Robinson's Podcast

200 - Sean Carroll, Daniel Dennett, & Steven Pinker: AI, Parapsychology, Panpsychism, & Physics Violations

22 snips
Mar 24, 2024
Sean Carroll, Daniel Dennett, and Steven Pinker discuss AI, parapsychology, panpsychism, and physics violations. They delve into debunking claims of parapsychology, explore artificial intelligence, discuss the ethical implications of counterfeit individuals created by large language models, and contemplate the evolution of jobs in the technological age. The conversation touches on the intersection of AI, creativity, and physics, evolution, chaos, decision-making, teleonomy, and emergence in consciousness.
Ask episode
AI Snips
Chapters
Transcript
Episode notes
INSIGHT

Counterfeit People Threaten Trust

  • Large language models can flood the world with convincing counterfeit people that exploit our instinct to treat communicative agents as human.
  • That threatens trust and fragile institutions like democracy unless rapid, intelligent countermeasures are deployed.
ADVICE

Prefer Systemic Defenses Over Paranoid Checks

  • Rely on technological and institutional countermeasures rather than forcing everyone to detect fakes manually.
  • Advocate quick, vigorous, and intelligent policy and tech responses to unmask counterfeit agents.
INSIGHT

LLMs Are Not The Same As AGI

  • Large language models are distinct from general AI because they lack propositional understanding and real-world grounded concepts.
  • This explains why they hallucinate and why different AI architectures may be needed for deeper intelligence.
Get the Snipd Podcast app to discover more snips from this episode
Get the app