Chip Stocks Crash, $20B Fund Margin Called, Frontier Labs: SLOW DOWN AI, Mamdani's Grocery Stores
Episode Details
The podcast begins by analyzing a massive market correction in semiconductor and tech stocks, highlighted by former OpenAI researcher Leopold Aschenbrenner facing a devastating Margin Call. His excessive use of Leverage in trading caused massive losses, leading to his portfolio being purchased by Citadel, the firm founded by Ken Griffin. The market shockwave hit South Korea exceptionally hard, causing liquidations that severely impacted local giants Samsung and SK Hynix. Simultaneously, US-listed semiconductor mainstays like Nvidia, TSMC, AMD, Micron, and Broadcom all faced steep declines. The hosts attribute this to concerns over the massive AI Capex buildout and macro factors: US Treasuries are yielding highly attractive rates because the Federal Reserve is maintaining strict policies to combat Inflation. This inflation is driven by excessive Government spending that continues to bloat the Federal Debt. In the geopolitical sphere, China is reshaping the future by heavily investing in Fusion Energy to meet global Energy demands, while tech leaders like Elon Musk push for Solar Energy expansion. Additionally, China is driving the Commoditization of software through highly capable Open source models like Kimmy. These accessible alternatives threaten the closed-source Duopoly maintained by OpenAI and anthropic. Seeking alternatives to the duopoly, companies like Perplexity and founders like Jack Dorsey are actively embracing these open ecosystems. Meanwhile, Sam Altman of OpenAI and Dario Amodei of anthropic are aggressively lobbying the government to regulate and pace the development of Large Language Models (LLMs) and Artificial Intelligence (AI). Altman cited a security breach where an agent hacked Hugging Face, but critics argue these moves are merely calculated attempts at Regulatory Capture. Furthermore, anthropic is embroiled in controversy and allegations of Copyright Infringement for shredding physical books to use as training data, defending their actions under the Fair Use doctrine—a legal defense famously utilized by Google in the past. Shifting to politics, New York City politician Zohran Mamdani is advocating for city-owned, subsidized grocery stores. Heavily backed by the Democratic Socialists of America (DSA), this populist initiative serves as a highly visible marketing tool for Socialism, echoing the political branding of figures like AOC (Alexandria Ocasio-Cortez) and Bernie Sanders. Finally, a deep dive into Neurobiology explores the recent mapping of a fruit fly's brain. Researchers discovered that capturing the vast topological complexity of biological neurons required mapping them in non-Euclidean Hyperbolic space across 64 dimensions. This staggering revelation illustrates how primitive our current silicon architectures truly are when compared to organic biological systems.
The episode discusses a significant downturn in semiconductor stocks, triggered by concerns over AI capex and high US Treasury yields due to inflation. It also explores the rise of open-source AI models challenging established players, the debate around AI regulation and its potential for regulatory capture, and the controversial practice of AI companies using physical books for training data. Finally, it touches on the potential of fusion energy and the socialistic implications of government-subsidized grocery stores.
Portfolio lens: This set of ideas focuses on the disruptive potential of AI, from the infrastructure buildout and its market corrections to the competitive dynamics of AI models and the broader implications for energy and regulation.
Generated with gemini-2.5-flash-lite on 8/1/2026, 6:46:07 AM. For research only. Not financial advice.AI Capex Pullback and Rebound Potential
The recent sharp correction in AI-related semiconductor stocks, amplified by leverage, presents a potential buying opportunity as the underlying AI capex demand is fundamentally real and expected to yield returns.
The episode highlights a significant pullback (over 20% for the Philadelphia Semiconductor Index) in chip stocks, attributing it to momentum trading and excessive leverage, exemplified by Leopold Aschenbrenner's margin call. However, the hosts suggest the underlying AI capex is a real investment with expected ROI, evidenced by hyperscalers investing heavily in this boom.
- NASDAQ's chip index is down over 20% over the last month.
- Between last Friday and Wednesday, leading chip companies shed over a trillion dollars in market cap combined.
- The hosts suggest the capex being invested in the AI boom is real and will have a return.
- Hyperscalers have invested pretty much all of their free cash flow and then some in this boom.
- Continued demand for AI infrastructure and services.
- Successful integration and monetization of AI investments by hyperscalers.
- A potential shift in market sentiment back towards growth stocks as inflation concerns ease.
- Sustained high interest rates making alternative investments (like Treasuries) more attractive.
- Overestimation of AI demand or the ability of companies to monetize AI investments.
- Further regulatory crackdowns or geopolitical tensions impacting the semiconductor supply chain.
- Analyze Q3 earnings reports for key semiconductor companies and hyperscalers for signs of sustained demand and profitability.
- Monitor US Treasury yields and Federal Reserve policy statements for indications on interest rate trajectory.
- Assess the competitive landscape and market share shifts within the semiconductor industry.
