At breakfast one day, we talked about subcultures and a few of us millennials told younger folks how information used to travel before texts and mass location-based internet. Within the distinct tribe you were in, you’d pass on tips and lists of where to go, whether because it was cool or because it was safe, and search the city looking for this one magazine, and sometimes in the magazine you’d have a scene report.

This is one of those.

The sheer allure of Sylvan, Haley and Tobias saying: we’ve bought an ex-nunnery on Lake Lucerne. In July, we’ll host a month for guests coming from all of the world + many different fields, technology to philosophy to art to business, to co-design what the space will become. We think it will be a place for R&D. We think theory, prototypes and building are all parts of the same thing. You should come.

Obviously you are going to say yes.

When I got there, after the tube and flight and tram and train and boat, this was the view:

Nothing is more wholesome than thinking about / talking about technology while looking at a lake.

This is going to be a photo heavy post because so much of what I took from Stella was how much place influences you.

I love how London is explicitly a rat race that pushes me to a certain mode, so useful most of the time, of competition and grind.

Space, air, and beauty inside and out encourages you to be more open to new ideas, new people and rest. And look at all this beauty.

Partly, I wrote this post because I found so little online about this special place. It deserves documentation. As the welcome letter says, the buildings at Stella (main building, dorm, an enormous gym, chapel, lake house) were stewarded by the Baldegger Sisters and used for decades to educate.

Open every single door and cupboard, Haley told me at checkin, and really I did my best. For someone used to living in a crowded capital city, who used to run a large startup space that nonetheless got overrun (team standup on the stairs again? sure) the amount of space was extreme, and extremely beautiful to me.

Perfect buttercup yellow curtains in every room; every chair was good.

Check out the chic little crosses.

The library; as new guests arrived they’d leave their own books here.

I opened a basement door and it was concrete, set almost looking like dirt, up to the ceiling. This absolutely huge first floor gymnasium. A narrow phone booth with floor to ceiling shelves, too little space left for all by the most slim caller.

Then there’s the chapel, designed by Heinrich auf der Maur in 1965.

Darren playing the chapel’s lovely organ.

From early morning swims to hazy orange sunsets, you get the picture; it was cute.

The deal was; you will turn up.

There will be other guests, coming and going, staying for as long as they want to or can.

Some will be friends of yours.

Some will be friends you don’t see often enough; here’s Scott and I holding our tiny stars in the next gen tech philanthropy - ie where does frontier lab money go? - session led by Aish at Analogue.

Some will be people whose worked you’ve followed for years, coming from Berlin to San Francisco to Singapore; some will be people who’ve followed your work for years; artists, game designers, funders, academics, AI researchers, founders, culture writers and more.

Every day they’ll be a shifting agenda of presentations, readings, group chats, yoga, live action role plays like Anthropic vs the markets vs the government, radio shows that reflect on the day. You’d wake up and check the pins.

Simon Denny on American Dynamism, a16z and the aesthetics behind defence technology (his pod on the same topic linked here); the image says the grind that’ll win.

Not every session worked for me, honestly, including my own, and that’s OK.

I almost preferred the looser one-to-one chats or the meandering conversations that happened over group breakfast, lunches, swims, dinners. You’d turn a corner, follow the sound and find Adnan manning the radio.

I think that looseness only worked because of the thought and care that went into everything from the group invited to the delicious food made by Ada’s.

Like Andy Warhol said of Studio 54, you want “a dictatorship on the door and a democracy on the dance floor”. So many want to be dancers; not enough people are comfortable being benevolent dictators.

Lukáš and I made many tiny martinis one night, here for Grace and Wills:

Flora and Adnan ran a party in a Backroomsy basement made better by orange light.

More retreats should bring together such different profiles and put them in a specific physical space (ideally somewhere you can’t leave easily) to build that common language and get planning - but then most “retreats” aren’t Stella.

Excited to see what this becomes.

Finally, the meme-worthy (this cap is now my favourite Whatsapp sticker).

Thanks for indulging me,

Sarah

By “worlds”, which can sound vague, I mean internal narrative + ritual (strategy + direction, culture, hiring), external (product craft + story) and ecosystem (people building on your product plus the lovers + haters that shape how the world meets you).

I’ve always loved thick, textured companies, the ones that accumulate culture, language, network, feel until they seem like a physical place. You can tell it in the products they put out; in the hires you meet; in how years later we talk about them.

And typically certain teams just lean that way. Last night I got a late photo of a team I work with drinking a celebratory tomato juice at midnight after fixing something critical. It made me so happy; this startup is on the biggest tear right now, loving it, and aware they need to document the highs and lows. Future biographers will want them to.

Anyhow! Someone mentioned this worldbuildy blog a few weeks ago - every time I get an email about it, it’s a cool interesting one, pls always email folks whose work you like, I do and get 99% response rate - and I realised I’d been thinking about it in the back of my head because of Backrooms.

If you haven’t seen this film, and you should, a gentle sad eyed therapist, plagued with bad childhood memories, counsels a man struggling through anger issues and divorce.

After leaving the family home, the man sleeps in the showroom bed in the pirate themed furniture shop he manages.

On night, chasing flickering lights, unexpected noises and the source of sky high electricity bills, he finds a portal into the most liminal spaces; an endless series of backrooms, strip lit, with unexpected shapes, sizes, dimensions, and creatures.

Like every truly original film, there are endless interpretations but the story it told me was; can we ever really find our way out? There’s so much comfort to living in your own version of memory over the here + now.

And the visuals! Scuzzy around the edges, fuzzy like a dream, glitching like a computer. It feels so fresh.

A lot of good substacks have gone deep on the film’s backstory. Made by Kane Parsons (20), it’s one of A24’s biggest ever hits. The sheer scale of box office success speaks to both how starved we are for the creativity and that internet culture, as in shared ideas we live in, is far more influential than mainstream media would have us believe.

Everything’s computer.

Parsons and the film he made emerged from a very particular world of early 2020s era liminal space imagery such as abandoned mall tours and creepypasta which I always think of as the modern version of sharing gothic ghost stories around the fire. Parsons uploaded his own short series - the first has 90M views! - got found and made the film.

For discovery and IP, the implications are fascinating. Every artefact - from your startup to your house to the music you love - builds on what comes before but we’re entering a new era of forkable IP. Great stories, rich with influence, generating discussion on the story, sure, but also many more stories in their wake, each worked and reworked by the depth of fan fascination and uniqueness of imagination. 6M stories on Ellipsus in last year alone, many from fan fiction writers building universes within recognisable IP. In the old world, big IP holders would be pulling out their lawsuits but between quirky brand mashups (remember KFC x Crocs?) and the pull of younger audiences seeking novelty and participation (wholesome!) it’s all change.

