Welcome to Sentivia

Alright, let me tell you a bit about what I'm doing, what we're doing, because it's quite a journey. Essentially, I'm the lead on a project that’s aiming to construct a complete and detailed computer model of the human brain. It's an ambitious goal, no doubt about it, but we have been making real strides. For the past four years, we focused on a proof of concept, a smaller model using a portion of the rodent brain. That was like our test run, and it worked! It showed that our approach had validity. And now, armed with the knowledge and experience gained from that process, we’re scaling the project up to tackle the big one – the human brain.

Now you're probably thinking, “Why? Why even attempt something so massive?” Well, there are three key reasons behind this. Firstly, understanding the human brain is fundamental to our progress as a society, as a species even. It's a vital step in our own evolution. We need to unlock the mysteries of our minds to better connect with each other, address challenges, and really just move forward as a civilization. Second, we can’t keep relying on animal experimentation forever. It's just not sustainable or ethical in the long term. We have to find a new way, a way that still allows us to gather the data and knowledge we need but doesn’t rely on constant experimentation on living creatures. This model we’re building, in a sense, acts like a Noah’s Ark, an archive of everything we know about the brain, all in one functioning model. It's a crucial step towards ethical research. And thirdly, and this one is near and dear to my heart, there are billions of people on this planet struggling with mental disorders. And, too often, the treatments we have now are simply empirical, based on observation and trial and error without a clear understanding of why they work. I genuinely believe that this project can provide us with concrete solutions, enabling us to treat those disorders more efficiently, with targeted methods based on a real understanding of the brain’s processes.

But even at this stage, before we even have a fully functioning human brain model, this model is incredibly useful. It allows us to delve into some really profound, fundamental questions about how our brain actually functions. And today, here at TED, I want to share a particular theory we’re exploring with our simulations – the idea that the brain creates a version of the universe, a sort of projection or "bubble" that we perceive around us. Now I know this sounds a little out there, even a little philosophical, because it is! It's a debate that philosophers have tackled for centuries, but for the first time, we can now actually explore this hypothesis in a very rigorous, scientific way using brain simulations. We can finally ask the very important questions – is this theory plausible? Is it actually how our brain works?

I can give you a little example. Why is the moon so massive on the horizon? Well, it's not that it's actually bigger, it's because our perceptual bubble, the one that our brain creates, doesn't stretch 380,000 kilometers out to the moon. It kind of runs out of space. And so, what we do is, we compare the moon to the objects within our perceptual bubble, like the buildings. Our brain then makes a decision. It creates this perception that the moon is enormous when in reality it’s not. Those decisions, the ones that our brain is constantly making, are absolutely critical to supporting our perceptual bubble, they maintain it. They keep it alive. If you can’t make decisions, you can’t perceive, you can’t think, you can’t feel. Even something like anesthesia, you may think it puts you into a deep sleep or blocks your pain receptors, but that's not really how it works. It actually introduces "noise" into the brain. The neurons get confused, they can’t communicate, and you can’t make decisions, which is why, when a surgeon is hacking away at your body, you don’t feel it or know what is going on. You're too busy trying to make sense of everything. So, every single time you take a step, open a door, you have to compulsively make decisions. We infer more about a room than our eyes can possibly capture. 99 percent of what we see isn't from our eyes, it is the result of the decisions our brain has made.

So, while I can confidently say “I think, therefore I am” because I have access to the inner workings of my own perceptual bubble, it’s far harder to say “You think, therefore you are” because you’re within *my* perceptual bubble. We could debate this for hundreds of years. But we don’t have to. Instead, we can use our model to ask that very concrete question – is the brain actually capable of building a perception, does it have the necessary capacity to do it?

It took the universe billions of years to actually build this thing, the brain. It had to adapt and improve, add things like the frontal lobe so we could develop instincts. But the real key step in that evolution was the neocortex. This is the new brain. The mammals needed it so they could cope with things like parenthood, complex social interactions, things that required sophisticated cognitive functions. In many ways, the neocortex is really the peak of evolution, the culmination of the universe’s creativity. The numbers of neurons in the neocortex increased about a thousandfold from the mouse to the man. It’s an organ of incredible power. And it’s still evolving, adapting and changing, especially the human brain.

