Certainly the prospect of an afterlife mediated through the telephone system conjures up some bizarre ideas of hell, or limbo. "Your reincarnation is being queued until a body is available to receive it. Please hold until a body is free. Thank you for calling Afterlife Central. Your reincarnation is being queued..." and so on for all eternity.
BT's artificial life research team says it hopes to record the electrical pulses that pass down our nervous system - for example from the eye to the brain - onto memory chips. The basic principle is that once these pulses have been recorded from one person's eyes, they could be played back into someone else's brain. BT researchers are now talking about extending this to cover all the sensory inputs to the brain, so that the raw material of everything a person sees, feels, hears, smells and tastes could be captured in some kind of black box, using almost unimaginably powerful memory chips. The scientists at BT's Martlesham Heath laboratories in Ipswich call the chip the Soul Catcher.
One member of the BT team loftily predicted: "This is the end of death." In perhaps three decades, he said, it would be possible to relive other people's lives by playing back their complete recorded experiences via a computer. With the Soul Catcher, a new baby could be given the digitised lifetime experiences of a dead person.
It is almost pure science fiction: in William Gibson's world half the characters wear electronically enhanced "eyes" of this sort. BT's engineers are assuming the capacity of memory chips will continue to increase a hundredfold every decade, as they have done for the past 20 years. But even if they don't it would simply mean that one chip was not enough to hold a lifetime's experience.
The really interesting problems are not altered even if the black-box chip has to be changed every year - or every day. For what BT's engineers are assuming is that human experience can be digitised, and that when we see, hear, or feel, our brains are manipulating data according to complex rules, just as computers do. They are proposing as fact one of the classic thought experiments of artificial intelligence: the brain in a vat.
The brain in a vat is a brain with electronic wires grafted into its nerve endings which supply it with all the sensations it would have had from the outside world. In the original form of this experiment, the substitution would be performed slowly, sense by sense, as it is in Frederick Pohl's science fiction novel Man Plus.
The BT researchers seem to be proposing something altogether more ambitious: not virtual reality, but transferable reality. These speculations are very hot at the moment. More than a thousand scientists and philosophers attended a conference in Tucson, Arizona, in the spring, entitled "Towards a Science of Consciousness", where all manner of speculations about the nature of reality and of experience were bandied about.
Most of the people there believe they are standing on the verge of a revolution as profound and exciting as that brought about by Crick and Watson's discovery of DNA. The brain, after all, doesn't use any spooky technology so far as we can tell. Electrical signals are carried round it with the help of chemical reactions, and though the machinery for this is very complicated indeed, it is still just electric currents and chemical reactions. There can be no reason, why the pattern of pulses in your retinal nerve while you read this could not be duplicated in silicon, even if the technology to do this is nowhere in sight.
Yet there was a surprisingly strong backlash at Tucson against the sort of confidence the BT researchers take for granted. The world of consciousness research is deeply split between those people who believe that consciousness and experience are fundamentally nothing more than a form of information processing, and those who believe that our brains are both more complex and more biologically determined than that. As far as I could see, the balance of advantage in the argument is moving away from the information processors. Even the strongest proponents of "strong artificial intelligence", who believe we could create a computer that thought and experienced in the same ways we do, now agree it will require an extraordinary degree of complexity. Danny Hillis, a former whizz-kid at the American Massachusetts Institute of Technology - who founded a supercomputer company named Thinking Machines - suggested that you would need a network of computers a thousand times larger than the present Internet to begin to do useful design work on an artificial mind.
Among neuroscientists, the scepticism runs a great deal deeper. The brain and its processes grow more complex the more closely they are examined; and it becomes harder and harder to draw firm boundaries between information and its processing. The BT scheme for transferable experience is based on the idea that there is at some stage a feed of raw data from the sense organs, which is then processed by the brain. But this raw data is harder and harder to find. The distinction between the brain and the surrounding vat begins to look unreliable.
Dr Doug Watt, a neuroscientist at Quincey Hospital in Boston, was one of the most eloquent defenders of the essentially biological nature of our experience at the Tucson conference. He points out that all our experiences have an emotional content: they are not neutral information. "It is really almost impossible to separate cognition and emotional worth. Emotional worth is embedded in just about everything we do. It is essential for working memory. A recent project at Yale found there was nearly no such thing as an emotionally neutral word."
The essential thing that makes experience tolerable, he says, is that it is ours: integrated in a world we understand because it is made of our memories. When this integration breaks down, the result is psychosis, or delirium, not the kind of simple transfer of information envisaged by BT.
What is more, the structures of working memory are dependent on physical connections inside the brain. These vary from brain to brain, both genetically and as a result of experience. And the sorts of experience that can shape the brain change at different ages. A child brought up in the dark for the first three years of its life will never see properly, even though its optic nerve is fully developed. It is well-known that babies brought up without any human affection or interaction will quite often die. They are genetically programmed to expect affection and response from the world around them, and when this fails to happen, says Dr Watt, "they are tortured in the most profound way. They are trying to make sense of a world which just will not make sense because it is emotionally wrong."
Dr Fraser Watts, a former president of the British Psychological Association who is now the Starbridge lecturer in science and theology at Cambridge University, says: "You can't just pipe experience from one individual to another. Our experiences are the result of the interaction of sensory input and the way we process it. This is a process involving continual feedback from the brain to the sensory apparatus.
"It is the translation of nerve impulses into experience that is the problem: the boundaries between science and science fiction are much more blurred than people realise."
The dreams of the BT scientists can be traced at least as far back as Frankenstein's monster. It is a curious fact that when scientists have gone in for science fictional speculation, they have gone much further than professional writers dared. In Brave New World, published in 1929, Aldous Huxley foresaw most of the Nineties: genetic engineering, recreational drug use; easy sexual manners - and, in a development strikingly like BT's, "feelies" films that contained what we now would call virtual reality. Huxley's contemporary, the Marxist geneticist JBS Haldane, was imagining much wilder futures, in which the human race would attain to immortality.
Similarly, the science fiction of William Gibson takes for granted the kind of melding between brain and silicon that the BT researchers strive for. People wear silicon implants to give them memory, and rent out brain space to hide corporate data in. Sober neuroscientists won't say anything about these possibilities. After a couple of beers, they will say that the only implant which shows any sign of improving human brain performance or curing Alzheimer's comes from aborted foetuses, which is why no one will discuss the subject publicly.
But Gibson's world is in many ways more realistic than the imaginings of scientists such as Frank Tipler, whose book, The Physics of Immortality, publicised the idea that the human personality is just so much software that can be digitised and stored, and then reloaded at will on to better hardware.
His characters wear silicon-enhanced eyes that allow their experiences to be shared by others. But they are not immortal, and they continue to inhabit a recognisable social world. When they share their experiences it is not with helpful policemen or amusing friends, as in the BT fantasies. In Gibson's world, the experiences are shared with a world-wide audience of voyeurs and the wearers of implants are highly-paid stars in whose life as celebrities there is no distinction between work and play
His silicon-enhanced eyes are a lot more like television cameras than optic nerves. They feed a pre-processed version of reality to an audience which uses its own eyes and brains to make sense of it. And if there is one thing that the reaction to the BT announcement makes clear, it is that this is a lot more realistic than any other forms of virtual reality. Pure information that could pass from brain to computer and back again unchanged is as impossible a commodity as eternal life.