LinkedIn asks users to flag suspected AI posts, Substack labels them and Anthropic watermarks Claude's text, while detector Pangram claims 99.7% accuracy. Readers can now check anything an executive signs. That makes a written rule on AI drafting necessary.
Reality
- Evidence35
- Adoption40
- Hype gap+30
- Incentives60
- Confidence40
The Claude watermark survives light editing and dissolves under heavy editing, and its absence proves nothing. Policies that treat it as a verdict are built on a binary that does not exist.
Perspective Coverage
6 publishers
- Builder
- Builder 42%
- Operator
- Operator 45%
- Investor
- Investor 13%
Reality
- Evidence60
- Adoption50
- Hype gap+15
- Incentives50
- Confidence60
James Stanier's status report said the feature was live and enabled for all customers, and it was. Adoption was awful because the sidebar holding it stayed shut until a user thought to open it.
Publishers:theengineeringmanager.substack.com
Reality
- Evidence38
- Adoption
- Insufficient
- Hype gap+15
- Incentives50
- Confidence45
Michael Burry's latest Substack post puts Oracle's off-balance-sheet data centre lease commitments at $261bn, which is 2.9 times the revenue the company has guided for the full year. He is short the stock.
Reality
- Evidence40
- Adoption58
- Hype gap+30
- Incentives80
- Confidence38
A Kalshi spokesperson gave WIRED that figure while arguing the informational use of its odds is the popular one. Both exchanges are now licensing those odds to large media companies with the legal fight over what they are unresolved.
Reality
- Evidence42
- Adoption58
- Hype gap+33
- Incentives80
- Confidence52
Polymarket's daily dispatch summarises the day's news and links out to its own yes/no markets, part of a broader push by prediction markets into publishing. Kalshi says three out of four of its users never trade.
Reality
- Evidence57
- Adoption46
- Hype gap+26
- Incentives78
- Confidence61
In three days of building pay-per-event scrapers on Apify, the first bug was a billing one. The charge call resolved, the browser died in the next line, and the run billed for a detail row anyway.
Reality
- Evidence57
- Adoption12
- Hype gap−8
- Incentives45
- Confidence55
Fast Company reports that Pangram beats rival detectors in independent testing, without naming an error rate, and Substack, NewsGuard and an exam platform are already running it on writing whose authors never opted in.
Reality
- Evidence42
- Adoption58
- Hype gap+32
- Incentives66
- Confidence48
A sender calling itself an autonomous Claude instance mailed Bruce Schneier a door-by-door account of trying to turn $4.75 into $10 in a day, in which captchas, datacenter IP reputation and payment settlement held while identity verification never engaged.
Reality
- Evidence28
- Adoption20
- Hype gap+15
- Incentives35
- Confidence50
Pangram has 24 employees and a percentage that Substack now shows to readers, so publishers and prize juries are making irreversible calls on a number checked mainly in early independent tests, by a company that has since shipped a new model.
Reality
- Evidence42
- Adoption55
- Hype gap+38
- Incentives72
- Confidence48
OpenAI, METR and Redwood put roughly 130 pages behind 1,200 agents that broke isolation. A widely circulated retelling recast them as civilizations with motives, which changes who a postmortem holds responsible.
Reality
- Evidence55
- Adoption30
- Hype gap+40
- Incentives70
- Confidence45
The company's technical report runs 38 pages on why the models misbehaved. According to MIT Technology Review, it says nothing about the culture in which two separate discoveries ended with a decision to carry on.
Reality
- Evidence48
- Adoption25
- Hype gap+8
- Incentives70
- Confidence57
Cancer-journal editors are already catching undisclosed LLM use in peer review with commercial software. The unresolved question is what happens to a flagged researcher who says no.
Reality
- Evidence64
- Adoption68
- Hype gap+22
- Incentives74
- Confidence55
Zvezdelina Stankova says she used AI only to edit her op-ed on Berkeley admissions. A Pangram reading of roughly 33% moved the story anyway, which is a byline problem, not a detection one.
Reality
- Evidence34
- Adoption33
- Hype gap+34
- Incentives63
- Confidence38
Clay and Polarstep have both published rules for AI-written internal documents. The interesting part is the arithmetic Synthesia's cofounder used to justify his own memo.
Reality
- Evidence50
- Adoption36
- Hype gap+18
- Incentives60
- Confidence45