Elon Musk Unveils Neuralink's Plans For Brain-Reading 'Threads' and a Robot To Insert Them (theverge.com) 201
Neuralink, the secretive company developing brain-machine interfaces, held a press conference today where it unveiled some of the technology it's been developing to the public for the first time. The first big advance is flexible "threads," which are less likely to damage the brain than the materials currently used in brain-machine interfaces and create the possibility of transferring a higher volume of data.
"The threads are 4 to 6 micrometers in width, which makes them considerably thinner than a human hair," reports The Verge. The other big advance that Neuralink unveiled is a machine that automatically embeds the threads into the brain. From the report: In the future, scientists from Neuralink hope to use a laser beam to get through the skull, rather than drilling holes, they said in interviews with The New York Times. Early experiments will be done with neuroscientists at Stanford University, according to that report. The company aims for human trials as soon as the second quarter of next year, according to The New York Times. The system presented today, if it's functional, may be a substantial advance over older technology. BrainGate relied on the Utah Array, a series of stiff needles that allows for up to 128 electrode channels. Not only is that fewer channels than Neuralink is promising -- meaning less data from the brain is being picked up -- it's also stiffer than Neuralink's threads. That's a problem for long-term functionality: the brain shifts in the skull but the needles of the array don't, leading to damage. The thin polymers Neuralink is using may solve that problem.
However, Neuralink's technology is more difficult to implant than the Utah Array, precisely because it's so flexible. To combat that problem, the company has developed "a neurosurgical robot capable of inserting six threads (192 electrodes) per minute [automatically]," according to the white paper. In photos, it looks something like a cross between a microscope and a sewing machine. It also avoids blood vessels, which may lead to less of an inflammatory response in the brain, the paper says. Finally, the paper says that Neuralink has developed a custom chip that is better able to read, clean up, and amplify signals from the brain. Right now, it can only transmit data via a wired connection (it uses USB-C), but ultimately the goal is to create a system than can work wirelessly. Currently, the company is testing the robot and threads on rats, but it's hoping to actually begin working with human test subjects as early as next year.
Story is developing...
"The threads are 4 to 6 micrometers in width, which makes them considerably thinner than a human hair," reports The Verge. The other big advance that Neuralink unveiled is a machine that automatically embeds the threads into the brain. From the report: In the future, scientists from Neuralink hope to use a laser beam to get through the skull, rather than drilling holes, they said in interviews with The New York Times. Early experiments will be done with neuroscientists at Stanford University, according to that report. The company aims for human trials as soon as the second quarter of next year, according to The New York Times. The system presented today, if it's functional, may be a substantial advance over older technology. BrainGate relied on the Utah Array, a series of stiff needles that allows for up to 128 electrode channels. Not only is that fewer channels than Neuralink is promising -- meaning less data from the brain is being picked up -- it's also stiffer than Neuralink's threads. That's a problem for long-term functionality: the brain shifts in the skull but the needles of the array don't, leading to damage. The thin polymers Neuralink is using may solve that problem.
However, Neuralink's technology is more difficult to implant than the Utah Array, precisely because it's so flexible. To combat that problem, the company has developed "a neurosurgical robot capable of inserting six threads (192 electrodes) per minute [automatically]," according to the white paper. In photos, it looks something like a cross between a microscope and a sewing machine. It also avoids blood vessels, which may lead to less of an inflammatory response in the brain, the paper says. Finally, the paper says that Neuralink has developed a custom chip that is better able to read, clean up, and amplify signals from the brain. Right now, it can only transmit data via a wired connection (it uses USB-C), but ultimately the goal is to create a system than can work wirelessly. Currently, the company is testing the robot and threads on rats, but it's hoping to actually begin working with human test subjects as early as next year.
Story is developing...
Why do you not want to help people? (Score:5, Insightful)
This should be banned.
If you read the article, you'd find that already primitive versions of a brain link (using a much scarier array of fixed needles) are allowing some humans with no motor control, to actually control devices.
Why would you want to ban such amazing and potentially useful technology?
The vision to essentially "merge"with AI systems is a powerful one - like deepfakes now where an AI can generate nearly anything you can imagine, being linked to such a system to give it instantaneous feedback as it is generating a result, to tweak to exactly what you would like to see could be astonished useful.
The tech the company is talking about here doesn't seem that out of bounds or even that scary, at the sizes they are talking about missteps would hardly cause any notable damage. This is just the kind of thing that is pretty much inevitable to develop over time, it seems like rather than being as fearful of it as some cave men were of fire, we should bask in the glow of the eventual progress it enables.
Re:Why do you not want to help people? (Score:4, Funny)
Just call this tech something like "functional piercings" and adventurous young people will want to try his out.
