Pandora's Brain (5 page)

‘Ha! No, although if you would like to go back to my place and watch it right now . . . ? No, I thought not. Actually, strictly speaking they don’t upload their minds into the matrix, because their minds remain inside their flesh-and-blood bodies, and if those bodies are harmed their minds die.’

‘Which is my point entirely,’ Alice said. ‘When they’re inside the matrix they’re not real. The real humans are inside those flesh-and-blood bodies.’

Matt was about to reply, when he saw that Alice had stopped paying attention. ‘Look, there’s Ned,’ she said. ‘We really should go over and say hello – thank him for inviting us.’

‘Inviting you, you mean. You go. I’ll be over there, saying hello to Jemma: I haven’t seen her for a while. I’ll catch you later.’ Lowering his voice, he added, ‘Anyway, I’m not sure that Ned would pass the Turing Test.’

‘I heard that, smart-ass,’ Alice said over her shoulder. ‘Suit yourself. Catch you later.’

Matt watched Alice’s shapely behind sashay towards the knot of people Ned was in. He hoped she was putting on that walk for him. His attention was focused on Alice’s receding posterior as Jemma approached him.

‘Why so glum, Romeo? She likes you much more than she likes those gorillas, you know.’

‘Hi Jemma,’ he replied, then looked back at Alice, now chatting happily with Ned and a couple of his friends. The postures of the young men showed their appreciation of Alice. ‘Yes, I tell myself that, but I’m afraid I’m not very convincing.’

‘You under-estimate yourself, you know,’ Jemma smiled. ‘And perhaps you under-estimate Alice too. She wouldn’t be very smart if she found Ned more interesting than you.’

‘Thank you for the compliment. Or was that a mild rebuke?’

‘Now you’re being over-sensitive. Never mind. In a few years you can upload your mind and have all that terribly attractive modesty wiped away.’

Matt gave her a sharp look. ‘Were you eavesdropping on our conversation just now?’

Jemma laughed. ‘No, silly. Carl was telling me about it. That
transubstantiation
thingy.’

‘Transhumanist. What did Carl say about it?’

‘How did he put it?’ Jemma placed a finger to her bottom lip, and then pointed it upwards in mock sudden inspiration. ‘That was it: a creed for people who have learned nothing from the history of artificial intelligence.’

Matt laughed. ‘That sounds like Carl: deliberately cynical, hard to argue against, and resolutely determined to throw out any babies that might be lurking in the bath water. Maybe he’s right, but I think there just might be more to it.’

‘That sounds fun,’ she teased him, rolling her eyes. ‘Meanwhile, I’m getting cold. Shall we rescue Alice from Ned and go back inside? It would be a shame if she got bored to death before you got to take her home.’

*

Alice and Matt were sitting next to each other on the sofa in his house. Sophie had gone to bed before they got home, leaving a note that she would see them in the morning.

Matt reached across and placed his palm against Alice’s cheek. She turned her head and kissed his hand. He ran his thumb slowly across her lower lip.

‘You’re too far away,’ she said. ‘Come here.’

Matt shuffled closer on the sofa and Alice leaned into him so that the back of her head was resting against his chest. He held her head in his hands and tilted it upwards toward his own. He felt an overpowering rush of tenderness as he leaned down and kissed her. She kissed him back and reached up to hold his hands against her cheeks. He moved his hands down to below her breasts and held her gently but firmly. She held his face as they kissed.

She pushed his face gently away and sat up. She smiled, and traced a line down his cheek until it reached the corner of his mouth. Then she placed the fingertips of her right hand along his lower lip, and held them there for a moment. She gazed into his eyes with a look that made him tremble.

‘Shall we go upstairs?’ she asked.

He grinned. ‘If you insist.’

She grinned in turn. ‘I do.’

Matt sat opposite Carl in a small cafe on the corner of the two main streets in their town. Carl was restless and agitated, looking down at the rough pine wooden table. He played with the paper sachets of sugar and didn’t finish his coffee. The streets outside were heavy with snow and the café was busier than usual, with local people unable to get to work and needing to escape their homes for an hour or so. A couple of families were sitting at the large table in the window. The window was steamed up, but it was still possible to see the cars outside that had made it to the crossroads but were unable to climb the hill to turn onto the High Street. More snugly wrapped children were coming in, laughing and chattering, and the cafe was doing a great trade in the sponge cakes that were on display under glass domes on the reclaimed shop counter, assisted by the appetising smells of coffee and sticky sugary things.

Matt had raised the subject of transhumanism again, and was beginning to regret it. Both of them had now read most of the Kurzweil book, and watched various videos on YouTube. Matt was finding the ideas intriguing; Carl was not.

