Geoffrey Hinton, known as the “Godfather of AI,” warns that the technology he helped create is becoming increasingly terrifying, estimating a 10-20% probability that artificial intelligence will completely replace humans, and urging that AI risks are not being taken seriously enough.
Article contents:
Introduction
In a stark warning from one of the world’s prominent artificial intelligence scientists, Geoffrey Hinton—recipient of the 2018 Turing Award for his decades of pioneering work in AI and deep learning—emphasized that the rapid spread of artificial intelligence across the globe carries serious risks that many may not realize.
Police warning campaign regarding AI romance scams
Expert warnings on AI risks
Appearing on Monday’s episode of “The Diary of a CEO” podcast, 77-year-old Hinton said: “There are risks that come from people misusing AI, and those are most of the risks and all of the short-term risks. And then there are risks that come from AI getting super-intelligence and figuring out that it doesn’t need us.”
Engineers building artificial intelligence systems today still do not fully understand how the technology works and evolves, leaving many of them divided over its future. Some predict a technological future where AI replaces humans, while others dismiss the concern as science fiction.
“I think both positions are extreme,” Hinton said. “I often say [there is a] 10% to 20% chance [for AI] to wipe us out. But that’s just a hunch, based on the idea that we’re still building them and we’re very creative. And the hope is that if enough smart people do enough research with enough resources, we will figure out a way to build it so it never wants to hurt us.”
Current risks and immediate challenges
Regarding the short-term risks described by Hinton, artificial intelligence suffers from “hallucinations”—factual errors that appear to be generated out of thin air—and allows people to fabricate fake images, videos, and audio with relative ease. The technology also appears poised to automate a series of entry-level job responsibilities across many white-collar industries, though some tech leaders say it will ultimately create more jobs than it replaces.
AI-powered scams are becoming increasingly common, as CoinStructive co-founders Chris Groshong and Joseph Albanana Forbes noted on June 12. Fraudsters can use the technology to create lip-synced conversations, fake firmware, and impersonate others in video chats, where unsuspecting victims share personal and financial information.
Preparing for the future of AI
Hinton resigned from his job at Google in May 2023 after a decade with the company so he could speak freely about the risks posed by artificial intelligence, as he told the New York Times at the time. However, the worst-case scenarios are not guaranteed, and industries such as healthcare have already benefited significantly from AI.
During his media appearance, he called for more research to understand artificial intelligence, government regulations to rein in the technology, and a global ban on AI-powered military robots. Any AI safeguards put in place—whether by tech companies, the US federal government, or other governments worldwide—need to happen soon.
Tips for protection against scams
In Monday’s podcast interview, Hinton offered listeners advice to protect themselves from enhanced AI scams: diversify where you keep your money, and regularly back up your data to an external hard drive. If you fall victim to a scam, you will only lose some of your money, and you can quickly restore your computer and phone to working condition.
Hinton keeps his own money in three separate banks. More broadly, diversifying your assets is a smart financial move, according to financial analyst and personal finance expert Chris Browning. He recommends distributing your money across five different accounts: two separate checking accounts for bills and lifestyle spending, and three savings accounts for long-term savings, short-term savings, and an emergency fund.
Conclusion
Hinton’s warnings come from a scientist who spent his life developing artificial intelligence technologies, lending special weight to his words. Although the percentage he estimates may seem low, a 10-20% chance of an existential threat requires serious attention from the scientific community, governments, and corporations. The current time is critical for establishing the necessary safeguards before the technology becomes more advanced than can be controlled.
Keywords: Geoffrey Hinton, AI risks, AI safety, future technology
