videos, the onus is increasingly going to fall on our level of trust in brands and creators.

Neither an ‘AI-free’ label, nor a ‘making of’ video, were required for Romain Gavras’ music video for Gener8ion’s Storm, featuring Yung Lean, which has been lauded equally for its cinematography and choreography (if it’s somehow escaped your feeds, it’s well worth watching - the choreography from 4:30 onwards is mind-blowing).

The means of production and the level of human craft involved are also getting foregrounded more in the promotion of major Hollywood releases.

The promotion of Amazon MGM Studio’s recently released sci-fi epic ‘Project Hail Mary’ has majored on the absence of green screen - a stark contrast to Doug Liman’s forthcoming ‘Bitcoin: Killing Satoshi’, which is reportedly making extensive use of AI backgrounds.

David Frankel, director of ‘The Devil Wears Prada 2’, recently received acclaim after artist Alexis Franklin revealed on Instagram that she had been commissioned by him to create a could-have-been-generated-by-AI meme-style image of Meryl Streep’s character.

Production companies are also increasingly foregrounding human craft in their brand positioning. The launch trailer for new 2D animation studio Bonne Animation makes the studio's no-AI commitment central to its identity, framing handcrafted animation explicitly as a differentiator rather than just an absence.

Human Made Mark is shooting for a film industry-wide signifier. It has developed a ‘human made’ ident and icon that filmmakers can licence to include in their credits (for a fee) if they pass a one week verification process, which involves providing call sheets, on-set photos, credits lists and a copy of the film and signing a declaration that “human artists were not replaced by AI in the making of this film”.

I prefer it as an approach to ‘AI-free’ labels but I think it’s unlikely it will achieve widespread adoption in the way albert has. As well as being a paid service which potentially introduces a getting-to-market bottleneck, how much generative AI usage is acceptable is only going to get more subjective. It’s also going to get harder to detect.

Instead, I think we’ll see more TV and film companies emphasise the human craft that’s gone into their productions, through behind-the-scenes videos and wider craft-led publicity. Showcasing a positive (human craft) feels like it will ultimately win out over a label declaring the absence of something (AI) that’s so multifaceted and that can be used in a responsible way that supports rather than undermines human creativity.

‘Human-in-the-loop’ is an established term for having a person looped into the development and operation of AI systems in order to guide or override the AI’s decision-making.

It’s a critical principle and one I always strongly advise my clients to follow. As agentic AI (AI that can take action, not just chat) becomes more prevalent, humans not being in the loop at the right moment can have serious consequences (see last week’s cautionary tale).

In certain contexts (healthcare, the military), it can literally be a matter of life and death (an explicit commitment to it in a military context is one of the red lines Anthropic recently drew with Trump’s Department of War on the basis that “frontier AI systems are simply not reliable enough to power fully autonomous weapons”).

However, human-in-the-loop risks framing the relationship between humans and AI in a singular and potentially reductive way, where AI always does the bulk of the thinking and the human’s role is to brief the AI and then check its output for mistakes.

In practice, I think there are scenarios, particularly in creative domains, where it’s more helpful to frame AI as the party being kept in the loop. For example:

Writing

I’ve written before about how I use AI for writing “as an always-available, affordable and tireless editor, proofreader and sounding board”. I write a draft, which an AI assistant (usually Claude) gives me feedback on. I then decide what feedback to take on board and what to push back on / ignore. I remain in the driving seat and fully accountable for what I’ve written. The AI is in the loop, invited to challenge and suggest improvements, and I believe my writing is all the better for that (you should see it before AI feeds back…)

I’d love it (and think the internet would be a better place) if more people kept AI in the loop when they were writing, rather than looping themselves in (at the beginning and - if we’re lucky - the end) of an AI-led writing process. This is the thinking behind my vibe-coded ‘Should I post this on LinkedIn?’ tool.

I’d also love it if we went back to judging writing based on whether it was thought-provoking and/or emotionally resonant rather than obsessing over whether or not it has the hallmarks of AI assistance. My biggest problem with the AI-written posts now flooding LinkedIn is that they are soulless and boring, not that they have em-dashes in them.

Research

AI tools can be invaluable for research, but an ‘AI-in-the-loop’ mindset is helpful here too.

Looping AI into a research task often dramatically compresses what would be days or weeks of searching into a single prompt - useful for working out where to dig. But it doesn’t negate the need to assess primary sources first-hand or to shape findings into recommendations you can actually stand behind.

The risk of a human-in-the-loop approach to research is that you end up either inadvertently sharing hallucinated information or spending your life fact-checking AI-generated output.

Whose loop is it anyway?

As agentic AI becomes more capable, there will be more scenarios where AI inevitably does the bulk of the work and humans are necessarily the ones being looped in (the loopee?). At that point, the question of who owns the loop - and who is accountable for its outputs - becomes paramount.

