Into AI Safety Sobering Up on AI Progress w/ Dr. Sean McGregor
8 snips
Dec 29, 2025 
Dr. Sean McGregor, a machine learning safety researcher and founder of several initiatives like the AI Incident Database, delves into the complexities of AI evaluation. He critiques the flaws in current benchmarking practices, emphasizing their vulnerability to training-data leakage and real-world misalignment. Sean introduces BenchRisk, a new framework aimed at improving benchmark trustworthiness. He also discusses the founding of AVERI, a nonprofit focused on frontier model auditing to ensure responsible AI deployment and navigate the tension between market and regulatory safety.
AI Snips
Chapters
Transcript
Episode notes
Measurement Broke While Models Improved
- Decades of ML optimization have eroded our ability to measure models reliably.
- Sean McGregor warns this destroys scientific understanding of systems' capabilities.
Accidental Move Led To Better Opportunities
- Sean moved to Orange County for an internship that was cancelled on his first day.
- He scrambled, found consulting and better-fit work, and says the setback helped his career.
If An Internship Fails, Hustle Strategically
- When plans collapse, network widely and take small gigs to stay afloat and visible.
- Sean recommends leaning on community integrity to find roles that fit your strengths.