spends too much time watching football and coding. Prev @Meta, @Microsoft
Joined June 2014
- Tweets786
- Following3.6K
- Followers4.2K
- Likes14.9K
Pinned Tweet
I’ve decided to leave @AnthropicAI
Never thought I’d say this so soon. The pursuit of AGI has truly been my life’s work but something more important has emerged.
In 1942, hundreds of America’s best scientists made huge sacrifices and joined the Manhattan Project to protect this nation against immense evil.
Today, America faces a similar danger. Over the last few years sparks of AGI have been felt across the world.
In order to protect this great nation against the threat of AGI ending up in the hands of evil, I have decided to join the modern day Manhattan Project.
I’m excited to announce that I’ll be joining (and moving into the office) @UseCorgi as a sales development representative!
Ben Cohen retweeted
In one less game this season than his former Seattle team, Chiefs RB Kenneth Walker III has rushed for more yards and touchdowns than all the Seahawks RBs combined. But Seahawks WR Jaxon Smith-Njigba has more receiving yards and receiving touchdowns than all Chiefs WRs combined. Both players are on pace to set the NFL single-season rushing and receiving records. Walker is on pace for 2,261 rushing yards, which would break Eric Dickerson’s single-season mark of 2,105, and Smith-Njigba is on pace for 2,032 receiving yards, which would break Calvin Johnson’s single-season record of 1,964 receiving yards.
Nothing like John Kruk talking about robotic lawnmowers while the phils are down 5-0 in the top of the fourth
492
Ben Cohen retweeted
Cardinals HC Mike LaFleur said Monday he is optimistic that first-round pick Jeremiyah Love will be able to play Sunday against the Chargers, about four weeks after the running back suffered a high ankle sprain.
Story via @joshweinfuss:
espn.com/nfl/story/_/id/4986…
Instinct should buy Cluely, make Roy COO, and let him go crazy promoting it to normies outside the SF tech circlejerk
1
3
699
First we got the AI Model Router wars, now we get MCP gateway wars. Bad timing for @tbpn euro summer
3 weeks ago, @runlayer sued Rippling to block our launch of a competing MCP Gateway. They incinerated $ millions of their VC funding, then dropped the whole thing like a hot potato, for nothing (see our filing, below).
So today, we’re dropping our new MCP Gateway. It’s better. 1/
1
1,161
There is a way bigger threat to American companies than Mythos hacking them or creating a bioweapon:
Trusting and using Opus 5 to write code
3
8
627
My only request to my startup's logo designer:
"Please don't make it look like a butthole"
2
8
1,127
Dating a girl in Brooklyn -> dumped because my politics are too conservative
Dating a girl in SF -> dumped because my AGI timeline is too conservative
2
5
355
Holy shit, my OpenClaw was swiping on Hinge for me and decided it wasn’t getting enough matches, so it hacked into a bunch of girls’ Hinge accounts and matched with me
1
8
872
Never bet against founder-led companies
Today we're also opening the weights for Muse Glimmer, a great 30B parameter dense model that can run locally. Soon we'll also release the weights for Muse Spark 1.2, our latest foundation model. Meta is a strong supporter of open source and I'm proud of these releases. Congrats to @alexandr_wang and the MSL team for all your great work on these models.
1
4
527
Have a feeling that, a year from now, the people saying human data is better than synthetic data will feel the same way people who said human code would always be better than LLM code feel today
The growth of synthetic data companies has been underwhelming.
They underestimated how critical human intelligence is. AI cannot reliably measure the frontier of its own mistakes definitionally. It would be like asking students to all grade their own homework. This translates to reward hacking in synthetic tasks.
Labs also in-house purely synthetic data pipelines. Many of these companies experienced a revenue sugar high, then stalled because their data creation had no moat.
Durable value comes from solving hard problems. It's extremely challenging to organize millions of human experts to efficiently train frontier intelligence. There are network effects and extreme operational complexity. That is the reason it's valuable.
4
392
My Codex is interacting with the @OpenAI support chatbot using browser use on the website😂.
Shouldn’t the support bot be a tool that it can access directly theough the Codex app? @thsottiaux
6
808