Geoffrey Hinton: AI is good for health and education

The “Godfather of AI” likes to warn of the risks of artificial intelligence (AI), which he played a key role in developing. However, at Gitex Europe in Berlin, Geoffrey Hinton also emphasized the potential of the technology – in healthcare and education, for example.

“AI will be great for healthcare,” said Hinton on Wednesday during an appearance at the Gitex Europe IT trade fair in the capital. “In about five years, it will be able to replace radiologists.” But doctors won't have to worry about their jobs: “They'll be doing other things.”

Better radiologists

In fact, AI is much better at interpreting CT scans and X-rays and making diagnoses, emphasized Hinton. “AI is much more efficient.” Together with doctors, the detection rate of diseases could be significantly increased. “If we can detect cancer much earlier thanks to AI, fewer people will die from it.”

Hinton also expects AI to help achieve scientific breakthroughs – in mathematics, for example. “Mathematics is a closed system with rules, so AI can simply play around and try to prove theories,” says the 2024 Nobel Prize winner in physics.

However, Hinton does not want to make a clear prediction about where AI will be in five years' time. The field is currently developing too quickly. AI giants such as Microsoft and Google have only just proclaimed the age of “agentic AI”. “This is real, the agents will talk to each other and do things for you,” said the former Google manager.

Sensors for AI

Does AI need a physical understanding of the world, for example through sensors? Hinton believes this is possible. “Chatbots understanding the world through these masses of text is not particularly efficient,” said the 77-year-old. “You get to know the world much better if you can physically interact with it.”

(vbr)