China, AI Immigration, Rare Earths & Chips, Tariffs, Markets | BG2 w/ Bill Gurley & Brad Gerstner

This episode unpacks the escalating strategic chess match between the US and China in AI, from data control and talent acquisition to critical resource dependencies like rare earths and chips, offering crucial insights for Crypto AI investors navigating geopolitical and technological shifts.

A 25-Year Reflection: Dot-Com Boom vs. AI Revolution

• Brad Gerstner kicks off the discussion by reflecting on his 25th HBS reunion, drawing parallels between the current AI boom and the dot-com era of 1999-2000.
• He recalls the overestimation of the internet's short-term impact due to limited high-speed internet access, contrasting it with the dramatic underestimation of its long-term transformative power.
• Brad highlights Amazon's volatile stock journey: peaking at $243 in 1998, dropping to $26 from a $150 start in 2000, yet achieving a 440x to 1,800x split-adjusted increase from its 2000 levels.
• "We thought we knew everything about the internet... And it turns out we overestimated what it was going to do in the short term... But what's probably even more surprising, Bill, is how dramatically we underestimated the long term," Brad notes, setting the stage for considering AI's trajectory.
• The NASDAQ, despite the bust, is up 5x from its 2000 peak and 10x from its trough, underscoring the long-term growth potential of transformative technologies.

Actionable Insight for Crypto AI Investors: The historical precedent of the internet boom suggests that while short-term AI market volatility is likely, the long-term disruptive potential, especially at the intersection of AI and crypto, could be vastly underestimated. Investors should maintain a long-term perspective, focusing on fundamental technological advancements rather than short-term hype cycles.

The Rapid Pace of AI and Emerging "Data Walls"

• Bill Gurley and Brad Gerstner discuss the relentless pace of AI development and a new, concerning trend: the emergence of "data walls."
• Bill points to recent instances where companies are attempting to restrict access to or use of data for AI training.
• Examples include Reddit suing Anthropic despite a deal, and Salesforce changing its terms of service to prevent training on CRM and Slack data, even for the customers who own that data.
• MCP (Multi-Cloud Platform) connector: A technology that allows AI to query data across different cloud services. Salesforce is using this but restricting training.
• Brad notes that OpenAI's Deep Research is now searching across GitHub, Google Docs, Gmail, etc., likely using MCP, highlighting the tension between open access and proprietary control.
• Brad draws a parallel to Zillow's past dependency on Google, recalling Bill's warning: "You can't build a brand in the underbelly of Google because over time they will have to take that territory back." This is now happening at "hyperspeed" in AI.
• OpenAI reportedly reached 400 billion annual searches 8 years faster than Google.
• Companies like Windsurf and Cursor are now building their own models to reduce dependency on platforms like OpenAI or Anthropic.

Strategic Implication for Crypto AI Researchers: The rise of "data walls" by centralized entities could significantly impact the development of decentralized AI. Researchers should explore and advocate for open data initiatives and technologies like zkML (Zero-Knowledge Machine Learning), which enables private verification of AI models and could be crucial for secure, decentralized data processing and model training without compromising user data ownership.

China's Competitive AI Strategy: "Let a Thousand Flowers Bloom"

• Bill Gurley introduces research (analyzed by ChatGPT 4.5) on China's strategy for fostering innovation, particularly in AI and related industries.
• Contrary to the perception of single, state-sponsored champions, China encourages massive competition, often seeding 500 to 1,000 startups in a target industry (e.g., over 500 EV startups).
• This "Darwinian competition" allows the best companies to emerge, fostering robustness and innovation. Bill, drawing on his experience with the Santa Fe Institute (a research center focused on complex adaptive systems), finds this approach sensible.
• Benefits of this approach include:
• Exposure to greater optionality and diverse approaches (e.g., Chinese solid-state LiDAR at $130/car vs. Waymo's at $5,000/car).
• Development of a more robust and competitive supply chain.
• Brad contrasts this with the Soviet model of inefficient state actors, noting China's success in producing globally competitive products.
• Government involvement in China is often at the provincial level, seeding these numerous startups, rather than direct federal picking of a few winners early on.