‘AI has been promising the earth since the 1960s,’ Carl grumbled, ‘and it has simply not delivered. Maybe we will create an artificial consciousness some time in the future, but not for thousands of years. When you watch those videos of the latest military-grade robots on YouTube, or when you ask your smartphone a question, do you detect anything like a conscious mind? The AI community said we would have them by now, along with flying cars and personal jetpacks.’

Matt raised his hands. ‘Come off it, Carl. It’s ridiculous to say that AI has made no progress. Self-driving cars are legal on public roads in parts of the US, and they will be legal over here soon too. Computers can recognise faces as well as you and I can: a lot of people said that would be in the ‘too-hard’ box for decades. Real-time machine translation is getting seriously impressive. This is all driven by the hugely increased processing power at researchers’ disposal, so they are going back to their original goal of developing a human-level intelligence which will pass a robust version of the Turing Test. A conscious machine.’

Carl wrinkled his nose and shook his head dismissively. ‘Never happen! At least, not in your or my lifetime. Just think about the scale of the task. We have billions of neurons in our brains, all wired to each other in incredibly complex ways. It will take centuries before computers can emulate that sort of structure. And even when you have the structure replicated, you still have to work out which pathways are the important ones, what order you connect things up, and exactly what they do when they are hooked up. Have you read about the
C. elegans worm?’

‘Can’t say I have,’ Matt replied.

‘Well look it up. Neuroscientists have had a complete map of its neuronal structure for several years now. It’s well documented, and there have been several projects to create a working virtual representation of the worm. They have all failed. The map doesn’t provide enough information about how to connect the neurons, and how each one is supposed to behave. Now this is just a worm, with about 300 neurons and 7,000 synapses. If they can’t model that, what chance is there of modelling a human brain, with a hundred billion neurons?’

Carl shook his head. ‘The computer scientists at Google and the rest of Silicon Valley think they can just build an analogue of a brain and that’s it – job done. But in reality that’s not even half the job. They haven’t got an adequate theory of mind, and there’s a little thing called psychology which they’ve completely forgotten about.’

He tapped the side of his head. ‘This took millions of years to evolve. It’s madness to think it can be replicated in a few years just because we have machines that can run a video game.’

‘Well of course you may be right,’ Matt conceded. ‘But you know, you’re being every bit as dogmatic that it won’t happen soon as Kurzweil is that it will. Think about what happens if Moore’s Law continues. By around 2060 an ordinary laptop computer (or whatever its equivalent is by then) will possess the same amount of processing power as all the human brains on the planet today – combined.’

‘But Moore’s Law won’t continue,’ Carl protested heatedly. ‘Exponential curves never do. It’s the oldest mistake in the book to take a rapidly rising trend and extrapolate it. You’d have thought that we would all have learned that way back when Malthus forecast that the world would run out of food because of the expanding population. Moore’s Law is going to run into the buffers soon because Intel can’t cram many more transistors onto an integrated circuit without setting fire to it.’

Matt frowned and smiled simultaneously. ‘Malthus
didn’t get the population extrapolation wrong so much as he failed to foresee the spectacular growth in agricultural productivity. Which was due to guess what . . . technology. Obviously you’re right that it’s dangerous to extrapolate an exponential curve. You
have to challenge the assumptions, and be sure that
there won’t be any changes which stop the progress.
But there are half a dozen technologies that can supplement and replace silicon chips and keep Moore’s Law on track. There are 3-D integrated circuits, optical computing, and ultimately there’s quantum computing.’

Carl snorted. ‘Fairy stories!’

Sensing that Carl was getting annoyed, Matt opened his hands in a placatory gesture. A blast of cold air made him shiver and he looked over to see the families in the doorway of the cafe, collecting their boots and coats. The waitress started clearing the window table and as the door swung closed the noise level dropped several decibels.

‘Tell me why this is irritating you.’

Carl was looking down. He flicked a tiny ball of paper along the table, away from them. ‘Because it’s so stupid!’ he said, sulkily. ‘It’s intelligent design for smart people. These Singularitarians, Transhumanists,
Extropians
, whatever they call themselves, they are all just guilty of a massive amount of wishful thinking.’

Carl leaned back, paused, and smiled sheepishly. ‘End of rant.’

Matt smiled. ‘No, it was a good rant. You give good rant, Carl. And as it happens I partly agree with you. There is a bit of a self-satisfied feeling about it all, as if they are initiated into a secret which no-one else knows, but everyone will be jolly grateful when they unveil it and bestow their blessings upon the world. And you may well be right about it being a delusion. Perhaps Kurzweil has got his calculations all wrong, and maybe the problem is many orders of magnitude harder than he thinks. Perhaps it will take thousands of years, not decades, before humans can upload into computers – if indeed they ever can.

‘But who knows, maybe they are right about the timing. Or nearly right. Maybe we are the last generation of mortal humans. Or maybe our kids will be. The mere possibility makes it one of the most interesting ideas I have ever come across. It makes me want to know more.

