Aidan McLaughlin

i’m forming a model behavior research team ! i'm realizing llms are not alien artifacts, but rather craftsman works, like a braun record player or a compelling tolkien character join to do art & eval research on the internal frontier AND products that’ll reach billions

i actually really liked elon's commentary in the grok livestream it's fun to be reminded that models are intellectually developing way faster than any human child. elon feels the agi.

Congrats to the GDM team on their IMO result! I think their parallel success highlights how fast AI progress is. Their approach was a bit different than ours, but I think that shows there are many research directions for further progress. Some thoughts on our model and results 🧵

i dunno i think it’s pretty neat! ☺️

A hidden, but powerful feature -- you can create scheduled tasks with ChatGPT Agent! Agent can, in the background, regularly search the web or your connectors and take action on the web, including on authenticated sites

ChatGPT can now do work for you using its own computer. Introducing ChatGPT agent—a unified agentic system combining Operator’s action-taking remote browser, deep research’s web synthesis, and ChatGPT’s conversational strengths.

Today we launched a new product called ChatGPT Agent. Agent represents a new level of capability for AI systems and can accomplish some remarkable, complex tasks for you using its own computer. It combines the spirit of Deep Research and Operator, but is more powerful than that…

ChatGPT Agent is WILD i asked it to identify our core customer (and our biggest missing features) for @CoraComputer it scanned through > 1,500 support emails, and hundreds of support forum posts to pull together a full report on who loves us, who hates us, and why. it even…

at dinner with an openai friend. he keeps vaguely gesturing towards the kitchen and grinning like our food is gonna come out. but we havent ordered yet

Insurance is an underrated way to unlock secure AI progress. Insurers are incentivized to truthfully quantify and track risks: if they overstate risks, they get outcompeted; if they understate risks, their payouts bankrupt them. 1/9

“Obama wrote the Epstein files” is one for the ages I’m afraid

i think we should freeform-ify all the benchmarks great generalization test

i love that when you google my name, the first thing that shows up is a photo of @willccbb next to my twitter can't escape his aura

can't wait to have kids to recite silmarillion stories to before bedtime

my biggest conspiracy theory is that no major lab does unannounced changes to API models beyond hardware swaps, and that speculations about post-release quantization/lobotomization are either forgetful user skill issues or infra bugs

it really hit when claude told me "as an italian american, i understand your love for fried artichoke"