Actionable Insight for Crypto AI Investors: China's strategy of fostering intense domestic competition before global expansion could lead to highly resilient and innovative Chinese AI and hardware companies. Investors should monitor these emerging players, especially in areas like AI-specific chips and robotics, which could impact global supply chains for AI compute relevant to crypto.

China's De-Prioritization of Market Caps and Implications

• The discussion shifts to China's differing objectives regarding corporate success compared to the West.
• Bill highlights research suggesting China de-prioritizes the market capitalization of its successful companies.
• "The Chinese government may not consider that [massive market caps] part of the objective function of what a win is," Bill states, suggesting a focus on employment, global competitiveness, and affordability for citizens.
• BYD's recent 30% price cuts, possibly government-encouraged, exemplify this focus on low-cost production and market share over high margins.
• Brad notes that once winners emerge, the Chinese government exerts more control (e.g., golden shares, preferential procurement, regulatory approvals), citing the post-Didi IPO era and actions against Alibaba.
• This contrasts with the US mindset where trillion-dollar market caps are seen as a positive sign of success.

Strategic Implication for Crypto AI Investors: When evaluating Chinese companies in the AI or related tech sectors, investors must understand that government objectives (e.g., global market penetration, domestic affordability, strategic control) might override shareholder value maximization in the Western sense. This could lead to different valuation metrics and risk profiles.

AI Talent, Immigration, and National Competitiveness

• The conversation turns to the critical role of talent in the AI race and US immigration policies.
• Brad and Bill reiterate their long-standing view on the need to recruit and retain global talent, referencing a previous discussion with President Trump on the All-In Pod about "stapling a green card" to diplomas.
• H-1B visa: A non-immigrant visa that allows U.S. employers to temporarily employ foreign workers in specialty occupations.
• They express concern over Marco Rubio's tweet suggesting the US might revoke visas of Chinese students in "critical fields" like AI.
• Bill warns: "When you take these kind of broad statements... they have the potential to be slippery slopes... that could be very dangerous to our long-term competitiveness."
• Brad emphasizes the damage to "Brand USA" and the risk of deterring global talent, noting that 40-50% of top AI researchers in the US are of Chinese origin.
• Bill argues against a "China hawk" mindset that leads to counterproductive policies, citing the Manhattan Project's reliance on immigrant talent.

Actionable Insight for Crypto AI Researchers & Investors: Restrictive immigration policies targeting specific nationalities in critical fields like AI could significantly hamper US innovation in both AI and crypto. The community should advocate for policies that attract and retain global talent, as this is fundamental to maintaining a competitive edge.

Rare Earths, AI Chips, and US-China Trade Tensions

• The discussion addresses the strategic importance of rare earths and AI chips in the US-China relationship.
• Brad highlights a Wall Street Journal article on China's leverage with rare earths (elements critical for magnets in electric motors and electronics), risking factory shutdowns for US companies.
• He draws a parallel between rare earths and AI chips: both are viewed as existential by each country.
• Brad proposes a hypothetical trade: the US offers China access to a deprecated Nvidia Blackwell 30 (B30) chip in exchange for guaranteed rare earth supplies.
• The B30 chip lacks HBM (High Bandwidth Memory) and CoWoS (Chip-on-Wafer-on-Substrate) packaging, making it less suitable for cutting-edge AI training but still competitive for other applications in the Chinese market.
• Arguments for this trade:
• Keeps Chinese developers in the Nvidia CUDA (a parallel computing platform and programming model created by Nvidia) ecosystem, slowing Huawei's ecosystem growth.
• Generates US tax revenue and profits for Nvidia to reinvest in R&D.
• Secures critical rare earth supplies, buying time for the US to develop its own sources.
• Potentially reduces the risk of China moving on Taiwan out of desperation for chip access if they perceive AGI (Artificial General Intelligence) as an imminent existential factor.
• Bill agrees, emphasizing the economic illogic of restricting trade in areas where the US excels and warning that current policies might inadvertently "build a wall around America" rather than China.

Strategic Implication for Crypto AI Investors: The geopolitical tug-of-war over AI chips and rare earths directly impacts the hardware supply chain crucial for both AI development and cryptocurrency mining/staking. Investors should monitor these trade negotiations closely, as disruptions or resolutions will affect hardware costs, availability, and the competitive landscape for AI compute providers. Diversification of supply chains for critical components is a key long-term trend to watch.