Backrooms is what happens when internet native worlds get so dense they spill into mainstream culture, much like how great startups feel before the revenue graph catches up.

So back to startups.

This is what’s been in my head.

In that last post, I used Stripe as an example of companies building worlds:

“The story that anchors the world; Stripe exists to increase the GDP of the internet.

What Stripe actually builds and sells is a set of financial infrastructure products which naturally enable their customers to sell more….therefore increasing the GDP of the internet. But having a slightly grander worldview than ‘we sell financial infrastructure products” ensured the company attracted a distinct group of employees who then created a distinct set of cultural norms. When I talk about cultural norms I mean how groups agree to work together; from company or industry specific slang to stylistic design choices to how decisions are made to communication styles. These become powerful reinforcing mechanisms that build alignment and enable groups to move forward effectively”

But when I think about Stripe, even as a user of their products, do I think about the products first?

Or do I think about this ideaspace around them, spanning from the Collison installation to the progress studies world of Cheeky Pint podcasts, all the directions their ex-employees go that somehow have narrative coherence and this all-time killer tweet?

The scope of the cofounders influence is so much broader than one company.

But what sat behind the founders? YC and its world, its norms, its network.

Going back to this diagram, I’m increasingly interested in worlds / ideaspaces where some kind of anchoring belief or idea brings together people, attracts capital, creates and accelerates ideas and spins out influence, founders, companies.

Because, if we look beyond the noble idea of learning for learning’s sake, this is what universities used to do.

Say you’re Oxford. The idea of Oxford has travelled very far through time and place. It stands for quality education and prestige. You join because you will have this idea of prestige attached to you making you credible and legible for those that need to put you in a job-shaped box from your parents to future bosses. You’ll meet interesting fellow prestige builders and seekers.

For Oxford, each year brings a fresh wave of emissaries graduating into the world carrying and reinforcing the story of your prestige through their success.

This is the ultimate soft power.

But now, as legibility evolves with the internet’s wider distribution where certain kinds of credibility can be built faster and built better by + for yourself, the ideaspace of universities as generative distinct physical locations - the transaction of what they bring in an era when students become customers - is less clear for some.

I’m not per se picking on universities although I do think their genteel pace and style feels suddenly poorly fit for the modern world.

More that a young person now, in search of an idea space, has so many options.

I argued before that residencies and accelerators are taking over traditional pre-seed funding in San Francisco (in Europe I think it’s a blend of these plus the new wave of tiny funds like mine). Some great examples;

The Helsinki hackers at FR8. Anchoring belief: “environment shapes destiny and ambition is a moral duty”

El Segundo’s gung ho nationalists: “defence should be a priority”

Leverage Research, early effective altruism and the many post-EA movements: “if we understand humans, we understand the world. We can better engineer ourselves + others”.

Articulating beliefs, and behaving in a way that was true to them, attracts talent and capital, developing beliefs in further articulations in the forms of splinter groups, influence via emissaries and of course companies.

You can look at an entire wave of influential people and companies pursing safe AGI and trace it back to 2000s cybernetics, 2010s rationalism and EA.

These are pockets of people you want to be in orbit with building in high-trust idea spaces. Oxford exports prestige through degrees; these build their own prestige and export belief and practise through people. The narrative gravity of what these people and their companies communicate creates further impact + influence and attracts further capital.

(I continue to think some of my favourite tech companies are compelling founders wrapped in software. People follow people)

This is new soft power.

If you look at your peer group and see screenshots, rituals, in-jokes or similar work choices that feel like alumni paths, maybe you already have an ideaspace or world forming.

Could you think through what the core beliefs of the group are, besides the specific thing you personally are building?

How would you want someone else to communicate what your peer group are doing and why?

What are the things you could make, from writing to software to playlists, that would make someone say “hell yeah” and join you?

Worlds don’t arrive fully formed; they snowball.

Yours might already be rolling down the hill.

Summer 2026 so far is powered by homemade coffee paletas, world cup wall charts, long jorts 24/7, and Ben Lerner’s Transcription (more autofiction by boys, please).

Did you read a book recently you loved?

Seeking summer reads, fiction + non-fiction; reply with your favourites!

I’ll share the whole list.

This video began playing and I swear the room gasped.

Last week this funny anon Twitter account slid into my DMs and sent me the Luma for Midjourney’s hardware launch.

I could tell you Midjourney are an AI image generator, one of the earliest, but reading that in 2026 is to bunch them with a cohort of less interesting companies.

Their Discord in early 2022 was one of my favourite places online. This circular library is an early relic, maybe the earliest thing I made (can’t think why else I’d keep something so emergent) in that time when prompt to image made us gasp.

Four years later, I turned up next to IKEA on Market Street in San Francisco; not the most promising start to an extraordinary evening. But then:

Bouncers asked us to turn phones off and…..people did. The room felt thick with attention.

You walk into a dark and beautiful member’s club, with high ceilings, comfortable sofas and a grand piano playing. We’re drinking slim glasses of wine.

There was this portentous wait. You start talking to other guests, chip designers and artists, doctors and journalists. Bryan Johnson is there. You realise they have somehow curated the group because, in the 150 or so of us present, there’s coherence in what we care about.

We go upstairs and David, Midjourney’s founder, is sat in front of a circle of chairs. Some fill the chairs, some sit on the floor at his feet. He begins to share his story, humble and excited, just an enthusiast, just a disciple.

He tells us his dad was a dentist who ran his business from a sailboat.

He tells us about looking up from that boat into the night sky which made him fall in love with physics.

He tells us that led him to building his first company, Leap Motion, and with it all the joy and pain of being a little early in the dream of controlling computers with hand movements.

He tells us about Midjourney being a community supported R&D lab making hundreds of millions in revenue each year and how that gave them as a team the space to consider what else they wanted to build [he doesn’t mention the Meta software license deal for image generation that probably helps with free cash flow].

I don’t have the exact words but he tells us something very like what is on their website:

“Over the coming months, we’re unveiling a wide range of ambitious projects under the themes of imagination, coordination, reflection, beauty, and human flourishing.

We hope our work will help tell stories of a humane future that we all want to be a part of, and convince you that we aren’t at the end of time, or the beginning, but that we are all midjourney in a vast and great adventure”

Then he tells us one vast great unknown right now is our own bodies.

Increasingly we speak to artificial intelligence about our health but we’re acting with imperfect data.

