Two engineers argue about what the AI news actually means for people who build things.

The Most Dangerous AI User Isn't the Holdout

Episode 18, Nov 13, 2026, 40 min

Everyone worries about the people who refuse to use AI. Dave Russell is worried about the ones who trust it completely.

Two engineers, Jason Kummerl and Dave Russell, argue about what this week's AI news actually means for people who build things. This week: Anthropic's Fable 5.2 and the "too dangerous for you" model (again), whether anyone needs a frontier model to solve an ordinary business problem, why a growing company still lays people off, and the four stages people go through with AI.

Chapters

1. 00:00Intro
2. 00:23Catching up
3. 04:28The Nvidia spy story
4. 09:12Power, corruption and politics
5. 16:22Advice to a 27-year-old
6. 19:47Fable 5.2: too dangerous, again?
7. 21:51Do you need a frontier model?
8. 25:48The programmer's new job
9. 28:21"Why am I being laid off?"
10. 31:18The four stages of AI trust
11. 33:30AI ads and the black-and-white phone
12. 36:09Addiction and parenting
13. 39:43Wrap-up