What does an 'AI supercycle' cost? Nokia is cutting Bell Labs research from 681 people to 408

Nokia plans to cut Bell Labs research headcount from 681 in 2024 to 408 by the end of next year, a reduction of about 46%, and to shrink its dedicated research centres from 19 to six. The company's stated reason is opportunity rather than automation: 'The AI supercycle presents a defining opportunity to rethink the role of networks in an AI-driven world,' with the remaining researchers concentrated on 'the highest-impact priorities' and on turning research into commercial impact. Marcus Weldon, who led Bell Labs from 2013 to 2021 and served as Nokia's chief technology officer, called the plan a 'top-down dismissal' born of 'corporate amnesia', and described research funding falling to 0.4% of Nokia revenues as 'shocking' for an institution with a century of history behind it. Nowhere does Nokia claim AI is doing the work these researchers did. This is the reallocation pattern in its starkest form: an AI boom cited as the reason to spend less on the research that produced the last one.