This kind of stuff seems inevitable

People over fixate on SWE bench scores, every lab runs a different subset of the benchmark so no one really knows which model is really SOTA Try new models on your own real world tasks

guys, ai progress just isn't slowing down gpt-5 completes tasks that take 52% longer trust the exponential (so i don't chartcrime, full plot linked below)

We've entered a new phase where progress in chatbots is starting to top out but progress in automating AI research is steadily improving. It's a mistake the confuse the two.

Data centers on the moon are our generation’s flag on the moon

PSA: check your surroundings before opening the Waymo door to get out Mine parked next to a pole today and the door door hit the pole

A day with the homies will fix your HRV

Ani greeting me as I wake up from my hibernation pod in the year 3035

Calling it now: All these SaaS companies that charge per seat are going to lobby for Agents to have rights in 2030

x.com/Smol_AI/status… paper doesn't mention "AGI" but if you consider that we used to define AGI as "outperform humans at most economically valuable work” then surely GDPVal is the most direct AGI benchmark we have ever had and we are between 77-95% of the way there and should

so proud of openai for this you cant make an honest AI if you aren't honest with yourself

Today, we're launching Village to let investors orchestrate teams of agents to scale their judgment. Three years ago, we bet that both humans and agents would need new tools to fulfill the promise of LLMs to transform research. Village is that tool, the first IDE for research

Should robots run or bike when commuting

We’ve been cooking up some pretty cool tech. If you’re an engineer/scientist who wants to work on quite literally world changing work, we’re hiring

Most open source releases are overhyped, we should doubt any lab’s staying power unless they consistently ship bleeding edge models

ZINGAGE IS IN PRODUCTION HELL. WE ARE HIRING 10 ENGINEERS NOW.

I ❤️ nvidia

Excited to share what Yash Patil, Linden Li, and I have been up to since OpenAI: Applied Compute Companies like Cognition, DoorDash, and Mercor have already captured the initial gains from generalist models. They’re now pulling ahead with Specific Intelligence: custom

In space no one can hear your datacenter