Back to All-In Podcast

The Fight Over Open Source AI, Anthropic's $1.5B Payout, NYC Socialists: Evictions = Violence?

All-In PodcastJuly 24, 20261h 33m
Topics94
Chinese Open Source AI and US Policy Debate0:00David Sacks' Position on Open Source2:00Criticism of Anthropic's Regulatory Strategy4:30Enterprise Adoption and Distillation Explanation7:30Commoditization and Value Shift9:00Economic Impact of Banning Open Source11:30Geopolitical Escalation Concerns14:30Distillation as Standard Industry Practice15:30Hypocrisy on Training Data17:00Free Speech and Enforcement Challenges20:00Historical Internet Analogy21:30Economic Diffusion Benefits23:00Market Value Capture Evaporation24:00AI Value Capture Speed and Market Dynamics24:24Anthropic Revenue Tracking and Headwinds25:02Open Source Model Adoption25:31Technology Market Compression26:32Intelligence as Ubiquitous Utility27:30Browser Analogy for AI Value Capture28:32Open Source vs Closed Source Market Winners29:00Kimi K3 Cost Analysis29:31GPT-6.0 Development30:32Revenue Growth Comparison31:02OpenAI Revenue Acceleration31:32Frontier Model Duopoly Performance32:30Government Protection Strategy33:00Application Layer Strategy33:31Non-Released Model Strategy34:00Open Source Adoption by Startups34:30Margin Compression Risk35:00Task Distribution Across Models36:01Terminal Value Mathematical Mistake36:31Customer Migration Examples37:01Model Performance Comparison37:30Enterprise Requirements Beyond Models39:00Open Source Ecosystem Impact39:30Derivative Work Examples40:32Open Source Process41:01IP Theft Accusation Impact41:30US Knowledge Economy Value42:02Molecule Conversion Technology42:32Knowledge Economy Compression43:02China Manufacturing and Energy Advantage43:33China Long Game Strategy44:01Core Dependency Creation44:30Open Source Market Differentiation45:01Open Source Implementation Work Reduction45:30Investor Disclosure and Bias46:03China Hawk Hypocrisy46:30Open Source Value Proposition47:01Enterprise Open Source Strategy47:31Anthropic Copyright Settlement48:00Content Provider Rights and Collective Action49:30Training Data Value and Market Dynamics50:00Fair Use vs. Piracy Distinction51:00Hypocrisy in AI Training Positions52:00National Security vs. IP Theft Arguments54:00Copyright Boundaries and Knowledge Diffusion55:30Direct Competition and Copyright Infringement1:00:00Proposed Industry Solutions1:01:00Fatal Mistake Analysis1:01:30Content Creator Strategies1:03:00Regulatory Capture Risks1:04:30Publishing Schedule and Future Books1:06:00Google and Tesla Earnings Results1:07:00SpaceX IPO Performance1:08:30Capital Expenditure Analysis1:09:00Tech Companies Shifting Capital Allocation Strategy1:11:51Massive Capex Investment in AI Infrastructure1:12:32Google's Multiple Business Lines and Anthropic Investment1:13:00GCP's Enterprise Data Advantage1:13:31Infrastructure as Backup Value Proposition1:14:01Apple Web Services Discussion1:15:01Five Nines Reliability Requirements1:15:30Apple's Capital Return vs Investment Opportunity1:16:00Socialism Corner Introduction1:17:00Zohran Mamdani's Rental Policies1:17:31Evictions Characterized as Violence1:18:00John Quincy Adams Property Rights Framework1:19:00Historical Context of Private Property Rights1:19:31Anarchy and Tyranny Connection1:21:00Moralistic Framing of Property Seizure1:21:31Impact on Building Maintenance and Neighbors1:23:00Luxury Beliefs and Working Class Impact1:24:31Economic Solution to Housing Crisis1:26:00Austin Permitting Reform Evidence1:27:00Property Rights Analogy1:27:31Argentina Rent Control Outcome1:28:30Landlord Response to Restrictions1:29:00Credit Check and Background Check Bans1:29:31Market Response to Restrictions1:30:00Tenant Payment Incentives1:30:30Ghost Apartments Phenomenon1:31:30Landlord Strategic Options1:32:00
In a Nutshell

US policymakers are debating whether to ban Chinese open-source AI models amid claims of industrial-scale distillation by Moonshot AI's Kimi K3, with David Sacks arguing this would be a strategic mistake that forces American companies to pay 50-100x more for AI while China gains the manufacturing advantage. Anthropic settled a $1.5B copyright lawsuit for pirating books to train Claude and is simultaneously pushing for government protection against Chinese distillation, creating hypocrisy since they argue training on copyrighted material is fair use but object when others apply the same logic to their outputs. NYC socialists are advancing rent control and eviction bans that treat property rights as violence, mirroring the same pattern of regulatory capture that undermines the knowledge economy's foundation.

AI-Generated Notes

These notes were generated by AI and may contain inaccuracies.

Kimmy K3, released by China's Moonshot AI, is the leading global story. The open source model performs on par with Opus 4.8 and GPT 5.6, not 6 or 12 months behind. It is approximately 50% cheaper, creating panic similar to the Deep Seek moment in early 2025.

The White House became involved after Michael Katzio stated Moonshot AI distilled Anthropic's model for K3 development. Axios reported Monday that the White House was considering banning Chinese open source models. Three weeks earlier, this was presented as a hypothetical. On Wednesday, Wired reported Howard Letic does not want to ban Chinese models, indicating internal differences within the Trump administration.

Polymarket gives a 45% chance the US government bans an open source model in 2026, up from 22% a few days earlier.

David Sacks stated there is no White House decision to ban open source models. The administration is having an ongoing conversation about Chinese distillation but no decision has been made. He noted President Trump listens to multiple voices and has shown impeccable instincts on tech issues, supporting lighter regulation and more open approaches.

Sacks argued it would be a tragic mistake for government to act against the open source ecosystem, as it would hurt America's position in the AI race. He stated American companies must be able to use everything in the public domain, because the rest of the world will use those things.

Sign in to read the full notes

Get access to AI-generated notes, topic timestamps, and more.