George Uzon/AFP via Getty Images
Rejoice, because the year 2025 is finally over.
During our planet’s latest and seemingly endless revolution around the Sun, the tech industry’s obsession with AI reached even more incredible heights. CEOs began openly boasting about replacing their subordinates with AI “agents.” The phenomenon of so-called AI psychosis became a national news story as more people were put on edge by their silver-tongued chatbot companions. “slope” took on a new meaningAnd the word “circular” suddenly started being used in the same sentence as “billions of dollars” or even “hundreds of billions of dollars,”
Will 2026 finally liberate us from this endless cavalcade of big language model madness? According to Geoffrey Hinton, computer scientist and “godfather” of AI, this is unlikely. He predicts that AI will continue to improve over the next year and will reach a point where it will free us from all our terrible low-paying jobs.
“I think we’ll see AI get even better,” Hinton said. during an interview But cnn State of the Union on Sunday. “It’s already very good. We’ll see that it has the potential to replace many jobs. It’s already been able to replace jobs in call centers, but it’s going to be able to replace many other jobs as well.”
Hinton was one of three recipients of the prestigious Turing Award in 2018 for his work on neural networks, which formed the basis of modern AI, earning him the nickname of being the “godfather” of the field.
In 2023, Hinton announced that he regretted his life’s work after stepping down from his role at Google, where he had been for more than a decade. Since then, he has become one of the most prominent destroyers of technology.
during cnn In the interview, Hinton was asked whether he was more or less concerned about AI since making his now infamous announcement.
“I’m probably more worried,” Hinton replied. “It has advanced even faster than I thought. In particular, it has also become better at things like reasoning and deceiving people.”
According to Hinton, AI is progressing so rapidly that approximately every seven months it can complete tasks that took twice as long as before. He predicted that it would be only a matter of years before an AI would easily complete software engineering tasks that would take a human a month to complete.
“And then far fewer people will be needed for software engineering projects,” Hinton said.
Hinton made similar gloomy predictions in a conversation with Senator Bernie Sanders last month, Saying Tech leaders are “betting on AI to replace a lot of workers.”
However, it still remains to be seen whether AI will actually progress that much. Many attempts to replace workers with semi-autonomous AI models have failed, while some new models like OpenAI’s GPT-5 have shown only weak improvements.
More on AI: After outcry, Firefox promises “kill switch” that will turn off all AI features
