Download Brain Into Computer
YOU’RE FALLING FAST, the rush of wind pressing into your face as you free-fall towards the Earth. Only a few seconds ago you were in a plane, soaring 3km above the ground. Soon it’s time to open your chute. You pull the chord and… phew! It opened, and it’s time to experience the glorious landscape below.
But what if this experience isn’t really yours? What if it was someone else’s and you’ve merely downloaded the data from the internet directly into your brain? The idea that we could all one day share our thoughts, actions and emotions online has been dubbed the ‘Mindscape’ by the science fiction writer Sarah Newton.
“Barring some kind of major calamity I believe the Mindscape – a cloud-based database of human experiences – is pretty much inevitable. It will mark the next stage of our evolution as a species,” she says.
The Mindscape is just one of the big ideas being proposed by participants in the Hieroglyph project, but could it ever really happen? What does modern science have to say about the brain and how information and memories are stored? And would we even want to do it in the first place?
Experiences are created by the firing of groups of electrochemical nerve cells called neurones. Neurones fire in response to stimuli – light falling on the retina, for instance, or a chemical message from another part of the body. They send signals to one another along connective tendrils called axons, so a single stimulus can trigger activity throughout the brain. Different areas of the brain specialize in different things: those at the back are concerned with vision, for instance, while those in the front are about thinking, judging and feeling.
Modern imaging techniques, such as functional Magnetic Resonance Imaging (fMRI) make neuronal activity visible, so you can look at an active brain and deduce what sort of experience it is having. If someone is confronted by an angry person shouting at them, for instance, their brain will probably show activity in a little bit of deep brain tissue called the amygdala (fear); in a spot on the cortex behind the left ear (word comprehension) and in an area above the left ear (noise).
Then there would be various other flurries of electricity – each one a different thought, emotion or perception triggered by the event. Our ability to ‘read’ these signals is still pretty crude – you can tell if a person is looking at a house or a horse, but not precisely which house or horse. Undoubtedly, though, mind reading of this sort will get better as scanning techniques are refined.
Recall is simply the repetition of the neuronal patterns that generated the original experience. Most recollections are contaminated by neuronal activity related to the present moment, so the memory is vague and hazy and you never totally lose touch with the here and now. An intense recollection, however, such as the flashbacks that happen to people with post-traumatic stress disorder, is literally a ‘re-living’ of the original experience – you feel the same feelings, see the same sights, hear the same sounds and have no awareness of the present.
Given this, it seems that it should be possible in theory to ‘read’ a person’s experience, encode it digitally, then ‘replay’ it in another brain. However, there are several obstacles to overcome before we can step into the Mindscape, and some of them may be beyond our current technological prowess. But that won’t stop scientists from trying…
The first challenge is technical: capturing someone’s experiences and memories in the Mindscape would mean recreating their brain in digital form and that is a vast undertaking. But this is already underway! The current effort to recreate a brain digitally is being carried out by The Human Brain Project. The scheme is set to receive over €50 million from the European Commission over the next two years.
But even the team behind the project do not claim that they will be creating a sentient being. It is worth noting that its forerunner, the Blue Brain Project, took more than five years to reproduce a system equivalent to a cortical column in a rat’s brain – an ordered structure of 10,000 neurones. A human brain contains some 86 billion neurones, each one connected with hundreds or even thousands of others. But with continued advances in processing power, given time the goal could be achievable.
The Mindscape idea would be technically much more difficult than the Human Brain Project. When you see a coloured-up diagram of the human brain with the functional areas neatly marked, it is easy to conclude that each brain is pretty much like any other. This is true on a macro level, but microscopically each brain is completely different and it is these differences that mould our experiences.
A simple stimulus – the sound of a pin dropping on glass, say – may trigger a billion temporal neurones to fire in my brain, and a billion different ones in yours. Some of mine might bring up memories associated with hearing a similar sound in the past; some may produce a visual image, others may stir the word ‘pin’ or ‘glass’ and so on. Yours will trigger different memories, different emotions, different words. One stimulus: two experiences.
To ‘map’ a person’s experience on to a digital brain therefore means that the artificial (or virtual) organ must be identical to the one in which the experience takes place, otherwise the result will be garbled. This involves working out every one of that brain’s minute distinctions. “The only way to re-create an individual’s brain would be to measure the exact topology of their neural connections – some 10″ of them -and then measure exactly how each brain cell fires in relation to all the others,” says artificial intelligence expert Igor Aleksander, Emeritus Professor of Neural Systems Engineering at Imperial College, London. “I’m afraid I’m a bit pessimistic.”
