CeladonAstrodon (@celadonastrodon) 's Twitter Profile
CeladonAstrodon

@celadonastrodon

ID: 847511474809458688

calendar_today30-03-2017 18:10:56

105 Tweet

43 Followers

538 Following

Andrej Karpathy (@karpathy) 's Twitter Profile Photo

It is hard to communicate how much programming has changed due to AI in the last 2 months: not gradually and over time in the "progress as usual" way, but specifically this last December. There are a number of asterisks but imo coding agents basically didn’t work before December

Joseph Viviano (@josephdviviano) 's Twitter Profile Photo

me: "can you use whatever resources you like, and python, to generate a short 'youtube poop' video and render it using ffmpeg ? can you put more of a personal spin on it? it should express what it's like to be a LLM" claude opus 4.6:

ZoeCyber (@zoecyber001) 's Twitter Profile Photo

I was today years old when I realized that we don't need to hack your database to see your data - we just need to be in the same building as your server. ​While you're busy rotating JWT tokens and hardening your firewall, your hardware is physically screaming your secrets into

ℏεsam (@hesamation) 's Twitter Profile Photo

AMD Senior AI Director confirms Claude has been nerfed. She analyzed Claude's session logs from Janurary to March: > median thinking dropped from ~2,200 to ~600 chars > API requests went up 80x from Feb to Mar. less thinking and failed attempts meaning more retries, burning more

AMD Senior AI Director confirms Claude has been nerfed. She analyzed Claude's session logs from Janurary to March:
> median thinking dropped from ~2,200 to ~600 chars
> API requests went up 80x from Feb to Mar. less thinking and failed attempts meaning more retries, burning more
Marius Hobbhahn (@mariushobbhahn) 's Twitter Profile Photo

The model now starts calling our scenarios out as "alignment test from Apollo". Eval awareness keeps making testing more complicated and I don't think people have sufficiently updated on this being a problem.