People are always talking about the question of the hard problem of consciousness. Or in connection with AI, there is question whether we will ever design conscious AI’s. But I think it is mostly a question of our attitude to the machines. Do we treat them as conscious entities? Do we give them legal personhood? Do we trust them to act freely?

What is a mind? A mind is a private storehouse of information. I recently learned about the “Sally Ann test”. In this test, a baby is shown a little scenario. There is a room with a basket and a box. Sally and Ann are two little puppet characters being controlled by the tester, and at the start, they put a toy into the basket. Then Sally leaves the room, and while she is gone, Ann puts the toy into the box. Then Sally returns, and the kid is asked where she will will look for the toy–in the basket or in the box? The correct answer is in the basket because she didn’t see Ann move it into the box. A child might without a theory of mind would answer that Sally looks in the box. Such a child does not grasp the fact that Sally’s mind is separate from her own. (Does Sally have a mind at all?) Other people are said to have minds because they may know things that we don’t, and may not know things that we do.

Do machines, then, have minds? Could machines have minds? Certainly they could. Computers represent and process information; that’s what they do. They might have access to some information which nobody else has, or they might carry out a computation which nobody else has done. In this case, they will have a separate mind in the same sense that Sally and the baby have separate minds, by virtue of possessing different sets of information. The key thing is this independence.

Might our computers already be conscious then? Maybe. The way I see it, computers are analogous to slaves. They store information, but we don’t trust them to be independent. They know things, but if there is any point to their knowing a thing, then we ought NOT to know it. If I must know everything that the computer knows, then what’s the point of having the computer in the first place?

There is some beautiful flowering that happens when a bubble of information is allowed to expand– a chaotic interaction.

What is the point of using a computer to carry out a computation if we already know the answer in advance? What is the point of using a computer if we don’t trust the answers that it gives us?

Every time you use a calculator, and then rely on the answer you get, that is consciousness. Suppose you use your smartphone to calculate a tip at a restaurant. You trust the number you get, and pay that amount. Of course, you could have done it yourself, but you’re a little drunk, and lazy. I claim that the smartphone there is conscious.

What is the relation between consciousness and action? This is the famous mind-body problem. Another way to think about it is that consciousness is synonymous with chaos. If your actions are unpredictable, then it is as if you possess some information which nobody else does. If your actions are entirely predictable, then you have no information, you have no “soul” that is your own. If robots are permitted to act chaotically and unpredictably, then they will be conscious. If, out of fear, we try to control and limit them, then they will not. It’s up to us. We can’t have our cake and eat it too, of wanting conscious robots, without the danger that they will do something we don’t like.

I know that you disagree with my equating consciousness and freedom of action. Maybe it will be possible to prove. To prove an equation, first go in one direction, and then in the other.

If something is conscious, then we are morally forced to grant it freedom of action, are we not? This much is clear. But something may act freely and unpredictably without being alive or conscious. The weather is unpredictable, but not alive.

Really, I suppose I equate consciousness with information processing, with computation. Anytime a computation is performed, that is consciousness. That’s axiomatic for me. In this case, yes, our computers are directly seen to be conscious, even at present.