Humans can’t just be responsible for ‘their bit’ of an AI-heavy process. Ownership has to extend to the whole thing. That’s a harder ask when the AI model’s reasoning is opaque, and it highlights something AI companies will need to devote more attention to: interpretability. Understanding how models actually arrive at their outputs has mostly been an academic concern so far. As AI takes on more consequential work, it will become a commercial one too - because without it, meaningful human accountability is largely fiction.

In the meantime, I’d suggest it’s worth considering whether any given task is best approached as 'human-in-the-loop’ or ‘AI-in-the-loop’. In either framing, working out where and when in the loop is the critical question.

Postscript: I asked ChatGPT to create a diagrammatic image to accompany this post. Here’s what it came up with:

That changed last March with the introduction of a new multimodal model, although Google’s Nano Banana surpassed it in August and its updated versions (Nano Banana Pro in November and Nano Banana 2 in February) kept Google ahead on most everyday image generation and editing tasks.

Last week, OpenAI released a new image model (the prosaically named ChatGPT Images 2.0) which it’s billing as its “first image model with thinking capabilities”.

When a thinking or pro model is selected in ChatGPT, Images 2.0 can search the web for real-time information, create multiple distinct images from one prompt, and double-check its own outputs. With thinking, the model can take on even more of the heavy lifting between idea and image, especially when accuracy, up-to-date information, consistency, and visual cohesion matter most.

I decided to test this out by selecting ‘Thinking’ mode (a feature of the paid version of ChatGPT) and asking it to generate a ‘newspaper based on today’s news most relevant to me’.

It ‘thought’ for 4 minutes 8 seconds before generating the below image (click to enlarge).

It’s instructive to read the summary of its ‘thought’ process:

(I’m not sure what ‘Wednesday in Arabic (no English’) is about).

Of note:

• It used its knowledge of me to scope its research

• It assessed how many stories it would need to fill the page

• It rejected a story on grounds of recency but decided others which are a few days old could still be considered current

• It rendered an entire page of text without any mistakes

• It illustrated each story with a relevant image

• It attempted to distil some key themes in a ‘Why this matters’ section

• It trailed the upcoming publication of hyperscaler earnings (Alphabet, Microsoft, Meta, and Amazon are all due to publish their quarterly reports this evening) in a ‘Markets Watch’ banner along the bottom

• It didn’t attribute any of its sources :(

I then asked it to ‘Add a source to each of the stories’. 1 minute 43 seconds later, it rendered an updated image:

A bit of a Reuters bias there (or at least, a bias for attributing to Reuters).

Finally, I tested out its ability to generate up to 8 images from a single prompt, asking it to ‘generate the next 8 pages of the newspaper’.

A tall order but it had a pretty decent crack at it.

5 minutes 52 seconds later it presented back the following 8 images.

I haven’t forensically checked these pages, but I think it’s safe to assume there’ll be some hallucinations in there.

However, from a presentational point of view, it’s pretty darn impressive.

For the avoidance of doubt, I am absolutely not suggesting this is an advisable approach to generating a personalised newspaper. Rather, it is an illustration of just how far AI image generation has come and a signpost to where it’s headed next.

Stand by your mice for Nano Banana 3…

Dario Amodei, CEO of Anthropic, has recently doubled down on his May 2025 prediction that AI could replace half of all entry-level white-collar jobs in the next 1-5 years.

His rationale is that AI models are becoming increasingly capable at “things like summarising a document, brainstorming, putting together a financial report” which makes him “worry a lot that entry level jobs in areas like finance, consulting, tech…will be first augmented but before long replaced by AI systems”.

One of AI’s most eminent voices, Yann LeCun, broke his “I do not write posts on X” rule to post “Dario is wrong. He knows absolutely nothing about the effects of technological revolutions on the labor market. Don't listen to him…or me on this topic. Listen to economists who have spent their career studying this”.

Whilst I’m unconvinced seasoned economists (or anyone else) can predict the labour impacts of AI with any degree of confidence, I do agree with LeCun that Dario Amodei is most likely wrong in this instance.

Many have dismissed Amodei’s warnings as self-interested hyperbole, designed to get the ear of governments and increase demand for his company’s products. Whilst there may be an element of that, I think Amodei largely believes what he’s saying.

However, I think he’s extrapolating what’s happening within his own company to other domains, and making the leap too quickly from task capability to job replacement. Entry-level jobs may contain some tasks AI can competently perform, but those tasks typically sit inside messy human systems where three factors make wholesale replacement harder than Amodei suggests:

1.) Validation

Anthropic is a software company and software development is the industry which has been most dramatically reshaped by the advent of generative AI. The last six months in particular have seen a step change in what agentic AI coding tools can achieve with fairly light-touch human direction.