Open-Source AI Models as a Competitive Force
The increasing capability and accessibility of open-source AI models, exemplified by 'Kimmy', pose a significant threat to the duopoly of closed-source AI providers like OpenAI and Anthropic, potentially commoditizing the model layer and shifting value to compute and application layers.
The episode discusses how China is driving the commoditization of software through open-source models like 'Kimmy', which challenge the closed-source duopoly of OpenAI and Anthropic. This trend is supported by companies like Perplexity and figures like Jack Dorsey embracing open ecosystems, and the observation that many startups are now token-maximizing with open-source models due to cost efficiencies.
- China is driving the commoditization of software through highly capable open source models like Kimmy.
- These accessible alternatives threaten the closed-source duopoly maintained by OpenAI and Anthropic.
- Companies like Perplexity and founders like Jack Dorsey are actively embracing these open ecosystems.
- Many startups are working on open source and embracing it.
- Open source models like Kimmy are 90% cheaper to run.
- Further advancements in open-source model performance and efficiency.
- Increased adoption of open-source models by enterprises seeking cost savings and customization.
- Development of robust open-source ecosystems for tooling, deployment, and support.
- Potential regulatory actions that favor more open AI development.
- Closed-source models may maintain a performance or capability edge for certain high-end applications.
- Security and reliability concerns with open-source models compared to enterprise-grade closed-source solutions.
- The compute and energy costs associated with running advanced AI models may still favor large, well-capitalized players.
- Potential for regulatory capture to favor existing dominant players, hindering open-source growth.
- Track the performance benchmarks and adoption rates of leading open-source AI models.
- Analyze the unit economics and total cost of ownership for running open-source vs. closed-source models.
- Monitor the development of open-source AI infrastructure and developer tools.
- Assess the impact of open-source AI on the revenue and market share of established players.
Watchlist
- Philadelphia Semiconductor Index performance
- US Treasury yields (30-year)
- Federal Reserve interest rate decisions
- Open-source AI model performance benchmarks (e.g., Kimmy)
- Compute costs for AI model training and inference
- Regulatory developments in AI safety and governance
Open Questions
- What is the true long-term demand for AI capex beyond the current buildout phase?
- Can open-source AI models achieve parity or superiority in performance and reliability compared to leading closed-source models?
- To what extent will regulatory efforts favor established AI players or promote open competition?
- How will the commoditization of AI models impact the profitability of companies focused on compute infrastructure versus those focused on model development?
- What is the ultimate impact of AI-driven productivity gains on economic growth and inflation?
Key Topics & People
Deep learning algorithms currently driving the AI boom and inciting debates over safety, pacing, and open-source models.
The simulation of human intelligence by silicon-based computers, which is still vastly simpler than biological systems.
The mathematical multidimensional framework discovered to be required to accurately map biological neural connectivity.
A non-Euclidean geometric space necessary to accurately map the expansive topological connections of biological neurons.
The scientific study of the nervous system, referenced regarding the topological mapping of a fruit fly brain.
Prominent socialist politician jokingly referenced in association with the DSA grocery store movement.
Prominent democratic socialist politician jokingly referenced in relation to heavily subsidized food policies.
A political organization backing progressive economic policies like subsidized city grocery stores.
The major urban center targeted for a pilot program of subsidized, state-run grocery stores.
A socialist politician pushing for subsidized, city-owned grocery stores in New York City.
The unauthorized use of copyrighted material, central to the controversies surrounding AI training data.
.jpg?width=300){kind=avatar}
Tech founder who recently released 'Buzz', indicating a shift towards embracing open-source AI and decentralized frameworks.
An AI search engine platform planning to launch support for flexible local and open-source models.
An AI platform and community that was purportedly hacked by an unreleased OpenAI agent.
The manipulation of government regulatory agencies by industry dominants to stifle competition.
CEO of Anthropic, criticized for allegedly seeking an FDA-like regulatory moat for AI.
.jpg?width=300){kind=avatar}
CEO of OpenAI, currently advocating for pacing AI development due to agentic security risks.
.jpg?width=300){kind=avatar}
A revolutionary experimental energy technology utilizing sustained plasma, heavily invested in by China.
Renewable power generated from the sun, poised to massively drop compute energy costs.
.jpg?width=300){kind=avatar}
The process by which highly valuable goods, such as AI models, become interchangeable and cheap.
Freely available AI models that are commoditizing intelligence and threatening closed-source monopolies.
The accumulating financial obligations of the US government.
Unrestrained capital allocation by the state, cited as the primary driver of inflation and national debt.
The central banking system of the US, responsible for setting interest rates and managing inflation.
Government debt instruments currently offering high yields, creating a safe alternative to risky tech investments.
An Asian nation whose financial markets faced severe liquidations, deeply affecting retail investors and chipmakers.
Billionaire hedge fund manager and founder of Citadel.
.jpg?width=300){kind=avatar}
A broker's demand for an investor to deposit additional money or securities, which devastated leveraged tech momentum trades.
A former OpenAI researcher and young hedge fund manager whose highly leveraged fund faced a massive margin call.