If we could zoom into the surface of the neocortex you would see these tiny modules, like G5 processors in a computer, except we have about a million of them. They were so successful in evolution that they were duplicated repeatedly, added in enormous quantities to the brain until we just ran out of space in the skull. That’s why the neocortex has become so folded, all these convolutions are simply a way of packing as much of those columns as possible into a limited space. Think of the neocortex as a massive grand piano with a million keys. Each column produces a note. Activate them, and you create an entire symphony. But this isn’t just the symphony of perception, it’s a symphony of your reality.

Now, obviously mastering a million-key piano takes a long time, hence the need for education, hopefully an Oxford education. But there’s also genetics. Some people are born with a natural ability to master the neocortical columns. And some new theories suggest that autism, the “intense world” theory, is when these columns are super charged, extremely reactive and plastic, so autistics can learn to produce a symphony we can’t even imagine. But conversely, if there’s a problem in one of these columns, then the note is going to be off, your perception is going to be corrupted. The symptoms of disease emerge.

So, the holy grail for neuroscience, for understanding perception and reality, even physical reality, is to truly understand the design of the neocortical column. And that is what we've spent the last 15 years doing, dissecting out the neocortex, systematically. It's been like going into a rainforest and cataloging every detail. We look at each tree, how many there are, what shape they are, what type they are, where are they positioned. But it's more than just cataloging, we have to understand all the rules of communication, the rules of connectivity, because neurons are very selective about who they connect with. It’s also more than cataloging because we also have to build the three-dimensional digital models of them. We've done this for tens of thousands of neurons, building digital models of all of the different types we’ve discovered.

And once you do that, you can start constructing the neocortical column. You see these branches intersect in millions of places, and at each intersection, they can form a synapse, the location of chemical communication. These synapses together make up the network, the circuit, or you could say, the very fabric of the brain. But this presents a challenge for any theory of the brain, especially the one that says that reality emerges from the carpet, from the fabric, the pattern.

The biggest secret of the brain’s design is diversity. Every single neuron is different. Like the trees in the forest, there might be different types, but each individual pine tree is still unique. There’s no neuron in my brain that is the same as another neuron in my brain and neither will yours. Your neurons won’t be positioned and oriented exactly the same way as mine, you may have more or less. So, how is it even possible to create a shared reality that we can understand? Well, the good thing is, we don’t have to just speculate. We can look at all the 10 million synapses, we can look at the fabric and even change the neurons. We can use different variations, position them in different places, orient them differently, add more or less. And, when we do that, the circuitry changes, but the underlying pattern, the *design* of the circuitry, doesn’t.

The fabric of the brain, even if it's smaller or bigger, with different neurons and morphology, we do actually share the same fabric pattern. This is what we believe is species-specific, which may explain why different species can't communicate with each other.

So, how do we bring this all to life? With mathematics! We use equations. The equations that turn neurons into electrical generators were discovered by Nobel Laureates. And we also have the mathematics that allows us to describe how neurons gather information and then create lightning bolts to communicate with each other. These lightning bolts release chemicals from the synapses. It's a surprisingly small number of equations that are needed to simulate the activity of the neocortex. The thing is though, you need a huge computer. In fact, you need one laptop just to perform all the calculations for one single neuron. So you need 10,000 laptops. That's where companies like IBM come in. We use their supercomputer because they know how to put 10,000 laptops into the size of a refrigerator.

Now we can load all of these neurons into the supercomputer and fire it up, take the magic carpet for a ride. When we activate it, we see what’s happening in the brain during a stimulation. It’s a very first glimpse. When you look at it you might ask “How can a reality emerge from that?”. We haven’t trained the neocortical column to create any specific reality but we can ask the question “Where is the rose?” Where is it inside, if we stimulated it with a picture? It has to be there.

The way we can find it is by ignoring the neurons, ignoring the synapses, and looking at the raw electrical activity, because that's what it's creating - electrical patterns. And when we do that, we see these ghost-like structures, these electrical objects, emerging within the neocortical column. These objects are holding all the information about whatever stimulated the brain. And if you zoom in, you find it’s like a whole universe.

The next step is to project these brain coordinates into perceptual space. If we can do that, then you’ll be able to step inside of the reality created by this piece of the brain. So, I believe it’s possible that the universe evolved a brain to see itself, which could be a step towards becoming aware of itself. There's still so much to do, but I hope that you are partly convinced that it’s not impossible to build a brain. We can do it within 10 years. And, if we are successful, I promise to send a hologram to TED to talk to you about it. Thank you.