Module 02 · signal fire · Verification ·
The Price of Being Sure
Generation is free, checking is not — epistemic hygiene becomes an allocation problem, and the allocators win.
Rotterdam, 2036. The due-diligence packet for the acquisition is flawless: audited statements, reference calls transcribed, a plant inspection video with drone footage. Your analyst agent assembled it in forty minutes and flags its own work: “All twelve exhibits are consistent with each other. Consistency is what fabrication optimizes for. Recommend spending your two physical verifications on the inventory count and the largest customer.”
The seller’s agents produced the packet. Your agent produced the skepticism. What is left of the diligence profession is the second sentence: deciding where to spend the checks that still cost something.
The mechanism
Two cost curves crossed sometime around 2023 and never uncrossed. Producing a claim — fluent, sourced-looking, internally consistent — now costs approximately nothing (baseline B2). Checking that claim against reality kept its price: an audit takes weeks, a reference call takes an afternoon, a replication takes a lab (baseline B3). When production is free and inspection is dear, the market floods with product. That takes no cynicism about people; the price gap does it on its own.
Hayek’s old point about markets (“The Use of Knowledge in Society,” 1945) was that they economize on knowledge: prices let you act on information you do not possess. The 2036 version inverts the emphasis — the scarce input is no longer information but warrant, the grounds for believing any given piece of it. Whoever allocates scarce verification well — spending expensive checks where the downside concentrates, accepting cheap risk everywhere else — holds the appreciating skill.
The skill is not believing less. It is pricing the check: matching your scarce verifications to the claims where being wrong is expensive.
Notice what this does not require: detecting fakes by eye. Detection is an arms race you should expect to lose on schedule. Allocation — stakes-weighted skepticism, provenance habits, knowing which claims your decision actually turns on — is a portfolio discipline, and it works whatever the fakes look like.
⏵ Toy 02 · The verification market
incoming claim · stakes: HIGH
“The supplier’s audit certificate is current.”
- For one week, before forwarding or acting on any claim, write one sentence: “the check this claim deserves costs ___ and I am / am not paying it.” The habit is the training.
- Build a provenance reflex: prefer sources that show their pipeline (data, method, chain of custody) and downgrade the rest by default, regardless of polish.
- In your team, name the two decisions this quarter where a false input would hurt most — and pre-commit the verification budget there, before the claims arrive.