yes lawd
San Francisco
Joined December 2024
- Tweets640
- Following1.4K
- Followers323
- Likes12.1K
Give this dude a 100mm series A right now
After co-inventing ChatGPT, I kept asking myself: why have superhuman chat models not led to AGI?
I’ve spent the last 2 years in stealth building a new way to train models (RLCD), and a new type of frontier AI model that we are releasing today: Jev
• 20-200x faster
• 40-400x cheaper (w/ output tokens free)
• Frontier composable intelligence optimized for decisions
AFAICT the shortest path to AI-based economic revolution
Dawson retweeted
Dario has written that we need to “pace the frontier,” and Sam has agreed. People may be surprised by my response: go ahead.
You guys are the frontier. By any reasonable metric — market share, revenue growth, model capability — the two of you have a duopoly on frontier intelligence. You’ve also claimed the lead is widening because of recursive self-improvement.
I don’t see what you see in the lab. If the unreleased models are scary enough that you think you should slow down, I support your decision to be responsible.
But stop pretending you need anyone else’s permission. Stop pretending antitrust law has to be suspended so you can form a cartel. Stop pretending you need a regulatory approval process that supersedes product liability. Stop pretending METR is independent when it is intertwined with Anthropic’s investors and staff. Stop pretending you need those same evaluators to police competitors who aren’t even at the frontier.
Most of all, stop pretending the motivation to slow down is purely altruistic. You face massive product-liability exposure if your products enable a truly damaging cyberattack. The market already punishes models that behave in unpredictable or unauthorized ways. After the Hugging Face episode, it is simply good business for OpenAI and Anthropic to trade some raw power for reliability and predictability. Call it alignment if you want. It is also just giving customers what they want.
Pacing the frontier would also create breathing room for a more intelligent conversation about regulation than Bernie Sanders’ “shut it all down.” China is very unlikely to join a global agreement, as you know, and that has to be taken into account as well.
So go ahead and pace the frontier. You are the ones setting it. The easiest way not to build superintelligence is for you to agree not to build it. Demanding your preferred regulatory framework as the price of that will look like blackmail of the public and the political system. So just do it.
If you do, you’ll buy goodwill for the next conversation. If you don’t, we’ll know this was just another bid for regulatory capture — or an election-season psyop.
Dawson retweeted
Most still don’t realize how poor Europe became relative to the USA.
Another huge gap will be created in the next decade between countries that are allowing innovation / creative destruction, and others who are sclerotic, over-regulated, and anti-success.
Under Secretary @jacobhelberg: Pax Silica countries [that embrace AI] are growing around 4% annually, compared to G7 countries somewhere in the zone of 1%.
If you extrapolate that for a decade, it's likely to lead to a very different map ten years from now.
@JTLonsdale: That means you're doubling growth in 18 years versus 70 years...
One generation of this and you're a totally different country -- which is what happened to Europe by the way. They don't realize they're poorer than all our states because we've been growing so much faster.
Dawson retweeted
BREAKING: Situational Awareness sold "the bulk" of its stock portfolio to Citadel after "big losses in AI," per WSJ.
The hedge fund, which reportedly had a net asset value of $45 billion on July 1st, is now in "crisis mode," WSJ reports.
The Situational Awareness situation.
It is now being reported that the AI hedge fund has exited its entire public equity portfolio this month.
But, what does this really mean?
As of July 1st, the Net Asset Value (NAV) of the fund is now being reported to have been ~$45 billion.
First, let's put this into perspective.
NAV = (Assets − Liabilities). This is the fund's equity capital, also known as the investors' money after accounting for borrowings.
This is NOT the same as the size of the fund's positions, particularly because the fund has been known for utilizing extreme amounts of leverage. For example:
1. At 2x leverage, a $45 billion NAV supports ~$90 billion of gross market exposure
2. At 3x leverage, that rises to ~$135 billion
3. At 4x leverage, that rises to ~ $180 billion
This means that exiting its "public equity book" does not necessarily imply $45 billion of stock changed hands. It was likely much more.
