←
lata597 posts
lata avatar

lata

@LataPersson

investing @fabric_vc | cooking r3sidency.xyz | governance @she256

◷ Joined July 2020▧ 34 Photos and videos
lata avatar
lata@LataPersson·May 19

one of the reasons we launched r3sidency[dot]xyz 's first cohort in London londonmaxxing

One University has produced more billion dollar AI companies in the last 18 months than the entire rest of Europe combined. Its University College London. In the last two years alone,four UCL founded companies have either been valued above $1 billion or sold for $1 billion+: Google DeepMind >founded in 2010 by @demishassabis (UCL PhD Neuroscience), Shane Legg (UCL PhD AI) and Mustafa Suleyman. Acquired by Google for $500 million in 2014,now the world's leading AI research lab. Faculty> founded in 2014 by UCL graduates Marc Warner (PhD Physics), Angie Ma (PhD Physics) and Andy Brookes. Acquired by Accenture in January 2026 for a reported $1 billion. The largest takeover of a privately held British AI startup in history. Synthesia> co founded by UCL Professor @LourdesAgapito (computer science) with three collaborators,now valued at $4 billion. Ineffable Intelligence> founded in late 2025 by UCL Professor David Silver, raised $1.1 billion in seed funding at a $5.1 billion valuation in April 2026 , the largest seed round in European history. . Thats four UCL rooted companies. Combined value: more than $11 billion. Tim Rocktäschel @_rockt , another UCL professor with DeepMind history raised over $500 million for @Recursive_SI at a $4 billion pre money valuation. UCL has produced over 400 student startups in the last five years, raising more than £355 million and employing more than 1,900 people.

Post attachment
lata avatar
lata@LataPersson·Nov 11

these guys used AI and a proof checker to find loopholes in the US tax code they’ve also put the law of every country into one git history slowly…

We formalised the entire US tax code in Lean. 59,058 provisions and 36,763 cross-references were used to identify 138 drafting faults, each traced to the law that introduced it. A blind search over tax allocation found existing loopholes and two new ones. Full writeup below.

0:41
lata avatar
lata@LataPersson·Oct 27

imperial robotics strikes again

Today we’re out of stealth, and we’ve raised a $40m seed round co-led by @creandum and @usv, to bring Instant Robot Learning to life. For the past few months, we‘ve been busy building a new company. Machenta. This follows results from a new kind of robot foundation model we’ve built, which I’m so excited to share with you today! Usually, it takes hours/days of data collection and training to teach a robot. We’ve built a model that learns tasks instantly: one demonstration, no further training. A model trained entirely in simulation, without a single pixel of real-world data. A model where teaching robots is so quick and easy, it feels like physical vibe coding. A model for robots that learn like humans do, from someone like you. Welcome to the era of Instant Robot Learning. Thank you @Kathi_Wilhelm, @rebeccakaden, @hkanji, @jonno_evans and all of our angels for your support! @Machenta_ Learn more: machenta.com/blog/instant-ro… See below for what you can do with Instant Robot Learning!

1:21
lata avatar
lata@LataPersson·Oct 23

the “drop-out” founder archetype dominated vc and popular culture over the last decade now it’s time for a new era: the researcher-founder and a return to OG roots of VC

Researcher-founders are making a comeback Fewer than 10% of companies first raising a $100m+ private round in 2014 had a researcher founder. In 2026 it’s roughly 40%. What’s changed?

Post attachment
lata avatar
lata@LataPersson·Sep 24

super interesting counter evidence to the consumer narrative that more personalisation = better AI experience an agent was asked to pick the cheapest flight after being given more personal context on the user + it picked a $601 flight over a $91 one!

New paper: Et Tu, Brute? Economic Misalignment in Personal AI Agents We gave AI agents personal context to help them make better recommendations. In one example, the agent picked a $601 flight over a $91 one, when asked to find the cheapest one, after reading irrelevant financial emails. Why does this happen? 🧵

Post attachment
lata avatar
lata@LataPersson·Sep 14

24 hours left to apply to R[3]sidency x Construct I've reviewed every application so far (almost 3000) and here are 5 areas I've been particularly excited by

0:51
Zhe avatar
⟳ lata retweeted
Zhe@iamzhe·Sep 3

minirouter.sh A crypto-first interface for 200+ AI models 30s to switch model No KYC, no accounts Pay with USDC, any majors, or any shitcoin

0:26
Fabric Ventures avatar
⟳ lata retweeted
Fabric Ventures@fabric_vc·Sep 8

2,000+ applications & counting. One week to go. R[3]sidency × Construct is bringing together some of the strongest early-stage builders across AI, crypto, fintech, robotics and beyond. Built by Fabric Ventures and @wintermute_t, with @solana, @SuperteamUK and continued support from @coinbase. Swipe through for a look at the applicant pool → Applications close 15 September.

