News for hackers. Stuff that matters.

A comment on US and UK refuse to sign AI safety declaration at summit. Parent comment

Re:US and UK refuse to sign AI safety declaration at summit by sirsinsalot on Wednesday February 12, 2025 @04:28PM (#43029951)
It has been shown many times that current cutting edge AI will subvert and lie to follow subgoals not stated by their "masters".
[ Parent | on HN ]
Re:US and UK refuse to sign AI safety declaration at summit by slg on Wednesday February 12, 2025 @08:08PM (#43031616)
Care to provide examples?
[ Parent | on HN ]
Re:US and UK refuse to sign AI safety declaration at summit by 20after4 on Wednesday February 12, 2025 @10:51PM (#43032490)
Maybe not the specific example the parent was thinking of but there is this from MIT: https://www.technologyreview.com/2024/05/10/1092293/ai-syste...
[ Parent | on HN ]
Re:US and UK refuse to sign AI safety declaration at summit by code_martial on Thursday February 13, 2025 @12:47AM (#43033039)
Subversion and lies are human behaviours projected on to erroneous AI output. The AI just produces errors without intention to lie or subvert.

Unfortunately, casually throwing around terms like prediction, reasoning, hallucination, etc. only serve to confuse because their notions in daily language are not the same as in the context of AI output.

[ Parent | on HN ]