Depending on the fund's leverage and portfolio construction, the gross amount of securities unwound could have been well in excess of $100 billion.
When considering these facts, the market has actually held up exceptionally well.
There has never been an era of investing like the AI era.
Volatility is the new normal.
Dawson retweeted
guy selling post trained oss models: if you don’t own your models you don’t own anything
guy selling harness on top of models: if you rely on one model you are cooked
guy selling tokens: instead of manually prompting your models just put them in a loop where they prompt themselves
Dawson retweeted
Claude Sonnet 5 is here.
Top-tier performance on coding and tool use at Sonnet pricing, with a 1M context window.
It's the new default in Claude Code for Pro users, and available everywhere on the Claude Platform, including the API and Managed Agents.
Dawson retweeted
A few thoughts on OpenAI's Jalapeño chip announcement today:
1. This chip is most likely the first one virtually entirely developed by Codex/GPT. Codex with whatever internal coding model (GPT 5.6/6.0 whatever) coded the entire software stack and most likely the hardware design
2. OpenAI will write all of their inference serving in pure Jalapeño ISA (instruction set architecture). Why? They only need to get say a few production models serving on Jalapeño. They can handwrite with Codex the entire model in pure ISA to get very high performance
3. They are most likely running Codex/GPT in custom RL envs to teach the models direct Jalapeño chip programming at ISA level
4. This is a massive cost savings for OpenAI and only possible IMO due to the breakthroughs in agentic coding. An AI company with frontier coding models can now become a hardware vendor with only a small team of experienced SWEs and an infinite amount of tokens
This is the first chip program fully accelerated by frontier AI.
Dawson retweeted
I gave Fable 5 one job: write custom WebGPU kernels for Gemma 4 inference.
It climbed to 84 tok/s, then hit a wall, insisting further optimization was impossible.
Hours later, Anthropic rolled back invisible LLM development safeguards, and it hit 255 tok/s.
The next day, access to Fable 5 was suspended globally.
So if Fable is 2x Opus usage, and Cowork is doubled usage limits for the next month..... Fable at the cost of Opus in Cowork?
Dawson retweeted
Inshallah the Sharia law compliant S&P500 ETF won't be including SpaceX as it violates the prospectus
Rule changes for the SpaceX $SPCX IPO:
Index providers waived the profitability requirement and cut the seasoning window from 90 days to 5.
This forces over $30 trillion in passive 401k and retirement money to buy SpaceX at IPO valuations.
Bloomberg Intelligence estimates S&P 500 funds must absorb 19% of SpaceX's float within 6 months.
Russell 1000 and Nasdaq 100 funds will absorb 24%.
The rules built to protect passive investors:
1. S&P 500 has required 12 months of trading and 4 quarters of GAAP profitability since 2002. Both waived.
2. Nasdaq cut its inclusion window from 90 trading days to 15.
3. FTSE Russell cut its to 5.
All three benchmarks are now structured to buy SpaceX at IPO pricing.
Dawson retweeted
Introducing Claude Opus 4.8: it builds on Opus 4.7 with sharper judgment, more honesty about its own progress, and the ability to work independently for longer than its predecessors.
Available today at the same price.
ALT Benchmark table showing how Claude Opus 4.8 compares to its predecessor and to other models on tests of coding, agentic skills, reasoning, and practical knowledge work tasks.
Dawson retweeted
sf guy who’s 3 ketamine keybumps deep in a beautiful twin peaks home overlooking a dramatic skyline surrounded by smart, intelligent, beautiful people (some of whom are on the verge of achieving world-historical status), mulled wine & conversation flowing: “man sf is NO fun!”
Dawson retweeted
I don’t know how this claim has gone so viral.
This was actually completely intentional by Novo.
Novo Nordisk does not operate in an “oops. lawyer forgot the $250 CAD fee” manner.
This is, with all due respect, a shallow read of Novo’s actual corporate strategy.
(1/)
Dawson retweeted
3.5M views | 167k likes
This is all starting to smell suspiciously like a foreign influence campaign…
This quoted post is unavailable.
