Working Intelligence: Making Americans Irreplaceable Beyond Agency: AI, Autonomy, and the Art of Self-Rule
9 snips
Nov 18, 2025 
Harry Law, a researcher at the University of Cambridge and former ethics lead at Google DeepMind, dives deep into the interplay of AI and human autonomy. He discusses how AI can either enhance or hinder our self-rule. Law distinguishes between AI that clears cognitive space for growth versus that which fills it superficially. He emphasizes the importance of user-selectable guardrails to maintain a pluralistic information environment and advocates for preserving our capacity for autonomous decision-making as machines become smarter.
AI Snips
Chapters
Transcript
Episode notes
Three Roots Behind Modern AI
- AI's current success comes from three long-running traditions: statistical, computational, and biological roots.
- The 1980s convergence of those traditions made scale the dominant driver of progress.
AGI Is About Proactive Agents
- AGI means reviving the original AI dream of a general system that can do anything a human can on cognitive tasks.
- The key shift is toward agents that act proactively, not just reactive chat tools.
Autonomy Versus Deference
- Autonomy matters because AI can either clear space for human flourishing or fill space and promote deference.
- Overreliance risks outsourcing intellectual and moral work to models instead of cultivating self-rule.