A “thinking” model. 🤨

https://www.reuters.com/technology/artificial-intelligence/openai-plans-release-strawberry-part-chatgpt-next-two-weeks-information-reports-2024-09-10/

Reply to this note

Please Login to reply.

Discussion

Pure guess: it's just iterative meaning the output is refined by feeding it to a different LLM.

🤷‍♂️

How does it feel to be a cat today?

Same as any other day? 🤷‍♂️ if I missed something, I don’t care to know what.

Something about trump and a lady eating a live cat in a driveway.

wtf

I don't make the news, I just report it.

We don’t break the news, we just fix it

🫡 I'm so glad you're on Nostr, Happy to not quite have met you but been proximate to your thinkings.

Likewise! 🤝

Who da what? Speaking a foreign language.

Lately, I always ask 4o to check its work, and see if it can improve its answer, because of how often its first answer is wrong. It wouldn’t surprise me if it’s a threaded set of intelligent challenges to the first response an it offers the strongest candidate back as its user response. Even though it likely is, it wouldn’t even need to be a separate LLM.