Ethan Mollick Questions Cumulative AI Safety Risks
Wharton professor observes that incremental safety compromises may form an unpredictable system.
TLDR
Ethan Mollick, associate professor at Wharton and author of Co-Intelligence, posted that society now faces many small decisions trading modest reductions in AI safety for clear gains in utility. Each choice looks sensible on its own, yet together they build a complex system whose overall risks remain poorly understood. He states it is unclear whether those risks compound. The post records his assessment of current AI development trends.
Combined views
19.5K
1 Source, first seen 29d ago