Big Technology Podcast · Alex Kantrowitz

Are We Screwed If AI Works? — With Andrew Ross Sorkin

·1 hr 5 min·4 clips

AI could create 25% unemployment without a market crash, Andrew Ross Sorkin says.

1. Big Technology Podcast centers this episode on whether AI working well could create economic damage instead of just growth. 2. Alex Kantrowitz hosts Andrew Ross Sorkin, CNBC and New York Times journalist and author of 1929, whose book on the 1929 crash frames the conversation. 3. The episode asks whether the bigger risk is not a market collapse, but AI producing 25% unemployment through massive productivity gains. 4. Sorkin points to $700 billion in AI capital expenditures and says last month software stocks lost $1 trillion in market cap. 5. The host argues that if AI can do 20% of the workforce, the economy could still get a production boom and create new demand. 6. Sorkin answers that the main question is who will have the money to buy the new goods and services. 7. The two talk through a painful transition period, comparing it with earlier technological revolutions that did not adapt cleanly. 8. Sorkin gives examples of research, model building, and paralegal work that Claude or ChatGPT may already be able to do. 9. He also says tokens are not free and asks how much cheaper AI will become than a human. 10. Software engineers become a major test case, because AI is already strong at code and yet software-developer job numbers on Indeed are still rising. 11. Sorkin says he still writes his own articles and wrote 1929 without AI, because the book started too early for those tools. 12. He predicts that people will likely co-author books and articles with AI, which raises the question of what the human role becomes. 13. The private business example is a contract that Sorkin put into ChatGPT, which redlined it and drafted a cover letter. 14. That example leads him to argue that small-business lawyers may lose routine work first, while higher-order work remains with larger firms. 15. Sorkin also describes cloud code building workflow tools for his media operation and connecting to services he would not have paid for otherwise. 16. The journalism discussion covers Narrative Science, ESPN, locker-room reporting, and the idea that AI can analyze a game but not replace human interviews. 17. The show keeps a conversational, back-and-forth interview style, with the host challenging Sorkin's assumptions and Sorkin revising or sharpening them in real time. 18. The tone is analytical and skeptical, with repeated references to prediction markets, an app, and the possibility that Sorkin could be wrong. 19. Listeners interested in AI, markets, software, and labor economics will get the most from this episode. 20. Listeners looking for a light tech chat or a quick product demo will probably skip it.

As heard by us

A sharp look at whether AI markets break by succeeding too fast.

Andrew Ross-Horkin turns the AI boom into a simple but uneasy question: what if the thing fails not because it is inflated, but because it works too well?

Read the full review in PlayNext →

Why you'd press play

When AI stops being hype and starts breaking the math, press play.

Read the full recommendation in PlayNext →

Listen to the show on

Apple Podcasts Spotify

Play clips from this episode More from Big Technology Podcast