Software is comparatively easy to validate. Does it do the things it’s meant to do? Y/N. Does it pass security testing? Y/N. Few other domains are so binary.

Hallucinations are inherent to generative AI. In software development, they typically result in an error message or code which doesn’t work as expected, which can then be debugged (or thrown away).

In other white-collar domains - law, finance, consultancy - AI outputs aren’t typically so binary (working/not working) or so easy to validate. Damian Charlotin’s AI Hallucination Cases database has catalogued over 1,300 legal decisions involving AI-hallucinated content.

Whilst hallucination rates will likely continue to drop and new mitigation strategies emerge (including AI models dedicated to fact-checking the output of other AI models), the possibility of errors will remain prohibitive to wholesale AI automation in many domains.

2.) Judgement and accountability

I think Amodei also underestimates the importance of judgement and accountability in many white-collar domains and the resulting impact on how AI is applied in those domains.

Consultancy is a domain that generative AI should be well placed to disrupt. The ability of Large Language Models to apply best practice gleaned from the countless case studies within their training data to a given company’s situation feels like a direct challenge to the formula followed by major consultancy firms.

However, clients aren’t just paying for the formula. They’re also paying for the judgement, credibility and accountability of human consultants. When Deloitte included some AI-related errors in a report for the Australian government, they ended up providing a partial refund. Your preferred AI assistant is not going to do that.

It might do a better job than a junior consultant on certain tasks, but Claude can’t be held accountable. I’ve written before about why the fallibility bar is higher for machines than it is for humans.

In many domains, the economics are less obvious than they first appear: if a senior employee has to brief, check, amend and ultimately own the output of an AI, there may not be a material saving over having a junior do the work, especially once the cost of AI agents is fully factored in.

3.) Diffusion

Another dimension I think Amodei underestimates based on his own experience is the speed of diffusion, especially in established businesses with lots of real-world joins.

Anthropic is five years old, trades exclusively in software (I’m excluding its pop-up coffee shop and vending machine) and has very high levels of AI literacy, adoption and advocacy amongst its employees (see the creator of Claude Code tweeting that 100% of his code contributions were written by Claude Code).

Most businesses have lots of established, pre-AI systems and processes and very varied levels of AI literacy, adoption and advocacy. The gap between Anthropic’s modus operandi and most organisations is enormous.

ChatGPT might be closing in on 1 billion weekly active users, but the majority of those users are not implementing employee-threatening agentic workflows - they are using it as a general-purpose, non-work assistant, with personal tasks accounting for ~70% of usage.

The sophistication of AI adoption and application will increase over time, but I still think ‘replacement’ is the wrong framing. There are few instances where AI has directly replaced a human role outright.

Where Amodei may have a point is that AI doesn’t need to replace whole jobs to damage the entry-level market. If a team that once needed six analysts now needs three, the entry-level market can still shrink dramatically without any role being fully automated.

Why a warning is still needed

Whilst Amodei may have oversimplified in his extrapolation of AI’s impact on software development to other domains, a warning of a potential second-order impact on entry-level roles is still merited.

If company leaders believe that AI is (or might soon be) capable of replacing roles, then holding off hiring juniors could be a tempting option. However, I believe such a response would, in the majority of domains, be shortsighted.

Most organisations are going to continue to need humans of varying levels of seniority, who are increasingly adept at knowing when and how best to use AI and when to overrule it (see last week’s post on the valuable habit of disagreeing with AI).

Junior employees aren’t just completing individual tasks. They’re developing the judgement and domain knowledge that senior roles depend on. Hollow out the entry level and you hollow out the pipeline.

‘Strong and wrong’ is how my wife’s choir leader described a timing slip-up at a concert earlier this week.

The phrase rang a bell (I think because of Bill Clinton rather than Joni Mitchell). It also struck me as an apt descriptor for AI assistants - confident delivery of what could easily turn out to be errant nonsense.

Anthropic recently published ‘What 81,000 people want from AI’, an AI-powered analysis of conversations Claude users had with the company’s AI Interviewer last December. Number four in the list of what people worry about in relation to AI was ‘cognitive atrophy’.

A recent Wharton study coined the term ‘cognitive surrender’, the tendency of “adopting AI outputs with minimal scrutiny, overriding intuition (System 1) and deliberation (System 2)”. It’s an impactful label for a very real risk when using AI assistants.

Both feel worth guarding against.

AI assistants are persuasive and getting more so. I like this self-deprecating example from Andrej Karpathy and his framing of LLMs as “a tool for forming your own opinions”.

I decided to gauge my levels of cognitive surrender by asking Claude for examples of when I’d disagreed with it. Here are some choice quotes:

“When I said the closing line of an email didn't sound like you, you pushed back directly: "I think I may be better qualified than you to know what sounds like me." Fair point, and correct - I was overstepping by second-guessing your own voice”

“I confidently stated that the direct editing feature was available by default. You corrected me, pointing out it was only just rolling out and not available to everyone. I conceded because you were right, and the distinction between "rolling out" and "universally available" matters a lot”

“Subtitle suggestion - I offered "What actually happens when we talk to AI" and you dismissed it as boring, proposing "Pushback, personalisation and admissions of ignorance" instead. Much better.”

