digest
Short Summary
Main recommendation: Kyle Reidhead of Milk Road argues investors should keep buying AI infrastructure exposure, including memory stocks, neocloud stocks, power, energy, and data center stocks, because demand for compute still far exceeds supply. He sees AI revenue growth at Anthropic and OpenAI as strong enough that he is not concerned about the companies' large obligations, and he expects AI-integrating software names such as ServiceNow and Salesforce to benefit.
Key Points
- Reidhead believes AI demand will continue regardless of leadership changes at the big labs, since current intelligence is still far ahead of how much the market has adopted.
- He sees a possible incentive for regulatory capture but doubts pausing development would work, because open source models and China would catch up.
- He argues AGI is close, but that AI using itself to train the next AI is not yet true ASI, which he thinks is still years away.
- Revenue growth is his main argument against concern: OpenAI reportedly rose 20 percent in July, and Anthropic reached a billion in revenue with expected enterprise revenue of 2.5 billion by year end.
- He notes that software stocks have rallied, and he thinks the application layer on top of AI is the part that has lagged and may now benefit.
Details Worth Noting
- Concrete instructions: Buy and hold a basket of AI infrastructure categories the speaker names: memory stocks, neocloud stocks, power and energy stocks, and data center stocks. He says he has held these over the past year and expects them to keep rising over the next year. For software, he says he owns Salesforce and ServiceNow and expects companies that use AI to rebuild their business and products to see growth over the coming years.
- He cites compute demand figures as the basis for this: He says SpaceX is trying to reach 210 gigawatts by the end of next year, and Meta is getting 15 gigawatts in the next few years (likely referring to xAI/Musk companies rather than SpaceX specifically, as stated in the transcript).
- He says the market's AI selloff in software was overdone, with companies such as loan servicing firms now positioning to integrate AI.
- He says the speaker's view on regulatory capture is uncertain and he may be missing risks the labs know about.
- Figures are the speaker's own claims and are not independently verified in the transcript.
Bottom Line
The speaker is bullish on continued AI spending and infrastructure demand, while acknowledging uncertainty about regulation and timelines for true superintelligence.
