Meta Shifts the Blame + Do Data Center Bans Work? + The Final HatGPT
Summary
Kevin Roose and Casey Newton discuss Meta’s $17 billion settlement over child safety, detailing mandated product changes like screen time limits and the company’s strategic push for competitors to adopt similar standards. They also analyze Meta’s internal restructuring efforts and shifting focus toward AI. Finally, Princeton professor Arvind Narayanan argues that local opposition to data centers will not significantly slow AI progress, as efficiency gains outweigh capacity constraints, suggesting better avenues for regulating AI deployment.
Topics discussed
Intro: Gemini Enterprise ad and hosts' wine spill anecdote
Meta's $17B child safety settlement details and context
Meta's new teen restrictions and the TikTok/YouTube leverage
Analysis of Meta's business strategy and workforce overhaul
Meta's shift to AI-native pods and harm reduction view
Intro to Arvind Narayanan on data center backlash
Why data center moratoriums won't slow AI progress
Compute scarcity, efficiency gains, and local opposition
Data center protests as collective bargaining by riot
Final episode of the 'At GPT' segment introduction
Vision Pro used for tear duct surgery procedure
GTA 6 leaks and Netflix exclusive trailer frustration
Amazon drone delivery fail and slow messaging apps
AI startup burnout and the 'meat proxy' phenomenon
Humanoid robots break Usain Bolt's 100m record
OpenAI's claim of being 80% to AGI and definition debates
Outro and sponsor credits
Listen ad-free on Castria