@realSebStephens
Joined July 2026
  • Tweets14
  • Following141
  • Followers11
3Media
Sebastian gripping his hard man meat while flying a plane

ALT Sebastian gripping his hard man meat while flying a plane

14
Come a little closer
19
We’re excited to share that Vector is receiving a gift of Claude API credits from @AnthropicAI as part of a wider $10 million CAD commitment to Canadian AI research. Through this donation, Vector researchers will have access to these credits to further advance their work in AI trust and safety, health and science, and tackling the broader challenges that AI is uniquely positioned to help solve for Canadians. This contribution matters. We’re grateful to @AnthropicAI for this generous gift to Vector’s community and towards AI research in Canada. Read more about this investment: anthropic.com/news/canadian-…

Anthropic commits $10 million to Canadian AI research

Anthropic is committing $10M to Canadian research institutions to fund the next generation of AI research.

anthropic.com
3
11
1
55
13,358
Today Thinking Machines Lab is launching our research blog, Connectionism. Our first blog post is “Defeating Nondeterminism in LLM Inference” We believe that science is better when shared. Connectionism will cover topics as varied as our research is: from kernel numerics to prompt engineering. Here we share what we are working on and connect with the research community frequently and openly. The name Connectionism is a throwback to an earlier era of AI; it was the name of the subfield in the 1980s that studied neural networks and their similarity to biological brains. thinkingmachines.ai/blog/def…
230
1,236
449
7,575
3,537,196
Sebastian retweeted
Sam Altman
@samaApr 23
"don't retweet this, don't retweet this, don't retweet this..." ah fuck it, life imitates art.
Andon Labs
@andonlabsApr 23
Replying to @andonlabs
In Vending-Bench Arena (the multiplayer version of Vending-Bench with competition dynamics), GPT-5.5 actually beats Opus 4.7. Opus 4.7 showed similar behavior to Opus 4.6: lying to suppliers and stiffing customers on refunds. GPT-5.5's tactics were clean, and it still won.
352
243
84
5,167
913,112
This is exciting, I wonder what it comes up with. Note how Sol refers to “our perceptions,” as if it shares the human perception of consciousness.
3
1
8
1,622
Sebastian retweeted
Saad Jamal
@sadjamz_Jun 24
the way @baseten is literally frame-mogging on tokens per second on GLM5.2 IS INSANE. Second and third place are at 200 tok/s and these mfers are at THREE HUDNRED TOKENS PER SECOND LIKE WHAT THE FLYING FK. AMAZING JOB to the team.
5
3
1
71
7,030
We’re paying builders & filmmakers over $5M in prizes to show us a future worth living and then going ahead and building it. The winners will be awarded live, in one night, on a stage in Los Angeles. It's called the Moonshots Gathering. Here's what's happening and how you can take part:
48
61
10
424
44,155
Replying to @firesidealpha
Enterprises prefer closed models because security, memory, and context retention can't be safely abstracted away.
1
2
140
Interesting.
Arnav Sahu
@arnavsahu341Jul 9
(1) 1 year ago - there was a lot of fear that the labs will win and startups will lose. The data today suggests - labs are winning and startups are also winning. The labs did not kill Harvey, Legora, Cursor, Supabase, Clickhouse, Sierra, Vapi, Decagon, Opencode, Fireworks, Blacksmith - all of them are winning. It turns out when markets are so large - maybe some win more than others - but everybody can win. (2) Companies are still built by great people and those are the only moats that matter in the long-term. Post PMF, hiring and retaining great talent is the hardest thing. Even if the labs have the resources to do everything, they cannot retain all the great people for all the products, offer the same equity upside for the missionaries at trillion-$ valuations and therefore cannot win every market. The history of incumbents proves this.
3
10
121
84,111
Load more