‘Oh, yes,’ said Rowe. ‘So far as I know.’

‘No one has ever suggested any other reason, have they?’

‘I don’t think so.’

‘No, of course not. But one needs to get these fundamental considerations straight before one builds on them. Anyway, if that is so, I think we can assume that a computer is a more efficient statistics-producing machine than any possible combination of horses, dogs, or muscular young men. But that’s not the end of it. For what have we shown? That any human activity which consists of repetition, or of manipulating variables identifiable in advance according to predetermined rules, or of manipulating a known range of variables at random, can in theory at any rate be performed by a computer.’

‘Yes,’ said Rowe. ‘That’s really the basic, er, the basic . . .’

‘. . . principle on which our work is founded. So it is. So we automate the bingo industry. But now consider the activity of the person who plays bingo. Isn’t his activity – covering the numbers on a board as they are called and signalling when a certain pattern is completed – precisely analogous? Couldn’t his work be done for him by a computer fed with a very simple programme indeed?’

‘Yes,’ said Rowe. ‘Yes, it could.’

‘And will be. No one is going to spend hours in a bingo hall if he can simply pay over his money and have a computer play for him instead. Filling up a football coupon is another job which a computer could easily be programmed to do. You could set it to fill the coupon at random, or to make a selection based on any system you liked, or to select its own system, or to choose between a random choice and a system of its own choosing. Once again we have a range of variables which can be identified in advance and manipulated according to predetermined rules. It’s programmable. Having a man to perform it is a waste of time.’

‘Y-e-e-e-es,’ said Rowe. ‘Y-e-e-e-es.’

‘And when you come to think about it, you could programme a computer to appreciate the cricket results for you – or even to appreciate the actual automated game as the playing computer played it. It would be instructed to register applause at amazing freaks – at, say, the announcement of a slip catch off a fast bowler in poor light. It would register annoyance when the side it was instructed to identify with suffered a reverse – annoyance mixed with reluctant admiration if it was the result of the other side’s skill. It would be programmed to register boredom when nothing unusual happened for some time – without any boredom being inflicted on any actual human being.’

‘But the point of watching cricket, surely,’ said Rowe, ‘is actually to see and appreciate the skill of the players.’

‘Then why are so many people content to listen to it being described on the wireless? In either case the activity of the spectator or the listener is the same – to register a selection from a range of reactions in correlation to the permutations of variables he is offered. So far as I can see, it’s a finite activity, which means it’s a programmable one. The spectator is eminently replaceable.’