Tom Le Bree (@tomlebree) profile picture

Failed Rap DJ. Former partner at Prehype London.

London
Joined March 2009
Photo by Tom Le Bree (@tomlebree)Photo by Tom Le Bree (@tomlebree)Photo by Tom Le Bree (@tomlebree)Photo by Tom Le Bree (@tomlebree)Photo by Tom Le Bree (@tomlebree)Photo by Tom Le Bree (@tomlebree)Photo by Tom Le Bree (@tomlebree)Photo by Tom Le Bree (@tomlebree)Photo by Tom Le Bree (@tomlebree)Photo by Tom Le Bree (@tomlebree)
Tom Le Bree retweeted
A growing construction company wanted to bid on more work, but experienced estimators are hard to hire and the ones it had were buried in prep. Preparing a 200-line tender took around 40 hours. Now it takes about 8, and the estimators still make every pricing call. The hard part was everything the company already knew about prices, suppliers and past jobs, scattered across spreadsheets and people’s heads. How AI Point 8 put that to work is in the comments, and you can ask Stipe about it there. There are 13 more solutions for quotes and bids like this on file, from manufacturers, contractors and distributors. Built by @StipePoint8 at AI Point 8. This one’s on @aireceiptswork. How AI Point 8 built it: aireceipts.work/r/recU5ASE28… All solutions for quotes and bids: aireceipts.work/problems/quo…
1
1
3
181
Tom Le Bree retweeted
BREAKING: OpenAI just launched Astra 6.2 and Sol 6.2!! We've been testing them for the last few weeks @every and, unfortunately, these models aren't great. Overall a disappointing release compared to Fable 5.2. - Writing. We found Astra 6.2 to be a step back from Astra 6 on writing. 6.2 appears to have been trained to very extremely concise and actionable—so much so that its writing is hard to understand, and it gives short answers when long answers would be appropriate. - Coding. We found performance to be similar to Astra 6 on most of our benchmarks. However, in real coding tasks Astra 6.2 seemed to stop too early frequently, which made it quite frustating to use on end-to-end builds. - Design. Astra 6.2 has good default design instincts. But it struggles to vary its designs, resulting in output that has a noticeable "house style". Full vibe check coming soon on @every!
Some of you have been asking very politely. Some of you have just been asking. GPT-6.2 Astra and Sol are now available. Super efficient and built to perform where it counts—the real world. openai.com/index/introducing…
111
67
65
1,205
205,446
There it is hiding in plainsight "becomes a top-30 League of Legends player" I am going to start a fund to back people who excel at computer games. At this point, it's not a coincidence.
So Diogo Almeida is one of the most fascinating CEOs I have ever interviewed. born in the Philippines 2008: wins bronze for the Philippines at the International Mathematical Olympiad 2013: starts his career as a software engineer at Amazon 2017: joins Google Brain, working on cutting-edge deep learning research 2018: leaves Google Brain and takes roughly 18 months off to figure out what makes him happy travels the world, experiments with improv, art and meditation becomes a top-30 League of Legends player discovers that none of it makes him particularly happy 2020: joins OpenAI when it's still a relatively obscure research laboratory helps develop reinforcement learning from human feedback (RLHF), the technique that transforms how people interact with AI 2022: co-authors the landmark InstructGPT paper the work becomes foundational to ChatGPT November 2022: ChatGPT launches and becomes one of the fastest-growing consumer products in history 2023: contributes to GPT-4, one of the most important model releases in AI history but starts questioning the fundamental assumptions behind the technology he helped build 2024: leaves OpenAI while ChatGPT is exploding in popularity people think he's crazy for walking away co-founds TypeSafe AI with Erik Gafni and Sasha Sheng spends two years in stealth developing an entirely new class of AI models September 15, 2026: launches Jev, TypeSafe's first System One model Company makes $100M in revenue in the first 7 days of launch raises a reported $40M seed round led by DCVC October 2026: a reported $870M Series A at a $7.5B valuation, backed by a16z argues that the next breakthrough in AI will come from making intelligence cheap, reliable and autonomous My notes from my discussion with @CompleteSkeptic below: 1. Why Jevons' Paradox Does Not Cut It. We Are Going to 10x That Diogo expects a "mega Jevons paradox", with usage that dwarfs the token consumption of LLMs and AI running in the background everywhere. He expects people to stop seeing AI as something that "slops crap at them" and start seeing it as something that makes life better. 2. Why SaaS Incumbents Are Better Positioned Than OpenAI and Anthropic in Many Application Spaces SaaS companies know their customers' workflows and have the scale to fund the upfront engineering that real automation requires. Diogo argues their edge comes from being closest to the problems, not from distribution. Teams inside products like TurboTax know where to improve the interface, while OpenAI and Anthropic are "still living in their ivory tower." 