Will AI Destroy the World?
Geoffrey Hinton, the Nobel laureate known as the “Godfather of AI,” estimates a 10–20% chance AI leads to human extinction within thirty years. Elon Musk insists AI is far more dangerous than nuclear warfare. Yoshua Bengio, a Turing Award winner, warns that AI could lead to “catastrophic” outcomes for humanity. Stuart Russell, author of the field’s most-used textbook, compares the AI race to “Russian roulette.” In 2023, hundreds of researchers and tech CEOs signed a statement equating AI risk with pandemics and nuclear wars.
We have heard this before.
In 1946, Einstein’s Emergency Committee of Atomic Scientists warned that nuclear weapons would “surely destroy our civilization.” They have not. In 1977, Digital Equipment Corporation founder Ken Olsen declared there was “no reason for any individual to have a computer in his home.” By 2000, over a billion PCs sat in homes worldwide. In 1995, astronomer Clifford Stoll wrote in Newsweek that the internet was overhyped “baloney” and that online commerce would never work. Three years later, Nobel laureate Paul Krugman predicted the internet’s economic impact would be “no greater than the fax machine’s.” In 2007, then-Microsoft CEO Steve Ballmer told Fortune the iPhone had no chance of gaining significant market share. It now commands roughly half the smartphone market.
The pattern is consistent: catastrophic predictions about new technologies overestimate harms and underestimate adaptation. The 1960s automation panic predicted mass unemployment; instead, new industries emerged and living standards rose.
Will “this time be different?” That is possible, but we would bet against it.
 | A guest post by
|
As usual, Jeff's common sense and historical research hit the nail right on the head. I would even add that when I was in college I did a lot of reading from the Industrial Age in Europe, where similar catastrophic predictions were made concerning new technologies, and even moreso during say the early years of the 20th Century in the U.S. What I would also love to hear from Jeff is how he feels about the less catastrophic but even more ubiquitous predictions that AI will lead to mass unemployment. When I hear this from people I point out that, while AI is a relatively new term, all computers, going back 70 years or so, are artificial intelligence, and that overall they have vastly increased employment and productivity. My off the cuff observation is that while say AI trucks may put a lot of truck-drivers out of work, the vast increase in goods delivered will mean huge increases in employment for those creating, designing, making, advertising, etc. those products. Not to speak of the high-tech jobs created in the programming of the AI and all the physical tech that will go with the new vehicles. I would guess that the net effect is more products, more jobs, more productivity, more wealth. Am I crazy to think so?