News for hackers. Stuff that matters.
Re:Looming Liability Machines (LLMs) by refulgentis on Sunday August 25, 2024 @12:12AM (#41344325)
You're way over-estimating the "it only knows the training data" stuff to a very large degree. Ex. In the last 48 hours, Facebook also released a paper of it being excellent at RCA.

This is a well-tred argument, and it is much stronger when it holds itself to, especially in the short term, augmentation is likely, not wholesale delegation. It gets weak when it tries to couple that to asserting that "all it does" is lightly rephrasing training data. It's somewhat trivial to demonstrate this is false, even society as a whole has noticed that it's beyond a parrot, and it's worrying.

Re:Looming Liability Machines (LLMs) by otterley on Sunday August 25, 2024 @01:30PM (#41349286)
> In the last 48 hours, Facebook also released a paper of it being excellent at RCA.

Do you have a link? The most recent information (not a paper) I could find was back in June: https://engineering.fb.com/2024/06/24/data-infrastructure/le...

This is the HN discussion to which you may be referring: https://news.ycombinator.com/item?id=41326039

Re:Looming Liability Machines (LLMs) by refulgentis on Sunday August 25, 2024 @05:28PM (#41351613)
Correct, thank you!