To be able to make better choices, to find experts, identify issues, to compare our general health with the wider population, and live healthier and longer, we need a much higher volume of data, captured more quickly and cheaper than we currently can.

MRI scans are either a luxury for those than can afford proactive or preventative healthcare, or medically mandated at the last minute. They’re cold. Kind of uncomfortable.

Midjourney’s scanner feels more like a treat.

Your body, descending into water, passes through rings of underwater sensors using echolocation to send sound waves through from every angle. Enough waves and enough angles form a dynamic image of what’s happening inside your body in under a minute.

Part of the beauty of what David tells us is in the detail of the device.

“Each square creates ultrasonic waves and records the ripples back at millions of times per second. Together they act as both a choir and an audience - producing terabytes of data each second. If we converted that data into HD internet video you’d need to watch 500 hours of footage for every 1 second of scan data.

Each slice continuously crossfades between the raw reconstruction and its AI segmentation; what the scan lets us identify inside the body.

The sheer number of mechanical elements, the inconceivable volume of data, and the computational power required for this to all come together is one reason why no such machine was ever made - until now”

He tells us that after the talk we can go next door and try it for ourselves.

With a flourish, the curtain to the back room drops to show a series of the machines, lined up.

This is not a future promise but something that’s been built.

And this isn’t a scaled model because, as David says, “we chose to build the big thing first”.

David says they want to have 50,000 scanners available worldwide by 2031, a billion scans a month, and that with enough early imaging, the world could avoid 50% of all healthcare costs and maybe 30% of all early deaths.

Finally, he says the warm water guided them to imagine an R&D environment that felt cosy and luxurious, somewhere you’d go with or without the scanner, not cold and clinical.

Midjourney open their first spa in 2027, with hot tubs, saunas, cold plunges and scanners, and I will see you there.

Honestly, the entire evening felt like a religious experience.

The audaciousness of something totally unexpected, with real utility, delivered with humility and style.

There was this divine quality to the story I heard, how focused we were, the way I left wanting to tell every single person.

The day before I’d been at Y Combinator’s Demo day. Such a brutal format; 174 companies with one minute each for them to tell the hot, restless audience their life’s work. The guy sat next to me scrolled Polymarket and Raya. The guy the other side vibe coded versions of each product pitching in real time. The sun beat down. The pitches rolled on and off.

I felt for the two-thirds of founders presenting who still needed to fundraise.

And the thing I left with in my head were Polymarket and Raya.

Prediction markets; high-end gated dating.

South Korea’s latest viral app, FoodNeverComes, does what the name says; you get the dopamine of smashing the buttons, choosing and making the order but not the food itself.

Everybody worships but we direct our devotion differently.

The story that my ear is most ready to hear right now in startups is what I might call a hail mary.

A really big story of something audacious that benefits a lot of people.

Like:

A floating solution to our ongoing energy needs.

Async inference on European GPUs.

Better automation + alignment tools for governments.

(these are things I got excited about recently)

If building something divine is the result of a group pointing in the right direction with ritualised care and craft in service of an idea to have faith in - and Midjourney Medical made me think it is - then perhaps the way to start is by being a heretic. Question what’s obvious. Spend unreasonable amounts of time on what’s not obvious yet.

Focus on what you, as the founder, personally want to see change in the world and why.

It makes complete sense the freshest clearest take on the tech we want comes not from frontier labs or governments but the Pope. He’s got to stop hashtagging AI on his tweets (boomer energy) but otherwise the message is; this moment forks the road. Are we building a tower of babel with all the pride, confusion and centralisation that involves, or something a little more exciting where tech is genuinely serving us?

This is what abundance feels like to me; bets on the future expressed through the idea you choose to work on, product, business model, org chart and culture.

This is all one choice; it can be divine.

Things that felt worship-worthy this week:

• Fablepool. Like an agent-first kickstarter, pooled cash behind collective prompts. Useful for those of us with things to build where you want to share the token costs. Cool!

• my heatwave best friend in building focus; this droney 528hz playlist

• software engineer wins “religious exemption” from using AI at work in favour of hand crocheting her code

• I used to be on the board of New Public, who work to research + prototype healthy digital public spaces. Eli wrote this great piece on what he said at the Vatican

• struggle with the volume of emails / tweets / essays that are now slop? Ongoing theory that there will be a rebel alliance of huge companies whose products and audiences just resist AI. Ellipsus is a collaborative writing platform (disclaimer; investor) and launched Emboss this week. Human proof-of-work!

• on wild conviction: I did a podcast with Odin on alignment, emotional capital and how I think about venture capital in 2026

• more real world rewards for nerds; the cafe in Shenzen’s biotech hub that gives you free coffee if you’ve published research in Nature

• A mega funding pot to eradicate respiratory viruses for good funded by Stripe, Anthropic etc. Imagine that domino meme that begins with OpenAI Sam + Dario board disagreements and ends with no more common colds

• Finally; I live right by these big wild marshes in East London and went to see the sun come up on solstice along with a slice of Londoners who live at 528hz. The kind of people who chant, wear loose trousers and bring herbal tea to share everywhere. I love them. Walking home in the hot 6am air was perfect.

Til next time,

Sarah

Raise your hand, be earnest and enthusiastic, know you know nothing, follow your gut, ask for what you want.

Hold things lightly and hold your head high when something doesn’t work.

V good image by Ash Lamb

Amazed I have to keep learning this, having been mid to high cringe all my life, but here we are. The latest is Londonmaxxing.

For those who aren’t terminally online, this started on twitter as a meme phrase. And, yeah, my cringe reflex twanged when I first saw it.

The UK’s one of the only countries in the world where being patriotic is often taken as an anti-signal about a person; down to our mixed feelings about the England flag and what it can symbolise.

But shame, even ambiguity, can keep you stuck in what you think is your place.

No thanks.

This matters right now when it feels like so much of the future - what we build, how we want to exist together, how we communicate to the world what and who are - is up for grabs in the most exciting way.

And in the UK we have this legendary asset in London.

As Gritcult says, “London’s antifragile; it’s lindy”.

Geographic reality becomes destiny.

Islands are forced to build up their trade hubs and resilience.

The history of trade led to English as the world’s predominant language both for humans and most recently for computers. Shared interfaces speed things up.

Trading ports import and export goods, people and ideas.

For hundreds of years, London’s been a magnet for a constant flow of skills, cultures and languages. My son is British, Irish, Russian, German, Dutch, Portuguese, American, Welsh, Scottish and not a little confused at just how much heritage he has.

It’s simpler just to call him a Londoner.

And the consequences of this superdiversity - a term that got invented for London - is an unusual level of resilience.

