Neuralink is a Fitbit in Your Skull With Tiny Wires to Fix Your Brain and Spine

Neuralink is recruiting for great engineers to make brain repair affordable and reliable. They implanted brain read-write devices into brains in pigs. They have received emergency FDA approval in July and will soon implant into humans in clinical trials.

The information for this article is almost exclusively from the Neuralink demo. The video of the entire demo is embedded below.

Neuralink needs chip designers, software people, product designers and robotics people. The team has 100 people now and eventually, they will have 10,000 people. These were statements from Elon Musk. Neuralink has a job board with job postings. There are currently about 20 job postings. They could hire multiple people for each position. Check with the company Neuralink for updates on hiring.

The sensors are like little microphones that “listen” for brain spikes. Brain spikes are action potentials in neurons that are the main means of communication in neurons.

The goal is to make it faster, safer, smaller with millions of brain sensors and provide to millions of patients with fully automated implantation.

The first clinical trials will be in people with severe paralysis.

Pigs are great models for research and can prove the robustness of the devices.
The Neuralink devices will approach the cost of smartwatches.

They have a full custom ASIC with an analog front end to amplify and sense the microvolt signals. They need to be very energy efficient.

An affordable, reliable and brain interface with thousands to millions of two-way probes will be able to fix the following:

* memory loss
* hearing loss
* blindness
* paralysis
* depression
* insomnia
* extreme pain
* seizures
* anxiety
* addiction
* strokes
* brain damage

All of your senses could be improved or restored.

It is less than the thickness of your skull. It is inductively charged so it charges like your smartphone.

They are making a really good surgical robot.

A coin-sized piece of skull if removed and the robot inserts the coin-sized device and all of the electrodes.

They have installed the 1024 brain electrode implant into a pig for two months and they used dual brain electrode implants into three pigs.

They have implanted the link into and out of pigs.

Right now they are limiting implantation depth to the first 3-4 millimeters of the cortex. Hearing, sight, and paralysis are solvable at this depth. They plan to go deeper. They will need longer wires and adjust the robot to go deeper. The wires are 7 mm long. They will need to solve sensing problems to avoid veins.

They think they can go sub-micron threads from 5-micron threads.

They will enable people to get upgrades of the devices.

They will be able to enhance senses. Sight and hearing could go beyond human ranges. We have those kinds of sensors now in planes and cars and the feeds could be converted and fed to the visual or hearing areas of the brain.

SOURCE -Neuralink
Written By Brian Wang, Nextbigfuture.com

60 thoughts on “Neuralink is a Fitbit in Your Skull With Tiny Wires to Fix Your Brain and Spine”

  1. When the Piggies in Tesla Semis emerge from the Boring tunnels well equipped in their tailored space suits with lit flamethrowers and solar powered ventilators humming, you know Elon is coming for those responsible for the ‘fascist’ coronavirus shutdown!

    Reply
  2. Btw, the simplest safety measure is just let people control the power switch to their own BCI. Turn off the power, and any torture or privacy concerns end right there.

    Barring that, just go in a Fataday cage to block external transmissions. The proverbial "tin foil hat" might be enough. If you think you're being tortured, stay in a Faraday cage for a week. See if it gets better.

    Bottom line, if it's too obviously bad, no one will want to install them. And there'll be a public outcry to fix it. But if it's subtle enough to not get noticed, then is it really that horrible?

    I expect there'll be a whole industry for BCI security, just as there is now for cyber security.

    Reply
  3. "This 'fire' thing is really dangerous. Imagine if it goes out of control – it can burn everything. We must limit this to one small log at a time. Some things just shouldn't be tried, not once."

    If this had been said when fire was discovered, we'd still be living in caves. Similarly, limiting BCIs to read-only will completely cripple their useful potential. And even read-only can already be pretty dangerous.

    People imagine all sorts of horrors with every new technology. Most of those don't pan out, or pan out not nearly as bad as imagined, because other people make sure to put in the necessary precautions.

    It's important to discuss the dangers, so that they can be identified and mitigated in time; but there's no need to panic.

    (Btw, we already have write-enabled brain implants. They're used to help people with epilepsy, parkinson's, depression, and hearing loss, to name a few. With read-only, we can't even do that much.)

    Reply
  4. they are investing in a *50,000 sq. ft campus* which is absolutely huge

    50,000 sq. ft? My unit converter says this is less than half a hectare. For a campus that's… rather…. not huge.

    Reply
  5. That plus the pig army will be connected to Starlink and will be able to coordinate kinetic bombardment from Starships in orbit.

    Reply
  6. These people are so obsessed with new "exciting" tech that they fail to see the obvious downsides of it. Living in their futuristic bubble that if someone dares to point out why it isn't a good idea, they'll instantly get upset.
    Imagine trusting billionaires, let alone humans to make these possible mind control devices unbiased and safe.

    Reply
  7. It seems like you're confused. This is a progress update for a company in early days, not a debut of a finished product. To me the most important things from the update are: a new surgical robot that seems to work, a working BCI which seems to be comparable to the Utah Array already even though it's an early prototype, and they are investing in a *50,000 sq. ft campus* which is absolutely huge and shows it is a serious investment.

    Reply
  8. I already have a rather high IQ. Not bragging, just saying.

    Aside from being time-limited and "single-threaded", while also acknowledging that it is entirely possible to be a genius at some things and, simultaneously, an idiot (or at least average) at other things, the problem is memory.

    While it's not precisely true that if you don't use it, you lose it, it can and does take awhile to get back up to speed on something you haven't been using. Yet if I could have something like a near-perfect long term memory . . . .

    Reply
  9. But, in 1959 the USA did have liquid fueled rockets, both home grown and captured from Germany, that had been developed to the point where they could achieve suborbital velocity, and orbital with multistage systems.

    The equivalent prototype brain interface is not obviously here yet.

    Reply
  10. Yes I have refuted your point. Unlike you, I listened to the presentation. The future – not the current – neuralink HW will allow connecting a powerful "co-processor" – what Elon Musk calls a "tertiary cortex" – to your brain. This information was not given in the slides, you had to listen to the presentation.

    What is left unspoken is that there will be future ANN's with superhuman intelligence irespective of neuralink HW. Neuralink is the "link" that allows expanding the human brain by a new part. But this is pretty obvious, isn't it? 

    The only question is when the singularity will occur; 10 years? 30 years? 50 years? Keep in mind that it is SW that is lacking, and we have an upper bound on the maximum complexity (human genome information) of that SW….

    Reply
  11. I always enjoy watching electrical storms on lightningmaps.org when they pass overhead. In a similar way, could Neuralink 'pinpoint' the position of each actual neuron that is firing, and its signal strength, by triangulation between several strands of neural lace? That way a receiver would not be needed for each individual neuron and much better detail could be observed.Just a thought…

    Reply
  12. Doesn’t matter how many electrodes you have in somebody’s head…. you still have no way of finding the correct neurons that represents the symbol that you want to read… the best you can do is put the electrodes close to the sensory inputs and outputs to the brain and decode the I/O before the network branches out massively…. also, you might be able to place electrode onto bottle neck points in the brain like the bridge between left brain right brain,,, if you can decode the left-right-bridge of the brain that might get a ton of useful information… beyond that…it exponentially more complex to decode what a neuron firing means in terms of symbols…

    Reply
  13. Agreed. Every technological revolution starts small.

    But here, I'm not sure they even have one. Or something that clearly differentiates them from the rest.

    Yes, they might eventually add much finer neuron access and writing. But that's pending to happen.

    Reply
  14. So, anyone that isn't a billionaire should just shut up?

    What I do is irrelevant (pun intended) to making a comment. He was touting something way above of what he showed.

    Musk tends to exaggerate his business ideas, even if he delivers a fair bit of them, but in this case I think he went too far.

    Reply
  15. The US didn't have much of a space-program in 1959, either. 10 years later there were men on the Moon. You can talk about ingenuity, you can talk about money, courage, and dedication as well, but what really made it happen was generating enthusiasm in the general public (whether it was emotions generated from feelings characterized by "Beat the commies" or "Space: The final frontier") is almost irrelevant.

    Reply
  16. How these things respond to a 100 year super solar flare is definitely going to need evaluation. On the other hand, if I was going blind, I'd probably be willing to swap my right arm for one if it would let me see.

    Reply
  17. Gay teens and even children are already being "cured" by being "transitioned" to the opposite sex, I mean, a gender, non-gender or pluri-gender opposite, parallel or tangentially at right angles to the one they are cruelly "assigned" at birth. It's the ultimate gay conversion therapy – but you won't find many "theo-conservative" types behind it.

    Reply
  18. So we're talking about technology that can "certainly help people with neurological diseases, disabilities and those suffering locked-in syndrome", but you are "not very impressed"?

    I can't wait to hear what you are working on if you look down on this research. Keep in mind this is just an early demo to try to help with recruiting efforts and company was only founded 3 years ago. This is very early to have working product in any form.

    Reply
  19. Well, see… you haven't really refuted or countered my point.

    NeuraLink is billed as brain-machine interface, not some kind of coprocessor.

    Reply
  20. Agreed. Though I would like to point out, I've never actually met anyone who wants to "cure" gays. I think these people are about as common as snow leopards, whereas the hard core social justice crowd makes up, say, 0.5% of the population..? So about 1.5 million in the USA..?

    Reply
  21. Not necessarily. 

    Musk argues that we have distinct regions that cooperate intimately. The fact that we have human intelligence and not chimpanzee intelligence is due to the advanced frontal cortex that executes the wishes of the primitive "monkey brain". I don't know how true or false this is.

    He furthermore argues that another "brain layer" made in silicon would extend our intelligence beyond human intelligence. And because you would join the brain with the "tertiary layer" in a vary intimate way, our "lower" layers would in fact be controlling the new (and higher) layer. The key here is a direct neural, broadband, interface. You don't have that with the computer screen.

    My comment to this is that you will probably not be remotely the same person any more. If any human would lose their frontal cortex that person would be gone. Similarly, if you would add human intelligence to a chimpanzee, that chimpanzee would no longer exist. And if you add superintelligence to a human, he/she would be human in name only.

    But if this is the only way to survive the advent of computer superintelligence, it might still be worth it…. Though I doubt that the cortex-to-computer interface can keep pace with the ever improving pure artificial intelligence…

    Reply
  22. I'm sorry to harp about this torture angle. What is someone implanted images of how you raped your mother in your mind? There would be no physical damages on you body and it might be very difficult to prove that it had even taken place. You would be completely destroyed, never the less.

    But it doesn't have to be so dramatic. What if an interrogator implants visions of how a citizen does minor, but bad actions. Push an old lady, steal from a beggar, ignore his kids etc. This would probably "brake" any civil rights activist. Make them doubt their own moral worth..

    Or even more subtle. Implant vast amounts of just boring memories.. How to prove that this has taken place, even if you could establish that someone had a great deal of such memories? And what would the effect be? Would the victim be able to separate true events from "projected" visions?

    If this was even a remote possibility, I think that a lot of people would live in fear..

    Reply
  23. I don't have that kind of faith in the justice system. Just take Germany for example. At first, they argued that it was reasonable to get cell phone records when the suspected crime was terrorism. Then they included child pornography and finally within 6 months, any kind of crime. Any, such as tax evasion, jay walking or littering…

    And did you read about the FISA warrant? Even the highest court in the USA (except for the supreme court, of course) was duped into allowing spying on the Trump campaign. So what do you think happens when the next Hillary Clinton comes to power, if it is technically and legally possible to extract someones thoughts?

    And what is to say that humans will even have an exclusive right to their own thoughts in the future, in say, 100 years?

    Reply
  24. I wonder if a future version will do something more like phased-array targetting of individual neurons. Phased array "read" might already be possible with their existing technology, though I heard no indication they're trying that.

    Neuron-targeted phased-array writing would be tougher. Neurons would be at different pre-existing excitation states, so even low signals from multiple electrodes and aimed at triggering a particular neuron might trigger other nearby neurons. Maybe trigger all nearby neurons to a ground state and then write a new excitation level to each? Not all patterns of excitation would be possible. Seems like a more promising "write" approach would be to aim to stimulate a single neuron per electrode contact.

    Long term, I think we'll need to move beyond this crude electrode approach. We really want something that can 'read' the current excitation level of individual neurons and modify (not necessarily trigger) that level up or down. That's a lot finer-grained approach. A lot of the visions Elon mentions probably can't really happen until we get that level of tech.

    Reply
  25. "how long before a conservative mind is seen as something that should be "corrected"?"

    Five Seconds.

    (As stated, it is already true, but I'll assume you meant "…corrected once this technology is available".)

    I'm sure someone, somewhere on the webs has already suggested using it to fix Trump – and to be fair, some theo-conservative has almost certainly enthusiastically suggested using it to 'cure' gays.

    Reply
  26. We already have very effective means of torture, so I don't see what the big deal is there. It's like worrying that we could do some very painful things now that we've invented surgical techniques. Yes, some people will make a movie about a human centipede and we'll all shudder a bit.

    As for the privacy bit… that's a little more interesting. If we already have legal precedent for getting a warrant and invading the privacy of your technology (phone, PC), then what's the difference between that and getting a warrant to read the bit of technology resting on your skull? This might make for a very different court system – lie detectors that reliably work. Justice would we simpler.

    Reply
  27. I don't know that it would increase innate IQ. In fact, I'd say it would no more increase IQ than had radio, TV, telephones, or computers.

    Reply
  28. Porn is easy – the cool new trick will be convincing your brain that what you're watching is what you're doing, or maybe have done recently.

    Reply
  29. I certainly get what you're saying, specifically the brute forcing part. A standard dictionary attack on a network password can potentially take millions of hours (even years), while using a finer grain approach (the names of which don't bear mention, here) can take less than a day. It's about the method. I'm not so sure that they won't find a way to get a finer grain of control. They very well may do that.

    Reply
  30. image kinda look like there is a coil in there … so weather channel is going to start mentioning when there is solar flare incoming so i can go into my basement faraday cage or i'm just taking a chance my brain will fry one day ?

    Reply
  31. And then we have the question about "correcting" personality disorders. You start with violent psychopaths, but then you also "correct" for lesser mental disorders… Looking at american campuses, how long before a conservative mind is seen as something that should be "corrected"?

    And how about when the state decides it needs to monitor our violent tendencies in real time? When you want to punch someone, the brain interface "zaps" you, kind of like an internal brain tazer. Or just freezes your muscles…

    Talk about making us into meat puppets…

    Reply
  32. I am surprised that no one in the audience brought up the obvious questions of privacy and torture.

    If, as one of the team members wanted, you would be able to transmitt thoughts without using language, you should be able to read thoughts of a subject without his/her consent, right? So if you have a warrant, you can look into the suspects brain. Or perhaps without a warrant…

    And torture… If you can transmitt ideas and images, what's stopping that from being used as the supreme version of torture? Horrible ideas, horrrible images, horrible pain…

    Reply
  33. Given that simple brain scans prove that Primals make a huge difference in brain state, not to mention the clearly obvious epigenetic differences, this tech may be able to find more evidence. It has been 50 years, so maybe it is time. Evidence of neurosis continues unabated, see news for details.

    Reply
  34. You have to start somewhere, fake it till you make it.

    Were people really expecting something akin to a culture neural lace sometime in the near future?

    Reply
  35. You can already offload computing from the brain to an external device. The simplest example is a calculator. But current interfaces are rather clunky.

    That can improve once they can read the motor cortex in more detail, but for the really advanced stuff, they'll need to decipher intent from other parts of the brain.

    Other than that, if/when they can write to short-term memory, that can potentially boost knowledge acquisition quite a bit (somewhat like the famous "I know Kung Fu" scene). Initially they may only load bare facts, and you'd still need to process them to learn anything. That could already be faster than having to read them. But eventually they might be able to load processed knowledge directly (aka "understanding"). The difference is similar to loading source code vs loading compiled binaries, respectively. (Note that inside your brain, you still need to load processed knowledge into short-term memory to use it. Here you'd just be loading the same thing from an external source.)

    But all of this is still a long way off.

    Reply
  36. Of course this not how the brain is supposed to be fixed. This is about another maniac who in disguise, made up excuses and already a substantial herd of blind followers is exploring his version of humanity control. He is also doing it with starlink, the million people colony on Mars and who knows how else!

    Reply
  37. You have to think of this as a pre-Falcon 1 prototype. There is still a long long way to something on the level of Starship or beyond. Their immediate concerns are:
    1. Increase electrode resolution (fewer neurons per electrode).
    2. Simultaneously increase electrode number. With current tech it's usually a trade-off: more electrodes or more resolution, not both.
    3. Improve the implantation technology.

    I don't know how they compare to state of the art, but they seem to have made some progress. We can reasonably expect them to make a lot more progress given time.

    Reply
  38. You are welcome to wire yourself up and become Elon's VR-porn addicted pavlovian pet.
    For the rest of us, it is urgent to make sure that the sponsors of this trillion$ venture would not have the power to end humanity as we know it by forcing this tech onto everyone (except themselves).

    Reply
  39. A warning: many people here are Elon Musk fans, for whom he can do no wrong. I'm only applaud when he funds and delivers something cool.

    And I was not very impressed with this either. Not after seeing the description of Neural lace objectives in some blogger's page.

    This is why: decoding the brain signals with deep learning is pretty much state of the art. Impressive the first time you see it, but not groundbreaking.

    And about the brain-writing part: they are trying to brute force their way into the brain like trying to program a computer hitting it with a monkey wrench. There's no precision nor fine grained model of the brain in what they are doing.

    The only part that has potential is the implant technology for getting brain signals out, and improving some neurological conditions that are already known to improve by zapping electricity to the brain. This can certainly help people with neurological diseases, disabilities and those suffering locked-in syndrome, but it's very far from the brain-connection nirvana they touted.

    Reply
  40. If they can develop something that is not so invasive, I am in.

    But if I'm dying, and somebody wants the turn me into Robocop, then wire me up!

    Reply
  41. You thought 2020 has been a rough year? 2023: year of the AI controlled pig army drone swarm delivered by autonomous Cybertrucks a dozen at a time…with lasers strapped to their backs and their ability to wade small ponds in a minute or two, zap you in the butt and find any mushrooms and other old emergency Covid food supplies you have hidden, our days are numbered. 
    You mean you did not see Elon's master plan coming? Obvious stuff 🙂

    Reply
  42. Can you imagine using this to connect multiple peoples brains together like a multi core processor. Or maybe using multiple pigs brains networked as one huge biological computer.

    Reply
  43. By just as much as sitting in front of a computer screen does. It's not a co-processor for your brain.

    At least not until they can externally emulate a part of the brain and send back and forth signals, or connect multiple people to share brain power.

    Reply
  44. Is the short bit at around 20:30 all they have on "writing" to the brain? I didn't watch the whole video, so maybe there's more, but this seems really unimpressive. All they show is that they can trigger neurons by sending small amounts of electricity. Thousands of neurons per electrode, which sounds impressive, but consider the brain as a complex circuit board; you would not want to send electricity to thousands of components on a circuit board–that would probably damage it, definitely would not make it work as designed.

    The "reading" seems impressive, too, but I'm sure you could make just as accurate predictions with a machine learning model and high enough fidelity data from *any* sensor. Researchers recently decoded the keying to open a lock using just the sound of someone opening a lock with a key. Computer activity can be deciphered with just a camera trained on the hard drive activity light. I have no doubt one could predict the position of a pig's joints in real time using just audio, probably even heart rate and galvanic skin response. They're not "reading the brain" with this device. And there's absolutely no evidence that they can "write to the brain" by sending electricity to it.

    This device may be useful for CBI control of advanced prosthetics or other devices, but probably not much better than existing devices.

    Reply
  45. Really exciting stuff. Brain-computer interfaces are, in my opinion, the crux to our future advancement as a species.
    And I really hope that we can get fully immersive virtual reality in my lifetime, through the next generations of this tech.

    Reply
  46. Could such a device improve the efficiency of workers on an assembly line or the harvesting of crops? Is is possible to make the proletariat more complacent with devices such as this implanted in their skulls?

    If so, it might be time to take profits in your Tesla or Amazon stock and reinvest in this technology.

    Reply

Leave a Comment