Sam Altman speaks on GPT-5: > we had this big GPU crunch > we could’ve made another giant model > people would want it… and we’d disappoint them lol > so we made a *really smart, really useful* model instead > optimized for inference cost LMAO https://t.co/A3TSnBqkUi
— NIK (@ns123abc) Aug 15, 2025
from Twitter https://twitter.com/ns123abc
August 15, 2025 at 02:36PM
via IFTTT
No comments:
Post a Comment