It's all about trust. How do we help machines (and humans) know what to trust? | Hacker News

Hacker Newsnew | past | comments | ask | show | jobs | submit

		johnecheck 5 months ago \| parent \| context \| favorite \| on: A small number of samples can poison LLMs of any s... It's all about trust. How do we help machines (and humans) know what to trust?

nativeit 5 months ago | [–]

See Tom Scott’s rather prescient lecture to the Royal Society titled, “There is No Algorithm for Truth”.

phs318u 5 months ago | [–]

We can't help humans figure out who/what to trust. Are chances with machines are slim.

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact