Neuralink broadcast
Elon Musk
2020-08-28
video: https://www.youtube.com/watch?v=iOWFXqT5MZ4
Introduction
Alright, welcome to the new product demo for Neuralink. Pretty excited to show you what we've got. I think it's going to blow your mind. The primary purpose of this demo is recruiting. I'll emphasize this at the beginning and then at the end. We're not trying to raise money or do anything else. Our focus is to convince great people to come work at Neuralink, bring the product to fruition, make it affordable and reliable so that anyone who wants one can have one.
What is our goal? Our goal is to solve important brain and spine problems with a seamlessly implanted device. You want a device you can put in your head, feel and look totally normal, but it solves some important problem in your brain or spine.
The reality is that almost everyone over time will develop brain and spine problems. If you live long enough, everyone is going to have some kind of neurological disorder. These range from memory loss to brain damage. The thing that is important to appreciate is that an implantable device can actually solve these problems. I think a lot of people don't quite realize that.
All of your senses, sight, hearing, feeling, pain, these are all electrical signals sent by neurons to your brain. If you can correct these signals, you can solve everything from memory loss, hearing loss, blindness, paralysis, depression, insomnia, extreme pain, seizures, anxiety, addiction, strokes, brain damage. These can all be solved with a magical neural link. This is an extremely fundamental thing. I don't think a lot of people understand that.
The neurons are like wiring. You need an electronic thing to solve an electronic problem.
Current medical research
Current medical research-- what's the state of the art? Current medical research has shown that you can read neurons in the human brain. It's something called the Utah array which has 100 channels per array. It's a bed of ridged spikes that is inserted with an airhammer literally. It's slightly disconcerting (discomforting?) I think. There's wires and a box on your head. There's some infection risk. It also looks pretty weird if you walk around with boxes on your head. In order to use it, you need an extra medical professional there. It's only been done to a few people. It's served as a point of proof of concept that this be done. It's not something the average person can use effectively.
Currently available tech
There's something called deep brain stimulation where they put a small number of electrodes into your brain and they will zap your brain with electrical current. It's useful, but cannot read or write high-bandwidth information. It's like a tv that works but not always, and it has limitation.
This has greatly helped over 150,000 people. Despite being a bruteforce approach, it has been very effective for a lot of people. This is what is currently available.
We want to radically improve this by multiple orders of magnitude, 100x, 1000x, then 10,000x.
Neuralink architecture
We have dramatically simplified the device over the past year. A year ago, we had a device that had multiple parts including a piece that had to sit behind your ear. It was complex, and you wouldn't look totally normal-- you would have the thing behind your ear.
We have simplified it to something that is about the size of a large coin and it goes in your skull, replacing a piece of your skull, and the wires then connect within a few centimeters or about an inch away from the device.
This is sort of what it looks like. It's sort of like a fitbit in your skull, with tiny wires.
Link v0.9
Our current version, link v0.9 has about 1000 channels which is 100x better than the next best consumer device that is available. It's 23 mm x 8 mm. It fits quite nicely in your skull, which is about 10 mm thick. It's flush with your skull, it's invisible, and afterwards you only see a tiny scar. If it's under your head, you wouldn't know at all. In fact, I could have a neuralink right now. You wouldn't know. Maybe I do.
It has megabit wireless data rate, post compression. There's a lot of functions related to monitoring health and monitoring things like possible strokes, heart attack, or convenience features like playing music. It's sort of like if your phone when into your brain. Not a great analogy.
Inductively charged
It's inductively charged, the same way you charge a smartphone or phone. You can use it all day, charge it at night, and have full functionality. It would be completely seamless. Yeah, no wires.
Getting a neuralink
You need to get a great device, and have a great robot that puts in the electrodes and does the surgery. You want the surgery to be as automated as possible. The only way to achieve the level of precision needed is with a great robot. We're looking for people who can help develop the device, and the robot.
We feel confident about getting the installation of the link procedure done in an under an hour. You can go in the morning, and leave the hospital in the afternoon. It can be done without general anesthesia.
You open a piece of skull, you remove a coin sized piece of skull, then the robot inserts the electrodes. We'll talk more about that later. Then the devices replaces the portion of the skull that was removed. Then we close it up with super glue which is how a lot of wounds are closed. Then you can just get up and walk around afterwards. Pretty cool.
Surgical robot
This is our surgical robot. We ultimately want this robot to do the entire surgery. Everything from incision, removing the skull, inserting electrodes, placing the device, and then closing things up and getting you ready to leave. We wanted a fully automated system.
This robot does actually work. We've used it for all the implantations.
Getting a link
This shows a close-up view. I don't think it's too gruesome, of the electrodes being inserted into the brain. If you look closely, it's counterintuitive: if the electrodes are inserted carefully, there's no bleeding. If you have very tiny electrodes, and if they are inserted very carefully, so that the brain is imaged by the robot to avoid the veins and arteries so that the electrodes can be inserted with no noticable neural damage in inserting the link.
You sort of think if you stab something with a wire it will bleed, but at a very small scale it won't.
Does it work? The three little pigs demo.
Does it actually work? I'm excited to show you the "three little pigs" demo. Let's bring out the pigs. We'll cut over and show you.
What we have in pen number one is Joyce, and she does not have an implant. Obviously, healthy and happy. This is how you know it's a live demo. She's trying to eat something in the corner of her pen... We'll get to her in a second.
Let's move on to Dorothy. The pigs are a little shy. Here is Dorothy. She used to have an implant, and then we removed the implant. This is an important thing to demonstrate, is reversibility. If you have a neuralink, and then you decide you don't want it, or you want an upgrade, you can get it removed in such a way that you are still healthy and happy afterwards. Dorothy demonstrates that you can remove this, and be indistinguishable from a normal pig.
Gertrude, are you serious? Okay. Can we get them all in the same pen? It will be a little crowded. Is Gertrude still back in the thing? Okay, we need to bring Gertrude out. This is a real, live demo.
Can we zoom in to Gertrude? She's clearly interested in something in the back of her pen.... Can you see her? Alright. Alright, it might take a sec. Well, this worked earlier. Alright, well.... Can we lift the curtain and then zoom in? Alright. There we go. Great.
This is .... Gertrude, thanks for coming out. What you're-- the beeps you're hearing are real time signals from the neuralink in Gertrude's head. This neuralink connects to neurons that are in her snout. Whenever she shuffles around and touches something with her snout, that sends out neural spikes which are detected here. On the screen, you can see each of the spikes from the 1024 electrodes. Then if-- if she shuffles around and finds some food, you can see the neurons will fire much more than when you're not touching the snout. That's what's making the beeping sounds.
As you can see, we have a healthy and happy pig. Initially shy, but obviously high energy. Had a loving life. She's had the implant for 2 months. This is a healthy, happy pig with a 2 month old implant and it's working well. Yeah.
Alright, cool.
I hope this works. We actually have--- we said, what if we do two neuralink implants? We've been able to do dual neuralink implants in three pigs at this point. We have a couple of them here. We're able to show you can have multiple neuralink implants. Again, healthy, happy and indistinguishable from a normal pig. It's possible to have multiple neuralinks in your head, all sending out signals, and be working well. Thank you.
Reading brain activity
We just showed you a demonstration of reading brain activity. As I was saying, each of those dots represent a neural spike. The blue chart at the bottom is showing the accumulation of neural spikes in that region.
In terms of initial brain reading activity, when we have say one of our pigs on a treadmill... pig on a treadmill... funny concept, really.... and we take the readings from the neurons, and we try to predict the position of the joints. So we have the predicted position of the joints, and we measure the actual position of the joints, they are almost exactly aligned. With a wireless neural implant, we were able to predict all the positions of the limbs in the pig's body, with very high accuracy.
Writing
In terms of writing to the brain or stimulating neurons... we obviously need precise control of electric field in time and space. We also need a wide range of current for different brain regions which require different amounts of current. And you don't want to harm the brain.
We analyzed stimulation of neurons with a two-photon microscopy. It's very impressive technology. You can see in real-time how the neurons are firing. The red things are the neurons-- reds or flashing things are neurons firing. I should say, the electrode firing. The red things are electrodes firing, and the green are the neuron bodies responding to the current from the electrodes. You can see them lighting up different brain regions. By carefully controlling the electron field, you can have a single electrode stimulating a thousand or 10,000 neurons. You can be influencing millions of neurons from a single electrode.
Here's a similar chart showing stimulation at different power levels.
Read and write on every channel
For the initial device, it's read/write on every channel. It's 1024 channels. It recharges over night. It has quite a long range... you can have the range being to your phone, I should say. That's an important thing. This would connect to your phone. Actually, the application would be on your phone and it would be communicating by essentially bluetooth low-energy to the device in your head. In a lot of ways, it is like a fitbit in your skull with tiny wires.
You will not be able to see the device at all; it will look completely normal and you'd have a small scar under your hair.
Clinical studies
We've received a breakthrough device designation from the FDA in July, thanks to the hard work of the Neuralink team. We're working closely with the FDA and we'll be extremely rigorous. In fact, we will significantly exceed the FDA's minimum guidelines for safety. We will make this as safe as possible.
With Tesla, while it's legally possible to ship a 1 star car, the cars we make are 5 stars in every category. We maximize safety at Tesla, and we'll take the same approach at Neuralink.
Our goals
The goal of this presentation is recruiting. We want people who are great at solving problems to join the company. Email engineering@neuralink.com -- write the software, create the chips, and productionize everything. We'd like robotics engineers. We also need people who have worked on and shipped products. If you've shipped a smartwatch or a phone, or any kind of complex electronics or device, or advanced medical devices. We'd love for you to contact us and consider working here. A very important point is that you don't need to have prior experience on brains. Some people think they can't work at Neuralink because they don't have bcakgrounds in brain; that's okay. We need electronics, mechanical engineering, product design, all the things that companies need. It's not just engineering- it's everything at Neuralink.
Questions
We have questions that have been asked over the internet. We'll do some live Q&A. Let's bring in a bunc hof people from the Neuralink team. Alright. If you have any questions, please submit your questions to the Neuralink twitter account. We'll try to answer as many questions as we can over the next hour. Feel free to ask hard questions. Let's move over to the team.