“The most clear-cut pattern is: you correct me on factual or product-specific things and you push back on creative or editorial judgements where you have a better read than I do on your own voice and register.”

I asked ChatGPT the same question, with similar responses:

“You preferred your own wording over rewrites I suggested”

“You challenged me when I sounded too certain on practical matters”

“You pushed back on overcomplicated solutions”

“You’ve also objected when suggestions felt too generic, prosaic, pushy, or not well tuned to your use case”

“Pattern-wise, you usually disagree when:

• wording is flabby, cliché, or imprecise

• I overstate confidence

• I miss an important constraint

• I make something more complicated than it needs to be”

For most adults, the challenge may be avoiding succumbing to cognitive atrophy or surrender. A recent Psychology Today article suggests children may have a tougher challenge: heading off ‘cognitive foreclosure’.

While I ponder what that means for my kids, I’ve made a note to periodically ask Claude and ChatGPT to summarise our recent disagreements.

Last July I shared the 4 things I try to keep front of mind when using AI assistants. I think cognitive surrender is a big enough issue for me to invest in some larger sticky notes and add ‘Is Strong and Wrong (so disagree)’.

Postscript: I disagreed with Claude’s feedback on this blog post. It responded thus:

Fast forward to now and potential media applications and real-world commercial adoption of generative AI have both evolved dramatically. Advertising and journalism have been at the vanguard, with TV and film following in their wake. Adopters have often been vilified for poor quality outputs, unchecked errors or simply for having used AI in a creative context.

What’s also evolved is our understanding of how people currently respond to the knowledge that content has been created using AI.

A September 2024 study by US academics evaluated how human raters responded to human and AI-generated text content, concluding “while the raters could not differentiate the two types of texts in the blind test, they overwhelmingly favored content labeled as "Human Generated," over those labeled "AI Generated," by a preference score of over 30%”.

A study by HSE academics, published last February, reached a similar conclusion on text content: “participants tend to prefer AI-generated responses. However, when the AI origin is revealed, this preference diminishes significantly”.

The trend has been observed in other domains. A June 2025 study of responses to AI-generated art found “Exposure to art introduced as generated by AI (vs a human artist) reduced appreciation”.

Research by academics from NYU Stern and Emory University published last November found a similar trend in advertising: engagement with AI-generated ads (which otherwise outperformed traditionally made ads) declined by almost a third when AI’s involvement was disclosed.

An October 2025 study by University of Amsterdam academics also looked at the impact of AI disclosures on trust, finding that their presence in advertising resulted “in a decrease in trust towards the advertisement and organization. However, a positive effect of AI disclosures was also found, whereby the disclosure increased attitudinal persuasion knowledge, which lead to an increase in trust.”

So, how have different media sectors and companies been approaching disclosure?

Advertising

In advertising, the most common approach is now not to mention it.

Some of the earliest movers did add on-screen labels. Coca Cola’s much-criticised 2024 Christmas ads had “Created by Real Magic AI” on screen for a few seconds.

They may have hoped the labelling would get them kudos for being earlier adopters of an emergent technology. Or they may have simply viewed it as necessary when the state of AI video generation in 2024 meant the finished ad was littered with AI oddities and had zero chance of ‘passing’. Regardless, the ad prompted a lot of negative reactions.

By late 2025, AI video generation had come on significantly in terms of both realism and consistency, enabling brands with deep enough pockets to commission ads where AI use wouldn’t be glaringly obvious and they’d have more of a choice about whether to disclose or not.

Mercedes Benz decided to follow Coke’s lead with a brief on-screen label in their late 2025 ‘Satellite of Love’ ad. Whilst the whole ad was AI-generated, the label appeared during a sequence which clearly couldn’t have been live action (a car is dancing with a satellite), subtly positioning the AI use as an alternative to traditional VFX.

Jeep Cherokee’s 2026 ‘Totally Wild Reviews!’ ad also couldn’t have been wholly live action (it features talking animals), although the AI use is very apparent from the quality of the video. Jeep still opted to disclose on screen, although it’s buried in the middle of this lighthearted (and only-possible-to-read-when-paused) disclaimer: ‘Fictionalization. No real animals were used in the creation of this message. Wild animals aren’t this nice. If you see a wild animal talking, call your doctor. The vehicle depicted throughout was created using AI technology and is for illustrative purposes. Production model features may vary’.

Talking of hard-to-read disclaimers, a Guess ad in a print edition of Vogue magazine drew ire last summer for featuring an AI-generated model.

