There was an interesting piece in the Wall Street Journal on neuroimplanting. The authors argued that it is not a matter of whether neuroimplants will break through, but when. Already, we have EEG-based applications for directing machines directly with brain signals, and as the article points out, some motor conditions such as Parkinson’s can be treated with microelectrodes.
Neuroimplanting is, however, a technology that is slightly further down the horizon owing to a very simple obstacle that many other soon to be seen technological breakthroughs don’t suffer from: we still don’t really know that well how the brain works.
The brain, according to current knowledge, has about a hundred billion neurons, with a whopping hundred thousand billion synapses connecting them. For the last few decades, it was thought that each synapse could function as something like a transistor, either letting current through or not.
This would put the amount of controllable variables in the brain in the ballpark of a maximum of one hundred thousand billion. Controlling such a huge amount of variables with microelectrodes or even neural dust is going to be quite challenging.
But the complexity of the neural net is only the beginning. Like I said, the above has been the received view for the last few decades. Recent advents in neuroscience have, however, pointed to an even more complex picture.
In a groundbreaking Stanford imaging study, it was found out that the synapse is in fact a far more complicated structure than a simple transistor, with individual memory systems and other components more typical to a processor than a transistor.
The neural complexity is not limited to the brain either. Also the gut has about five hundred million neurons to further stir up the stew, so to speak. Add to this the more than a hundred neurotransmitters that either excite or inhibit neural signals, the complexity of variables to be controlled or analyzed by an implant is mind-blowing. I won’t even go into glia cells.
So the brain is likely to be tremendously more complex than we think. But so what? Like Marcus and Koch rightly point out in the WSJ paper, every breakthrough technology has looked at such challenges at some point.
I do believe we will have garden variety neuroimplanting in the future. But I suppose that future is slightly farther beyond than is implied in the article. That being said, I think we should not downplay the upcoming advent of HUD’s and other augmented UI devices.
After all, we do have highly sophisticated neural interfaces in place already: our eyes, ears and skin. In a classical experiment Paul Bach-y-Rita used the skin and the tongue to restore vision – yes, actual visual imagery! – to congenitally blind people. In fact the EEG devices that we now have are not that different from what Bach-y-Rita did. By picking up relevant EEG signals we can, for example, guide a robotic arm. Just like we could by hooking the electrode to a facial muscle, for example.
Before we get to the point where a memory module is hooked up directly to the brain – and we will get there, by the latest in a century or two – we can already significantly boost our mental faculties by the devices we have right now.
By being able to tap into relevant information and sorting out the wheat from the chaff, we can augment our mental capacity dramatically as we speak, with nothing fancier than a smartphone.