Neuralink 2020 update — highlights and comments
Dear Readers, have you been waiting for Neuralink’s updates on brain-computer interfacing? I suppose that the answer is yes!, as you are here with me, taking a closer look at the Neuralink-related stuff.
After the last year’s Elon Musk & Neuralink presentation, I prepared a brief article for you on the basis of the Neuralink’s white paper [1], indicating the novelty of their neurotechnological achievements and providing you with my short comments. If you are here for the first time or you’d like to refresh some information before you start with the current blog post, you can find the previous one here [ CLICK HERE].
The structure goes like this — we’ll start with some highlights from the yesterday’s Neuralink Progress Update, Summer 2020 [2] presentation with some explanations if they’d be needed. The last, but not least, would be a bunch of my comments targeting what has been said during the presentation. Let’s start the journey then!
Neuralink presentation highlights
(Presentation is available at [2] and [3])
Experimenters, lab coats, robotic arms, testing, biosignals. A short introductory video of Neuralink’s R&D work invites us to participate in the presentation of Link V0.9 — a small, implantable device with a growing potential to link humans and technology. Literally.
Elon Musk started his speech with an invitation to collaborate on the project and to develop Link (BTW. If you work in such tech field, you may be interested, see [3]). Afterwards, the idea behind their research efforts was highlighted. Their aim — to solve brain-related problems — was further exemplified with a number of brain disorders or mental health disorders, such as memory loss, blindness, depression, brain damage or stroke. You may ask — how? How could it work to solve any of these problems? The long-term goal is to develop, as Musk said, a generalizable device, which can help with each of these disorders.
A brief explanation here: brain-related disorders are mostly very complicated. They can involve not only a local, small part of the cortex, but even distant regions and subcortical structures, which cooperate together and transmit electrical signals to each other. If you consider that there is approx. 86 billion of neurons in human brain [4, 5] and a few orders of magnitude more neural connections, then you can start thinking about it as a really computationally challenging problem.
Next, Musk discussed the state of the current research in that field, by using an Utah array [6] as example. This is a microelectrode array (having, as it was said, 100 channels per array), which can be placed on the cortex surface and the needles are therefore inserted into the cortex [6]. The electrodes can record the neural activity of the communicating cells, which have the ability to generate an action potential. What is the advantage of Link V0.9 presented by Neuralink during the meeting?
Since the presentation from July 2019, the concept of the device has changed in at least one aspect — previously it was designed as an external element attached to the scalp, and now it was modified in order to fit in our skulls. A coin-sized chip can be mounted in the skull bone and basically be invisible to other people.
Technical description of Link V0.9 was provided: 1024 channels (note: 10 times more than in the array used by Musk as an example!); some additional sensors, such as temperature sensor; wireless; having a rechargeable battery.
It’s sort of like if your phone went at your brain or something
E. Musk, Neuralink 2020 presentation [2] (8 m 18 s)
(No, it isn’t, but I’ll discuss it later)
Next topic: the surgical robot. This one is pretty impressive! Last year I really enjoyed the idea and I still consider it as their greatest achievement in the project. The design of the robot has changed, however I didn’t notice whether any of its mechanical properties did. You can read about it in my last year’s blog post here.
The Three Little Pigs Story
Three pigs: Joyce — doesn’t have an implant; Gertrude — currently with an implant; Dorothy — had an implant earlier. What is the story? The pigs were presented to show that they’re in a good condition with an implant or after its removal.
(Note: I don’t want to discuss here any animal testing-related topics. There were representatives of Neuralink’s animal care unit present during the Q&A session.)
There was a presentation of Gertrude’s brain activity, as she was going around and touching the ground with her nose. What we could see was that the number of spikes (recorded neuronal events of activity) was greater, as the pig was actively using her nose. There was an online presentation of the recorded signal. It’s really interesting!
Up to that moment, we could get the information about the recording process. How do you think, can a recording-only device help us to prevent any of the disorders listed above? The answer is: rather not. So, next to the recording part, there is a stimulation part needed in order to induce some changes in the brain tissue. And, as you may intuitively expect, the last part of Elon Musk’s presentation covered the topic of brain stimulation, as he shared some microscopic images of the neurons being stimulated with the electrodes.
Q&A session
The presentation was not long, and most of the meeting was dedicated to the Q&A session with Elon Musk and Neuralink team leaders. I don’t want to bring all the questions, as most of them was basic / technical / futuristic, I think. However, there was a single question which presumably most of us see very seriously: what could be the first application? How do you think?
They answered that Neuralink’s aim is to first help the patients with paraplegia or tetraplegia, which are the disorders related to the spinal cord injury. Neuralink’s ambitious goal is to somehow bypass the injury and to help the patients to regain their motor abilities.
My brief comment
Well, I have to admit that I wanted to make this blog post even longer by adding some futuristic-transhumanistic thoughts, so if you got here — big thanks, guys! If you got here and would like to read my opinion — even bigger thanks!
Generally, I enjoy the overall idea of invasive brain-computer interfacing, as this technology is promising in many ways, e.g. by providing a direct brain read/write channel with good electric and mechanical properties and targeting the specific location, which may be relevant for its clinical application. However, the research is still in the early stage, and from the current point to the point in which one can cure or at least control a brain-related disorder, there’s still a lot of work that has to be done.
During the presentation a number of the disorders to help with was listed. As far as I know, most of them differ in etiology, neural pathways, target locations and ongoing biochemical processes (such as the levels of neurotransmitters or products of the neural tissue damage). Each of these disorders is being carefully examined by researchers around the world, even for decades. Many mechanisms are still unknown or mysterious, and therefore in most of the cases it is difficult to estimate what ought to be recorded and stimulated in order to restore specific brain functions. This is why I don’t enjoy this easy-going style of presentation, which may make the layman feel that all this research is basically easy to implement. Not yet, it’s much more difficult. But okay, you may say that this is only a pop-scientific explanation and I shouldn’t worry about that. Maybe, but I don’t see this as an excuse to simplify the issue.
To briefly sum up, I do enjoy the idea from the neurotechnological perspective, I really like this and I keep fingers crossed for Link V0.9 and the further versions. But the promises that we will be able to stream the music to our brains [7] or have ability to save and replay memories [2 (46 m 50 s)] are still quite far from now to be achieved.
References:
[1] Musk, E., Neuralink (2019) An Integrated Brain-Machine Interface Platform with Thousands of Channels. (white paper)
[2] https://www.youtube.com/watch?v=DVvmgjBL74w&feature=youtu.be (Access: 29.08.2020)
[3] https://www.neuralink.com/ (Access: 29.08.2020)
[4] Azevedo, F. A., Carvalho, L. R., Grinberg, L. T., Farfel, J. M., Ferretti, R. E., Leite, R. E., … & Herculano‐Houzel, S. (2009). Equal numbers of neuronal and nonneuronal cells make the human brain an isometrically scaled‐up primate brain. Journal of Comparative Neurology, 513(5), 532–541.
[5] Herculano-Houzel, S. (2009). The human brain in numbers: a linearly scaled-up primate brain. Frontiers in human neuroscience, 3, 31.
[6] https://en.wikipedia.org/wiki/Microelectrode_array#In_vivo_arrays (Access: 29.08.2020)
[7] https://futurism.com/the-byte/elon-musk-neuralink-stream-music-brians (Access: 29.08.2020)
[8] https://pixabay.com/illustrations/brain-blueprint-thinking-analysis-1845941/ (Access: 29.08.2020)
Originally published at https://annastroz.com on August 29, 2020.