Market Analysis: Tariffs, Debt, and Economic Outlook

• Brad provides his current market outlook, balancing recent optimism with underlying uncertainties.
• The NASDAQ is up 20% from recent lows, with the S&P also showing recovery.
• Key factors for future market direction:
• Tariffs: The outcome of US-China trade talks is critical. Brad hopes for a "Bessant consensus" leading to blended global tariffs around 10-15%, avoiding trillions in new costs.
• Reconciliation Bill: Passage is needed to extend tax cuts and provide new stimulus (e.g., no tax on tips/overtime, Social Security tax deductions). Failure could mean a $4 trillion tax increase.
• National Debt: While acknowledging Elon Musk's concerns, Brad points to the Ray Dalio/Bessant "333" plan (3% GDP growth, 3% debt-to-GDP target over time) as a more viable path than immediate drastic cuts.
• Interest Rates: Brad is not overly concerned by the 10-year Treasury yield fluctuating between 4-5%, a range it has held for two years. He anticipates Fed rate cuts if PCE (Personal Consumption Expenditures), a key inflation indicator, continues to moderate.
• "Don't take yourself out of the game, but I think it is a wait and see approach," Brad advises on market participation.
• Bill expresses continued uncertainty given these unresolved major issues.

Actionable Insight for Crypto AI Investors: Macroeconomic stability, influenced by tariff resolutions and fiscal policy, will impact investment sentiment across all asset classes, including crypto and AI ventures. Favorable outcomes could fuel growth, while negative developments might trigger market downturns, affecting funding for early-stage Crypto AI projects.

Delaware Incorporation and Corporate Governance

• The conversation touches upon corporate governance issues, starting with Delaware's changing legal landscape.
• Bill reiterates his concern that Delaware is no longer as predictable for corporate law, citing a Fortune article on lawyers winning excessive fee multipliers.
• He believes high-profile companies are most at risk from "activist judges."
• Crucially, Bill clarifies that litigation risks in Delaware apply to private companies, especially large unicorns with secondary share trading, not just public ones.
• "Just because you're not public doesn't mean this shouldn't matter to you," Bill warns.
• Brad confirms he's had companies ask about reincorporating out of Delaware.

Strategic Implication for Crypto AI Companies: Crypto AI startups, particularly those achieving significant private valuations or facilitating secondary trading of equity/tokens, should re-evaluate the risks of Delaware incorporation. Exploring alternative jurisdictions with more predictable corporate law may become a necessary fiduciary consideration.

Proxy Advisors and Shareholder Interests

• The final topic is the duopoly of proxy advisory firms, ISS and Glass Lewis.
• Brad highlights Senator Hagerty's criticism of these firms, which control ~97% of the market and allegedly "hijacked corporate governance."
• Bill recounts his past interaction with ISS, finding a lack of first-principles thinking on aligning board actions with shareholder interests.
• He questions why large index funds (e.g., BlackRock) heavily rely on these advisors, suggesting they may have agendas beyond pure shareholder value (e.g., 80% foreign-owned, pursuing other corporate philosophies).
• Ironically, super-voting shares, often seen as poor governance, can be a defense mechanism for founders against the influence of these proxy advisors.
• Bill proposes an AI-driven alternative to ISS and Glass Lewis that could model good governance based on aligning shareholder interests.
• "Maybe we need an alternative to these two things... maybe be good to see something like that pop up," Bill suggests, with Brad enthusiastically agreeing to "incubate that."

Actionable Insight for Crypto AI Researchers & Investors: The potential for AI to disrupt traditional corporate governance mechanisms, like proxy advisory, is significant. Researchers could explore AI models for assessing governance quality in both traditional companies and DAOs (Decentralized Autonomous Organizations). Investors might find opportunities in funding startups that leverage AI to offer more objective and shareholder-aligned governance solutions.

This episode underscores the intricate dance of AI supremacy, global talent, and resource control, particularly between the US and China. Crypto AI investors and researchers must keenly observe these geopolitical and strategic maneuvers, as they will profoundly shape data access, hardware availability, and the very framework of innovation in the converging AI and crypto landscapes.