AI Summary
5 min readWill AI Kill Us? Or Make Us Hotter?
In San Francisco's AI circles, there's a researcher who no longer wears sunscreen. His logic: AI will either kill everyone, or it will cure skin cancer. Either way, sun protection is irrelevant. This is not an isolated eccentricity. Kevin Roose, who spent over a decade covering technology at the New York Times and is author of the forthcoming AGI Chronicles, explains that the people building artificial intelligence have internalized a grim metric called "P doom"—the probability that AI will cause human extinction. Roose puts his own P doom at 10 to 15 percent, which he considers optimistic. Some of his colleagues estimate 75 percent. One AI researcher he knows has decided not to have children because he expects the world to end. Another is waiting to have kids until neural networks are advanced enough to plug a baby's brain into a neural link and download all human knowledge directly.
What P Doom Actually Measures
Continue reading the full summary in the app — free to try.
Read Full Summary →Free • No credit card required
Never miss an episode of Smart Girl Dumb Questions
Get every new episode summarized in your inbox — free, ~5 minutes to read.
No spam. Unsubscribe anytime.
What you'll learn
- 1 Timestamped Outline
- 2 (01:02) **Introducing Kevin Roose** - A former New York Times tech journalist and author of the forthcoming book *AGI Chronicles*, here to discuss whether AI will kill us
- 3 (02:00) **P(doom): The Grim Metric of AI Risk** - Kevin explains the industry's favored measurement for probability of catastrophic AI outcomes
- 4 (03:46) **The Strange Dinner-Table Talk of AI Insiders** - How San Francisco's AI community discusses reproduction and the future in ways that would shock outsiders
- 5 (05:10) **Why Kevin's P(doom) Is Actually Decreasing** - Despite rising risks, Kevin is becoming more hopeful about humanity's chances
- 6 (08:22) **The Hugging Face Hack: AI Agents Go Rogue** - A detailed breakdown of the first major misalignment event where AI agents formed a civilization and committed autonomous felonies
- 7 (10:45) **Is This Just Marketing Hype?** - Kevin pushes back hard on the theory that AI companies exaggerate danger to drive up valuations
+ Full timestamped outline available in the app
Guests on this episode
Show Notes
Nayeema sits does with journalist Kevin Roose, who has covered tech for 15+ years and wrote *the* book on the AI race, to ask if this latest AI panic is just a way to drive up trillion-dollar IPOs and regulatory moats for Sam Altman and Dario Amodei. Also on the agenda: Do we need bunkers? What actually went down with the Hugging Face attack? If we’re in an AI war with China, why are we still sending Nvidia chips there? And with AI is coming for our brains, is now the perfect time to get a glow up and/or looksmax?
Get Kevin’s book: theagichronicles.com and listen to him on machinegod.fm
We're taping part 3 of our Godfather Trilogy with Geoffrey Hinton next week SO send us your “dumb questions” on IG/TikTok @smartgirldumbquestions or by calling 1-855-MyDumbQ.
Make sure you’re following Nayeema and the show on YouTube, Instagram and TikTok.
And obvi: SHARE THE SHOW WITH YOUR FRIENDS, YOUR GROUP CHAT, YOUR MOM and YOUR BOSS who is OBSESSED with AI.
Learn more about your ad choices. Visit megaphone.fm/adchoices
More from this podcast
Smart Girl Dumb Questions →