Friday 20 March 2015

Would You Worry About Robots That Had Free Will?

Its perhaps a scary thought, a very scary thought: an intelligent robot with free will, one making up the rules for itself as it goes along. Think Terminator, right? Or maybe the gunfighter character Yul Brynner plays in "Westworld", a defective robot that turns from being a fairground attraction into a super intelligent robot on a mission to kill you? But, if you think about it, is it really as scary as it seems? After all, you live in a world full of 7 billion humans and they (mostly) have free will as well. Are you huddled in a corner, scared to go outside, because of that? Then why would intelligent robots with free will be any more frightening? What are your unspoken assumptions here that drive your decision to regard such robots as either terrifying or no worse than the current situation we find ourselves in? I suggest our thinking here is guided by our general thinking about robots and about free will. It may be that, in both cases, a little reflection clarifies our thinking once you dig a little under the surface.




Take "free will" for example. It is customary to regard free will as the freedom to act on your own recognisance without coercion or pressure from outside sources in any sense. But, when you think about it, free will is not free in any absolute sense at all. Besides the everyday circumstances of your life, which directly affect the choices you can make, there is also your genetic make up to consider. This affects the choices you can make too because it is responsible not just for who you are but who you can be. In short, there is both nature and nurture acting upon you at all times. What's more, you are one tiny piece of a chain of events, a stream of consciousness if you will, that you don't control. Some people would even suggest that things happen the way they do because they have to. Others, who believe in a multiverse, suggest that everything that can possibly happen is happening right now in a billion different versions of all possible worlds. Whether you believe that or not, the point is made that so much more happens in the world every day that you don't control than the tiny amount of things that you do.

And then we turn to robots. Robots are artificial creations. I've recently watched a number of films which toy with the fantasy that robots could become alive. As Number 5 in the film Short Circuit says, "I'm alive!". As creations, robots have a creator. They rely on the creator's programming to function. This programming delimits all the possibilities for the robot concerned. But there is a stumbling block. This stumbling block is called "artificial intelligence". Artificial Intelligence, or AI, is like putting a brain inside a robot (a computer in effect) which can learn and adapt in ways analogous to the human mind. This, it is hoped, allows the robot to begin making its own choices, developing its own thought patterns and ways of choosing. It gives the robot the ability to reason. It is a very moot point, for me at least, whether this would constitute the robot as being alive, as having a consciousness or as being self-aware. And would a robot that could reason through AI therefore have free will? Would that depend on the programmer or could such a robot "transcend its programming"?

Well, as I've already suggested, human free will it not really free. Human free will is constrained by many factors. But we can still call it free because it is the only sort of free will we could ever have anyway. Human beings are fallible and contingent beings. They are not gods and cannot stand outside the stream of events to get a view that wasn't a result of them or that will not have consequences further on down the line for them. So, in this respect, we could not say that a robot couldn't have free will because it would be reliant on programming or constrained by outside things - because all free will is constrained anyway. Discussing the various types of constraint and their impact is another discussion though. Here it is enough to point out that free will isn't free whether you are a human or an intelligent robot. Being programmed could act as the very constraint which makes robot free will possible, in fact.

It occurs to me as I write out this blog that one difference between humans and robots is culture. Humans have culture and even many micro-cultures and these greatly influence human thinking and action. Robots, on the other hand, have no culture because these things rely on sociability and being able to think and feel for yourself. Being able to reason, compare and pass imaginative,  artistic judgments are part of this too. Again, in the film Short Circuit, the scientist portrayed by actor Steve Guttenberg refuses to believe that Number 5 is alive and so he tries to trick him. He gives him a piece of paper with writing on it and a red smudge along the fold of the paper. He asks the robot to describe it. Number 5 begins by being very unimaginative and precise, describing the paper's chemical composition and things like this. The scientist laughs, thinking he has caught the robot out. But then Number 5 begins to describe the red smudge, saying it looks like a butterfly or a flower and flights of artistic fancy take over. The scientist becomes convinced that Number 5 is alive. I do not know if robots will ever be created that can think artistically or judge which of two things looks more beautiful than the other but I know that human beings can. And this common bond with other people that forms into culture is yet another background which free will needs in order to operate.