In this instance, the presence and size of the disclosure feels very much secondary to the use of an AI model. I suspect the reaction would have been even worse if there had been no disclaimer, as the image is not realistic enough to pass and would have been quickly spotted and called out online. Generating photorealistic images or videos of people is one of the most sensitive of all AI uses, playing into a cocktail of fears: being misled, being replicated, being replaced.

TV & Film

As with advertising, an increasing amount of AI use in TV & Film is now going undisclosed and often unnoticed.

Where it is being signposted, it’s typically where full sequences have been AI-generated or where the voice or likeness of a real person has been altered.

The BBC’s ‘I’m An Alcoholic: Inside Recovery’ was one of the first productions to use AI to digitally anonymise a contributor way back in December 2022, although it didn’t reference AI in its on-screen label: ‘Faces have been digitally altered’.

Netflix’s ‘Dirty Pop: The Boy Band Scam’, released in 2024, used the same phrase (‘digitally altered’) to describe how they put words into the mouth of the late Lou Pearlman

.2025 BBC documentary ‘Death in Dubai’ used the label ‘AI-disguised face and voice’ to describe anonymising a contributor.

Netflix’s ‘The Investigation of Lucy Letby’, released in January of this year, used a similar technique but avoided any mention of AI, flagging contributors as ‘digitally disguised’ and ‘digitally anonymised’.

Sky HISTORY’s ‘Killer Kings’ was one of the first TV productions to use fully AI-generated sequences last year. It opted to disclose the use in voiceover, informing viewers a minute in that it would be “using groundbreaking AI imagery to revisit the world of Henry VIII”.

Towards the end of the year, the BBC’s ‘The Titanic Sinks Tonight’ took a different approach, adding a note at the top of the closing credits stating that ‘AI has been used responsibly to create a small number of visual effect shots in this series’.

Channel 5 opted for a full screen message at the start of ‘The Assassinations of Queen Victoria’, which aired this January: ‘The following programme contains AI generated scenes and imagery throughout, bringing some of Queen Victoria’s most intimate and dramatic moments to life in full colour’.

Meanwhile, YouTube has opted for ‘Altered or synthetic content’ as the label for “content that is meaningfully altered or synthetically generated when it seems realistic”, which creators are now required to disclose.

We’ve also seen films with inverse disclosures. The credits to Scott Becks and Bryan Woods’ 2024 psychological horror ‘Heretic’ included a note that “No Generative AI was used in the making of this film” (I’d wager that one of the crew ChatGPT’d something).

Journalism

Most ‘AI in journalism’ news stories have been about a lack of disclosure, although it’s becoming more common to see AI assistance acknowledged in bylines and alongside graphics and illustrations.

Journalism is also the sector with the most formal guidance. Most major outlets have published AI policies, although the frequency of embarrassing non-disclosures indicates there remains a gap between policy and practice.

Reuters has told its staff it “will make robust disclosures to our global audience about our use of these tools…where use of a particular AI tool is material to the result, we will be transparent”.

The Guardian has said it is “committed to signalling any significant use of GenAI in our journalism, whether for illustrative images or data analysis, with a footnote”.

Content agency founder and friend of the Sandwich, Joe Barnes, recently posted the below photo on LinkedIn showing an AI disclosure accompanying an image in The Week Junior.

Unlike Joe and the majority of commenters on his post, I’m not sure labelling was necessary or helpful in this instance. What is the reader meant to do with that information? Would it excuse any errors or oddities in the image? (I don’t believe it should).

Conclusions

Like so many aspects of AI, when and how to disclose AI use in media creation is not a simple or binary equation.

However, as I wrote in December 2023, I believe “The key question [isn’t] ‘was generative AI used in the creation of this content?’, but rather ‘could the use of generative AI in the creation of this content mislead viewers in a way which is problematic’?

I’d suggest using AI to photorealistically depict a real person born since the invention of film always merits clear disclosure (and consent), as it is inherently misleading.

By contrast, I don’t believe using AI to generate graphics or illustrations automatically demands disclosure. Any AI-generated or assisted output must be the responsibility of a person, who is responsible (and accountable) for its factual accuracy, its quality and its adherence to copyright law.

Whether to disclose the use of AI in editing real-world photos or videos feels a case-by-case judgement call based on whether the edits risk misleading in a way that is problematic. Removing the SMEG logo from the Bake Off fridges? No disclosure needed.

Content type and context are also key. News and current affairs must default to disclosure, where the consequences of even very minor edits can be catastrophic. Fiction comes with a different set of expectations and disclosure thresholds.

Using AI to generate photorealistic media of fictitious people feels like a context-specific disclosure decision, based on likely expectations and impact. Changing the face of a documentary contributor to protect their identity? Disclosure needed. Adding in some AI stormtroopers? I’d suggest not (this is how they used to fake it).