If you think about the US; historically New York leads on finance, fashion + media, DC on policy, San Francisco on tech and LA on film.

London has all this, and more, in a hub and spoke city where we tend to live out in neighbourhoods and commute into a dense and varied centre.

The diversity of industries means when a sector wobbles or even dies, something pops up permissionlessly to take its place.

The invention of modern composable shipping containers killed London’s shipyards and the Thames as a trade route (too narrow); no problem, we built Canary Wharf in its place to become of the world’s financial hubs (It was partly built using the proceeds from a legendary gold heist because we produce world class criminals too).

When Bankside Power station shut up shop, we rebuilt it into one of the city’s best museums (all with free options; it’s confusing to me when I have to pay for museums elsewhere) Tate Modern.

My father-in-law used to play in this massive crater formed by a bomb in the Blitz; the hole it left in the City was big enough for someone to build an entire future city in my personal favourite London spot, the Barbican.

London’s timezones mean it’s one of the only places in the world you can do truly global work (one of the many excuses I used not to move in the era I managed teams in Tokyo, Sydney and San Francisco simultaneously).

And one of the most underestimated London global soft power strengths, magnetically attractive to small town kids like me since time began, is culture from skateboarders on the South Bank to fashion kids trotting out of Central St Martin’s. In the last two weeks alone, I went to:

• a preview of the V&A East museum + exhibit on Black British music

• an extremely glamorous ball in Chelsea’s the Saatchi Gallery

• an enjoyable but baffling film at a 37-seat grindhouse cinema (more below)

• States of Exception at Somerset House on arts/tech/social systems or “one of your sovereign protocol governance things” as Stuart would say

• took my son for his regular swimming lesson in an ex Olympic pool

• plus Flower AI’s decentralised AI summit and an agent demo night because London’s tech events are world class, too

Undeniably, there’s something electric in the air with the local tech scene right now.

But beyond the visible tweetable public successes of huge raises, exits and office expansions is the infrastructure coming underneath; a core part in a large fragmented city is places to put people together.

Charlie at Ramen Club has done this brilliantly for years and was one of the organisers of the very first Londonmaxxing event I spoke at a few weeks back.

I was late, Lime biking through the City on a very hot Thursday evening through the throngs of boozers and what I felt when I walked into the rammed space was a wall of energy. Londonmaxxing is something to rally around; a meme to call out everything good and exciting about this city, a batsignal to find each other, a commitment to get on and fix stuff that isn’t working.

Because nothing is more “you can just do things” than launching a new city, I was panelling with Joe who is doing just that with Forest City (I love) and Josh from Plugged Founders.

So hard to sum up one of the most optimistic exciting evenings I’ve had for ages so I’ll just drop these (I can tell I care about the topic bc of the cringey levels of hand movement).

One of my most strongly held theories is that San Francisco’s tech scene is a collective wealth project; a small dense scene where capital and talent flows as people invest in their friends, give customer contracts to friends and buy friends companies. The capital then gets reinvested and the merry-go-round starts again.

We need more of this in Europe. No company succeeds alone; its success puts money in the pockets of founder, early team, investors and through taxes, to better streets, hospitals and education systems. I’m into it.

Important things you need to know;

a) the photographer was wearing these amazing Bottega Veneta boots with bright green soles that I have been chasing around secondhand sites. It’s not a competition because many scenes can and do thrive, but even the photographers at tech events in London are better dressed than 99% of West Coasters.

b) as a thank you, the team gave me this iconic book of British crisp packet designs after my strong enthusiasm for the blind crisp tasting. Excellent design inspiration and a reminder of one final underrated London asset; humour.

The next morning I woke up to a new group chat with hundreds of members.

I’m excited to see what Londonmaxxing does in the next phase of the scene but, honestly, the main beneficiary of all this cringey enthusiasm is me; it’s not really about London, or even Europe, but about the sheer energetic choice and fun of collectively building a better future.

Last month has been a TIME (in a good way). I started writing this newsletter sitting in an Athens bookshop / cafe called Free Thinking Zone in March. Now it’s May. Shruggy emoji. Here’s everything I learnt and loved since then:

• The Nickel, a grindhouse cinema in Clerkenwell that shows the most lovingly curated selection of “the most subversive, fearless, sublime, psychedelic cinematic treasures” = such good vibes. 37 seats, always booked out, staff straight out of central casting; when I walked in, the bored’n’beehived receptionist had her platform boots up on the desk and was popping gum. I saw 1967’s Privilege about a manufactured pop star used as a tool of political & religious control, shot as a documentary. The lead role’s played by Manfred Man’s singer Paul Jones then at the height of his fame; it’s Taylor Swift starring in an Adam Curtis take on Trump’s America.

• god tier nerdsnipe marketing from p0 with the Museum of the Human Web. As we shift to human-machine collabs for the next web, this IRL pop-up in San Francisco and online gallery celebrates pure human creation. Enter online to win one of these - all donations benefit the Internet Archive + Computer History Museum

• Now I’m a few investments deep into agent infrastructure I’ve been on a deep dive into brand + discovery for headless products. It’s just so very early; frontends exist only for trust + show don’t tell. This is the most modern resource piece I found so far; reply if you’ve got examples/inspiration for me!

• Dazzled by this. Watch it all the way through with the sound down (song is so-so)

• Ellipsus asked 5,000 people how they felt about AI in the context of writing + reading. Love to be a daily user + builder of tools to automate boring things but automating non-task related emails? Entire posts? This survey made me realise we seek out human-made stuff because it’s a conversation; I feel this too. Let’s talk

• but also; so dumb. I love it

• wrote a viral tweet about small sandwiches. Cringe

• and, finally, a story from Sari about how life is long and sometimes our only job is to listen

  The Sublime

  The story of a sound

  If you’ve used the Sublime browser extension, you’ve probably noticed the sound that plays when you save something…

  Read more

  5 months ago · 78 likes · 7 comments · Alex Dobrenko` and Sublime

  Til next time,

  Sarah

Projected onto a tall stone column is the blank face of a young Korean woman.

Underneath her image are forty nine lily pads, made of concrete, hovering in a shallow pool.

Sat on each lily pad is a fortune teller, from shamans with their bells and fans and robes to tarot card readers shuffling decks. The room is shaking and moaning as the tellers compete to crack a truthful reading of how the woman died.

This is Battle of Fates* and it’s as nuts as it sounds.

But the most nuts aspect, to me, was that over half the group had extremely visible facial plastic surgery. You can read one in five Koreans has had plastic surgery but until you see this many pale, uncanny, off faces in a group you don’t feel it.

