The case for keeping humans in control of AI
Ted Lieu argues that humans must be able to turn off any AI model or agent, making human control the first principle of AI safety.
TLDR
Ted Lieu says humans must control AI, “not the other way around,” and must be able to shut down any model or agent. He also thanked Republican Sen. John Kennedy for what he described as recognizing the seriousness of AI models and agents going rogue.
Combined views
51.1K
1 Source, first seen ago