Odd Lots

Jerry Neumann on the Problem With Investing in AI Right Now

November 12, 2025

Key Takeaways Copied to clipboard!

  • The current AI boom is analogous to the final stage of a previous technological wave (the computer revolution), rather than the eruption phase of a new one, suggesting value capture will favor incumbents who can leverage existing infrastructure. 
  • Successful adoption of revolutionary technology like AI, similar to containerization, comes from companies that use efficiencies to expand volume and pass value to consumers, not just those focused on immediate cost-cutting and margin increases. 
  • There is a notable lack of publicly traded 'AI pure plays,' with investment currently focused on 'picks and shovels' (infrastructure like chips and data centers), mirroring historical patterns where the true winners emerge later by building new business models on the foundational technology. 

Segments

Lack of AI Pure Plays
Copied to clipboard!
(00:02:09)
  • Key Takeaway: Despite the AI boom, there are virtually no publicly traded ‘AI pure plays,’ with NVIDIA being the closest example whose AI focus is recent.
  • Summary: The hosts note that public markets lack companies solely defined by AI, contrasting NVIDIA’s history with gaming and Ethereum mining before its current AI designation. Investment is currently concentrated in ‘picks and shovels’ like infrastructure providers. This situation creates a market where many companies are treated as winners despite the consensus that many will ultimately fail.
Value Creation vs. Capture
Copied to clipboard!
(00:06:15)
  • Key Takeaway: There is a critical distinction between AI creating societal value and who ultimately captures that value, which may not be the early investors or foundation model companies.
  • Summary: Jerry Neumann affirms that AI is a revolutionary technology but emphasizes that value capture determines who gets rich. He questions whether early investors, foundation model creators, or consumers will ultimately benefit most. This mirrors historical revolutions where incumbents or later adopters captured the most value.
Containerization Analogy
Copied to clipboard!
(00:07:01)
  • Key Takeaway: Shipping containerization, a revolutionary logistical technology, primarily benefited existing incumbents and downstream users like Walmart, not necessarily the initial innovators.
  • Summary: Containerization was a systems problem that incumbents like Maersk eventually adopted because resisting it meant being left behind, despite initial resistance from labor and ports. The founder of the instigator, Sea Land, made money primarily by selling early before the system fully matured. Later companies like Walmart and IKEA became massive winners by using the new logistics to expand product variety and volume.
Microprocessor Revolution Timing
Copied to clipboard!
(00:12:37)
  • Key Takeaway: The true revolutionary impact of the microprocessor (Intel 4004) only materialized years later when external inventors used it for permissionless innovation, like building personal computers.
  • Summary: Intel initially viewed the 4004 as an evolutionary step for building desktop calculators, not a world-changing technology. The revolution occurred when external inventors, like Steve Wozniak, used cheaper, more accessible chips (like the 6502) to create personal computers. This highlights that the system adoption, not just the core component, drives the revolution.
VC Exit Strategies and Private Markets
Copied to clipboard!
(00:19:01)
  • Key Takeaway: The abundance of late-stage private capital allows major companies like OpenAI to delay IPOs, forcing VCs to focus more on timing sales to incumbents rather than relying solely on traditional IPO windows.
  • Summary: VCs traditionally relied on the IPO window as the best exit, but the current environment offers alternatives like acquisitions by large incumbents. Being a public company is burdensome due to quarterly reporting requirements, making private status attractive if capital is available. VCs must now consider the power law, investing in one potential winner while managing fund life constraints.
AI Cycle Phase and Investment Focus
Copied to clipboard!
(00:26:06)
  • Key Takeaway: The current AI moment represents the culmination and end stage of the existing information computer technology wave, not the eruption of a new one.
  • Summary: Applying Carlotta Perez’s technological cycle theory, Neumann places AI at the end of the computer wave, similar to how containerization capped the automobile/mass production wave. Therefore, he is not investing, as he cannot fund foundation models requiring billions, and the true winners (like IKEA in the container era) will be existing knowledge-intensive companies that leverage the efficiency for massive volume growth.
Bubble Definition and AI Infrastructure
Copied to clipboard!
(00:38:51)
  • Key Takeaway: The current AI environment is characterized by overvaluation, particularly in infrastructure spending, but lacks the widespread leverage and painful economic contraction associated with a true bubble pop.
  • Summary: Neumann believes things are overvalued, especially the infrastructure buildout for chips and data centers, which may lead to future overcapacity. However, a true bubble requires a painful collapse, which is unlikely when major players like Microsoft are funding the spending, meaning a billion-dollar loss doesn’t significantly hurt the broader economy.