Dystopia Now AI Safety Hits the Mainstream
Sep 24, 2026
Kate Willett and Emile Torres unpack the sudden mainstream focus on AI safety, from Silicon Valley’s effective altruist and transhumanist ideologies to doubts about imminent superintelligence. They examine cybersecurity failures, AI industry economics, government bailouts, media-driven doomerism, and the political dream of machine rule. The discussion also turns to documented labor and environmental harms, narrow AI, and building a grounded, pro-human alternative.
AI Snips
Chapters
Transcript
Episode notes
AI Safety Serves A Post-Human Utopia
- AI safety emerged from a worldview seeking aligned superintelligence, not merely harmless software.
- Its promised future includes mind uploading, cosmic colonization, and radically redesigned post-human successors.
Internet Training Does Not Create A Machine God
- Current models trained on internet data do not obviously approach godlike intelligence.
- Emile Torres notes that an alleged mathematical breakthrough may have reproduced material already available in researchers’ online work.
Aligned AI Can Still Mean Human Extinction
- In this ideology, aligned AI can still make biological humans obsolete and eventually eliminate them.
- The movement treats extinction as acceptable once post-human successors inherit civilization and resources.