Statement on AI Risk

Re:Statement on AI Risk by revelio on Tuesday May 30, 2023 @05:26PM (#36131084)

No, lots of important AI researchers are missing and many of the signatories have no relevant AI research experience. As for being the cats whiskers in developing neural architecture or whatever, so what? It gives them no particular insight into AI risk. Their papers are mostly public, remember.

> Has it occurred to you what happens if you are wrong?

Has it occurred to you what happens if YOU are wrong? AI risk is theoretical, vague and most arguments for it are weak. The risk of bad law making is very real, has crushed whole societies before and could easily cripple technological progress for decades or even centuries.

IOW the risk posed by AI risk advocates is far higher than the risk posed by AI.

[ Parent | on HN ]

Re:Statement on AI Risk by arisAlexis on Wednesday May 31, 2023 @04:43AM (#36135987)

In order to make your argument clear to people reading:

If you are wrong there are no humans left.

If I am wrong,inequality,societies will suffer like they have always in the hands of the strong.

[ Parent | on HN ]

Re:Statement on AI Risk by hatefulmoron on Wednesday May 31, 2023 @06:32AM (#36136689)

Wouldn't you need to illustrate how likely each outcome is? I mean, there are lots of possible ways humans could be eradicated.

[ Parent | on HN ]

Re:Statement on AI Risk by arisAlexis on Wednesday May 31, 2023 @01:03PM (#36140673)

First of all, you agree that the probability is non zero right?

I am not a world known expert on xrisk to estimate this. You are not either. We have all these people claiming the probability is high enough. What else is ti be said? HNers shouldn't be reminded "trust in sciense".

[ Parent | on HN ]