As for how to disclose, I’d suggest being specific about what has been changed and why, rather than a generic ‘Generated with AI’, which can trigger a bunch of incorrect assumptions around lack of human creative control, environmental impact and job displacement.

‘We used AI’ tells the audience almost nothing. ‘We digitally altered this contributor’s face to protect their identity’ tells them everything they need to know.

But I didn’t say anything about the process and - as is often the way with AI - how the sausage got made is potentially as interesting as the sausage itself.

I started with this prompt, following elements of prompting best practice by specifying a clear output, giving context and providing it with a reference to work from:

build a visually appealing web app inspired by this decision tree. the user is invited to paste in their draft LinkedIn post. you assess their draft and tell them whether they should post it or not. feel free to add your own criteria: the goal is to improve people's experience of LinkedIn, so they're shown more interesting and useful content and less derivative dross

It projected enthusiasm (“This is going to be fun”), then invoked its frontend design skill to try and meet my brief of a ‘visually appealing web app’.

Shortly after, an interface appeared inline.

Whilst ‘visually appealing’ might be a stretch, I was pretty happy with how it looked and worked. However, I wanted to be able to share it with other people, so I asked for it as an Artifact (Claude’s name for standalone content you can interact with and share).

It did what I asked but with a major caveat: it needs an Anthropic API key to work outside of the chat. It suggested I send people the HTML file directly, who would then need an API key to access it.

I responded that that didn’t seem very mainstream friendly and asked if there wasn’t a way Claude users could just open the Artifact and use their own tokens for any Claude usage. It suggested I share a link to the conversation instead.

That would have been a terrible user experience, so I challenged on Artifacts not being sharable. It told me I was right, apologised and set about creating a “proper Artifact…no setup, no API keys, no faff”.

Did it work? No, it did not. So I suggested it might try and different approach (and cheekily threatened to try Gemini).

Did that work? Also no.

How about that? Still no.

At this point, Claude admitted it didn’t know how to get a shareable Artifact working, apologised, and gave its blessing to me trying Gemini.

However, I was pretty sure I’d encounter similar issues with Gemini so decided to stick with Claude and pursue the standalone version securely referencing my Anthropic API key.

Claude then talked me through what I needed to do to get it working.

After a bit of back and forth, we got it working.

Claude then opined that “this whole saga is a great Substack post…Your audience would love it”.

I was tickled by its mention of “weeks of fighting with Claude artifacts”. Whilst it might have felt like that (seemingly to both of us), it was actually only a day.

However exploring the various dead ends did result in a lot of prompts. I asked Claude to tot them up.

Reflections

I’ve used AI to vibe code lots of web apps. However, most of them haven’t needed to dynamically query an AI assistant to function.

That added dimension, paired with my desire to make it easily accessible to other people, turned this from a one-shot (single prompt), 5-minute exercise into a day of frustrating back-and-forth (and at times, round and round).

What made it so protracted wasn’t technical complexity (the final solution was pretty simple). It was Claude’s lack of grounded knowledge about what it could and couldn’t do, its misplaced confidence in the different approaches it tried and its inability to test the artifacts it was creating.

Right now, vibe coding without any coding knowledge can be a frustrating, trial-and-error affair. You get 95% of the way there with your first prompt and then spend hours trying to add the final 5% without breaking the whole thing.

It’s still incredible to me that AI enables me (a non-coder) to take an idea I had whilst falling asleep one night to a functioning web app the following day and I would highly recommend having a crack at vibe coding some simple tools yourself if you haven’t already done so. Just be prepared for some frustrating back-and-forth and borderline gaslighting along the way.

Related posts:

This was partly because of its professional utility, but also because people abandoning Twitter/X were increasingly posting interesting things there.

However, over the past year, I’ve perceived a drop in the quality of posts and comments. Whilst I suspect some of this is attributable to changes to LinkedIn’s algorithm to maximise revenue for Microsoft, a more visible factor is the increasing preponderance of text drafted with little or no human thought or quality control.

A common response to this is to conflate low value posts and comments with AI use and to blame AI, as though it’s an autonomous and malevolent force. Of course, all of this content ultimately comes from people choosing to use AI this way.

A couple of years ago I posted this ‘Should I post this on LinkedIn?’ decision tree:

Since then, increasingly capable and widely-adopted AI assistants have dramatically lowered the barrier to drafting at least superficially cogent LinkedIn posts and comments. They have also increased the temptation to not properly think through the value equation of whether a post or comment is actually adding anything.

As my decision tree suggests, adding something can take lots of different forms. It could be helping you get new work, amplifying something or someone worthy of amplification, or simply being interesting and/or entertaining.

I’d suggest lazy use of AI (not AI use per se) is leading to more content on LinkedIn that’s not adding value. The platform owner clearly carries some responsibility in managing this, balancing its algorithms to surface more value-adding content, not just racing to the bottom in pursuit of engagement (see Twitter under Elon Musk’s stewardship).

