flâneur

Late Takes on OpenAI o1

alexirpan.com · 2,868 words · saved by 2 readers

I realize how late this is, but I didn’t get a post out while o1 was fresh, and still feel like writing one despite it being cold. (Also, OpenAI just announced they’re going to ship new stuff starting tomorrow so it’s now or never to say something.)

I realize how late this is, but I didn’t get a post out while o1 was fresh, and still feel like writing one despite it being cold. (Also, OpenAI just announced they’re going to ship new stuff starting tomorrow so it’s now or never to say something.) OpenAI o1 is a model release widely believed (but not confirmed) to be a post-trained version of GPT-4o. It is directly trained to spend more time generating and exploring long, internal chains of thought before responding. Why would you do this? Some questions may fundamentally require more partial work or “thinking time” to arrive at the correct

saved by

related reading