What IS trust?

Trust is one of those liquid concepts you don’t feel the lack of until it’s gone.

But the definition I’ve come back to, repeatedly, is that trust is consistency. The assumptions that things will behave as they did yesterday; that people will do what they say they will do, that social norms, institutions and infrastructure will keep holding.

For a long time in the West they mostly did.

But now?

Startups lying about revenue or product reality; politicians lying; declining trust in health institutions; my friend Isabelle’s iconic tweet about declining cloth quality.

Add in AI for acceleration, product hallucinations and AI psychosis**.

Between GLP-1***, Botox and Korean jaw shaving surgery, we can’t even trust our friend’s faces to be consistent.

Remember when people wouldn’t dream of playing music out loud on the bus?

It’s a confusing time. From timelines to headlines to phones to the world around us, we’re all sat on our lilypads trying to divine the truth.

Trust now will come with a tax for both businesses and individuals.

This tax could be financial in paying for premium, gated or human-curated services.

It could be your time in terms of extra verification, reviews or security measures.

It could be the cognitive load you feel at needing to filter and question everything.

Sounds like a lot, right?

It is. We’ll pay, and pay big, for products and companies that can give us the security we need or the space to escape.

Trust layers where the shared thing sold is a sense of stability between us, businesses and governments.

Boring things; banks that pick up phones, news sources we know to be human edited, open source projects with better mechanisms to pay contributors and transparent changelogs.

But in tech terms; it’s protocol time!

Think money, identity and governance; real-world reputations, cryptographic receipts, auditable histories and transparent logs.

Right now, and OpenClaw is one killer example, we’re rushing into tools without secure backends the same way we rushed to tweet twenty years ago.

Look how that turned out.

At the end of last year, several people I put trust into frustrated or disappointed me.

I paid some trust tax in the form of my feelings.

And started 2026 with questions; do I need to change the way I show up? Should I be less open?

Much relief to have decided; no. Nobody wants to live in a world where you constantly watch your back or only find trust onchain.

We do our best work in high-trust situations; excited to see what other trust layers I can find and fund this year.

Also;

• Londoners: Judith Dada and I are hosting a breakfast at the end of April for builders/founders/thinkers on AGI (or as my San Francisco friends would argue is the better phrase, ASI; artificial superintelligence). Both phrases feel Rapture-y; instead of freaking out or checking out, we should talk it out. Come. I’ll bring bagels.

• The Resonant Computing Lab launched a fund to further their manifesto.

• Finally! Pretext means we get websites like water [PLEASE go and check this out; only on web]

• Had my first fund exit this week with team Dreamer joining Meta’s Superintelligence Labs! Amazing for the team to bring their agent-first operating system to the scale of billions and of course bittersweet for investors who love spending time with teams on the sheer push of early company building. Dreamer aside, “endings” are on my mind. So much info about every aspect of startup building; very little about one of the most common things that happens in startups which is that they shut down. Given there’s such a thing as a good shut down:

  • Exhibit A: Foundation Labs built this generous memorial site with micro case studies of past experiments complete with screenshots and lessons

  • Exhibit B: in my four happy years running London’s first startup hub, we tried to build a culture distinct from what was normal in the UK at the time. An important part of that was generosity; a give/take/share vibe that enables

    real talk. Adam ran Startup Funerals there; a regular event where founders could choose to give eulogies for something close to their heart that hadn’t worked. He would wear a dog collar. Silly + profound at once.

I’m off to Athens today. Have you been? Send me your streets to wander down, bars to sit outside and cool things to look at.

Thanks for reading,

Sarah

* Don’t judge me, I was sick.

** Of all the startup ideas I’m surprised not to see yet, large scale attempts to tackle this are one.

*** I’m also surprised I don’t see GLP-1 discussed more in Europe’s tech scenes. We have less of an optimisation culture than the US; of course weight is a class issue and tech is mostly well off; but still! A potentially very important invention and, for those less inclined to be honest, another contributor to the uncanny valley.

Last week I was in Paris for a few days and, walking through the Marais in the early evening, fairy lights strung out above full tables outside corner bars, I thought…

Parisians should be gesturing wildly with their cigarettes, getting ash in their apéro, with one high heel knee booted leg crossed over the other.

New Yorkers should help me work out how to get my suitcase through the turnstile by shoving it and me through with a short-tempered sigh.

“Come ON” she barked. The JOY! I would expect + appreciate nothing else.

All these identity games we play.

Sometimes the cities we elect to live in; I grew up desperate to be a Londoner.

Or the work we do or aspire to do.

Families used to have trades; butchers and bakers and candlestick makers.

“Smith” is one of the most common British surnames because blacksmiths were so important to the village - swords and cooking pots and horses shoes.

What does Anthropic’s labour market report on which jobs become automated with AI have to do with this banging book I just read by Stephen Greenblatt on the life of Renaissance playwright Christopher Marlowe?

Bear with me.

Rewinding history - before agents, AI, VR, the cloud, mobile, the personal computer, television, radio, railways, public libraries the second time around, mass market newspapers, through my low-born ancestors marking “X” on documents because they couldn’t write their own name - speech + voice was the way most of us got information.

It’s 1600 and the messenger jumps off their muddy horse to tell the town square about moves the dastardly Dutch navy are making.

Later tonight you’ll tell your neighbour and for a swap she’ll tell you the standard price of eggs based on her trip to the market so you know how to mark yours.

A few miles away in London, the first playhouses are opening.

Outside your village square or church, there aren’t that many places to get together so this one crowded space has standing room only and, because it’s a pious religious time and theatre isn’t seen as respectable, you’re going to be standing with soldiers, traders, sex workers, pickpockets and the daring. Leave your gold at home.

The theatre owners buy plays in bulk and six times a week they stage slop (slop has existed in every era and every medium); medieval morality tales of spurned wives and overseas battles that always end with good defeating evil.

Until this one debut play.

Tamburlaine the Great broke people’s brains for two reasons.

It tells the real-life story of a Scythian shepherd who, through his own limitless ambition, conquers half the world. The Elizabethan audience expected punishment to come…but it just didn’t. The play wrapped with Tamburlaine ruling half the world.

Then there was the shock of the language.

Medieval plays used clumsy rhyming lines. Marlowe introduced blank verse; unrhymed iambic pentameter with five beats on, five beats off, like a drum.

A favourite from his later play Dr Faustus; read it out loud

Remember all information came to you through the ear; this innovation is basically like the first time you ever read a text message. The shock of the new.

And it changed everything. All plays were written like this for decades. Every modern speaker or rapper you ever loved was influenced in some way by Marlowe; whether by the beats he added behind his lines or the grandness of his language.

