Let's go back for a second because you're saying as much about humans as you are about machines. Is there anything special about human beings -- our consciousness, or what people call our souls?
I'm hypothesizing -- and I think most biologists hypothesize the same thing -- that we are nothing more than bio-molecules interacting. Now, within that organization, there's obviously a specialness to us which gives us consciousness, which a rock doesn't have. But, again, in principle, I don't see at this point why we couldn't build a machine that had those attributes. Whether we are smart enough to build such a machine is another question.
Perhaps we really will be sharing the earth with conscious humanoids
I believe that's where we will end up. But even though a raccoon has good manipulation capabilities, nobody thinks a raccoon is smart enough to build a robot raccoon. And maybe we're just not smart enough to build a robot human. That could be.
Flesh and Machines: How Robots Will Change Us
By Rodney A. Brooks
Pantheon
288 pages
Nonfiction
Would such robots have the desire to survive and the capability to reproduce themselves?
We certainly don't know how to build such machines, but I don't see why that shouldn't be possible.
On the flip side of things, there's this sci-fi hope out there that human beings will be able to download their minds into machines to live beyond their mortal bodies.
I don't know whether that's going to be possible. It may be that our individual consciousness is so tied up with our own individual brains and development that in the foreseeable future, i.e., the next three or four hundred years, we're not going to be able to do that. That's an unknowable for us.
What problems or challenges are taking up your own brain space these days?
This is not something I am personally working on, but the big open question is in computer vision or robot vision. Over the last few years, our vision systems have gotten really good at tracking moving objects, recognizing faces and recognizing human bodies. But they are still quite lousy at things that a 2-year-old can do: tell whether someone is old or young, tell whether that's a cup in front of them or a tape dispenser or a telephone. They just can't do those things. And we've been trying to [teach robots to] do them for 40 years. So, I'm looking for two or three young Einsteins to come along and figure out what needs to be done there. Because I think we just haven't got it.
While that search is on, what are you doing?
You know, people have been asking me for a long time these questions about whether robots can really have emotions, what's really lifelike and what's living. And so I'm interested in a more fundamental question: What's the difference between living matter and non-living matter -- way down at even the bacterial level? What are the organizational patterns that make something alive? My hypothesis is that there's some deep scientific understanding that we haven't yet hit upon. That's what I've been working on over the last year or so.
Essentially: "What is life?"
Yes, what is life? Now, I recognize that one or two people have worried about this before. This is not a new question. But I hope we're coming at it from a few new angles. I have a research group devoted to this and this is what I'm working on. I think that until you actually do something you don't know how close you are.
Needless to say, there are a lot of people who already have their answer, namely God.
Of course a lot of people will think that, but as an atheist I am convinced there is a material explanation. In earlier times God was responsible for moving the sun across the sky every day, but later we learned that asking how the "sun moved" wasn't even the right question. I expect that there is a similar "answer" to the difference between living and non-living matter.