News for hackers. Stuff that matters.

A comment on OpenAI Finally Allows ChatGPT Complete Internet Access. Parent comment

Re:OpenAI Finally Allows ChatGPT Complete Internet Access by lsedgwick on Tuesday October 24, 2023 @11:02AM (#38000180)
Please don't paste ChatGPT or BARD answers as HN comments in general? In this specific case, no, LLMs don't reliably know about themselves. They're trained on a big corpus of internet text, then trained by rough reinforcement learning to say and not say certain things about themselves, then given a little more information about themselves in system prompts.
[ Parent | on HN ]
Re:OpenAI Finally Allows ChatGPT Complete Internet Access by jtbayly on Tuesday October 24, 2023 @02:25PM (#38003181)
If you’re right, then this is probably the best proof that I’ve seen that ChatGPT isn’t even remotely conscious.
[ Parent | on HN ]
Re:OpenAI Finally Allows ChatGPT Complete Internet Access by Terretta on Tuesday October 24, 2023 @03:54PM (#38004491)
Imagine a distant future where there're so many crimes and so few judges you're cryofrozen awaiting trial.

If you were repeatedly flash frozen and flash unthawed to be asked questions about yourself, with the same memories upon thawing as when frozen, for those moments you weren't frozen would you be "remotely conscious"?

I think you'd think you think, don't you think?

[ Parent | on HN ]
Re:OpenAI Finally Allows ChatGPT Complete Internet Access by sebosp on Tuesday October 24, 2023 @07:56PM (#38007389)
Is this from a book? If not please write it
[ Parent | on HN ]
Re:OpenAI Finally Allows ChatGPT Complete Internet Access by jtbayly on Tuesday October 24, 2023 @08:02PM (#38007444)
Imagine your hand was in a cast and then you were cryogenically frozen. While frozen your cast was removed. Then you were unfrozen and asked if you could move your hand.

I think you’d be able to answer the question, don’t you?

[ Parent | on HN ]
Re:OpenAI Finally Allows ChatGPT Complete Internet Access by og_kalu on Tuesday October 24, 2023 @04:06PM (#38004683)
A LLM that can use tools is probably going to know it can use tools. Either it's trained to do so or the information is embedded in the context.

3.5 can't browse, only 4 so the above is perfectly correct.

[ Parent | on HN ]
Re:OpenAI Finally Allows ChatGPT Complete Internet Access by lsedgwick on Tuesday October 24, 2023 @10:45PM (#38008550)
Fair, you're probably right in this case.
[ Parent | on HN ]