‘If it is at all possible, then it will happen, that’s for sure. There is too much to be gained for it not to happen. And there are some seriously smart people taking these ideas seriously. You’re always banging on about how smart Oxford Professors are – well one of them is at the forefront of transhumanism.’

‘Yes, I know,’ Carl admitted lugubriously. ‘Nick Bostrom is rather letting the side down.’ Carl’s face brightened as he spotted an opportunity to change the subject. ‘But Bostrom is also famous for something completely different, and it’s great fun. Have you heard of the Simulation Hypothesis?’

Matt shook his head.

‘OK, well the argument runs like this. Technologically advanced beings will create increasingly realistic virtual worlds, eventually populated by characters which believe themselves to be natural people living lives just as natural as ours. They will do it to re-enact what happened in their past, or to experiment with alternative histories, or just for fun. We are heading in that direction ourselves, with computer games like The Sims, so this seems a very reasonable supposition.’

‘OK,’ Matt said, ‘I’m with you so far.’

Carl nodded and held up his right hand with three fingers extended. ‘Therefore one of the following three statements must be true. First, no life form has ever progressed to the level where they could create simulations containing conscious beings – either because it is impossible, or because we are among the very first intelligent civilisations in the entire universe, or because all civilisations blow themselves up before they reach that point.’

‘Ouch,’ Matt said.

‘Second, many civilisations do progress to that level, but for some reason everybody refrains from creating such simulations.’

‘What, they all obey the
Star Trek Prime Directive
or something?’ Matt asked.

‘Yeah. Not very likely, is it. Which leaves us with statement three,’ Carl said with a magician’s smile, ‘which is that we ourselves are almost certainly living in a simulation.’

Matt jumped slightly, startled. ‘Come again?’

‘Think about it,’ Carl said. ‘If it is possible and permitted to create a lifelike simulation, then it will happen a lot, because there are so many good reasons to do it. Therefore every civilisation that develops the technology to build such simulations will go ahead and build a lot of them. Advanced civilisations with plentiful resources will create a great many of them. Therefore the number of simulated civilisations will be vastly greater than the number of naturally-developed civilisations. That means that statistically, we are much more likely to be living in a simulation than in a universe which evolved naturally.’

Matt was frowning. After a while he shook his head and conceded, ‘I can’t fault the logic. It’s bit spooky, isn’t it? I mean, you’re saying we are very likely to be a simulation created by a 31st-century teenager who knocked us up for a bit of homework. He, she or it could pull the plug on us and that’s it, game over.’

Carl nodded. ‘It’s my favourite solution to the Fermi Paradox.’ He noticed Matt’s puzzled expression. ‘You know: the observation that the universe should be full of life, but we don’t see any evidence of it.’

‘Yes yes,’ Matt said impatiently. ‘I know about the Fermi Paradox; I just don’t see how this Simulation thing resolves it.’ His eyebrows rose and his mouth formed an ‘O’ as realisation broke. ‘Oh I get it. Whoever created the sim we are in didn’t bother to programme in any aliens.’

‘Exactly!’ Carl grinned. ‘Creating one plausible world with a full history, flora and fauna was enough to get the job done, whatever that was. To programme in a whole bunch of other worlds complete with alien races and habitats was . . . well . . . an unnecessary use of time and resources.’

‘Which suggests that their resources aren’t infinite.’ Matt said.

‘Nobody’s resources are infinite,’ Carl replied. ‘Even if you turned all the matter in the universe into computing hardware, you would still end up with a finite amount of computational capacity. Actually someone has done that calculation. . .’ Carl raised a finger as he tried to remember, but then dropped it again, shaking his head. ‘Nope. I can’t remember who it was.’

‘So do you actually believe this
Simulation
theory?
Do you think we are in a far future version of
Call of Duty
?’

‘Do you know, I think I do,’ Carl said, sounding as if he was still developing his own thinking on the subject as he spoke. ‘As spooky as it seems when you first hear it, the
Simulation Hypothesis
actually feels right intuitively once you get used to it. Well, it does for me, anyway. For one thing, it explains why the universe appears to be digital rather than analogue.’

‘What do you mean, the universe is digital?’ Matt asked.

‘When you drill down to the smallest possible scales, the universe appears to be quantised. It used to be thought that atoms were the smallest units, and that they were indivisible. Now we know that you can go a lot smaller than that. But it seems that when you get down to 10 to the minus 35 metres – which is pretty bloody small, admittedly – you can go no further. It’s called the Planck length, and in both string theory and quantum loop gravity theory, a dimension smaller than this literally makes no sense. The Planck time is the time it takes light to travel the Planck length, and again, you can’t get a shorter step in time.

‘Apparently there is a machine in America built by a physicist called Craig Hogan which is starting to test this. It’s called a holographic interferometer, or a holometer. It’s going to test the hypothesis that the universe is a hologram.’