Marvin Minsky - 'I want an anti-fad fad' (151/151) скачать в хорошем качестве

Marvin Minsky - 'I want an anti-fad fad' (151/151)

To hear more of Marvin Minsky’s stories, go to the playlist:    • Marvin Minsky - The amazing videophone fro...   The scientist, Marvin Minsky (1927-2016) was one of the pioneers of the field of Artificial Intelligence, having founded the MIT AI Lab in 1970. Since the 1950s, his work involved trying to uncover human thinking processes and replicate them in machines. [Listener: Christopher Sykes]; date recorded: 2011] TRANSCRIPT: Now, what I want is an anti-fad fad. And if you look at my book The Society of Mind or the later one called The Emotion Machine, the idea there is that it’s nice to find a simple theory of how to make a machine smarter and there are a lot of those. What... the trouble is that none of those simple theories can solve very wide classes of problems. Each kind of learning machine or conditioning machine is good for solving some problems. Each prediction machine is good for solving other problems. And a system that simulates Darwinian evolution... trial and error and copying and breeding and so forth... evolutionary machines can solve other problems, but they don’t have the billions of years that Darwinian evolution had to... to develop human brains. So, what I’m looking for is just to have a few more people working on theories of how to combine other theories. It seems to me that the secret of the human brain and... why are we so much more resourceful than all the other animals put together?... is because we... we don’t work according to some particular clever... clever way of thinking. We’re not purely logical. We’re not purely memory based. We’re not purely rule based. But we have maybe 20 or 30 different strategies for learning and thinking and problem solving. And we have higher level management... we’re like committees of competing interests and competing ways to do things. And so, I’d like to see more people funded so that they can spend five or 10 years trying to develop these higher level, more complicated theories. There's just... and I see that this is very important for the future because when we have three or four billion older people and very few younger people, we’ll need these smarter machines to... to facilitate what they do in their everyday lives and... robotic vacuum cleaners are already pretty good, but there’s no machine that can make a bed and there’s... there's just so many things that every normal person can do. I don’t think any robot has ever put a pillow into a pillowcase. It would be fun to try, but I... I think probably everyone would lose that contest today.