W Daniel Hillis – How I see artificial intelligence now (248/248)
1,557 views · Published 30 August 2017 · 3:16 · Indexed 14 November 2026
Channel: Web of Stories - Life Stories of Remarkable People · 2017 · Science & Technology
To listen to more of W Daniel Hillis’s stories, go to the playlist: https://www.youtube.com/playlist?list=PLVV0r6CmEsFzo005LM6PTERy4uxq92sWe Born in 1956, US inventor, scientist, engineer and visionary W Daniel Hillis pioneered the concept of parallel computers and co-founded Thinking Machines and Applied Minds, which marked a new era in computing and established Hillis as a computing legend. [Listeners: Christopher Sykes and George Dyson; date recorded: 2006] TRANSCRIPT: I've changed my idea of what I think artificial intelligence is going to be like. I think when I first started thinking about it, I kind of imagined the science fiction I Robot of an embodied artificial intelligence of being very much like a human, but made of electronic parts or something like that. And it's not that I think that's impossible to build, but I don't think that's the interesting thing. I mean one of the great things about an intelligence that's freed from the bounds of a human body is it can do things that humans can't do. It can have much larger memory capacity. It can be in many places at once. It can use as its eyes all of the sensors that it's connected to. It can learn something in one experience and apply it to the others. So if there are self-driving vehicles, I don't think of it as an intelligent vehicle. It's a system of vehicles. So if one of them learns a lesson and realises there's a pothole someplace or a dangerous intersection, there's no reason all of them can't know that. And so I think we fall into the trap of imagining an artificial intelligence as being too analogous to human intelligence. So people ask what seem to me silly questions sometimes, like they say, 'Well, there's a self-driving vehicle. Would it be willing to sacrifice the driver to prevent running over a pedestrian?' As if it's going to have a little brain in it like the driver does. But actually, that decision is going to be made the system, by society, by the laws we set, by all the experience of all the self-driving vehicles. And so in some sense, that's a much better way of making a decision than having the driver sitting in the driver's seat make that decision. It's going to be a much more informed decision that a lot more participate in and is done for much less selfish reasons. So because that decision is going to be made by the system of all the artificially intelligent vehicles, not by a single individual self-preserving AI brain. So I now think of intelligence as kind of extended intelligence. It's the intelligence of all the machines working together and all the people working with them. So it's the kind of intelligence, in some sense, like is in Google right now. Google isn't really machine intelligence, it's not really human intelligence, it's some combination of humans and machines that are able to find things that you're searching for. And I think AI is going to be like that. It's going to be a combination of machines and people and so it's that extended group of intelligences that's going to be the thing that does the thinking. So that's how I see artificial intelligence now. And that's the thing I want to be proud of me. [CS/GD] So you're pretty far along. We're getting there
More from this channel
-
0:44
David Weatherall - Entering medical school
-
1:33
Katharine Whitehorn - Embodying the 'Sluts' article
-
6:55
KG Subramanyan - Mohandas Gandhi's views on art
-
2:22
Baruch Blumberg - The link between some cancers and viruses
-
5:03
Carl Djerassi - How Picasso sculpture unveiled problems in my marriage
-
6:39
Renato Dulbecco - Three Nobels, just one class
-
2:26
Marek Edelman - Different criteria for measuring the value of human life
-
2:45
Jules Engel - So who is responsible for a great animation?