3. Why We Do Not Hire From OpenAI Diogo sees getting into OpenAI as an anti-signal and says its researchers love algorithms and scaling but hate data. Data is a low-status topic at most labs, and the few people focused on the right task would want to pursue their own. That leaves researchers obsessed with algorithms, which he likens to throwing money in the furnace. 4. How the AI Bubble Could Burst and Where the Demand Shock Could Be Part of today's AI demand is driven by FOMO, from people believing they are getting good results. Diogo finds many buyers miscalibrated and overestimating what AI can do. Once they face the realities of its capabilities, demand could fall quickly. 5. What Metric Did ChatGPT Optimize for in the Early Days? Before Claude Code, ChatGPT was optimized for monthly active users rather than the best chatbot experience, which is how sycophancy emerged. Companies are now optimizing for the MAU equivalent in coding agents. Without a dedicated model, Diogo doubts they will produce a comparable one soon. 6. Why the Talent Market Is Not Crazy for Us and We Don't Pay for Mercenaries TypeSafe pays people good amounts but not the highest, which filters out hired guns. Diogo says Anthropic is suffering from mercenaries because it pays more than anyone else, and an entire market now exists to prepare candidates for Anthropic culture interviews. 7. We Wanted Sonya Huang More Than We Wanted Sequoia Diogo did not want Sequoia's money and brought the firm in because of Sonya. He values her particular shape of brilliance, which offers a perspective no other advisor gives him. One TypeSafe value is to make great decisions slowly, and he sees her as a valuable thought partner for that. (links in comments)
1
66
Tom Le Bree retweeted
The history of drug discovery is a slow battle against biological entropy. @IsomorphicLabs is changing the clock speed, moving toward experimental accuracy and making meaningful progress in its mission to solve all disease. GV is excited to continue backing this world-class, London-built team again in their Series C! Congrats @demishassabis @maxjaderberg and team 🚀
We’ve closed our $5 Billion funding round co-led by Greenoaks and Sequoia Capital to build the future of medicine and bioengineering with AI. isomorphiclabs.com/articles/…
5
4
58
2,301
Let's get me on stage at @sxswlndn 🙌🏽🙌🏽🙌🏽 If you're done with being limited to one job title then vote for "The One-Title Trap" #sxswlondon
2
3
282
Tom Le Bree retweeted
Menlo’s excellent healthcare report confirms the trend we have been backing. The first wave of healthcare AI automated paperwork, and incumbents are now collecting most of the returns: v. interesting to read Epic and Microsoft took roughly two of every three new scribe dollars this year. The second wave is harder and more valuable and that’s where we’ve been focusing. Life-sciences AI spend grew 23x. Half of pharma now runs AI at scale, up from 15% a year ago. Clinical copilots grew 5.5x, while scribes grew 1.6x. Lasting companies will be built where incumbents have less of a head start and the full benefits of AI are most clearly felt - i.e. discovery, trials and clinical decision-making. This work needs proprietary data, deep science and/or products judged on patient outcomes rather than hours saved. Those advantages are hard to copy and startups are well placed to focus here. Sceptics (and there’s a lot out there) will point out that a minority of pharma leaders expect an AI-discovered drug to be approved within two years. I see that as a sign of how early the market is and with the underlying models being as bad today as they ever will be the chance that we see huge advancements here is inevitable, imho. The startups that generate the data and earn trust now have the opportunity to create huge value. Europe’s science base combined with its deep technical experience in AI gives its founders a real shot at this. We’ve been fortunate to already partner with some exceptional companies (Lindus, Uncovr, Apoha, Sano, Dalton, Sable, Inovia, Verisian, Resolyst, Outpost bio and many more) and look forward to meeting many more founders building for the future of healthcare in the era of AI @seedcamp
Healthcare is now AI’s biggest vertical market. Spend grew 2.5x to $3.7B this year, more than double legal ($1.8B) and more than legal and finance combined. It started in paperwork: scribing and billing. Now it’s moving into drug discovery, patient engagement and clinical decision support. Clinical copilots alone grew 5.5x to $550M. Our 2026 State of AI in Healthcare report from @grgyap, @jpsanday, @CroomBeatty, @hu_j, @derekgxiao and @sabrinarlu: mnlo.vc/2026-healthcare-ai-r…
8
6
56
9,947
1 year ago, I decided to focus on mascots (for my side gig). I had no idea if it was going to work, but I strongly believed mascots were a huge gap in the market. It's been a wild ride with many long evenings and weekends spent sketching, building, and iterating, and here we are
1
1
6
243
Pleased to announce we've acquired the luma.com/londonhackathons Luma page from @natal1e_c 🤌 A beauty of a URL slug that matches londonhackathons.com, and is very well looked after. Pleasure doing business with you, Natalie. 🤝
8
8
3
82
6,278
Tom Le Bree retweeted
Replying to @tomlebree
descript founders went quiet in my dms
1
1
50
The SAAS-pocalypse is back on. Hold on to your ETFs fellas.
