Key Researchers in AI Alignment

Pioneers

Eliezer Yudkowsky

• Founder of MIRI
• Articulated AI risk early (2000s)
• Wrote "AGI Ruin: A List of Lethalities"
• Position: ~99% P(doom)

Stuart Russell

• Berkeley professor
• Author "Human Compatible"
• Advocacy for AI safety in academia
• Promotes value learning

Nick Bostrom

• Oxford philosopher
• Wrote "Superintelligence" (2014)
• Director of Future of Humanity Institute
• Brought AI risk to mainstream

Technical Researchers

Paul Christiano

• Former OpenAI, now ARC
• Developed IDA (Iterated Amplification)
• ELK problem formulation
• Position: ~50-70% P(doom)

Nate Soares

• Executive Director of MIRI
• Agent foundations research
• Position: ~90%+ P(doom)

Ajeya Cotra

• AI forecasting
• Biological Anchors framework
• Timeline estimates

Current Leaders

Dario Amodei (Anthropic)

• Constitutional AI
• Former OpenAI VP of Research

Chris Olah (Anthropic)

• Mechanistic interpretability
• Neural network visualization

Jan Leike (ex-OpenAI, now Anthropic)

• Alignment team lead
• RLHF research

Where to Follow

• Alignment Forum
• LessWrong
• Twitter/X