It's All About That Pace
Summary
Hosts Jason and Jeff Jarvis analyze the AI safety debate, arguing that "doomer" narratives and regulatory capture by tech leaders distract from real issues like environmental impact and labor. They critique the recent push for a global AI slowdown as performative, while discussing Apple’s new Watch features that raise legal questions about conversation recording. The episode also covers China’s strategic use of open models, the potential for swapping AI models in iOS, and the broader power dynamics shaping the future of AI governance.
Topics discussed
Intro: AI safety inflection point and Dario Amodei's call for slowdown
Critique of AI leadership and media coverage of doomerism
The 'Test Rail' acronym and messianic dreams of tech leaders
Distraction from real AI issues: environment, labor, and education
Regulatory capture and the need for radical transparency in AI
AI in cybersecurity: finding vulnerabilities vs. plugging holes
Analysis of Anthropic's slowdown proposal and performative safety
Mark Zuckerberg's pushback and the lack of foolproof guardrails
Open source ethics and the incentives of AI company leaders
Skepticism toward extinction risk narratives and media balance
Reid Hoffman's proposal for free AI agents for every citizen
Community power over data centers and AI regulation
Geopolitics: US vs. China and the value of open models
The need for diverse stakeholders in AI governance decisions
Sponsors: Notion and Wired's Uncanny Valley podcast
Apple iOS 27: Swappable LLMs under Siri and DMA compliance
Apple Watch memory feature and the history of recording technology
Meta's camera-less smart glasses and privacy concerns
Humanoid robots in customer production lines and OSHA standards
AI solving 87-year-old math conjecture and the value of new knowledge
News: OpenAI hardware, Gemini 3.8, and Anthropic-Cloudflare merger
Book promotion, upcoming events, and show credits
Sponsors: Dropbox Working Smarter and Wired Uncanny Valley
Listen ad-free on Castria