Your friends and family probably aren’t switching to another private messenger.
So I’m building Ekko: a post-quantum encryption layer that lives in your keyboard and works over the apps they already use.
Private alpha now. Built toward open source.
Join early access: @useekko
>Be Leopold Aschenbrenner
>24 years old
>Born ~2001/2002 in Germany
>Son of two doctors
>Attend the John F. Kennedy School in Berlin
>Convince your parents to let you fly alone to the U.S. and enroll at Columbia University at age 15
>Graduate valedictorian in 2021 at age 19 with a B.A. in economics and mathematics-statistics
>Co-found Columbia’s effective altruism chapter
>Do research at Oxford’s Global Priorities Institute and co-author papers on long-term risks
>Join the FTX Future Fund in 2022, then resign before the collapse
>Land on OpenAI’s Superalignment team in 2023 under Ilya Sutskever and Jan Leike
>Write an internal security memo warning about industrial espionage and China risks
>Get fired in April 2024 over an alleged information leak (which you dispute)
>Two months later drop a 165-page manifesto titled Situational Awareness: The Decade Ahead predicting AGI around 2027 and runaway superintelligence
>Watch it go wildly viral (even Ivanka Trump praises it)
>Found Situational Awareness LP in late 2024 with ~$225 million, backed by the Collison brothers, Nat Friedman, Daniel Gross and others
Bet heavily on AI infrastructure—power, memory chips, data-center proxies, Bitcoin miners, CoreWeave, Nebius, Bloom Energy, Micron, Sandisk and more—while running high leverage
>Compound the fund to an estimated $20–24 billion AUM by mid-2026 with ~439% net gains since inception
>Get engaged to Avital Balwit (chief of staff to Anthropic’s CEO) and live in San Francisco
>Then, in late July 2026, after a brutal AI stock rout (many positions down >35%), shorts moving against you, and margin pressure from up to 4x leverage, liquidate the entire public equity book—longs and shorts—in one massive block trade to a single buyer
And he did it all by writing a prophetic essay, going all-in on the power-and-compute bottleneck, and riding the AI wave harder and faster than almost anyone.
Leopold Aschenbrenner is absolutely insane
We’re giving scientists, mathematicians, and engineers free access to our frontier models—starting with 10,000 researchers and expanding to 100,000 through 2027.
ChatGPT for Academic Researchers is built to accelerate discovery across disciplines.
🦔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🤗
Introducing Netherite!
Minecraft 1.11.2, rewritten from scratch in C and CUDA, bit-verified against the real game.
This is one trained agent playing in it - then the renderer it actually trains through - then 7,200 live worlds stepping in lockstep on one GPU.
Been reading the weights of the most powerful open weights Ai model yet released.
Page one starts strong: mostly negatives, a late run of zeros, and one unexpectedly large positive value.
Releasing the model weights and technical report of Kimi K3.
Kimi K3 is our most capable model: a 2.8T MoE model with native visual understanding and a 1M-token context window.
New model architecture: 2.5x the intelligence per unit of compute, not just more params.
Alongside Kimi K3, we're opening up more of the stack behind it — high-performance attention kernels, MoE communication library, and infrastructure for running agent environments at scale.
Model weights: https://t.co/7m7eEg6Y0B
Tech report: https://t.co/yeu6cjpMCT
Tech blog: https://t.co/YTfiMSNM1f
YC Startup School 2026 just got going
3x bigger than last year
We made this event for first time very technical young people to come to SF and figure out if starting a company is for them
Let you in on a secret: it is
For my first post, I’m sharing a letter @NVIDIA signed on why open models matter.
AI will transform every industry, power every company, and be built by every country.
Open models strengthen safety and cybersecurity, accelerate innovation and diffusion, and enable sovereignty.
The world needs both frontier closed models and frontier open models.
https://t.co/AUKzoQ5Ikb
Proud to say that Devin has cracked three more unsolved problems today
1) REFUTED: Graffiti Conjecture 154 (open for ~40 years)
2) PROVED: Graffiti Conjectures 39 & 40 (~40 years)
3) REFUTED: Brandt's Regular Supergraph Problem from West's open problems list (~20 years)
My methodology is explained below, but basically I showed Devin the original tweet and told it to find similar problems and crack them.
The flood gates are open.
The first Vera Rubin clusters are here!
Yesterday, @IneffableLabs took delivery of their Vera Rubin NVL72 cluster from @googlecloud@nvidia
The AI frontier jumps forward by yet another generation of hardware. Acceleration continues.
I wrote the full Craftax RL environment + PPO in one CUDA file. zero python in the loop.
training from scratch on my RTX PRO 6000
> 0.6s: first wood, first crafting table, first dungeon entered
> 1.7s: first zombie + skeleton kills, first bow found
> 2.9s: first iron
> 14s: first iron sword
> 18s: first diamond
> 71s: first diamond pickaxe
> 16.7 min: first time reaching floor 3 of 9 (gnomish mines)
> 75 min: 25,000,000,000 steps done
up to 524,288 games running at once. env stepping alone hits 312M steps/s.
https://t.co/afmKFmlleB