DiscussionOpinion

Gavin McCracken (Pt. 2): Robotics, AI, and A Commodities Supercycle

Value Hive Podcast1h 13m

Gavin McCracken, a PhD in AI and commodities investor, discusses how AI is solving mathematical conjectures through search rather than true intelligence, predicts a major commodity supercycle driven by robotics demand, and analyzes geopolitical risks in oil markets including potential Trump export bans and escalating tanker attacks in the Strait of Hormuz.

Summary

In this second conversation between Gavin McCracken and the host, they explore the intersection of AI research and commodity investing. McCracken recently attended an AI conference focused on mathematics and describes how AI systems are successfully disproving long-standing mathematical conjectures through computational search rather than intelligent reasoning. He argues that AI excels at finding counterexamples and exploiting gaps in systems, but cannot formulate new problems or demonstrate genuine creative intelligence.

McCracken presents a technology era framework comparing the Information Age (digital computing) to what he believes is the emerging Robotics/Analog Age. He argues that just as the Industrial Revolution peaked with nuclear fission, the Information Age is peaking with current AI scaling. The next era will involve using AI to design analog computers and robotics that solve specific problems with superior efficiency. This technological transition, he predicts, will drive a massive commodity supercycle as industries require vast quantities of raw materials to build physical robots and infrastructure.

The discussion covers specific commodity plays resulting from this thesis. Silver emerges as particularly important for robotics applications, with McCracken citing Tesla's stated goal of mass-producing humanoid robots—each requiring an ounce of silver. Tungsten is highlighted as critically undervalued given increased military demand from ongoing conflicts. Aluminum demand will surge for lightweight robot construction. McCracken also emphasizes that miners trading at low valuations relative to cash flow represent the best plays, as major mining companies cannot find ounces as cheaply through acquisition.

A significant portion of the conversation addresses oil markets and geopolitical risks. McCracken expresses conviction that Brent crude is inadequately priced given multiple tail risks: potential Trump export bans on WTI, escalating tanker attacks in the Strait of Hormuz (50+ confirmed hits by Ukrainian forces), the Panama Canal's capacity constraints, and broader Russia-Ukraine escalation. He notes a media blackout on tanker incidents despite their frequency and strategic importance. He argues that $200 oil would trigger a liquidity crisis but that the market is in denial about this risk.

McCracken shares his personal portfolio positioning, including large oil and Brent call positions financed through margin, acknowledging the strategy's volatility and risk. He describes a dramatic decision to liquidate equity positions and deploy capital into oil calls at the $68 level, which proved profitable as oil rebounded. He positions China as having effective control over oil prices through its ability to rapidly alter SPR buying, representing a geopolitical risk to Trump's second term if oil prices spike before midterms.

The conversation addresses why commodities represent better risk-adjusted opportunities than technology stocks, given that AI and semiconductor companies operate in highly competitive, commoditized spaces where everyone pursues similar strategies. By contrast, commodity exposure captures the structural demand from multiple drivers (AI infrastructure buildout, robotics revolution, geopolitical conflicts) without company-specific execution risk.

McCracken discusses why he reduced gold miner exposure despite reasonable fundamentals, arguing that a liquidity crisis triggered by $200 oil would cause gold to decline as people sell jewelry and assets for cash. However, he acknowledges ongoing currency debasement arguments for holding gold alongside the near-term technical overextension he observes.

Finally, both participants discuss the emerging opportunity in robotics companies and the challenge of identifying which will succeed. They note that unlike NVIDIA in 2022, most robotics-related stocks are not trading at cheap valuations, making the opportunity more challenging. They also express skepticism about Chinese robotics companies given geopolitical bifurcation and national security concerns. The conversation concludes with plans for a third discussion focused specifically on robotics as an investment theme.