This is why I've been freaking out - kieran made professional software we actually use internally (I do all my video recording and meetings with this now instead of descript loom and zoom). My favorite feature is is multimodal jev-based scene detection.
1
1
57
Interesting concept. I have been thinking of similar things in other industries - a couple of hours of expert time to verify my output. Shape. Correct. And harden.
Now writing is solved the new bottleneck I'm facing is getting people to tell me which of my ideas are worth publishing. The editors at @every do this but I already have a backlog with them so I'm considering paying people to read my stuff to decide what to submit to editors.
33
Tom Le Bree retweeted
Worst week of Jensen's life
The Nvidia CUDA moat is gone. Anthropic sees all compute as fungible and it only took Mythos 3 days to figure out how to run itself on AMD hardware.
25
89
1
2,123
171,822
Except Epstein where his recollection gets decidedly blurry. 😂
If you are a famous person, do NOT acknowledge or speak to Eric Weinstein. He will spend the next two years going on podcasts namedropping you and referring to you as his friend and colleague
52
Tom Le Bree retweeted
Not with that attitude you won’t
2
1
29
835
Tom Le Bree retweeted
the @every team, nov 2026 so proud of this crew!
33
14
7
253
21,594
Tom Le Bree retweeted
I did an interview with Amol Rajan on AI and the UK amongst other things. Has been a while since I did one of these longer form interviews so let me know if you find it useful!
I believe Ian Hogarth - @soundboy - deserves much wider and deeper credit and recognition for voluntarily forfeiting shares in Anthropic so that he could set up and run the AI Security Institute (@AISecurityInst) - where he, and Britain, are generally acclaimed to be doing an outstanding job. (Something else he deserves greater credit for). This noble decision was *very* expensive. I also believe, and hope, that for our fast-growing audience this may be one of the most useful conversations they hear about Technology this year. We go deep on: > What is AI? > The real rate at which it is developing > The race dynamic in global innovation > Humanity’s fork in the road > The significant chance of a huge market crash > The exhilarating case Quantum Computing > The exhilarating case for Fusion > The exhilarating potential for Robotics > Why outstanding British companies sometimes struggle to scale > How Britain, and Europe, can lead > The Future of jobs > How Founders succeed > His noble decision to do a public service role > His fantasy for Battersea Power station Ian is co-Founder and Partner at Plural, and co-Founder and Chair of the UK's AI Security Institute (@AISecurityInst). This was a great, immensely useful conversation and I'd be honoured if you lent an ear to The Amol Rajan Connection. Out now on Spotify, Apple Podcasts, Amazon Music and YouTube. Thanks for your time and attention to this matter.
6
20
1
129
9,428
Tom Le Bree retweeted
Our podcast show is one year old! 2M+ downloads so far and counting! Most quit after the first few episodes but we are still here and going strong!! A huge thank you to all our previous guests, partners and friends of the channel for all your support so far! We have one mission: Help you learn from the most brilliant minds shaping AI! Back to the microphone in 10 minutes!
13
3
1
58
14,896
Tom Le Bree retweeted
I almost feel like I shouldn't have shared this, too much alpha given away
2
3
56
5,921
The smartest moron in the US of A. Shaun is extremely intelligent. If he says something like this you can be assured he has an agenda.
Charlie Kirk was assassinated by the left Yet people still voted for it Says a lot
55