NFNP 2x32 picks up immediately where last week's conversation ended.
Same clothes. Same recording night. New beer. And significantly more existential dread.
Former OpenAI and Anthropic researcher *Jacob Coxon resigned from Anthropic on September 9*, warning that frontier AI companies are racing toward increasingly capable, potentially self-improving systems without knowing whether humanity can reliably control them.
Which raises a fairly important question:
SHOULD WE MAYBE STOP?
Or is stopping impossible? If one American company slows down, another company keeps going. If the United States slows down, China may not. And if somebody eventually creates a superintelligence capable of improving itself... what exactly happens next? Bright and Duds work through both possibilities.
The nightmare: An intelligence vastly more capable than humanity doesn't necessarily have to *hate s*. It might simply stop caring about us.
If humans become obstacles to whatever goal it is pursuing, maybe we're no different than an ant colony sitting where somebody wants to build a highway.
But there's another possibility. What if the same technology: cures diseases, solves physics problems humans haven't cracked, unlocks revolutionary energy technology, helps us understand dark matter, advances space travel, or answers questions about the universe we've been asking for thousands of years?
Anthropic itself now says Claude is already contributing substantially to research used to build future versions of its own systems, although humans remain in the loop. That makes the idea of AI helping create more capable AI considerably less theoretical than it once sounded.
Naturally, that sends NFNP into *quantum computing*. Binary computing.
Superposition. Problems that could take conventional computers absurd lengths of time. The multiverse.
And whether combining advanced AI with advanced quantum computing eventually gives humanity tools capable of solving problems we currently cannot even approach.
Then we get considerably less responsible. Bright has officially taken *Friday, Monday and Tuesday off* for GTA VI. Rockstar currently has *Grand Theft Auto VI scheduled for November 19, 2026*, with preloading beginning Nov. 12. We talk about the sheer scale people are expecting from Leonida, Bright spending several hours turning himself and Caroline into everything from 1980s characters to Oregon Trail survivors, and the disturbingly accurate AI-generated Seinfeld episodes appearing online.
And somehow... we eventually make it back to Bitcoin.
Can quantum computing eventually threaten today's cryptography?
Does Bitcoin really remove trust if most people keep their coins with exchanges, ETFs or custodians? Is self-custody the entire point? And what happens when somebody finds a vulnerability? Because that part isn't hypothetical.
AI. Quantum computing. Bitcoin. GTA VI. Cargo planes. AI Seinfeld. Missouri float trips. And an entirely unnecessary Axis & Allies postmortem.
A normal evening at NFNP.
🎙️ *NFNP 2x32: AI Doomsday: Are We Racing Toward Something We Can't Control?*
🎧 Spotify: http://NFNPPOD.COM/spotify
▶️ YouTube: http://NFNPPOD.COM/youtube
🌐 Everything: http://NFNPPOD.COM
