The stock market sell-off triggered by a Substack article about a doomsday artificial intelligence future is the latest signal that technological hype is morphing into hysteria in America.

The article, published on Sunday by Citrini Research, a little-known research outfit in New York, set out the hypothetical scenario in which by June 2028, the bullish case for AI becoming more intelligent than humans leads to a wipeout of white-collar jobs, declining consumer spending, software-backed loan defaults and a struggling US economy.

It went viral on social media and by the end of trading on Mondaybillions of dollars had been wiped off the value of software and financial firmscited in the article, including Visa, Mastercard, DoorDash, ServiceNow and Blackstone.

The sell-off, triggered by animagined scenario, highlights the huge uncertainty around what the AI market means for the broader economy and how little understanding there is about what the advancing technology means for companies and the way we work.

One reason for this uncertainty is the gap between tech executives’ fantastical predictions about the capabilities of AI and the reality of how most people are experiencing the technology so far.

The latest proclamations fromleaders at the AI Impact Summit in Delhi last weekincluded Sam Altman declaring that his technology could soon replace chief executives and would do a better job than him at running OpenAI, and Dario Amodei, the head of Anthropic, predicting that we were only “a small number of years” away from AI models “surpassing the cognitive capabilities of most humans for most things”.

Should we believe the salesmen who are motivated to exaggerate the transformative powers of their products as they seek to raise money at ever-higher valuations to build out AI infrastructure? Not to mention obvious questions such as how could an AI chatbot, albeit with impressively broad knowledge, replace the depth of wisdom and judgment derived from human experience.

Back in our present reality, a survey of about 6,000 executives by the National Bureau of Economic Research published this month found that more than 80 per cent of companies reported no effect on either employment or productivity from AI in the past three years.

Another cause of uncertainty is the vacuum of US government regulation of AI which seems unlikely to hold indefinitely.

•Stuart Russell: We can coexist with superintelligent AI — but who wants to?

Source: Drudge Report