Except for that one weird species of jellyfish that can pop back and forth between larval and adult states, all living things are built to die. So as unpleasant as it is, it's only logical to expect that someday you will also, ah let's put this delicately uh, stop being alive. Now I'm going to be honest I personally believe that a dramatically expanded lifespan would be awesome, but it's always been a long way off. And it's probably a longer way off than any of our lives will be but there is this vague idea that in the distant future, humanity might figure out a way to cheat death. 

So nobody's getting their hopes up but some researchers are trying to close the gap by figuring out how to map the human brain and simulate it on a computer. Essentially in the long term, they want to figure out how to upload the human mind. Now the idea of mind uploading has been floating around science fiction for decades but it's only recently that it started to look like actually maybe a little bit in the distant future possible. 

In theory all you'd have to do is program a map of your brain onto a computer and once things are all set up you just run the program and voila! An uploaded working brain. The question of course, is how to actually do that because like brains, are very complicated. 

The average human brain contains 86 billion neurons and each of those neurons can form thousands of connections with each other. We're talking hundreds of trillions of synapses or pathways that exchange information. So if you want to build a brain, the first thing you're gonna need is a map. The goal is to eventually put together a detailed database of each neuron in the human brain and how they all connect. 

But first, we're starting small. Millimeter long roundworm small. C. elegans is a tiny worm that is about as simple as you can get, which is why it is often used for research. In 2012, researchers completed it's connectome - the map of how all of its neurons interact with one another. Now it does help instead of 86 billion neurons it has 302. Other groups are trying to put together the fruit fly connectome, that's taking a little longer, small it may be but the fruit fly's brain has more than 30,000 neurons. It's taken so long to figure out the roundworm and even small parts of the fruit fly brain that it'll be a while, possibly forever, before we can map out something as complicated as a human brain. 

But even if we did have a perfect map of the human connectome, that wouldn't really be enough. Remember we're talking about hundreds of trillions of connections, and some of those are transmitting messages at more than 400 kilometers an hour. No matter how fancy your new gaming laptop is, even it doesn't have enough processing power to manage all that. But someday, it might. 

Back in 2012, for instance, a team of scientists at IBM was able to simulate 530 billion neurons and a hundred trillion synapses using the Blue Gene/Q supercomputer. Now those numbers are on the same scale as the human brain but it was the simplest model possible, just a bunch of processors forming a huge network. They weren't following any prescribed pathways or processing input the way a human brain would. Plus, the program was running more than 1500 times slower than a human brain would. 

But computers are getting faster all the time so speed is an issue that we can solve, even if it takes some decades. Problem is, we kinda have no idea how to get a network to act like a brain even if we know what the connections are supposed to look like. Because even after all these years of developing the field of neuroscience, we still don't know how brains work, or understand how consciousness happens. 

But maybe, it doesn't matter. We have billions of working human brains at our disposal, all we have to do is take 'em apart then reverse engineer a computerized version. Like a paint by numbers kit, you don't really need to understand the complexities of color theory as long as you have instructions. We just don't know how detailed those instructions need to be. 

For example, a program that models the entire brain at the molecular level is gonna be much more accurate than one that just simulates the signals between neurons. But as the simulation gets more complicated it becomes even more demanding to program and run. It's hard enough to model even a single cell right now, let alone 86 billion of them. 

At this point we don't know how deep we're gonna have to go to get a brain simulation up and running, especially if we want it to be a conscious brain. So we have a lot to learn about our minds before we're able to upload one to a computer but so far we don't know anything that indicates that it is definitely impossible.

Meanwhile thank you for using your natural brain, I assume, to watch this episode of SciShow which was brought to you and inspired by the movie Selfless, starring Ben Kingsley and Ryan Reynolds, which explores some of the moral and philosophical questions of this potential future, and also was super fun to watch. We got to see it here, all the SciShow team. Selfless will be released on July 10th and there's a link to the trailer in the description.