The Robots Are Getting Closer - 11
1/13/2003

Home

Future Index

The Long-Term Roles of Robots
If It Seems to You That I'm Rambling....
    If it seems to you that I'm rambling, that's because I am. In contrast to the usual arrangement, in which someone writes up their thoughts after the thoughts have had time to settle, I'm writing up my thoughts as I go. 
Enabling Technologies
    It seems to me that there are what I've called "enabling technologies", which really represent coping skills, and there are "core technologies" that address adaptive learning, motivations, mind states, emotions, the setting of goals, volition, and so forth. Examples of mental skills would be speech recognition, speech synthesis, computer vision, et al. Examples of "core technologies" would be visual pattern recognition, or data mining--detection of correlations among data--and the resulting synthesis of predictive patterns. 
Core Technologies
    It's worth noting that "core technology" capabilities are advancing rapidly in the form of intelligent computer programs. For example, Microsoft Word runs an unobtrusive spelling checker that redlines spelling errors. Recent versions of Windows try to adapt to their users' preferences. And there are many programs that attempt to analyze stock market trends for predictive purposes (leading to major investors playing against one another). Adaptive software is used for optical character recognition. This probably pales into insignificance in comparison with what mammalian brains accomplish, but it may be laying the groundwork for such higher level functions.
Most Devices Will Become More Flexible But Not Sentient
    My guess would be that over the near-term, robotics devices will utilize various skills that are orchestrated with simple,  pre-programmed IF-THEN types of control programs. I could imagine that cognitive brilliance won't be sought in many kinds of robotic devices. We're not going to pay extra for lawn-mowing robots that will argue with us over whether or not we should invade Iraq, or that can understand what's really going on with us, or that are secret members of an "equal-rights-for robots" underground. Dumb brutes will be just fine. They can easily be given the flexibility to avoid running over the dog, or a letter on the grass.
There Will, Perhaps, Be Applications for Sentient Robots
    There will be applications in which higher cognitive functions will be desirable. Inorganic companions come to mind. However, I think this will be gradual. I suspect that the development of complex robotic minds and bodies will be a highly challenging undertaking, spread over decades. 
The Key Issues Are Emotions, Motivations, Self-Awareness, and Volition
    I think the key issues are those of emotions, motivations, self-awareness and volition. There are those who question whether or not it can be done, but if it can be done, I think that it should be done with great caution. Of course, I don't think that one maverick robot could take over the world, particularly if, as the possibility draws closer, we erect safeguards and barriers to assure that it can't happen.
Emotions, Motivations, Self-Awareness, and Volition Arise Well Below Humans on the Phylogenetic Ladder
    It doesn't require high intelligence to develop emotions, motivation, and volition. Consider a mule, or a horse trying to get home to its stall. If we want to experiment with volition, it probably ought to be done with an isolated computer that doesn't have much brainpower--e. g., a-a-a   v-e-e-e-e-r-y   s-l-o-o-o-o-w   th-i-i-i-i-nki-i-ng  c-o-o-o-mp-u-u-t-e-e-er. 
We Might Gradually Graduate to Organic/Inorganic Life
    Another scenario is one in which we gradually and voluntarily draw upon computer capabilities to a greater and greater extent. This could lead to a gradually increasing reliance upon "tame" computers. We might meld with artificial intelligence to a greater and greater degree as extensions of our own minds. For example, Google is giving us an enormously enhanced ability to perform library research. I'm using Google as a super-encyclopedia, among other things.
If Sentient Robots Existed, They Would Probably Be a Laid-Back Bunch
    Even if they were endued with emotions, motivations and volition, I would expect robots to be free of organic neurological imbalances and content. I understand this. I've mentioned elsewhere that I have no desire to travel, to possess a vacation home, or to own an RV, a boat, or a plane. I have no urge to produce a lot of offspring. I don't hanker after power or additional money. I'm content with what I have. I can imagine that robots would be that way, too. Robots would have no physical needs except fuel, repairs (spare parts), and air to burn the fuel (or electricity, if they ran on batteries). Robots wouldn't need to feel pain, and wouldn't need to worry about modest temperature variations. They wouldn't need clothes to keep warm. 
Robots Needn't Have Expansionist Agendas
    Ideas like the notion that inorganics (robots) would expand throughout space raise the question: why? To what purpose? I could imagine that inorganics might choose to explore space. There's something to be said for exploring, but not much for multiplying one's kind throughout space. Expansion into space won't solve an overpopulation problem. One's occupied volume of space can't expand faster than the speed of light, whereas unrestrained population growth is exponential. There might be some value to hedging one's bets by spreading one's seed, but that doesn't need to be done broadcast. A few colonies in widely separated stellar systems, or even in different galaxies, should insure long-term survival. Even if inorganics could achieve simultaneous transfer, there would be no incentive for colonizing every stellar system in sight (and even the interstellar medium). Colonizing space is a guy thing. If there are resources elsewhere that aren't to be found locally, then an inorganic civilization might choose to establish outposts where these resources present themselves, but logically, I would see an inorganic society choosing to use and re-use the raw materials in its native system while it explored scientific, rather than interstellar,  frontiers. (I would also see building larger and larger telescopes for the passive exploration of space, and the exploration of space near the Solar System, and of space near, but not too near another stellar system. I could see seeding stars without planets before approaching any planets that might have intelligent life.)
    It's interesting to consider our relationships to three-year-olds, to our pets, and even to wild animals. Even though we're many times smarter than they are, most of us nurture them, and are fond of them. Our treatment of wild animals is to give them space, and let them alone. We may peek at them and watch them, but we want them to live normally in their natural habitats. I would hope that if inorganic life ever does become ascendant over organic life, it will conserve us as we are conserving other species.
    In any case, it will be a while before these questions become imminent (Don't worry. The problem will receive plenty of attention before it draws close.)
I Think the Most Likely Scenario is "Business As Usual"
    As I've mentioned previously, today's supercomputers have 10 billion times the speeds, and 10, billion times the storage capacities of 1952 computers, and they can perform tasks that we could only dream about back in 1952, but they're still mindless machines. My hope and expectation is that, as they have over the last 25 years, computers will become tens of thousands, or even hundreds of thousands of times faster and more capacious over the next 25 years, and that, as they have over the past 25 years,+ they will become cleverer and cleverer, but that they won't be given the gift or curse of sentience. 
    Eventually, it might be different, but eventually, we might be different.

Back                                  1   2   3   4   5   6   7   8   9   10   11   12                                   Next