AI doomsday? New thought experiment warns of social and economic upheaval by 2028

Euronews · Feb 26, 2026 · Collected from RSS

Full Article

Published on 26/02/2026 - 11:13 GMT+1 Artificial intelligence (AI) could soon replace human labour at a scale and speed that society cannot handle, according to a new scenario that imagines life in 2028. The prediction envisions a “global intelligence crisis” that hits the world by 2028 due to the replacement of high-skill labour with AI. James Van Geelen, chief executive of Citrini, an American investment research firm and Alap Shah, an AI entrepreneur, wrote the paper as a fictional memo dating from June 2028, looking back at how the crisis unfolded. They envision that the layoffs “due to human obsolescence” began in 2026, as companies started employing AI agents to do tasks without human supervision. Van Geelen and Shah refer to the firings that have happened in the tech sector in January, with Amazon, Expedia and Pinterest all announcing AI-related job cuts. However, experts have previously said that it’s difficult to determine whether AI efficiencies were to blame. Tech job cuts start a vicious cycle: as companies invest more in AI, the models become more capable and justify more layoffs, the authors wrote. “Each company’s individual response was rational. The collective result was catastrophic,” the prediction reads. Businesses had to adopt the latest technologies to stay competitive, so those “most threatened by AI became AI’s most aggressive adopters,” the authors said. According to them, displaced white-collar workers moved into lower-paying roles and blue-collar trades, while the few who kept their jobs faced wage stagnation. By mid-2027, the US economy tips into a recession and by 2028, unemployment exceeded 10 percent. AI agents The paper envisions that by 2027, AI agents run in the background of people’s devices. They write all computer code, they handle weeks-long research projects and optimise how users spend their money. While AI does create some jobs in the new economy, such as prompt engineers, safety researchers, and infrastructure technicians, it renders dozens more obsolete. The new roles also pay a fraction of what the old ones did, the scenario predicts. The shift of white-collar workers to lower-paying jobs further depressed wages as the labour supply surged. Many households are forced to turn to credit cards or retirement accounts to cover their mortgages. By 2028, the US economy is on the trajectory for another mortgage crisis. To mitigate this, the government needs to transfer more money to households, even though it will be collecting less taxes from them due to their wage losses. The authors note that this could put additional strain on the economy. The authors also predict an “Occupy Silicon Valley” movement similar to the one on Wall Street in the 2010s, where demonstrators in May 2028 blockade entrances to Anthropic and OpenAI’s San Francisco offices for weeks. Van Geelen and Shah caution that these scenarios may not unfold exactly as described, but they argue that AI is already changing the economy faster than institutions can adapt.