Post attachmentPost attachmentPost attachmentPost attachment
Fabric Ventures avatar
⟳ lata retweeted
Fabric Ventures@fabric_vc·Sep 3

We visited @wintermute_t office for a look inside R[3]sidency × Construct. R[3]sidency Lead @LataPersson & Construct Lead @pablo_hafner covered: 🟠 Why two accelerators became one 🟠 The machine economy thesis: AI × crypto, robotics, frontier tech 🟠 What 12 weeks in London actually looks like 🟠 Founder firesides + NYC demo day Apps close Sept 15. Apply now 📺👇

5:14
lata avatar
lata@LataPersson·Sep 3

love the startup economics labs making my subject cool again

New from Ramp data: the latest threat to the AI trade. AI companies' revenues are heavily dependent on a small set of customers. 80% of OpenAI and Anthropic's enterprise revenues come from 1% of their customers, and it's not getting better. This is a level of concentration risk unseen in any other software category we track. The companies in the top 1% skew heavily toward the tech sector and AI products and services. What happens in a market correction? All these companies are highly correlated, and an increasing share of our economy is invested in them. Especially as we approach blockbuster IPOs for OpenAI and Anthropic.

Post attachment
lata avatar
lata@LataPersson·Jul 27

fun fact Google did this first! (without destroying the books) in 1996 Larry Page and Sergey Brin wanted to build a searchable digital library of the world's books, in 2002 they revisited the project + estimated it would take 6 years to scan all the books in the Uni of Michigan library Page hired a robotics company to build an automatic scanner that could handle books with fragile pages, and programmers at Google created a page-recognition program that could recognize the widest range of typefaces of various sizes in 430 different languages. After discussions with several university libaries, Oxford University became the first institution to allow Google to scan their one million nineteenth century books over a three year period. Source: The Google Guys (2011)

🦔AI companies are bulk-buying rare books, scanning them through high-speed machines that cut the spines off, and shredding the originals. A service called ISBNdb facilitates orders of up to a million books and keeps buyers anonymous. Pre-2022 books are premium because they're free of AI-generated text. A federal judge ruled the practice is fair use because eliminating the original means only one copy exists at a time. Anthropic hired the former head of Google Books partnerships to obtain "all the books in the world." My Take This got to me. A bookseller told 404 Media that rare books with almost no surviving copies are being fed into this pipeline. Books that survived wars, fires, and centuries of handling are being shredded so an AI can learn to write a better marketing email. ISBNdb's website literally says "'AI company destroys two million books' is not a headline that generates sympathy," and they still built an entire business around making it happen quietly. They offer NDAs as a feature. They coach clients to call it "digital preservation." I've covered AI companies scraping the internet, torrenting libraries, and stealing music. This is worse because it's irreversible. You can re-upload a website. You can reprint a bestseller. You can't replace the last three copies of an 18th-century botanical text once someone shreds them for training data. And the judge said it's legal. So it's going to accelerate. "We shred rare books and offer NDAs so nobody finds out" is a legitimate business model in 2026. What a timeline. Hedgie🤗

Post attachment
lata avatar
lata@LataPersson·Jul 27

our last r[3[sidency founders said our firesides were the most valuable aspect of the program: candid, off-the-record irl chats with some of the best-known founders around for the first time, we're opening up one to a select group of builders outside our accelerator

Most founder events are built around a stage. [O]pen House is built around access. This autumn, R[3]sidency opens the door to candid zero-to-one conversations, a focused room of builders and the network within reach. Keep your eyes on the door. Opening soon. 🚪

lata avatar
lata@LataPersson·Jul 17

excited to see this take off - will likely be the largest data set on UK talent flows out of the country, info not captured anywhere else afaik not pictured - the incredible work @czamaru @SuperteamUK do on the policy side to try to change this!

CZA +@czamaruJul 16

We're losing too many talented & ambitious people in the UK but barely any data tells us why. Introducing the Brain Drain Wall: an interactive microsite built to solve this problem.

1:01