Working on something new. Prev. co-founder of @niftygateway, sold to @gemini. Occasional investor
Joined April 2009
- Tweets3.6K
- Following3.1K
- Followers13.9K
- Likes36.6K
Pinned Tweet
The truest form of privilege is to be born with a desire for constant change.
Most people find change uncomfortable. But in the modern world, becoming wealthy and happy depends on embracing constant change and growth.
If you naturally enjoy change, count your blessings!
We need a word to describe when an idea or belief is wrong; but societally useful when everyone believes it to be true
I keep encountering this phenomenon
AI doomerism is mostly an overcorrection to a total lack of Internet doomerism
No one thought the Internet would be bad and it turned out to be quite bad in many ways
Everyone has way overcorrected and is now unnecessarily afraid of AI
The worst part of living in an autocracy has got to be the fact that you don’t get to enjoy the absolute hilarity of post Election Day Twitter
Prediction markets as a source of news basically didn’t exist 5 years ago, and now they’re overwhelmingly the best way to figure out election results quickly.
It’s remarkable!
Money; power and fame are all destabilizing, but fame seems to be the most destabilizing of the 3 by far
For example, the life expectancy of a famous musician is 5 years(!) shorter than a non-famous one
Maybe there’s something about how fame distorts social relationships that makes it so unhealthy
This is very likely the best time in history to start a crypto company.
The financial system is moving onchain - the biggest TAM imaginable.
Meanwhile competition is near all time lows as so many people are leaving crypto
At @ycombinator, we're excited to fund more crypto startups solving financial needs including AI capital raising, stablecoins, trading, and agents.
Here's our latest crypto request for startups. Our next application deadline is Monday.
BREAKING: Russia has just announced a 'special military operation' to 'denazify' Nasdaqistan
Reports are flooding in of large troop movements and explosions
The U.S. really went all in on AI, like a gambler betting their entire life savings on black, and it hit
I’m really starting to suspect that the simulation is in fact rigged in the U.S.’s favor, lazy writing
Breaking News: Growth in the U.S. economy accelerated over the summer, as consumers and A.I. companies kept spending despite high prices. nyti.ms/48qLTME
Absolutely abysmal week for the Anthropic bears
Announcing Artificial Analysis Intelligence Index v5, featuring 2 new benchmarks and 3 upgrades. For the first time, the majority of the Index is built on private datasets
The Artificial Analysis Intelligence Index is our composite measure of model capabilities across tasks reflecting real-world usage today. v5 is our biggest upgrade since v4 launched in January. With v5, 60% of the Index is now based on evaluations with private test sets or held-out answers.
Alongside Intelligence Index v5, we’re launching the Artificial Analysis Progress Index v1, which tracks model capability over time on one continuous scale.
New to the Intelligence Index:
➤ FrontierCode 2.0: Cognition's newly-launched private benchmark of real-world coding tasks, graded on whether a repository maintainer would merge the agent's change
➤ Terminal-Bench-Science: 70 agentic scientific research tasks, run in a terminal environment
Upgraded:
➤ HLE-Diamond (replacing Humanity's Last Exam): a refined 1,000-question version of Humanity's Last Exam, split evenly between reasoning and knowledge questions
➤ AA-Omniscience v1.1: An updated dataset with saturated and public set questions removed, new questions added, and ambiguous or erroneous questions revised
➤ CritPt v2: a private, 40-question hard subset of CritPt v2, revised by the CritPt authors and run with a code-execution tool
Staying in the Index: AA-Briefcase, AutomationBench-AA, GDP.pdf and Terminal-Bench 4.0
Removed from the Index: GDPval-AA v2.1, SciCode, and AA-LCR v1.1
Key results:
➤ Claude Opus 5.5 (max) leads at 58, ahead of GPT-6 Astra (max, 57) and GPT-6.1 Sol (max, 55): Anthropic and OpenAI hold the top five models. Gemini 4 Argon (high) is next at 50
➤ The open weights gap widens from 3 points to 12: Kimi K3.1 Pro (weights not yet released) is the top open weights model at 46. Among released weights, DeepSeek V4.1 Flash 1017 (max) and GLM-5.3 (max) lead at 32 - 26 points behind the leader and below Claude Haiku 5.5 (max, 36)
➤ Cost per task varies by more than 13x among the leading models: of the 10 highest-scoring models shown, GPT-6.1 Terra (max) is the cheapest at $0.96 per Intelligence Index task and Claude Fable 5.1 (max) the most expensive at $13.04
“Reject France” is sending me lmao 😂
We should start saying WSB as shorthand for "We're so back", like how LFG has become ubiquitous as shorthand for Let's Fucking Go
Would save so much time typing
This is the point that Ben Thompson has been making repeatedly and is the perfect counterargument to the idea that intelligence becoming 'commoditized' is bad for Anthropic
They have the structurally lowest cost to serve. If intelligence becomes a commodity, they win!
A key reason for America’s success is that we have very patriotic and competent elites
Many countries have an elite who is mostly dedicated to extraction followed by a move to Switzerland
But American elites are builders who stick around
New SBF interview in @TheEconomist just dropped. He is still fully in denial
The parallels with Elizabeth Holmes are remarkable - neither of them can accept reality about what went wrong
*SPOILERS*
I’m shocked at how bad Billy and Elizabeth’s judgement is
Asking for $ from a movie that you’re making to show people you don’t want $ is a bad decision
CHANGING YOUR MIND about it later on is catastrophic!
All of the downside, none of the upside
I think Elizabeth definitely knew that Billy was asking for money, and when she said ‘I’m not benefiting from this movie financially’ that was one of her classic pedantic lies - she technically isn’t benefitting bc her and Billy have separate finances, but she knew Billy was