Hard Fork · Podcast summary
Meta Shifts the Blame + Do Data Center Bans Work? + The Final HatGPT
Princeton computer scientist Arvind Narayanan returns to the show to discuss Meta's record-breaking settlement, the efficacy of data center bans, and the final HatGPT.
Listen on the show ↗Key takeaways
- ⚖️Meta has settled a multi-state lawsuit for $17.1 billion over child safety violations, the largest settlement in the company's history.02:38
- 🤝Meta's settlement includes a clause that increases the payout if TikTok and YouTube agree to similar penalties, which Casey describes as a "weird hostage situation."06:50
- ⏰The settlement includes a default two-hour daily time limit on Facebook and Instagram, a default block on app usage between midnight and 6 AM, and muted push notifications between 8 AM and 3 PM.09:00
- 🚫Princeton computer scientist Arvind Narayanan argues that data center bans will not meaningfully slow AI progress, estimating a one-year moratorium would only delay AI efficiency by 5 to 10 hours.25:32
- 📈Narayanan suggests that efficiency gains in AI software and hardware are an order of magnitude greater than capacity gains from new data center construction.32:02
- 🍎The Apple Vision Pro is being used in surgery, with one study showing 19% shorter operating times and 100% functional success for tear duct procedures.48:07
Chapters
- 3:19Meta's Record-Breaking Settlement
- 8:29The "Hostage Situation" and Industry Leadership
- 40:30Data Center Bans: A Futile Attempt to Slow AI?
- 1:17:06The Final HatGPT
The digest
This week on Hard Fork, Kevin Roose and Casey Newton delve into Meta's massive $17.1 billion settlement over child safety violations, the surprising ineffectiveness of data center bans in slowing AI progress, and a bittersweet farewell to the beloved HatGPT segment.
Meta's Historic Child Safety Settlement
Kevin kicks off the show with the breaking news of Meta's $17.1 billion settlement, the largest in the company's history, stemming from a multi-state lawsuit involving 47 states, Washington D.C., and U.S. territories. Casey, who has been following the case since its filing in 2023, notes that while he initially found the case less than compelling, the release of an unredacted version of the lawsuit revealed damning evidence against Meta.
"This company is in trouble," Casey recalls thinking, highlighting that Meta absolutely knew millions of kids under 13 were using their platforms without parental consent, a direct violation of existing privacy laws. The settlement addresses two major issues: Meta's collection of data from children under 13 without parental permission and the company's use of addictive features designed to maximize user engagement.
As part of the settlement, Meta has agreed to implement several changes: a default two-hour daily time limit across Facebook and Instagram, a default block on app usage between midnight and 6 AM, and muting push notifications between 8 AM and 3 PM. Kevin points out that while the financial penalty is significant, it's not crippling for Meta, especially considering the company reportedly expected to make $10 billion this year from ads related to scams.
Keep reading in The Digest
The full 4-minute article, 3 more sections, and every new episode from Hard Fork and the creators you follow.
- The "Hostage Situation" and Industry Leadership
- Data Center Bans: A Futile Attempt to Slow AI?
- The Final HatGPT
More from Hard Fork
The Ezra Klein Show: Jensen Huang Thinks A.I. Alarmism Has Gone Too Far
Jensen Huang, NVIDIA CEO, discusses AI's industrial revolution, job impacts, safety engineering, and geopolitical dynamics.
The Ezra Klein Show: The A.I. Revolt Is Here
Jasmine Sun discusses her reporting on the backlash against AI data centers, exploring the economic, social, and political dimensions of this growing opposition.
OpenAI’s Two-Week Pause + Jill Lepore on the Threat of the “Artificial State” + Train of Thought
OpenAI paused training on its new model, Astra, after a security incident, and historian Jill Lepore discusses her new book, "The Artificial State," which explores the threat of AI to democracy.