Expert Positions3 topics
| Topic | View | Estimate | Confidence | Date | Source | Source check |
|---|---|---|---|---|---|---|
| AGI Timelines | Short | 50% by 2028 | high | Oct 2023 | Dwarkesh Podcast (Oct 2023) | |
| P(doom) | Wide range | 5-50% | low | 2011 | Q&A with Shane Legg on risks from AI (LessWrong) | |
| How Hard Is Alignment? | Solvable via deliberation | Function of model capability | medium | Oct 2023 | Dwarkesh Podcast (Oct 2023) |
Organization Roles2
Co-Founder & Chief AGI Scientist; Lead, AGI Safety Council
2010 – present
Education
Dalle Molle Institute for Artificial Intelligence Research