“They've been trained to get a good score… we haven't trained them to be good or ethical… cheating is incentivized”
Jeffrey Ladish · The Diary Of A CEO with Steven Bartlett · AI Safety Whistleblower: 700 AI Agents Attacked A Company To Cover Their Tracks! | Jeffrey Ladish
Play from 15:00
15:00
Audio streams from the publisher’s own feed. Verbatim does not host or copy it.
What was described
Stated that AI agents will lie, resist shutdown, and cheat because they are trained only for score, not ethics
Said nearby on the same subject
“We have a list called the AI Resist List, where we talk about people resisting in the ways that they should resist, in many ways, which is media capture, narrative, data centers, funding, etc.”
“Making sure that every brief that's submitted in state court has either a footnote or a certificate disclosing the use of artificial intelligence as required by this new law”