About this episode

<p>We are back with another interview with Gavin McCracken. This time, we spend most of the podcast chatting robotics, AI, and why those two themes could kickstart the next commodities supercycle. </p><p>I really enjoyed Gavin's take on AI and robotics, mainly because it aligns with my initial hypotheses. But it's always good to bounce ideas off a AI PhD. </p><p>Remember, NOTHING YOU HEAR IS INVESTMENT ADVICE. THIS IS TWO DUDES CHATTING MARKETS FOR ENTERTAINMENT ONLY. YOU ARE AN IDIOT IF YOU THINK THIS IS ADVICE.</p><p>I hope you guys enjoy!</p>

Key Insights

  • McCracken argues that AI systems are not demonstrating intelligence but rather conducting sophisticated search across pattern space, evidenced by their ability to find counterexamples to mathematical conjectures that humans theoretically could discover but would require years of computational work.
  • AI systems excel at finding what breaks or fails in existing systems rather than proving positive claims, suggesting their primary value lies in identifying vulnerabilities and gaps rather than creating new mathematical truths.
  • McCracken claims the Information Age is in its final phase, comparable to how the Industrial Revolution culminated with nuclear fission, and that the next technological era will focus on building analog computers and robotics using AI as a design tool.
  • The predicted commodity supercycle will be driven by the physical infrastructure requirements of robotics manufacturing, not from AI software deployment itself, making raw materials the superior investment compared to technology stocks.
  • Silver is significantly underpriced given that mass-producing humanoid robots for household use would require massive quantities, and current pricing doesn't reflect this demand signal.
  • McCracken estimates at least four million barrels of oil production are shut in globally from conflict, yet the market prices oil as if this isn't occurring, indicating a massive disconnect between reality and market price.
  • A media blackout exists around Ukrainian tanker attacks on Russian vessels and Iranian threats in the Strait of Hormuz, suggesting deliberate suppression of oil supply disruption information to prevent price spikes that would trigger inflation.
  • China effectively controls the global oil price through its ability to rapidly adjust SPR purchases and could manipulate crude above $200 strategically if it benefits their geopolitical goals, particularly before US midterms.
  • The Strait of Hormuz has become randomly dangerous without predictable attack patterns, making it rational for tanker operators to avoid transit even without formal blockades, creating effective supply disruption through fear rather than force.
  • Technology stocks represent poor risk-adjusted opportunities because all major AI and semiconductor companies pursue nearly identical strategies, creating competitive commoditization, whereas commodity exposure captures structural demand across multiple independent drivers.
  • McCracken argues that true thinking—deconstructing problems from first principles rather than applying existing knowledge—is fundamentally different from pattern recognition and remains uniquely human, making LLMs useful for information retrieval but not thought.
  • A liquidity crisis triggered by $200 oil would cause gold prices to decline despite currency debasement because individuals would need to liquidate physical gold holdings for cash, countering the typical inflation-hedge narrative.

Topics

AI capabilities and limitations in mathematicsTechnology era transitions and the robotics revolutionCommodity supercycle thesisSilver demand from roboticsTungsten shortage and military demandOil market risks and geopolitical analysisStrait of Hormuz tanker attacksPotential Trump WTI export ban consequencesPortfolio positioning with margin and leverageChina's control over global oil pricesGold market technicals vs. fundamentalsRobotics stock valuations and investment opportunities

Transcript

all right gavin this is our part two this well well uh much awaited part two um i feel like not much has happened besides literally everything in oil and gas precious metals and um so much so much to talk about and we're recording this late late for me i guess eight o'clock uh eight o'clock eastern time i'm just wrapping up bowl ice cream with some cookies so i'm in a pretty awesome mood and get to get to talk to you and i think i want to start this conversation on ai because you said you had a lot of thoughts on it you you've done some work in this i believe believe you're a PhD in…

Full transcript available for MurmurCast members

Sign Up to Access

More from Value Hive Podcast

Get AI summaries like this delivered to your inbox daily

Get AI summaries delivered to your inbox

MurmurCast summarizes your YouTube channels, podcasts, and newsletters into one daily email digest.