
Sep 28, 2026 · 1h 19m
Cal Newport shifts AI debate toward harms already here
Kids Should *Not* Use Chatbots! (You Should Be Wary, Too…)
The episode argues that emotional dependence, misinformation, and child exposure to chatbots demand more immediate scrutiny than hypothetical superintelligence.
- 1Chatbots should be treated as fluent tools, not people, to reduce emotional attachment and misplaced trust.
- 2Enterprise demand and coding tools have helped Anthropic build a business distinct from OpenAI’s consumer-centered model.
- 3Sustainable productivity requires limiting overload while confronting career setbacks without mistaking credentials for genuine interest.
Don't miss
Newport’s clearest intervention is his argument that adults should address chatbots as tools, not conversational minds, to resist emotional dependence and confident misinformation.
The brief
Cal Newport opens by challenging the fixation on hypothetical superintelligence, arguing that chatbot misinformation, emotional dependence, and child exposure are concrete harms demanding attention now.
His practical rule is simple: use chatbots as tools, with concise task-oriented prompts, rather than conversational partners whose fluent replies invite misplaced trust.
The discussion then turns to Anthropic’s enterprise strategy and Claude Code, before widening into the social costs of dumb phones and the case for sustainable knowledge work.
Newport also addresses a fired web developer, arguing that deliberate skill-building can fail to produce career success and that credentials should not substitute for genuine interest.
The episode closes with a response to Ezra Klein’s AI essay and a candid tour of studio changes, from lighting and headphones to riskier creative choices.
Featuring
Listen to the full episode and explore every guest, topic, and moment on PodLume.

Eliezer Yudkowsky
Nick Bostrom
Dario Amodei
Anthropic
Samuel Harris Altman
Mark Elliot Zuckerberg