this post was submitted on 17 Sep 2024
46 points (100.0% liked)

TechTakes

41 readers
19 users here now

Big brain tech dude got yet another clueless take over at HackerNews etc? Here's the place to vent. Orange site, VC foolishness, all welcome.

This is not debate club. Unless it’s amusing debate.

For actually-good tech, you want our NotAwfulTech community

founded 1 year ago
MODERATORS
dgerard@awful.systems
46
OpenAI does not want you delving into o1 Strawberry’s alleged “chain of thought” (pivot-to-ai.com)
submitted 2 months ago by dgerard@awful.systems to c/techtakes@awful.systems
4 comments fedilink hide all child comments
top 4 comments
sorted by: hot top controversial new old
[–] AVHeiro@awful.systems 17 points 1 month ago (1 children)

It's gonna turn out to be an filipino call center isn't it.

[–] dgerard@awful.systems 14 points 1 month ago (1 children)

we have finally achieved A Guy in India

[–] YourNetworkIsHaunted@awful.systems 2 points 1 month ago

That's one way to put a sentient being on the other end of the request, I guess.

[–] zbyte64@awful.systems 6 points 1 month ago* (last edited 1 month ago)

That’s OpenAI admitting that o1’s “chain of thought” is faked after the fact. The “chain of thought” does not show any internal processes of the LLM — o1 just returns something that looks a bit like a logical chain of reasoning.

I think it's fake "reasoning" but I don't know if (all of) OpenAI thinks that. They probably think hiding this data prevents cot training data from being extracted. I just don't know how deep the stupid runs.