If that was all that was required to produce the Mindscape the idea could possibly be doable. But even if our memories could be stored in digital form, there remains the problem of transferring them to a recipient’s brain.
This is the really tricky bit. Just as the virtual brain has to be identical to that of the original experiencer in order to get a true copy of what they are going through, so the person who draws down the experience from the Mindscape must also have identical neural architecture. If they don’t, the received experience will not be correctly ‘mapped’. If, say, you transmitted the experience of something coloured blue to a brain deficient in blue-sensing neurones, the result might be a green experience, or no experience at all, or – who knows? – a vision of flying pigs.
It would not therefore be enough for the Mindscape to read the information from a person’s brain and simply relay it. Rather the system would have to interpret the original data, and reformulate it, like translating text.
Let’s suppose that a technophile called Jane is in an art gallery in Italy. She looks at a painting of an animal and decides to transfer her experience to her friend Lara, in London. Both of them put on brain-reading devices – something like a very sophisticated EEG helmet – and the data from Jane’s brain is transmitted to the cloud-based translator.
Jane’s neural data shows that the painting first triggers activity in her visual cortex. This is transmitted along the neural pathway concerned with object recognition, where it is classified as ‘dog’. Fine so far. But Jane once had a bad experience with a dog and the event created a pathway from the neurones that recognize ‘dog” to her amygdala, the part of her brain that feels fear. As she looks at the painting, the activity from the object-recognition pathway therefore zooms off down the old path and produces fear.
Lara, on the other hand, does not have a connection from the ‘dog’-recognising pathway to her amygdala. In fact she loves dogs, so were she to see the painting first-hand, her neural activity would whizz from the object-recognition pathway to her pleasure circuit. Given that the Mindscape is about transmitting experiences, not stimuli, the signal it would send would include Jane’s amygdala activity as well as her sight of the painting.
So here we have a problem: the sight of the painting would normally produce pleasure in Lara, but the Mindscape is feeding in fear. This would not replicate the precise experience of Jane, though, because Lara would surely be puzzled by her own, unfamiliar reaction, whereas Jane’s original experience would not include puzzlement at all.
The only way around this would be to alter Lara’s entire self-identity. The information beamed into her would include all Jane’s active memories down to every physical detail. It would be a full-on hallucination. Furthermore, even when it ended, Lara could not just go back to being herself because the experience will have altered her brain. She may have retained her happy response to dogs, but she will also, potentially at least, have Jane’s fearful reaction. A little bit of her will have become Jane.
A more modest proposal is to create a receiver that can extract and then relay only those bits of neural data that are related to sensational stimuli – the neural activity of visual neurones, say. So Jane’s experience would be reduced to the sight of the picture, without the emotions and thoughts that accompany it.
But this would be a tricky task for many reasons. One is that what we think we sense is not what we actually sense. Experiments show that our experiences mainly consist of tiny bits of sensation ‘filled out’ by our imagination. One, carried out by Loughborough University, involved fitting eye-tracking devices to visitors at the National Gallery in London, and recording exactly where they looked when viewing a painting. It showed that they focused narrowly on areas of interest – faces, hands and the odd bright detail of clothes. Even when they looked for 10 minutes they neglected most of the canvas. When asked what they saw, though, all of them thought they had examined the entire picture – they had no idea that the parts they had really looked at were just a tiny bit of the whole.
This is typical of the way we view the world – we focus in on tiny areas of interest and ignore the rest. So although Jane may experience seeing the whole picture in the art gallery, she will probably only register a few patches of it, and Lara would therefore receive something like a half-obliterated photograph. Jane would do better to snap the picture with a smartphone and transmit the image itself.
There is also no guarantee that a reading of neuronal activity associated with sensation would actually capture sensation. For instance, the visual neurones of blind people have been found to encode sound, so their activity can produce different types of experience. It is not even clear that ‘visual’ and ‘auditory’ neuronal activity produces conscious experience at all. Some studies suggest that perception only occurs when activity from sensory neurones is transmitted to other parts of the brain.
Whether or not the Mindscape is possible, it challenges our deepest ideas about identity and consciousness. If it ever is realised, the change to our lives would be dramatic. As Sarah Newton says: “It will be utterly profound. We’re already becoming suspicious of the very concept of identity: it’s beginning to become clear that, just as the mystics said, the self is an illusion.”