#218: Anthropic IPO, Trump AI Executive Order, Rising AI Costs & OpenAI Merges Codex Into ChatGPT (1h 25m)

← Back to week overview

• Release date: 2026-06-09
• Listen on Spotify: Open episode
• Episode description:

  Anthropic filed for an IPO and published a paper warning that recursive self-improvement may arrive faster than anyone is ready for. Paul and Mike break down both, then cover Trump's AI executive order, government stakes in AI labs, and the corporate scramble to control AI token costs. Rapid fire: Apple WWDC previews, OpenAI's Codex-ChatGPT merger, Brockman's super PAC, AI rolling up the accounting industry, Stanford law professors losing to AI 75% of the time, and product updates from Google, Microsoft, Meta, and Anthropic. Show Notes: Access the show notes and show links here AI-Pulse Survey: Fill out this week’s AI-Pulse Survey here. Timestamps: 00:00:00 — Intro 00:05:53 — Anthropic IPO & Talks Recursive Self-Improvement 00:25:52 — Trump's AI Executive Order & Government Stakes in AI Labs 00:37:34 — The Soaring Cost of Intelligence, Part 2 00:57:34 — Apple WWDC 01:01:36 — OpenAI Is Merging Codex and ChatGPT 01:06:19 — OpenAI Distances Itself from Brockman's Super PAC 01:08:55 — AI Roll-Up Targets the Accounting Industry 01:12:23 — AI in Higher Education 01:16:29 — AI Use Case Spotlight 01:20:29 — AI Product and Funding Updates This episode is brought to you by AI Academy by SmarterX. AI Academy is your gateway to personalized AI learning for professionals and teams. Discover our new on-demand courses, live classes, certifications, and a smarter way to master AI. Learn more here. Visit our website Receive our weekly newsletter Join our community: Slack Community LinkedIn Twitter Instagram Facebook YouTube Looking for content and resources? Register for a free webinar Come to our next Marketing AI Conference Enroll in our AI Academy

Summary

• 🚀 Recursive Self-Improvement Accelerates: AI labs are already using models to build successors at unprecedented speed, with Anthropic reporting 52x code speedups and task lengths doubling every four months.
• 💼 100-Person AI Companies Emerge: Legacy firms face disruption as AI-native startups can match the output of 1,000-person organizations through agent orchestration and automation.
• 🏛️ Government Stakes and Oversight Rise: Trump’s executive order and bipartisan bills signal growing state involvement, including potential ownership stakes in frontier labs.
• 💰 Token Costs Force New Strategies: Companies are capping spend, routing to cheaper models, and rethinking pricing as AI usage burns through budgets faster than expected.
• 📚 AI Literacy Becomes Essential: Effective token use, model selection, and first-principles rethinking of workflows require widespread upskilling to avoid waste and unlock value.

Insights

• How will recursive self-improvement in AI research labs fundamentally reshape every knowledge-work profession within the next 18 months?
  • Time: 0:00 – 18:26
  • Answer: Anthropic’s internal data shows AI now writes over 80% of merged code and completes tasks that once took humans days, with task length doubling every four months. This ‘canary in the coal mine’ effect means the same productivity multipliers will soon hit law, HR, marketing, and other fields once capital flows in.
• What happens to company valuations and competition when AI-native startups can do the work of 1,000-person legacy firms with just 100 people?
  • Time: 18:26 – 25:49
  • Answer: Anthropic’s three scenarios highlight that even without full recursive self-improvement, 100-person AI-powered companies could already match or exceed the output of much larger organizations, forcing rapid restructuring of staffing and cost models across industries.
• Should governments take ownership stakes in frontier AI labs, and what risks does that create for long-term innovation and political swings?
• How can organizations practically control exploding AI token costs while still capturing the productivity gains that justify the spend?

Want the full treatment? Download expanded insights—timestamps, complete answers, and course materials from AI Weekly Insights.