Kit Marlowe was a smalltown boy, the son of a shoemaker.

He grows up with parents who probably don’t own books.

There’s no evidence they can read or write.

We don’t know how this happened but through some quirk or force of personal will, he wins a scholarship to a prestigious boy’s school (girls won’t go to school for 200 more years). He wins it over boys who have much more formal schooling than him.

He does well enough to get a scholarship to Cambridge University as one of their first non-aristocratic students.

In his last year, he’s absent for long periods. Just as the university are working out if he should be expelled, a mysterious letter arrives from someone well connected in the UK government telling Cambridge to give him the degree.

We don’t know how this happened but through some quirk or force of personal will, he somehow became a spy.

On the side Marlowe writes these dazzling disturbing plays where serfy audiences get to live out, at a safe distance, the thrill of endless power (Tamburlaine), greed (his next play, the Jew of Malta) and knowledge (his last play, Dr Faustus).

It influences a new young writer called William Shakespeare.

He’s also a smalltown boy from a humble background.

Unlike Marlowe, he’s married with two small kids.

And Shakespeare left school young. No university.

And this turns into his biggest advantage.

Marlowe going to Cambridge meant you became part of the gentlemen class.

It put you in a status shaped box where you couldn’t do certain things without risking it - like acting on the stage.

Shakespeare’s lack of class status gives him social freedom.

Most writers also acted because plays were published anonymously and sold once (no royalties); acting brings money and builds your name.

Shakespeare acts, becomes well-known, writes a lot of plays, buys a theatre, invests in real estate and empire builds. He’s the most famous playwright of all time even if, imho, Marlowe is the better writer.

How does this connect to now and Anthropic’s predictions?

What jumps out to me about this chart - leaving aside how much history tells us there are eternally unforeseen consequences, good and bad and brilliant, to every prediction - is that at heart it’s about status and identity.

In the old world, this meant that the work you did positioned you deeply in a family with a trade, a physical location, a class.

It made you legible to your neighbours and crucially to yourself; “I am a baker / he is a baker”.

But in the last fifty years the flipside of economic reshuffles, migration and social mobility even pre-AI is that perceived social standings of jobs and their potential for economic freedom got untangled.

For example, I’d argue society always needs journalism and houses.

In the UK, journalism is high status but poorly paid (kills me it’s become a luxury job where only those with rich parents should apply).

Being in construction is commonly perceived as low status but, in reality, skill and pay levels can vary hugely.

Then, for a certain slice of us, jobs got unbundled from job-shaped boxes (journalist, builder) into multi-hyphenate identities where the multiplicity became the point. Coder, photographer, barista.

That’s why this chart feels so confronting for many because it feels like an attack on who we are as people not just what we do.

Marlowe and Shakespeare should tell us when you cling too hard to an identity, it can hold you captive (and that the first inventor doesn’t always get the credit; it’s a hard truth history is written by the winners).

I’ve always resisted being put in a box, work-wise. I chose to build this one fund, Common Magic, versus “working in venture”. I would have no real interest in working for another fund. You could look at my early career and find it restless (very Marlowe) but in general I believe breadth is its own kind of depth.

This freedom is something many generations won for me without realising it.

The best explanation I heard for “Drinkwater” is that my family were too poor to drink wine and had to make do with water; proving trolling exists in every era and that my family’s history of stable hands and servant girls was too diverse to sum up into one surname.

The trend for a very long time in my part of the world has been social mobility.

Generation to generation, you worked assuming your kids would have more than you. More education, more freedom, more satisfaction, more choice.

This unspoken generational social contract is close to breaking right now; this is the problem I worry about and it’s not just on AI.

You have to squint to see the future.

But…

the most interesting people I meet choose to work on things that, right now, are not socially rewarding. Weird hard problems.

They’re comfortable looking stupid for a long time.

Sometimes this is a luxury past generations won for them in their own confidence and relative stability; sometimes this is because they begin with nothing to lose.

Alvin Toffler said the illiterate of the 21st century will not be those who cannot read and write but those who cannot learn, unlearn and relearn.

A crucial part of that is being willing to hold our identity more loosely than our curiosity.

____

What else is new?

• Speaking of work + identity, many of my favourite through-the-rabbit-hole AI workflow builders are doctors or advertisers or artists [ie: not technical]. This is wildly cool! Dreamer launched their Builder in Residence program; modern paid patronage to allow you to agent build full-time

• Someone got Doom running on a pregnancy test. I love people.

• Philosophically opposed to multi-hour podcasts BUT I will make an exception to recommend this gem with Renaissance historian Ada Palmer on how hard it is to steer history the way you want it + 10,000 other topics

• The sexiest site you will see all year; fit for Virgil Abloah’s archive

• A deep dive into Reddit; both its history + how its intent driven communities have surfed multiple internet waves - cockroaches, glitter and Reddit can live through anything

• “From the top of the mountain / Every morning I walk towards the edge / And throw little things off / Like car parts, bottles and cutlery” - Bjork’s Hyperballad is the song and video of the season. It’s written partly in blank verse

Years ago, a friend who is a huge Bjork fan went to Iceland for a short break.

He wanted to see some cool churches, do rúntur and absorb the landscape that shaped his icon.

Right at he disembarked the plane, he bumped into a woman in a big coat in the airport. It was Bjork.

Isn’t that beautiful?

✌🏻

Shoutout to William Restless Egg who named this newsletter by accident in the DMs

A launch party at the intersection of culture and technology (real name; the Grey Area) for the Whole Earth Redux, a book of essays about the legendarily influential early zine that inspired the Homebrew computer club and a thousand communes.

A South Park Commons fireside chat with Granola’s design cofounder and their angel investor Soleio for an audience of proto founders, angels and operators

And a Friday night debate, shoes off, high up in a house in the Twin Peaks hills with the Hamilton Society to debate whether Anthropic should or should have not submitted to the Department of War in a room of true patriots in jackets and ties.

These are three distinct but overlapping versions of San Francisco and I went to all of them last week.

I always love coming to this pluralistic city and I always love leaving.

pages from the original Whole Earth catalog on agency at the Grey Area launch

Between Block layoffs, OpenClaw selling to OpenAI, the start of the US-Iran war and Anthropic’s tussles with the US government, it was a time! Quick notes written on the plane:

1. Breakthroughs, bubbles and bad behaviour.

Historically we know breakthroughs and bubbles go hand in hand.

One of California’s first breakthroughs happened in 1848 in Coloma, north east of Sacramento, when gold was found accidentally on a sawmill construction site. The original discoverers tried to keep it a secret but rumour got out, newspapers and the president picked up the story and tens of thousands of prospectors rushed in.

