The Problem With...

The Problem With The AI Apocalypse: Connor Leahy

10 snips
Dec 9, 2025
In this insightful conversation, AI researcher Connor Leahy, founder of ControlAI, dives into the chilling implications of advanced artificial intelligence. He discusses how AI is already outsmarting humans and the potential dangers of emotional manipulation through AI relationships. Connor highlights the arms race between nations, the misinformation epidemic fueled by AI, and the urgent need for regulatory measures. He warns that AGI's rapid evolution could lead to uncontrollable superintelligence, igniting a call for immediate action to safeguard our future.
Ask episode
AI Snips
Chapters
Transcript
Episode notes
INSIGHT

AI Is Grown, Not Coded

  • Modern neural‑network AI is grown from massive data and compute rather than explicitly written code.
  • We have the numbers but still lack a deep understanding of how these models actually work.
INSIGHT

Opacity Means Loss Of Control

  • As AI becomes autonomous, society loses visibility and control over critical decisions.
  • Increasingly opaque systems will make it harder to intervene in or audit automated outcomes.
INSIGHT

Fast Leap From AGI To ASI

  • ASI means a system smarter than all humanity combined and would end human control if misaligned.
  • Once AGI can do research it can self‑improve quickly, potentially producing ASI very fast.
Get the Snipd Podcast app to discover more snips from this episode
Get the app