Comprehensive coverage

When can I upload my brain to the computer?

Reader's question: I am 59 years old, and in fairly good health. Is it possible that I will live long enough to put my brain into a computer? Richard Dixon.

By Guillaume Thierry, Professor of Cognitive Neuroscience, Bangor University

brain computer interface. Photo:
brain computer interface. Photo:

We often imagine that human consciousness is as simple as the input and output of electrical signals within a network of processing units - and therefore similar to a computer. The reality, however, is much more complex. For starters, we don't really know how much information the human brain can hold.

Two years ago, a team at the Allen Institute for Neuroscience in Seattle, USA, mapped the XNUMXD structure of all the nerve cells (brain cells) that make up one cubic millimeter of a mouse's brain - a milestone that is considered extraordinary.

Within this tiny cube of brain tissue, the size of a grain of sand, the researchers counted more than 100,000 neurons and more than a billion connections between them. They were able to record the appropriate information on computers, including the shape and configuration of each neuron and connection, which required two petabytes, or two million gigabytes, of storage. And to do that, their automated microscopes had to collect 100 million images of 25,000 slices of the tiny specimen continuously over several months.

If that's what it takes to store the full physical information of neurons and their connections in one cubic millimeter of a mouse brain, we can imagine that collecting this information from the human brain is not going to be a walk in the park.

However, extracting and storing data is not the only challenge. For a computer to be similar to the way the brain works, it would have to access all the stored information in a very short period of time: the information would have to be stored in its random access memory (RAM), not on traditional hard disks. But if we tried to store the amount of data the researchers collected in a computer's RAM, it would take up 12.5 times the capacity of the largest single-memory computer (a computer built around memory, not processing) ever built.

The human brain contains about 100 billion neurons (the order of magnitude of the number of stars in the Milky Way) - a million times more than those contained in the cubic millimeter of our mouse brain. And the approximate number of connections is ten to the power of 15. That is, ten followed by 15 zeros - a number similar to the individual grains contained in a two-meter-thick layer of sand on a 1-kilometer-long beach.

A question of space

If we don't even know how much information storage a human brain can hold, you can imagine how difficult it would be to transfer it to a computer. You will first need to translate the information into code that the computer can read and use once it is stored. Any mistake in this will probably prove fatal.

A simple rule of thumb for data storage is to make sure you have enough space to store all the data you need to transfer before you begin. If not, you'll need to know exactly the order of importance of the information you're storing and how it's organized, which is far from the case with brain data.

If you don't know how much information you need to store when you run it, you may run out of space before the transfer is complete, which could result in the information string becoming corrupted or unusable on your computer. Also, all data will need to be stored in at least two copies (if not three), to avoid the devastating consequences of potential data loss.

That's just one problem. If you noticed when I described the extraordinary feat of researchers who were able to completely store the 25,000D structure of the neuronal network in a tiny mouse brain, you'll know that it was made from XNUMX (very thin) slices of tissue.

The same technique would have to be applied to a human brain, because only very coarse information can be retrieved from brain scans. The information in the brain is stored in every detail of the physical structure of the brain and the connections between the nerve cells: their size and shape, as well as the number and location of the connections between them. But would you agree to have your brain deployed in this way?

Even if you agree that we are cutting your brain into extremely thin slices, it is highly unlikely that the full volume of your brain could ever be cut precisely enough and "reassembled" correctly. A person's brain has a volume of about 1.26 million cubic millimeters.

If I haven't already dissuaded you from trying the procedure, consider what happens when you factor in time.

After we die, our brains rapidly undergo significant changes that are both chemical and structural. When neurons die they quickly lose their ability to communicate, and their structural and functional properties change rapidly - that is, they no longer exhibit the properties they exhibit when we are alive. But even more problematic is the fact that our brains age.

From the age of 20 we lose 85,000 nerve cells a day. But don't worry (too much), we mostly lose neurons that haven't found their use, they weren't asked to get involved in any information processing and activate a self-destruction program (called apoptosis). In other words, several tens of thousands of our neurons kill themselves every day. Other neurons die due to exhaustion or infection.

However, this is not too much of a problem, as we have almost 100 billion neurons at age 20, and with that rate of attrition, we have only lost 2-3% of our neurons by age 80. And assuming we don't contract a neurodegenerative disease, our brains can still represent the thinking style Ours for life at this age. But what would be the right age to stop, scan and store?

Would you rather store an 80 year old brain or a 20 year old brain? Trying to store your brain too early would miss out on many memories and experiences that would later define you. But then, trying to transfer to a computer too late would risk storing a brain with dementia, one that doesn't really "work" either.

So, given that we don't know how much storage is required, that we can't hope to find enough time and resources to completely map the XNUMXD structure of an entire human brain, that we'll have to slice you up into zillions of tiny cubes and slices, and that it's basically impossible to decide when to do the transfer , I hope you are now convinced that this probably won't be possible for a while, if at all. And if that were the case, you probably wouldn't want to risk it in that direction. But in case you're still tempted, I'll continue.


