Sheila Warren is a lawyer who has been at the cutting edge of technology governance throughout her career - from Crypto to AI. She compares tech billionaires rewriting their own histories to a TV show retconning its first season, and they are not doing a convincing job of it. I met Sheila at the Conference on World Affairs in April, where she shared some of her thoughts about friction that I found incredibly intriguing. She has worked across civic tech, crypto, and AI governance, and she now chairs the Advanced AI Society, which works on ways to verify what AI systems do. We talk about why the real AI risks are process failures and not Skynet. We also talk about what happens to responsibility when we start treating AI agents like people. Then we get to friction, and Sheila makes the case that the struggles we automate away, from tying shoelaces to fighting it out in a boardroom, are how we essential for us to learn and to stay human. I tried to convince her of the joys of adaptive cruise control, and I'm still not sure where that argument ends.
","author_name":"Alexis Halkovic"}