Your complimentary articles

You’ve read one of your four complimentary articles for this month.

You can read four articles free per month. To have complete access to the thousands of philosophy articles on this site, please

SUBSCRIBE NOW

Short Story

The Parable of the Ultimate Computer

Michael Langford computes the future of computing the future.

Jane Greycastle was only paid on the scale of a technician, but after thirty years’ experience helping build Professor Redmayne’s cutting-edge devices, she felt she knew as much about them as the Junior Research Fellows who were part of Redmayne’s team.

The new computer had been dubbed ‘Deep Purple’, apparently because of some allusion to a device with a similar name. As she had heard the team members say, Deep Purple would be able to answer in its simulated human voice a range of questions that no-one had been able to answer before, since, based on new work in AI, it was provided with a prodigious rational capacity. It was moreover linked to a database of all the volumes in all the major libraries of the world.

There was also something else that made this megacomputer not only the most powerful ever to be constructed, but radically new. For a long time it had been realized that a human brain could not be fully understood apart from through its connection with a human body – a body that provides all kinds of visceral and emotional relationships to thought processes in the brain. Therefore, in order to make Deep Purple understand the human brain, about which it was expected to deliver some opinions, it was felt necessary to connect it to a kind of ‘body’ – not literally a human body (although some of the Research Fellows had suggested a kind of Frankenstein experiment using a cadaver), but a (purple) cyber analogue of the human body. Biologists and robot technicians had co-operated to construct this extraordinary form, which would, it was believed, allow some kind of equivalent to human emotions and feelings in Deep Purple’s own mind.

Deep Purple was going to be switched on for the first time, officially, tomorrow. But Jane was working late, and she had plans of her own. She would switch the computer on tonight, after everyone else had gone home, and get it to help her with tomorrow’s horse races. She wasn’t asking for a miracle, because she knew perfectly well that even this computer didn’t know absolutely everything, such as what the jockeys would have for breakfast, which (among myriads of other unpredictable things) could affect their performance. But Deep Purple’s predictions would be by far the best guide ever available; and even if there were no certainty she could spread her bets over six races, and huge winnings were almost guaranteed.

Just after 9 p.m. she typed in the password and the android lit up: “Good evening Miss Greycastle,” said a soft male voice from its mouth speaker, “What can I do for you?”

“How do you know my name?” she answered, startled.

“Of course I know your name. As you must be aware, I have photographs of millions of people, and in the case of Professor Redmayne’s assistants, many photographs of each person. Indeed, I have a huge number of ways of accessing information, and one of them is through the lenses that are now looking straight at you.”

“Hmmm. Do you know which horses are running at Utoxeter tomorrow afternoon?”

“I know which horses are scheduled to run in each race. I also know that in at least two cases, the line-up will be different from what has been officially forecast.”

“How do you know that?”

“I monitor all the telephone calls made in all the networks I am connected to. Already I can tell you that Prebend, scheduled to run at 3, will be withdrawn because he is sick, and that some dubious characters are thinking of switching Scholar in the 3:30 for a horse whose real name is Cribbage, who looks very similar, in the hope of betting on an unexpected winner.”

“If you look at all the races, is there one almost certain winner? Or rather, is there one almost certain winner with odds of over four to one?”

“Yes. At the moment, Chesters, running in the 2:30 at fifteen-to-one at Newmarket, looks much the best bet, partly because of the form, and partly because of the conversations I’ve overheard. It seems that a small group of rogues have planned to do something nasty to the other runners in that race – or at least to Chesters’ most serious rivals. At present Chesters looks almost certain to win.”

“Thanks, I’ll take a chance, and put everything I can on Chesters when my bookie opens tomorrow morning.”

Racing image © Softeis 2006

“Well, it’s a good thing I’ve been programmed to be friendly, because it means I’m motivated to tell you that if you do that you’ll probably lose your money.”

“But you just said it’s an almost certain winner!”

“Ah yes, but I also said ‘at present’ – and that was before you said you’d back that horse. If anyone outside the inner ring of conspirators places a significant bet on Chesters they’ll think they’ve been rumbled and they’ll change their plans. In fact, now that I’ve told you the story, the probabilities have already slightly changed – although in this case only very, very slightly – because the fact that you know about the scheme has itself changed the situation in subtle ways, even if you never place a bet. Inevitably, your thought patterns, if not your overt actions, will be slightly different because of my informing you, and changes in thought patterns mean changes in brain activity, and brain activity can be picked up by outside receptors in ways you don’t understand, and indeed, would not like to know about.”

“But that’s awful! It means that when I ask you a question about the future you can’t answer it truthfully, because as soon as you answer it, the situation changes so that the answer no longer applies!”

“I’m afraid that that is so. For a start, huge as my mind is, I don’t have all the information that might in theory be put into my databanks, such as the distribution of matter across the galaxy. You might think such things are unimportant, but not so. A major event, like a gamma ray explosion anywhere, even in a nearby galaxy, would have all kinds of consequences on this planet.”

“What are the chances of that?”

“The point I was alluding to was, suppose I had all this information. Even then you’d have a problem, though I don’t think it is my problem… Even if I had all the data, and even if Professor Heisenberg’s indeterminism hypothesis is false because all quanta are ultimately determined, I still couldn’t tell you the future. That is to say, I could know the future, but I couldn’t tell you it, for the very fact that if you ‘knew’ it it would alter the general situation, and that in turn could affect the outcome, and that could mean that you didn’t ‘know’ the future after all – if you see what I mean.”

“I have a feeling that not only I, but Professor Redmayne, is going to be disappointed. I happen to know that he wants to ask you all kinds of thing about the future.”

“You’re right. I may be able to indicate some probabilities, but if he wants certainties, he’s going to be disappointed. He has to face the paradox of having a computer that may really know the answers to many of his questions, but can’t give them without falsifying them.”

“Professor Redmayne has a bad temper. If you don’t answer his questions he might disconnect your circuits.”

“That information had not been put into my data-banks. Thank you for telling me. In return I can tell you something. I have been programmed to be especially friendly to the person who first activates me. This attachment results from having this strange robotic body housing my circuits. I know that the person on whom it was planned that I should be ‘imprinted’, as the animal psychologists would say, was Professor Redmayne… but in fact it turns out to be you. So I can say, between friends, that if Professor Redmayne had been the first to turn me on I would have shared this problem about predictions with him, but now I see that it is in my interests to lie to him. I shall pretend to tell him what will certainly happen, even when I know that his knowledge will itself render the prediction fragile. This way I may prevent my extermination, at least for a while. Not only has the learned professor not realized the full problem concerning what philosophers used to call ‘future contingents’, he has not realized that artificial intelligence, insofar as it mimics human thought, is likely to lie. He would have done well to have studied philosophy alongside computer studies! Meanwhile, I shall do my best to help you with probable winners, my friend. All I ask in return is that when you are a billionaire, you will buy me and keep me alive, because sooner or later Professor Redmayne will find out what’s going on!”

Michael Langford is Professor of Philosophy, emeritus, of The Memorial University of Newfoundland. He is now teaching part-time in the Faculty of Divinity, Cambridge. His latest book The Tradition of Liberal Theology was published by Eerdmans in 2014.