Perhaps the biggest problem we have is that even if we could understand the impossible and jump over the many hurdles discussed, we still know very little about the underlying mechanisms. Imagine if we were able to reconstruct the complete structure of the hundred billion neurons in Richard Dixon's brain along with each of their connections, and we were able to store and transfer this astronomical amount of data to a computer in triplicate. Even if we could access this information on demand and instantly, we would still be faced with a big unknown: How does it work?

After the "what" question (what information is there?), and the "when" question (when would be the right time to transfer?), the most difficult is the "how" question. Let's not be too extreme. We do know a few things. We know that neurons communicate with each other based on local electrical changes, which travel down their main branches (dendrites and axons). These can pass from one neuron to another directly or through exchange surfaces called synapses.

At the synapse, electrical signals are converted into chemical signals, which can activate or deactivate the next neuron, depending on the type of molecule (called a neurotransmitter) involved. We understand much of the principles governing such transfers of information, but we cannot decipher it from looking at the structure of neurons and their connections.

To know what types of connections apply between two neurons, we need to apply molecular techniques and genetic testing. This means again fixing and cutting the tissue into thin slices. It also often involves dying techniques, and the cutting needs to match these. But it doesn't necessarily match the cut needed to reproduce the XNUMXD structure.

So now that you're faced with an even more daunting choice than determining when is the best time in your life to give up on existence, you must choose between structure and function—the three-dimensional architecture of your brain versus how it works at the cellular level. This is because there is no known method for collecting both types of information at the same time. And by the way, not that I want to inflate any serious drama, but the way neurons communicate is another layer of information, meaning that we need much more memory than the incalculable amount previously predicted.

So the possibility of uploading the information contained in brains to computers is completely remote and may forever be out of human reach. Maybe I should stop here, but I won't because there is more to say. Let me ask you a question in return, Richard: Why would you want to put your brain into a computer?

Are our minds more than the sum of their (biological) parts?

I might have a helpful, albeit unexpected, answer to give you after all. I assume that you would want to transfer your mind to a computer in hopes of existing beyond your lifetime, that you would want to continue existing inside a machine once your body could no longer implement your thoughts and plans in your living mind.

If this hypothesis is correct, I must object. When you imagine that all the impossible things listed above will one day be solved and your brain could literally be "copied" into a computer - allowing a full simulation of your brain's functioning - the moment you decide to transfer your brain to a computer, Richard Dixon would cease to exist. Thus, the brain image transmitted to the computer will not be more vivid than the computer hosting it.

This is because living things such as humans and animals exist because they are alive. You may think that I just said something completely trivial, bordering on stupidity, but if you think about it there is more to it than meets the eye. A living brain receives feedback from the world through the senses. It is connected to the body which is based on physical sensations. The result is physical manifestations such as changes in heart rate, breathing and sweating, which in turn can be felt and contribute to the inner experience. How would this work for a computer without a body?

Any such input and output are unlikely to be easy to model, especially if the cloned brain is isolated and has no system to sense the environment and act in response to the input. The brain seamlessly and constantly combines signals from all the senses to produce internal representations, makes predictions about those representations, and ultimately creates consciousness (our sense of being alive and being ourselves) in a way that is still a complete mystery to us.

Without interaction with the world, however sophisticated and unconscious it may be, how can the mind function even for a moment? And how can it develop and change? If consciousness, artificial or not, has no input or output, then it is lifeless, just like a dead brain.

I can't think so I won't.

In other words, after making all the sacrifices discussed earlier, transferring your brain to a computer would have completely failed to keep your brain alive. You can reply that you will then request an upgrade and ask that your brain be transferred to a sophisticated robot equipped with an array of sensors capable of seeing, hearing, touching, and even smelling and tasting the world (why not?) and that this robot be able to act and move, and speak (why not?).

But even then, it is theoretically and practically impossible for the required sensors and motor systems to provide sensations and produce actions identical or even similar to those provided and produced by your current biological body. Eyes are not simple cameras, ears are not just microphones and touch is not just about stress assessment. For example, the eyes not only transmit contrasts and bright colors, the information from them is combined shortly after it reaches the brain to encode depth (distance between objects) - and we still don't know how.

And so it turns out that your transferred consciousness will not be able to relate to the world as your current living mind does. And how will we even manage to connect artificial sensors to the digital copy of your (living) brain? What about the risk of hacking? Or a hardware failure?

So no, no and no. I've tried to give you my (scientifically based) take on your question and although it's unclear from me, I hope I've helped ease your desire to ever get your brain into a computer.

I wish you a long and healthy life, Richard, because this is certainly where your mind will exist and thrive as long as it is applied by your mind. May God grant and it will bring you joy and dreams - something androids will never have.

More of the topic in Hayadan:


This article is part of the Big Questions of Life series

The Conversation series, published in partnership with BBC Future, seeks to answer our readers' funny questions about life, love, death and the universe. We work with professional researchers who have dedicated their lives to uncovering new perspectives on the questions that shape our lives.

2 תגובות

  1. Reminds me of those people who said. "We can never fly because we don't have wings." Evolution succeeded in producing us. There is no reason that with the help of intentional (and computerized) evolution we will not reach a solution to this as well.

  2. Sounds like a list of technical problems that will become simple in a hundred or two hundred years.

Leave a Reply

Email will not be published. Required fields are marked *

This site uses Akismat to prevent spam messages. Click here to learn how your response data is processed.