Sep 8, 2026 · 1h 28m
GPT-6 Astra brings AI job disruption into view
#237: GPT-6 Astra, NYC Bans AI in Schools, Trump Goes All-In on Data Centers & Lawyers Under Pressure to Pass on AI Savings
The episode connects computer-using AI to the future of professional work, while examining the political, educational, legal, and economic responses already taking shape.
- 1GPT-6 Astra’s ability to complete multi-step workflows could accelerate AI’s impact on professional work and employment.
- 2Data-center politics, school restrictions, and copyright lawsuits show institutions struggling to govern AI’s expanding footprint.
- 3Law firms and educators face early pressure to redesign pricing, teaching, assessment, and AI literacy around new capabilities.
Don't miss
Mike Kaput describes using computer-using AI on a home computer to collect, expand, organize, and analyze replies to a LinkedIn post.
The brief
Paul Reitzer and Mike Kaput examine GPT-6 Astra’s ability to navigate complex, multi-step computer workflows, asking whether this release marks a turning point for professional work and public awareness of AI displacement.
The model’s demonstrations in financial modeling, software development, and other knowledge work sharpen the central tension: capability can improve efficiency while also making employment disruption harder to dismiss.
The conversation widens from software to politics as data-center opposition, New York City’s school moratorium, and copyright litigation reveal competing demands for growth, protection, access, and accountability.
Law firms face pressure to share AI-driven efficiency gains, while MIT debates how teaching and assessment should change when guided and unrestricted AI produce different outcomes.
The most concrete example comes from using computer-using AI to collect and analyze LinkedIn comments, showing how quickly a general model can turn a public discussion into structured feedback.
The episode closes on a forecast of a visceral public reaction when visible job displacement becomes impossible to ignore, alongside a call for constructive, logic-based AI debate.