Geoffrey Hinton is credited with “inventing” AI and he won the Noble Prize in machine learning in 2024 in the process. He tells a CBS interviewer that he has real concerns about advancing AI and the risks that are being ignored with potentially trillions of dollars to be made. Hinton is a very credible source, not only because he was the one who started it all and knows the concept intimately, but because he comes across as a very sincere and rationale person.
Hinton was a professor working on research to model the brain. In an experiment in the mid-80s, he performed an experiment that was the precursor to AI, producing the very concept that large language models have at their foundation.
Hinton compares the rapid increase in the evolution of AI to getting too friendly with a cute tiger cub. If you continue around the tiger as it gets older, you have to have confidence that the beast will not kill you. Hinton says “you should worry” describing the pace of AI development and its potential consequences. He says that there is a 10-20% risk that AI will “take over” from humans.
Hinton believes that the rush to make trillions on AI is blinding the developers to the risks and causing them to push safety measures to the side, allocating a small percentage of computer time and budgets to safety considerations compared to the amount required. He is glad he is 77, presumably because an AI future is not one he wants to hang around and see.
With advances in quantum computing, the pace of AI might soon be on steroids.
