AI Mania Is Eviscerating Global Decision-Making
Simon Willison's Weblog Simon Willison ● Covered by 4 sources
A consultant says AI hype is warping decisions at big companies, with execs faking expertise and fudging productivity claims. Nobody wants to be the one who says the emperor has no clothes.
Based on reporting by Simon Willison's Weblog, Simon Willison — read the original for the full story.
Summary, retelling and take written by AI under human oversight; images are AI-generated illustrations. How we work · Report an error
Nik Suresh, who consults with large companies, has put together a pretty entertaining rundown of just how unmoored corporate AI decision-making has become. It's not a dry policy paper. It's a collection of anecdotes, sourced anonymously, that read like they came from a satire piece — except they didn't.
One story stands out: an executive at an organisation pulling in more than $2 billion in revenue admitted, right after unveiling a technical strategy built entirely around AI, that they had never actually used ChatGPT or any AI tool themselves. Not once. That's the strategy leading the tactics, rather than the other way around.
Then there's the engineer at a company that apparently tracks some kind of internal token leaderboard — a metric, presumably, meant to measure AI usage or output. Their solution to keeping up appearances: quietly checking out a parallel copy of the company's Go repository and having an AI rewrite the entire thing in Zig on the side, purely to look busy and stay employed. It's less a productivity hack and more a survival tactic.
The most revealing bit, though, comes from a skeptical executive at a vendor company who got asked why wildly inflated AI productivity claims — talk of 100x gains — kept getting repeated internally without anyone pushing back. The answer wasn't simply that it was sales fluff aimed at an eager audience, though there was some of that. The real issue was structural: customer executives were the ones making the outlandish claims, and if a vendor executive publicly doubted them, it would look like an attack on the customer, risking the relationship and the contract. Calling out nonsense that doesn't even touch your own company's mission is a fast way to lose your job, so nobody does it.
What comes through in all these anecdotes is not that AI itself is the problem, but that the incentive structures around it have gone sideways. Executives are performing conviction they don't have, engineers are performing productivity that doesn't exist, and vendors are performing agreement with numbers everyone privately knows are absurd. It's a hype cycle sustained less by belief and more by nobody being willing to be the person who breaks the spell.
My take — AI-written commentary, not fact-checked reporting
None of this is really about AI — it's about what happens when fear of looking behind becomes more powerful than fear of being wrong. Executives faking AI fluency and vendors nodding along to 100x claims they know are nonsense is a governance failure dressed up as innovation, and companies that eventually have to unwind these AI-driven strategies are going to have a very bad time explaining how they got there.
Read more about this at: Simon Willison's Weblog
