Listen

Description

Séb Krier is very AGI-pilled, but he rejects many popular narratives about the AI revolution. He joins Dan Williams and Henry Shevlin to explain why.

Séb previously led frontier policy development at Google DeepMind, recently joined the DeepMind Institute, and is a Senior Research Fellow at the Cosmos Institute. He and Henry both speak in a personal capacity.

We discuss:

• What Séb calls the “Sand God” view of AI, and why he rejects it

• Why recursive self-improvement looks more like a complex, jagged industrial process than an intelligence explosion

• What the Hugging Face incident does and does not show about alignment

• Why institutions and scaffolding matter even if AI systems are benevolent

• Whether the bigger danger is misaligned AI, AI companies, or concentrated state power

• p(doom), AI consciousness, and “pacing the frontier”



This is a public episode. If you'd like to discuss this with other subscribers or get access to bonus episodes, visit www.conspicuouscognition.com/subscribe