1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
|
Mikhail Shapiro
Thank you David. Thank you for having me. I am excited to be a part of this. There are a many a Saturday morning where I wake up thinking about how to enhance myself as human beings. But my wife is like a captive audience, so she doesn't count.
So, I think my favorite thing, my perspective here comes from the point of view of a scientist, but also someone who has spent some time on the commercial side, innovative companies that pursue this kind of tech. This potential as human beings that we have- in 80 years, to really enhance ourselves. Brain-machine interfaces, uploading, learning more efficiently and quickly, greater sensory powers. There are lots of other things that can be done with life extension, stem cells, etc., they seem very far away. There are multiple things standing in the way. There are multiple things that are going to be needed.
Performance requirements for making healthy humans even better, the benefit needs to be high, and the safety margin needs to be large. Depending on individual customers, it's a difficult tradeoff. The performance of these systems needs to be robust. Whether you're thinking thinking of investment from venture capital, societal, VC is part of my perspective, it's a long and expensive investment proposition. Risk adjusted and time adjusted, it's just not justified.
What I would like to talk about today, and I think a lot of speakers have touched on this, let's think of diseases as stepping stones to the ultimate technological goals. This is based on neuroscience. There are millions of people who are suffering from these diseases. They are suffering, as well as their loved ones. I am going to use my experiences as an example, and initially we're applying these to diseases.
Brain-machine interfaces. When I think of this, robocop always comes to mind. You take a brain, you put it in, motor and sensory inputs and outputs allow you to be a part of this motorized system. I would argue that we would fruitfully focus on locked-in syndrome, spinal cord syndrome, where people have lost the ability to control movement, and we have the ability to regain them. There's a nice perspective on this condition (movie reference).
In 2001, 2002, at Brown University. But really, the problem with spinal cord syndrome, the brain generates movement signals. The signals do not get approved, because there's a transection in the spinal cord. The idea behind brain-computer interfaces as applied to people with these diseases, and you can do this with technology like a chip, and you can re-route them for going directly from the brain to the limb that you are trying to control. There's a lot of science over many decades, there's a relatively simple mathematical code for controlling a cursor on a screen, and things like robotic arms and wheel chairs. This was some of the work, just some. I guess not. Some of the time. 2001. Happening at Brown, you can have brain-computer interfaces in their cortex, and they can play video games directly with their thoughts. You can look it up online in Nature. The dot appears in different parts of the screen. What we did was form a company called CyberKinetics that would take this tech into human beings.
A lot of people would be familiar with the BrainGate story. There was a gentleman with a spinal cord injury. There's a patient with Leugs Erik Disease. All of these people were able to gain significant amount of control. There was a demonstration of a robotic arm, and a wheel chair. How this tech will progress, well, against healthy humans, we need to think about what it offers at different points of time. We focused on people that were very paralyzed- people who couldn't control their human head movement, or maybe an eye blink, or a control of your sphincter. Some people have made interfaces based on that, control of your sphincter. Ultimately, if you want to expand people who are less paralyzed, you need to offer more, multi-dimensional control, performance, etc., movement in another arm. Ultimately, if you think about healthy people and brain interfaces, you have to ask, what are you offering? There has to be a practical.. I don't know how many of you would put a brain implant in your head just for the hell of it.
We talked about brain reading and uploading. You want to look at specific molecules in the brain, human being, ideally in real-time, and you want to read them out, and recompose what that human is doing. There are other approaches and some people next will talk about it. Imaging specific molecules in the brain, that's important. There's a dye to look at dopamine levels. Epilepsy, and rare genetic diseases. You want to look at specific molecules with non-invasive imaging. You don't want significant operations. Viral vectors, sensors that's encoded in the gene, inject it, travels up to the brain, and then beams out information about dopamine. We've started to do this, like at MIT.
We borrowed a concept from fMRI. A lot of the results that you see with fMRI are accidental: the oxygen protein (hemoglobulin) has iron in the center. Whether or not it has oxygen attached to it determines whether or not it sends a signal. They just happen to have an oxygen sensor in all of us. What we did was to create another protein that also had an iron center, and it's sensitive to something other than oxygen. It's a sensor of dopamine. We can sense dopamine activity in the brains of rat. We want to deliver this protein to the brain. There's all sorts of tech for delivering this. There's lentivirus, there's so on. We recently invested in a company called Genetics, and do glial cells that migrate to the brain as the vector. Now, through this tech, patients that die at age 5 or 7 with this rare disease, can be cured.
You can make real progress with rigorous science that will lead to the kind of ideas that we want to talk about. The tech is being developed. When you stimulate parts of the brain, the learning session, you can enhance learning. I think Ed Boyden's work with channelrhodpsin is a wonderful example of how tech can be useful, like by making blind people see again. It helps us gain additional information, orthogonal to our normal visual reception.
This is my last slide. Success factors for commercial clinical ventures, this is the best way to finance this. Look for a disease that isn't served by existing tech. What population is willing to try out just about anything? And you need to have high quality science, it can be from people in the garage, or where-ever- MIT or Harvard. You need rigor. Scientific rigor. You need to get onto the market in 3 to 5 years. You need strong backing, and great management to make it happen.
Let's look for those appropriate stepping stones. Develop rigorous science. I want to get to those goals.
|
