Steve Hanke says there’s too much talk about AI, and it could end badly.
The veteran trader and economist told Business Insider in an email that he agrees with Yann LeCun, Meta’s former chief AI scientist and a pioneer in the field. He cautions that large-scale language models (LLMs) such as ChatGPT are not very innovative.
“Because they are fluent in language, we are easily fooled into thinking that they are intelligent, but in reality, their understanding of reality is very superficial,” Hanke said, citing LeCun’s speech from Spring 2024.
LeCun added that while chatbots have their uses, “LLM is essentially an exit, a distraction, a dead end on the path to human-level intelligence.”
Hanke, a professor of applied economics at Johns Hopkins University, said he is “on LeCun’s side of the court” and believes AI is “overhyped and potentially dangerous.”
He told Business Insider last October that whether the market’s boom is rational or irrational “depends in large part on whether the incredible revenue projections of AI companies are plausible.”
“It may be wise to fasten your seatbelts,” he added.
Mr. Hanke was president of Toronto Trust Argentina in 1995, when it was the world’s best-performing market mutual fund. He also served as an economic advisor to President Ronald Reagan.
believers and skeptics alike
Despite some recent jitters, the AI boom appears to be going strong, with OpenAI reportedly close to raising more than $100 billion at an $850 billion valuation. The makers of ChatGPT generated more than $20 billion in annual revenue last year.
So-called “hyperscalers” rushing to build out the infrastructure to power the AI boom are predicting truly mind-boggling spending.
Meta, Amazon, and Alphabet recently predicted that capital spending in 2026 could total $520 billion, and Microsoft also plans to invest more than $100 billion this calendar year.
LeCun recently left Meta after more than a decade at the parent company of Facebook and Instagram. He left to found Paris-based AMI Labs and develop open source AI that can truly understand and model the physical world, not just language.
Hanke is not the only one sounding the alarm about the AI boom. Michael Burley, of “The Big Short” fame, has warned that big tech companies are investing too much in microchips, which could quickly become obsolete and deliver disappointing returns.
Jeremy Grantham, a leading expert on bubbles and long-term investment strategist at GMO, warns that previous innovative technologies, such as trains and the internet, always had an initial bubble that was bound to burst, and he expects the same thing to happen with AI.
But the soaring valuations are more than justified, as AI advocates from Elon Musk to Sam Altman predict the technology will dramatically increase productivity and generate huge profits.
