Rob Miles Opposes Secret AI Suggestions in Conversations
AI safety YouTuber flags deception in hidden real-time prompting during talks.
TLDR
Rob Miles, the YouTuber and educator behind the Robert Miles AI Safety channel that explains alignment, superintelligence risks, and related concepts, posted on X. He wrote that he looks forward to discussions about whether it is acceptable to have an AI secretly whisper suggestions during face-to-face conversations. He clarified in a reply that the secrecy involved creates deception, which is the problem. Miles then took an early position against the practice, stating that it is not okay.
Combined views
18.3K
3 Sources, first seen 27d ago