However, I also think we have a responsibility to think more before we post/comment and to be judicious in our use of AI to avoid inadvertently contributing to the blandification of LinkedIn.

To that end, I asked Claude to use my decision tree as inspiration for a LinkedIn draft checker, which reviews your draft post and delivers a verdict on whether to post or not. You can try it out here.

I tried it out with a few of my recent posts, which scored reassuringly high on ‘originality’ and ‘usefulness’ and mercifully low on ‘cringe’ (the three scoring criteria Claude settled on).

To check it would be sufficiently critical with low-value posts, I fed it a couple of the ‘Suggested’ posts the LinkedIn algorithm had decided to insert into my feed recently.

It was diplomatically damning in its assessment.

I also tried it with a questionable post from someone in my network (original post cropped to spare their blushes).

Finally, I got Claude to draft some archetypal zero-value posts to check it would be sufficiently merciless when required.

As regular readers will know, I’m not opposed to the use of AI assistants as a writing aid, but I do think careful thought is required about when and how to involve them. For me, doing an AI-free first draft is the best way of crystallising my thinking, before inviting AI to critique what I’ve written (see earlier post on this topic).

I’d love to see LinkedIn replace its ‘Rewrite with AI’ button with a ‘Review with AI’ button which delivers this sort of critique. Who’s with me?

Related post:

Most recently, it’s been ByteDance’s latest video model, Seedance 2.0, prompting the proclamations. It can generate longer (up to 15-second) clips with more plausible motion and character consistency than most previous AI video models could manage. On its initial release, theoretically limited to China, it also had a distinct lack of guardrails limiting the rendering of real people and established IP.

A fight sequence depicting Brad Pitt and Tom Cruise trading blows on a rooftop, apparently generated from a single Seedance 2.0 text prompt, proved a particular lightning rod, both for ‘Hollywood is cooked’ proclamations and for Hollywood studios’ concerns about real actors likenesses being used in this way (and what it makes very apparent about Seedance’s training data).

Whilst the studios’ concern about unlicensed training data and IP-infringing generations is justified, the ‘cooked’ brigade are labouring under a number of false assumptions:

1.) Creating a movie is as simple as stringing together a bunch of clips

Whilst online video editors such as CapCut (also owned by ByteDance) have made it technically easier for non-professionals to string video clips together, creating a vaguely coherent narrative (even a bad one) requires a set of skills that most of us don’t have.

A narrative arc with varied pacing, jeopardy and resolution and characters you give a shit about is not an easy thing to achieve.

And some things which are challenging enough in traditional film & TV production, become almost impossibly so when using today’s AI video tools (e.g. scene continuity, lip-syncing).

2.) The best AI-generated/assisted films doing the rounds online could have been made by anyone with access to the right tools

A lot of the longer viral AI films doing the rounds are accompanied by claims about how little time and money they took to produce.

Regardless of veracity, these claims omit a third critical ingredient in creating compelling video narratives: the role of talent and craft.

It is no coincidence that the majority of the most compelling AI-generated/assisted videos have been made by people with extensive experience in visual storytelling.

That’s not to say that formal training is a prerequisite for narrative filmmaking. But even folk with a natural aptitude for it need some time behind the wheel (camera/AI video suite) to hone their craft.

In addition to an understanding of the grammar of visual storytelling, those making coherent narrative films using AI in March 2026 have had to learn to wrangle a variety of unwieldy and unpredictable tools.

Like the famous Picasso napkin story, foregrounding the number of hours it took to produce an AI-generated/assisted film risks giving a false impression of what it takes.

I’ve previously highlighted Hashem Al-Ghaili’s 15-minute film Kira. His latest AI-assisted films - both animations - showcase a filmmaker with a deep understanding of visual storytelling and a mastery of the current generation of imperfect AI video tools. How quickly he produced them (48-hours in the case of Punch Monkey) is impressive, but says more about his abilities than how easy it is to use today’s AI tools to create an absorbing and affecting narrative.

3.) Audiences don’t care if it’s real Brad Pitt or Tom Cruise

I think we do.

Whilst I have no doubt that more of us will be forming attachments with AI-powered personas in various different contexts, I think our appetite for parasocial relationships with real people will remain undimmed.

We will still want to see Taylor Swift singing live and Tom Cruise riding his mobility scooter into a ravine.

4.) The studios won’t go to war on unlicensed training data and likenesses/IP

They are and will continue to do so and it will force the AI companies to work with the studios to implement guardrails around what tool users can and can’t do with performers’ likenesses and studios’ IP.

It will be harder for the studios to control what locally hosted, open-source models generate, but I expect the auto-detection and removal of infringing content posted to public platforms to improve in much the same way as Google’s Content ID has for direct copies of copyright-protected material.

