Bloomberg logo, reverse (Wikimedia Commons, public domain)
The co-CIO of the world's largest hedge fund compared the current AI moment to February 2020, right before COVID-19 upended everything.
Greg Jensen, co-chief investment officer of Bridgewater Associates, delivered one of the bleakest assessments of the AI boom yet. Speaking on Bloomberg’s Odd Lots podcast on September 11, Jensen argued that society is unlikely to meaningfully address the risks of artificial intelligence until something genuinely catastrophic happens, potentially involving deaths.
His analogy of choice: February 2020, the brief window when a handful of people were sounding alarms about a novel coronavirus while the rest of the world carried on booking flights. Jensen sees the AI moment in similar terms. Lots of warnings, very little urgency.
The numbers behind the warning #
Jensen oversees Bridgewater’s AI strategy and lab, making him one of the most influential figures investing directly in the technology he’s cautioning against.
Bridgewater’s own analysis forecasts that as much as 18% of US jobs could be displaced within five years due to AI advancements. To put that in perspective, the US labor force comprises roughly 160 million people. An 18% displacement rate would affect tens of millions of workers across industries.
Jensen also flagged that AI-related capital expenditure has become responsible for roughly one-third of recent US economic growth. He described the current phase of AI investment as “more dangerous” than earlier stages, with demand for physical infrastructure, think data centers, chip fabrication, and energy systems, accelerating at the same time that the broader economy is becoming increasingly dependent on AI-driven productivity gains.
A ‘token tax’ and developer liability #
Jensen authored an op-ed in the New York Times proposing what he calls a “token tax,” a mechanism designed to offset the economic disruption caused by AI-driven job displacement. The concept targets the computational resources that power AI systems, essentially taxing the tokens processed by large language models and other AI architectures.
Jensen also argued that AI developers and the corporations deploying their systems should face liability when AI causes harm, including criminal liability for AI-induced crimes. That position puts him at odds with much of the tech industry, which has generally lobbied for lighter regulatory touches and safe harbor protections.
Disclosure: This article was edited by Editorial Team. For more information on how we create and review content, see our