I do not think that there is any more reason to worry about a robot that would have free will than there is to worry about a person that has free will. It is not freedom to do anything that is scary anyway because that freedom never really exists. All choices are made against the backgrounds that make us and shape us in endless connections we could never count or quantify. And, what's more, our thinking is not so much done by us in a deliberative way as it is simply a part of our make up anyway. In this respect we act, perhaps, more like a computer in that we think and calculate just because that is what, once "switched on" with life, we will do. "More input!" as Number 5 said in Short Circuit.  This is why we talk of thought occuring to us rather than us having to sit down and deliberate to produce thoughts in the first place. Indeed, it is still a mystery exactly how these things happen at all but we can say that thoughts just occur to us (without us seemingly doing anything but being a normal, living human being) as much, if not more, than that we sit down and deliberately create them. We breathe without thinking "I need to breathe" and we think without thinking "I need to think".

So, all my thinking these past few days about robots has, with nearly every thought I've had, forced me into thinking ever more closely about what it is to be human. I imagine the robot CHAPPiE, from the film of the same name, going from a machine made to look vaguely human to having that consciousness.dat program loaded into its memory for the first time. I imagine consciousness flooding the circuitry and I imagine that as a human. One minute you are nothing and the next this massive rush of awareness floods your consciousness, a thing you didn't even have a second before. To be honest, I am not sure how anything could survive that rush of consciousness. It is just such an overwhelmingly profound thing. I try to imagine my first moments as a baby emerging into the world. Of course, I can't remember what it was like. But I understand most babies cry and that makes sense to me. In CHAPPiE the robot is played as a child on the basis, I suppose, of human analogy. But imagine you had just been given consciousness for the first time, and assume you manage to get over that hurdle of being able to deal with the initial rush: how would you grow and develop then? What would your experience be like? Would the self-awareness be overpowering? (As someone who suffers from mental illness my self-awareness at times can be totally debilitating.) We traditionally protect children and educate them, recognising that they need time to grow into their skins, as it were. Would a robot be any different?




My thinking about robots has led to lots of questions and few answers. I write these blogs not as any kind of expert but merely as a thoughtful person. I think one conclusion I have reached is that what separates humans from all other beings, natural or artificial, at this point is SELF AWARENESS. Maybe you would also call this consciousness too. I'm not yet sure how we could meaningfully talk of an artificially intelligent robot having self-awareness. That's one that will require more thought. But we know, or at least assume, that we are the only natural animal on this planet, or even in the universe that we are aware of, that knows it is alive. Dogs don't know they are alive. Neither do whales, flies, fish, etc. But we do. And being self-aware and having a consciousness, being reasoning beings, is a lot of what makes us human. In the film AI, directed by Steven Spielberg, the opening scene shows the holy grail of robot builders to be a robot that can love. I wonder about this though. I like dogs and I've been privileged to own a few. I've cuddled and snuggled with them and that feels very like love. But, of course, our problem in all these things is that we are human. We are anthropocentric. We see with human eyes. This, indeed, is our limitation. And so we interpret the actions of animals in human ways. Can animals love? I don't know. But it looks a bit like it. In some of the robot films I have watched the characters develop affection for variously convincing  humanoid-shaped lumps of metal. I found that more difficult to swallow.  But we are primed to recognise and respond to cuteness. Why do you think the Internet is full of cat pictures? So the question remains: could we build an intelligent robot that could mimic all the triggers in our very human minds, that could convince us it was alive, self-aware, conscious? After all, it wouldn't need to actually BE any of these things. It would just need to get us to respond AS IF IT WAS!


My next blog will ask: Are human beings robots?


With this blog I'm releasing an album of music made as I thought about intelligent robots and used that to help me think about human beings. It's called ROBOT and its part 8 of my Human/Being series of albums. You can listen to it HERE!

No comments:

Post a Comment