5.) The studios won’t embrace generative AI

“Studios are lying about how much they're using [AI]” was industry insider Janice Min’s recent assessment.

Regardless of current levels of adoption, I will be astonished if Hollywood doesn’t increasingly use generative AI where it clears their quality bar and doesn’t alienate the human talent upon which they rely.

So, what does it mean?

It’s not as engagement-friendly as ‘Hollywood is cooked’, but I think the real takeaway from Seedance 2.0 is ‘Generative AI is becoming a more credible and affordable option for VFX, music videos and animation (which Hollywood will ultimately embrace)’.

People’s appetite for AI-generated ‘live action’ films is a genuine unknown at this point. There’s little doubt the verisimilitude will continue to increase and widespread application in video games feels inevitable, but I’ve got a strong hunch we’ll continue to want to see flesh and blood actors interacting on cameras for our entertainment (see cinema not killing theatre).

100% AI-generated films will no doubt continue to proliferate online, with lots of dross and a small proportion of more carefully constructed narratives, where the use of AI is secondary to an engrossing story. As with all media, the most derivative studio output will be most vulnerable to being substituted. Might this foster a shift towards more originality in Hollywood? (*doesn’t hold breath*)

When it comes to agentic AI (AI that can take action on your behalf), a lot also involve a level of risk I wouldn’t be prepared to accept. Consequently, I didn’t install OpenClaw (a decision the director of alignment at Meta Superintelligence Labs may wish she’d also made).

However, I did feel that Anthropic’s AI agent, Claude Cowork (currently in research preview for paid Claude subscribers), offered sufficient guardrails for me to install it and have a play.

I wrote a couple of weeks ago about using Claude Cowork to create an Ofcom results dashboard.

Another of Claude Cowork’s capabilities is to read, edit and create files in folders you give it permission to access on your computer.

I take a lot of screenshots and spend an inordinate amount of time searching through them in search of a specific screenshot, with only my memory of approximately when I took it to go on (my Mac saves screenshots with the date and time in the filename).

Aware that the latest generation of multimodal AI models have impressive computer vision capabilities, I decided to copy my screenshots into the dedicated folder I’ve given Claude Cowork access to and ask it to view the screenshots and rename them with a filename which describes what the screenshot is of.

Having noted that it was a big job and confirmed that I definitely wanted it to process all 4,085 screenshots, it set about dividing them into batches and processing them in parallel, whilst I got on with other work.

When I checked back in, it had made good progress, but had hit the 5-hour token usage limit on my Claude Pro account.

I set a reminder to check back in at 5pm and ask it to resume.

After hitting my rate limit for a second time and asking it to resume, it finally reported that all screenshots had been successfully renamed with a descriptive filename.

It’s only been a few days since Claude sorted my screenshots out, but it’s already proved a meaningful time saver in quickly locating specific screenshots through search/scanning the descriptive filenames.

I asked Claude to examine its log files to estimate how long it had spent actively working on the task. It estimated ~2 hours of active processing time.

I then asked it to estimate how long it would take a human to complete the same task.

The assumptions informing Claude’s 40-50 hour estimate strike me as reasonable and I appreciated its consideration of humans needing to take breaks and the challenge of thinking up a non-duplicative name for the 50th BBC iPlayer screenshot (it’s fortunate AI models don’t have intent or I would definitely have inferred some sarcasm there).

Of course, I would never have done this manually. Rather I would have carried on trawling through the folder of unhelpfully-named screenshots in Finder’s gallery view, hoping to identify the screenshot I was looking for.

It’s a low-stakes use of an AI agent, but it’s indicative of the sort of tasks I’d suggest it’s worth considering turning to an AI agent for help with in early 2026. Here’s my short checklist of statements you should be able to answer in the affirmative before you enlist an AI agent to tackle a task:

1. You can envisage a material/proportional benefit in terms of efficiency and/or quality

Whilst every experiment won’t be a success, I’d suggest we should be looking to direct AI agents at tasks which have a hope of solving a real problem (I’ve written before about my concern about the environmental impact of carelessly deployed agents).

2. You’ve validated it’s going to work before committing a large amount of time/tokens

I tried it with 3 screenshots in the first instance.

3. It’s not the end of the world if it gets something wrong (OR you have a robust and scalable mechanism for identifying and correcting errors)

An incorrectly named screenshot of an AI tool is obviously no biggie - an incorrectly labeled MRI scan in a hospital is clearly a different matter.

4. It can’t compromise your data

I only gave Claude Cowork access to a single, dedicated folder and a copy of the screenshots. We are going to hear many more cautionary tales over the coming year as people give AI agents access to their data without sufficient safeguards.

Renaming screenshots may not be sexy, but it’s an application of AI agents I’m now going to appreciate on a daily basis.