The clearest cluster is a set of documents rather than a single book. The 20VC panel said the Poolside and Stripe letters were the two documents it most enjoyed reading in preparation, while its discussion of a storm book centered on the idea that the “absence of choice” is a danger signal. 20VC with Harry Stebbings

Standout recommendation

The Storm — Sebastian Junger’s storm book

• Type / creator: Book by Sebastian Junger; the transcript renders the surname as “Younger” and calls the book “The Storm.” 20VC with Harry Stebbings
• Link:20VC episode in which it was recommended
• Recommended by: A 20VC panelist identified as Jason, who called it “my favorite” and “one of the best.” 20VC with Harry Stebbings
• Key takeaway: The panel recalled the book’s line that “if danger can be described as the absence of choice,” then the loss of options is itself the warning. They applied it to OpenAI’s increasingly constrained strategic position. 20VC with Harry Stebbings
• Why it matters: It is a clean optionality test for founders and investors: when a move is forced by shrinking choices, the constraint—not just the move—should be treated as the risk signal.

Two letters for understanding AI’s capital regime

Poolside investor letter

• Type / creator: Investor letter from Poolside.
• Link:20VC discussion
• Recommended by: The 20VC panel. One panelist said the letter “gave me the feels,” called it “excellent,” and said he had read and reread it. 20VC with Harry Stebbings
• Key takeaway: As quoted in the discussion, Poolside had correctly anticipated a market for a U.S. open-source model and built one, but could not raise the capital required for the next compute-intensive step. The panel’s positive reading was that useful work in a hypergrowth market can still create substantial acquirer value even when standalone financing fails. 20VC with Harry Stebbings
• Why it matters: Read it as a financing case study, not as proof that frontier bets are safe. It separates two questions that are easy to conflate in AI: whether the product is moving in the right direction, and whether the company can finance the next leg.

Stripe letter

• Type / creator: Company letter from Stripe.
• Link:20VC discussion
• Recommended by: The 20VC panel, which called it “really really good” and later named it alongside the Poolside letter as one of the two documents it most enjoyed reading in preparation. 20VC with Harry Stebbings
• Key takeaway: The panel’s paraphrase of Stripe’s framework is that intelligence is fungible capital: it has to be priced, allocated, and managed more like money than like seat-based software licensing. 20VC with Harry Stebbings
• Why it matters: This gives operators a concrete budgeting question: if employees can draw on effectively uncapped token usage, who gets the budget, and what other spending must change? The episode’s CFO discussion makes the tradeoff explicit rather than treating AI usage as a free productivity layer. 20VC with Harry Stebbings

A compact investing prompt

Jane Street’s growing pains

• Type / creator:Financial Times article; the author is not visible in the accessible capture.
• Link:Read the article in the Financial Times
• Recommended by: Morgan Housel, who linked the piece after writing, “Good line,” and quoting its observation: “The longer your holding period, the more chance that something you can’t forecast occurs.” @morganhousel
• Key takeaway: Longer duration means greater exposure to events that no forecast captures. @morganhousel
• Why it matters: It is a useful check against treating a long holding period as merely a longer period of forecastable certainty. The available FT capture exposes the headline but stops at a subscription gate, so no broader thesis is attributed to the article here. ft