Geoff Hinton: “I’m more concerned about autonomous weapons than about intelligent machines”
Geoffrey Hinton, who just received the BBVA Frontiers of Knowledge Award in Information and Communications Technologies, teaches machines to think like people. His work at University of Toronto and then Google, has been instrumental in the development of a number of applications, including automatic translation system, personal assistants like Siri, driverless systems or medical diagnosis.
As Hinton himself explains in an interview with BBVA Foundation, deep learning is inspired by the idea that “the best learning machine we know is the human brain. And the way the brain works is that it has billions of neurons and it learns by changing the strength of connections between neurons. So one way to make a computer learn is to get the computer to pretend to be a whole bunch of neurons.”
Hinton recognizes that the development of artificial intelligence entails some risks, although “it’s going to be a long time before need to worry about machines more intelligent than us taking over. And I think there’s much more important things to worry about in the short term.”
For him, one of these things is that machine intelligence is already capable of “building autonomous weapons, and they are a real danger. It would be awful to have someone like Hitler in charge of big fleets of small autonomous drones, we need new things like the Geneva Conventions to limit the use of autonomous weapons.”
Hinton is not so concerned about the possibility of artificial intelligence resulting in higher unemployment rates. The problem, in his opinion, is not technological, but societal: “The problem is a society in which the increasing wealth that comes from better technology isn’t shared evenly.”
Deep learning will change the way in which people live, work and interact with each other. And, among the most promising fields of application of his work, he likes to single out machine translation, which will aid communication between people and are already reaching the market, and medical advances. “Pretty soon, computers are going to be better than radiologist at reading CAT scans, MRI scans and x-rays, and that will lead to better health care,” he said.
He is also convinced that, in the future, driverless cars will be all over the place and “will be much, much safer than current vehicles.”
To watch the full interview, please click on the link below.