跳转至

Products are out, brains are in

Ch01.026 Products are out, brains are in

📊 Level ⭐ | 9.4KB | entities/products-are-out-brains-are-in.md

Products are out, brains are in

摘要

As the marginal cost of producing software trends toward zero and capable tools multiply within every category, software's price may fall — but, the author argues, it will not death-spiral. What changes instead is why people buy software: not because they can't build it, but because they choose not to, and the deciding factor collapses to one thing — the judgment of the people running the software. Value migrates from the packaged product to the quality of thinking that sits above it, so the durable moat shifts from "what you ship" to "how well your team thinks."

核心要点

  • Software is becoming like bottled water or pre-washed lettuce: something that "should be free" is expensive, and that gap is now closing for code. The cost declines but doesn't crash — the purchase motive shifts from "I can't build it myself" to "I don't want to build or maintain it myself."
  • Enterprise buyers in heavily regulated industries won't build internal tooling even when software is fully commoditized; their evaluation shifts from ROI to trust: who won't get me in huge trouble, who can I trust with my customer data, who operates with enough foresight that I won't rip this system out later.
  • The twenty-first century's new factor of production is decision quality at scale: the cost of executing decisions keeps falling (automation, then AI), but the cost of making the best decision hasn't fallen at all — it's bounded by a clear understanding of complex, uncertain situations, which cannot be automated away.
  • "Thought leadership" is a dead phrase; what works is founder content in the mold of Buffett's Letters to Shareholders and Munger's Poor Charlie's Almanac — a public demonstration of prediction quality, decision transparency, epistemic honesty, and speed/economy.
  • Competition is a tug of war: two firms with identical stacks and markets are separated only by the cumulative quality of many small decisions over time — the strong-and-smart puller beats both the strong-and-dumb and the weak-and-smart.
  • Even if AI does almost everything, value keeps migrating upward (hardware → OS → middleware → applications → judgment above the app), and as long as LLMs are the primary AI-driven OS, buyers will want a human in the loop for the near future.

深度分析

Decision quality as the new non-fungible asset

The author extends classical economics — land, labor, capital, plus the twentieth century's technology/total-factor-productivity — with a twenty-first-century addition: decision quality at scale. The asymmetry is the point. Execution has become nearly free because of automation and now AI, but the cost of deciding what to execute has not fallen, because it is bounded by a clear understanding of a complex, uncertain, novel situation. You can automate the execution of a decision but not the judgment of which decision to make under imperfect information. History supports only "adequate" decisions; the best decision shifts constantly with dozens of small contextual differences. A company handing part of its business to another is therefore really betting on which team will make the most right calls most of the time — "companies that can see the future still have something valuable to sell."

The inversion of enterprise procurement

Even with software fully commoditized and democratized, large enterprises in regulated industries will not build their own tooling. Their questions shift from pure ROI to trust: who will probably not get me in huge trouble; who can handle my customer data safely (so my reps never look stupid or let anyone fall through the cracks); and who operates with enough foresight that I won't have to rip out this system, with all the data I invested in it, and repeat a long, painful re-evaluation and re-training cycle. The author draws a direct analogy to institutional investors evaluating fund managers: the fund's strategy can be replicated, but the alpha cannot. After decades of painful experience, investors learned that past returns are a poor predictor of future returns while decision-making process and intellectual rigor are somewhat better predictors — which is why the best funds pitch philosophy and mental models, not track records, and why software companies should dedicate themselves to doing the same.

Founder content as evidence, and the tug-of-war model of competition

Degenerate "thought leadership" — posting on LinkedIn, having a content marketer churn out AI-slop blog posts under your name — signals nothing to discerning buyers. What actually works is content that publicly demonstrates a track record of making better predictions and decisions than the median, at lower cost and higher frequency. Four ingredients matter: prediction quality (specific, confident, falsifiable forecasts that are both anomalous and rational, visibly updated when wrong — "nobody likes a cowardly founder"); decision transparency (show your reasoning, not just your conclusions, so buyers can judge how you think); epistemic honesty (acknowledge uncertainty while still making bold claims); and speed/economy (reaching good decisions faster and cheaper yields more decision capacity and first-mover edges before markets saturate). All of this is evidence for the tug-of-war metaphor: two identical firms, one rope, everyone in the mud — the winner is stronger, grips the rope better, and rallies the team to pull together at precisely the right time. The strong-and-dumb and the weak-and-smart pullers are both suboptimal; you want the strong-and-smart puller. Discerning buyers watch your film and see your swing before they place their bet.

The value-migration story, and why it isn't doom

The author pushes back on the "tech industry is dying" narrative as missing the forest for the trees. Markets have always been a history of value migrating as things commoditize — it happened to hardware, then operating systems, then middleware, and now the applications layer — and that is not a catastrophe, because plenty of markets are build-vs-buy (you hire a maid rather than clean your own house). What is genuinely new is where the value migrates: to the quality of thinking and judgment above the application that makes it productive. Even if AI does almost everything, as long as LLMs remain the primary AI-driven OS, people will want a human in the loop for the near future. The #1 question businesses must answer well for buyers becomes: how many strong, strategic people are on your team, and how do they work together?

实践启示

  1. Stop leading with the product. As feature parity rises, the product becomes less of a make-or-break factor; lead equally with your philosophy and brand (site design, founder content, product design principles). Brand and product side by side, or product second to brand — the author likes a plain .txt-style site that foregrounds ideas over flash.
  2. Build a public track record of specific, falsifiable claims. Write down where the market is going, decisions you made and why, and times you were wrong. Being right is evidence of your edge; being wrong is evidence of your epistemic honesty — and there are ways to be wrong in the right way.
  3. Make the team visible. These are the other people pulling the rope. Discerning buyers care about your key people too, especially your lead engineer and product lead.
  4. Compete on the quality of your predictions about your own roadmap. Show buyers a historical ledger of what you said you'd build versus what you actually built. Buyers who understand long procurement cycles and the pace of AI development will weight this heavily.
  5. Avoid gimmicks and use humor sparingly — a safe tack when selling into enterprise, where gimmicks attract the wrong crowd and cause misalignment. Like Amazon's documents-only culture, give people pure information: be quiet but right, not loud and wrong.
  6. Treat transparency as selection and churn reduction. Honest founder content gives prospects a clear view of whom they're trusting with part of their business and selects for customers who resonate with your decision-making philosophy — a shot at less churn and a bigger brand.

相关实体

→ 原文存档