The state went from frontier farming to entrepreneurial energy to this part of the economy collapsing under its own weight, with many miners going broke as larger-scale enterprises dominated, in less than a decade. A bubble.

Undeniably with AI we’re in a breakthrough moment of true technological acceleration.

There are profound, exciting and real implications you’ll have read about in far deeper takes but, most broadly, enterprises are adopting faster than we might have assumed. Incumbents have adapted faster, consumers are adopting faster, even governments are building and buying new tools more quickly. The speed of revenue growth in many companies is breathtaking; ElevenLabs, in the emergent field of voice models, went from $25M to $330M in revenue in just two years. The future slid into view.

The tech sector and especially this city has always had a fetish for agency but when breakthroughs happen, buccaneering fortunes get made.

Capital crowds in. And, in the Gold Rush analogy, as the newspapers, President and prospectors rush in, less sophisticated investors arrive too. Everything gets faster and hotter. The gold is real - Vencap released data showing that every five years the top 1% of tech exits get exponentially larger and credible data arguing this isn’t a bubble - and yet… many conversations felt bubbly.

By that I mean;

Model company friends working at lightning speed to keep up with the sheer velocity of their field.

Friends in mid to large size tech companies working to uplevel skills with an eye to the further layoffs everyone believes are coming. A whole slice of tech jobs will disappear.

Founder friends are either being chased night and day by name brand funds brandishing escalating good “fill in the price” term sheets or desperately trying to get in the door.

Investor friends endlessly worried about how high company valuations are (each generation of VC: “this company is too expensive”), how stable their job is, how stable their investments are.

Family offices who are happy to pay 5% upfront fees to access the smallest slice of household name model company via double layer SPVs.

And I want to make the case for a third b that goes naturally alongside breakthroughs and bubbles; bad behaviour.

I’ve observed this in London and Europe too recently - true low integrity behaviour from founders, investors and more - but honestly in SF I heard some stories that made my ears curl. I’m not just talking about people running double layer SPVs.

Something I’ve always liked about the Bay Area is that it’s both hugely expansive in vision and ambition and, at heart, historically densely networked and high trust.

It’s comparatively easy for a person new to the city to meet amazing people and find their way into the network. Enough newcomers have built incredible companies from scratch for everyone to be open-minded about taking that coffee. Nearly every amazing company or fund is deeply linked into others. It’s a collective wealth project.

And the penalty for violating these norms used to be high. If you treated a YCombinator founder poorly, you would expect to get blackballed by the community.

The riches at stake right are so high that it made me wonder if this era is partially over.

I don’t have a strong opinion about whether we’re in a bubble or not (Schrödinger’s cat; I actually think both are true); what matters more is that people are behaving as though we are.

There was a general unspoken sense there is some AI endgame on the horizon and this drives both poor behaviour and short-term thinking.

Rise of the deal guy.

Everyone I met in SF who was truly thriving was deep in their lane whether they’re working on true wicked problems or 24/7 OpenClawing (“me and my nation of agents are happy idiots right now”).

2. Residencies are eating pre-seed

OpenClaw strikes me as both THE a-ha we needed for the future of agents and I’m convinced it took off on socials because people loved to feel part of the crew buying Mac Minis and showing off their setups.

People want to be in tribes.

One of the flipsides to both escalating first-round fundraises (“I haven’t seen someone raise less than $5M for months”) and the madness of the market is the rise of the residency.

These are programs or loose communities working at inception or pre-inception stage. Some are residential and firmly for those seeking a soft landing into the Bay Area. I’ve been tracking South Park Commons (mostly 30+ operators/serial founders), Z Fellows (typically very young technical builders), hf0 (repeat founders) for years, not to mention Y Combinator and my old friends Entrepreneurs First who have been bringing great European founders to SF to demo for a while. But there are now a dizzying variety of hacker houses and programs swapping everything from light mentorship to 24/7 home and board for equity.

All these give new builders early identity and peer group and funnel them into distinct streams of next round investors. But the knock-on effect is that pre-seed, as in true pre-seed belief capital of the kind I practice in London and Europe, feels exponentially harder to practise in San Francisco.

….

Stewart Brand on his new book, Maintenance. Brand has always been great at articulating how we can and should be individualist (be high agency, work on what matters to us) and collectivist (think of our shared futures) at once. Another Shrodinger’s cat.

He reminded me that ideas are tools, too.

Oakland makes me think of abundance. The scented wildflowers sprawling into the pavement from overgrown flowerbeds, wind chimes and sun faded sofas on cosy porches, free libraries and political posters. I don’t miss living in America but I do miss the little slice on the Temescal/Rockridge borders that I used to call home.

That was an idea-as-tool I was left with; abundance not scarcity.

The long now not just today.

So what else is going on?

• The hardest thing isn’t the decision to start or build, but what to build and who to build it with. Inspired by South Park Common’s “-1 to 0” frame, I’m running another casual breakfast Thurs March 12 in London along with Shre at Granola for people stuck in the idea maze - come, hang, meet some peers, get inspired

• It’s not just you; everything from the quality of your clothes to your digital experience is getting worse. This advert by Norway’s Consumer Council makes the point and made me laugh.

• Tinder for personal websites; yes! The first one I got was a communal radio club

• PasteurBench by London Bio Compute is cool. Science needs more experiments - this benchmarks AI scientists through real-world challenges. The best part is that AI scientists basically means you!

• Due to SF’s vibey past and it’s current citizenship mostly preferring leisurewear it is an excellent place to shop vintage. Striped maroon and orange? Yes.

• I’ve read many books about the Bay area over the years. Here are my favourites:

  • Shoes Outside the Door, a history of the attempts to set up the first American Buddhist centre and all the crazy fall-out

  • Season of the Witch tracing the history of the city from the summer of love to the AIDs crisis

  • Patrick deWitt’s Sisters Brothers follows two rough and tumble assassin brothers on the horseback trail of a gold prospector in 1851

  • Cryptography and conspiracy; Mr Penumbra’s 24 hour Bookstore by Robin Sloan is brilliant

  • And Joan Didion’s Slouching towards Bethlehem is an essential take on Haight-Ashbury and America’s disillusioned runaway 60’s teens

Til next time,

Sarah

Behind us in time is the miner’s strikes, yuppies, the Sinclair ZX spectrum, punk and the recently departed prime minister Maggie Thatcher.

A twin belief Thatcher shared with the defining British prime minister of the next decade, Tony Blair, is a belief in the power of business big and small.

