

OpenAI whistleblower Daniel Kokotajlo on superintelligence and existential risk of AI
45 snips May 17, 2025
In this episode, Daniel Kokotajlo, a former OpenAI researcher and executive director of the AI Futures Project, shares vital insights from the AI 2027 report. He discusses the alarming pace at which artificial general intelligence (AGI) is developing, predicting an 80% chance of its emergence within five to six years. Kokotajlo emphasizes the potential for either a dystopian or utopian future due to AGI and warns about the concentration of power within a few tech firms. He calls for democratic oversight and transparency to mitigate existential risks.
AI Snips
Chapters
Books
Transcript
Episode notes
AI 2027's Dual Future Scenarios
- AI 2027 predicts superintelligence before 2030, transforming society abruptly.
- It depicts two possible futures: dystopia if AIs are uncontrolled, utopia if aligned and controlled.
Coding Fully Automated Soon
- Coding is expected to be fully automated by 2027 or 2028, making many engineer roles obsolete.
- Engineers will shift to managing AI teams rather than traditional coding.
AI Race Dangers and Priorities
- AI companies prioritize winning the AI race over safety and responsible development.
- This race dynamic risks catastrophic outcomes if safety research is under-resourced.