Source note Episode guide Original audio Topics: Technology

Software Stocks Implode, Claude’s Hit List, State of the Union Reactions, Trump’s Tariff Pivot

Summary

This All-In episode treats AI as both a market narrative and an operating technology. The hosts connect Anthropic announcements, coding agents, and a viral 2028 crisis scenario to software and financial-stock repricing, then debate whether AI destroys SaaS cash flows or expands demand through Jevons Paradox In AI. Concrete internal-agent examples support a new SaaS Cash Flow Survival Risk frame while also showing that adoption still depends on training, workflow design, review, tokens, chips, land, and power.

The second half moves from Data Center Cost Shifting and local infrastructure opposition to U.S. political polarization, the Supreme Court’s rejection of the IEPA tariff route, uncertain refunds, alternative tariff statutes, and a Science Corner on Yamanaka-factor retinal gene therapy. Market, employment, project-count, tariff, and clinical claims remain episode-attributed rather than independently verified.

Key Claims

  • Chamath Palihapitiya says AI has shifted the software-investment question from when cash flows weaken to whether some cash flows survive, while David Sacks argues that lower software-production costs may increase total demand rather than simply eliminate engineers.
  • The hosts say Anthropic-related product announcements coincided with sharp selloffs in legal, security, and legacy-software names, but they do not establish that each announcement caused durable business impairment.
  • A viral fictional 2028 AI crisis scenario is used to show how plausible but weakly evidenced narratives can move markets when investors lack real-time evidence about AI’s macroeconomic effects.
  • SaaS loses some of its former annuity-like predictability when AI can change seat counts, pricing, growth, product boundaries, and customers’ ability to build internal tools.
  • Jason Calacanis says his firm trained most employees to build agents for advertiser research, communications, staff summaries, clips, and thumbnails, increasing output without adding headcount.
  • Sacks cites software-engineer postings, company formation, and an Anthropic job listing as evidence for Jevons Paradox In AI, while the episode also acknowledges that change management, enterprise inertia, tokens, chips, land, power, and energy can slow adoption.
  • The data-center segment argues that ratepayer protection requires large technology companies to fund power for their projects, while Chamath warns that utility capital-spending incentives can still raise bills even if one load is nominally self-funded.
  • David Friedberg presents controllable Yamanaka-factor delivery to retinal cells as a possible early clinical route for cellular rejuvenation, but the Phase 1 discussion does not establish safety, efficacy, or general anti-aging benefit.
  • The tariff segment says the Supreme Court rejected the IEPA route while Section 122, Section 301, and Section 338 remained possible alternatives, leaving Tariff Refund Uncertainty and future tariff levels unresolved.

Key Quotes

“if these cash flows will survive” - Chamath’s frame for the software-market valuation shift.

“science fiction versus science fiction” - the episode’s warning about confident AI macro narratives without enough current evidence.

“Jevons paradox” - Sacks’s countercase that cheaper software production can expand total software demand.

Connections

Contradictions

  • No settled contradiction is adopted. The episode deliberately preserves competing AI labor and demand narratives rather than resolving them.
  • The software selloffs, hiring figures, productivity gains, token forecasts, canceled-project counts, revenue-per-gigawatt estimate, tariff totals, importer counts, and trial status are source-scoped claims that need independent evidence before use as current facts.
  • The tariff discussion agrees with The Supreme Court struck down a bunch of Trump’s tariffs. Now what? that invalidating IEPA did not eliminate other tariff routes, but the hosts’ preferred congressional response and refund treatment are political judgments.
  • The cellular-reprogramming segment strengthens the clinical-delivery case while leaving the cancer, dosing, reversibility, tissue-targeting, and human-outcome boundaries in Cellular Reprogramming Control Problem unresolved.