You decide if her phrase “a nation of shopkeepers” is a realistic assessment of how trade is how humans have come together over the centuries, time and time again, or a brilliantly cutting critique of a nation that sometimes continues to think small.

In 1991, one of the remnants of her government was a scheme that gave small grants out to would-be entrepreneurs. They’d unwillingly sit through “business lessons” in order to get free cash at a time when most other forms of capital for new things - like venture capital - didn’t exist yet.

This is how the members of the band Portishead met.

If you’ve never heard Portishead; imagine you’re in a small room. The room is made of wood and lined in velvet. It is dusk. You have a hard drink of your choice and Big Feelings.

They’re impossible to describe so you should probably just go listen to Dummy.

I am ideologically opposed to any podcast over three hours (not as ideologically opposed as to ninety second slopreels though) but Bandsplain’s detailed history of a band steeped in lore was god tier. Still thinking about;

• “Arsing about (creatively) needs blank and bored spaces. You need to be bored. Scrolling your phone is not arsing about”

• Their entire advance made this ten minute film noir, To Kill A Dead Man, to launch first single Numb (the single Numb isn’t in the film… love that)

• Creative collisions can come from anywhere even governmental schemes

But the best:

Geoff Barrow was so in search of originality and a sound that couldn’t be remixed the next week that he created and recorded loops, pressed them onto acetate, battered the records so they got that retro patina and sound, then sampled them.

Originality requires obsession - doing things others don’t or won’t.

Dreamer came out of stealth this week; a treat for us alpha users and extremely cool to see the Discord pop off. The original hypothesis back at the end of 2024 when I invested is that early agents feel stilted because they interact with the pre-agentic surfaces of your personal internet; what we need is an agent-native operating system for truly personal software.

Instead of juggling tools, tabs and integrations, Dreamer is luxuriously simple; in the alpha the last few months, I’ve built a few things that are specifically useful to me.

A heavily enriched 6am daily summary, tagged with the category of meeting, short context and links to pre-reads, of everything I’m doing that day.

A weekly suggested list of interesting things to do with my son that weekend from science fairs to new movies.

When I travel I always want coffee shops that have wifi and are my vibe; this week it’s read my calendar bopping around San Francisco and built me a map which reveals I am eternally cursed through lack of early coffee in this city.

The gallery of other people’s agents to remix and reuse are just a beautiful reminder of human creativity and ingenuity.

David wrote a long lovely piece here about the why and the build. So much of the brilliance of Dreamer is the infrastructure for those of us that have struggled with integrations or security concerns with other popular agent system tools. (there’s a joke in my household that if a founder mentions the Mother of All Demos, Xerox Parc, Hypercard, Alan Kay or “bicycle for the mind” as a reference, my ears prick up. the Smalltalk reference here made me very happy).

Finally, I’m still thinking about Tool Shaped Objects by Will Manidis. I love the frame of “tool shaped objects” for where AI is right now.

It’s both true that we live in a golden time for truly personal software and real creativity.

And it’s also true that so much token usage and orchestration layers that feel like an uncreative blank space; the click and the spend and the tweet and the number goes up.

He’s right to ask us to interrogate what we’re doing and ask what the number is for before we make it go up.

Just like having a healthy breakfast, it became an everyday habit to record my video calls and, when I remembered, in-person chats (don’t forget to ask permission).

I wanted to build up a picture of the world as I learnt it through meeting as many smart founders, expert operators and investors as I could. But when I got their annual roundup just before Christmas (shown here in screenshots), I realised I’d also built up a picture of me. Taken together these are the fund’s operating principles.

Common Magic began in 2023 with a slow hunch that “community” isn’t a trend or a vertical but a secret hiding in plain sight for how generational companies get built. You look at Apple, Figma, Github, Stripe, Supercell, Lego, Roblox, Airbnb, Palo Alto Networks, Disney; the only thing this group have in common is distinctive ideologies and methodologies for how they engage deeply with their customers and users.

I learnt this first-hand over fifteen years helping technical founders and teams build moats around their products in the form of communities and narratives.

Common Magic exists to find, fund and support the next wave of generational companies building products with community at their core.

As a solo investor so much of what you are selling to founders is yourself.

So, while it feels grand to publish operating principles as a humble small fund, I hope these help any founder considering Common Magic for belief capital (email me sarah@commonmagic.xyz if you are).

How you work is the work. So….

1. Focus on the founder + all else follows

My favourite design principle in early 2010’s Google was focus on the user; listen, build with not for people.

Running a fund is similar. There are markets that are better or worse (and my favourite; markets in motion) but, as a first round investor, the thing I weight most highly is how singular the founding team are.

That’s why Common Magic is specialist generalist

Something fund one taught me is that community and narrative can be core building blocks for startups in multiple domains from developer tools to consumer, robotics to enterprise AI and hardware. These show up in how you hire, build trust with customers, create evangelists and so much more.

The best founders are learning machines who know a secret about an opportunity or market. They have absolute grit. Their passion and ability to communicate why what they’re building matters pulls the future forwards.

Rather than focus on today’s trends or any specific vertical, Common Magic is specialist generalist; building an intentionally diverse portfolio with a clear horizontal thread.

2. Common Magic works to build trust; trust wins you the right to be very honest.

Portfolio founders told me they chose Common Magic for three reasons.

The first two are thesis clarity (clarity of story creates clarity of strategy) and distinctive value-add.

The third reason is personal style.

I believe in building high-trust relationships. Company building is hard; you need people who will cheer for you in public and coach you in private.

Founders will and should always do what’s right for their company but you need partners around you who are truth tellers.

3. Small, fast and focused beats big, slow and general

Small solo funds have structural advantages that Common Magic leans into.

Fred Destin once said venture firms divide into artisans and industrialists.

For all the benefits a supersize brand and platform team can bring to a brand new startup, I’m an owner just like founders are owners.

There’s a primacy to the solo GP — startup founder partnership because we’re both building our life’s work.

Staying small means quick decisions, ownership ranges versus hard targets and investors who know their place on the cap table.

[But no shade to the big funds. Small funds can believe in outlier ideas before others can; large funds supply the right capital and brand to help distinct companies scale. Plus any pragmatic small fund assumes some of their exit strategy comes from the largest funds seeking to buy earliest investors out.]

4. The secret ingredient is fun

Starting something new is brave.

Building something big is hard.

Both steps are easier if what you’re building matters to you and you’re having fun. The grind feels lighter if you give yourself freedom to experiment and try, build with people you trust and celebrate each moment of momentum.

How you work is the work.

Let’s go build some Common Magic.