Re: (Score:2)
"The vision to essentially "merge"with AI systems is a powerful one - like deepfakes now where an AI can generate nearly anything you can imagine, being linked to such a system to give it instantaneous feedback as it is generating a result, to tweak to exactly what you would like to see could be astonished useful."
Is basically a dream of destroying humanity. It is dream of human beings overcoming limits that exist naturally and are good , it is a vision of humans ceasing to be humans and 'evolving themselv
Re: (Score:2)
Why would you want to ban such amazing and potentially useful technology?
I suspect like human gene editing, driving around in ford nucleons one could easily come to the conclusion its simply too dangerous to support.
The vision to essentially "merge"with AI systems is a powerful one - like deepfakes now where an AI can generate nearly anything you can imagine, being linked to such a system to give it instantaneous feedback as it is generating a result, to tweak to exactly what you would like to see could be astonished useful.
A VISION? Is this how you believe technology should be evaluated based on narrow one-sided best case scenarios?
Re: Why do you not want to help people? (Score:1, Interesting)
> Do no harm
So that rules out setting broken bones.
Re: Why do you not want to help people? (Score:5, Insightful)
Re:Why do you not want to help people? (Score:5, Insightful)
2) That’s why we have tests, strict protocols, and trials on humans for whom this is the last chance at a better life. That way we learn what effect this may have on the brain, before it becomes a common procedure. “We should ban this because we don’t know” is a neanderthal’s reaction to fire.
The problem isn't that we dont know (Score:1)
the problem is that someone else does and he might be as trustworthy as a chines chipmaker.
Re: (Score:1)
Re: (Score:2)
People tattoo their eyeballs, willingly.
You don't think this line of research would continue via less ethical/controlled channels in spite of a ban?
The analogy remains apt.
Re: (Score:2, Insightful)
1) If the harm is already there, doing nothing *does* harm. I find "do no harm" a very curious mantra for a profession that's proven to do harm 20% of the time.
https://www150.statcan.gc.ca/n... [statcan.gc.ca]
Oh if we redefine "harm" as an "adverse event", well I guess it's all good, it sounds like a stubbed toe or a paper cut.
How about removing a tumor from a person that reports no pain from the tumor? Is that harm?
How about chemo? Designed to kill cells. Is that harm?
How about if we knew a better alternative to this
https [clevelandclinic.org]
Re: (Score:2)
Re: (Score:2)
I'm pretty sensitive regarding how invasive the procedure is to install whatever cyborg kit I might decide to roll the dice on, the less invasive the better. This one is approaching my range. Stay out of it.
"IF YOU DON'T LIKE NUCLEAR WEAPONS, DON'T USE 'EM," that's my stance on the issue. - Rev.Dr. Ivan Stang
Re: (Score:2)
Re: (Score:2)
1) Do no harm. Physically damaging the brain is the opposite of this.
2) We know very little about the mechanisms of the brain and even less about consciousness.
And these people still call themselves 'progressives.' What a joke.
Neuralink is an experiment, not an established treatment for anything. It's precisely because we "know very little about" brain functionality that this tech is being made available to experimenters.
Re: (Score:3)
To be fair it does say it's "less likely" to give you brain damage. Those are pretty good odds to give Musk a direct wire into your head!
Re: (Score:2)
Do no harm is bull. Every single medical treatment does some harm or has some risk of harm. Generally you aim for "do no *net* harm", but then you have to argue about how you do the calculus.
Regardless, brain electrodes are implanted all the time for all sorts of things: control of seizures and Parkinson's tremors, diagnostics etc. If Musk develops less invasive electrodes, that's reducing harm, not increasing it.
Re: (Score:2)
There was once a time where transplanting organs was experimental, and people probably made the same idiot arguments.
Guess what - we got better at it, and now it happens all the time and there are many people still alive because of it.
Re: (Score:2)
Re: (Score:2)
Everyone knows your obsession with me. You've got some dark shit going on between your ears. You haven't worked in 25 years, and now you're claiming that you work for yourself. Posting forum spam in the hopes of maybe someone will see your Hosts File shit and throw you a bone is not a job. It's not even a hobby. It's an obsession and an e
Re: (Score:2)
You are asking the wrong questions.
The Internet was created for knowledge sharing. Then "people" got onto it and figured out how to spam, troll, scam, and advertise. The natural reaction was to pile up band-aids on all these issues and it's been a game of cat-and-mouse ever since.
There had better be a pretty god damn good firewall in place before I hook up. I will not accept the shit show that "the experts" have come up with for email spam filtering, for example. Layers of shitty standards on top of a b
Re: Why do you not want to help people? (Score:1)
Also the best way to present crime. Crime became even more impossible than using a cell phone to commit crime.
It lead to great advances in precrime enforcement too.
Anyone read The Terminal Man? (Score:3)
Sounds familiar. But after this your brain will be a DOS terminal and all your thoughts will have file names with 7 dot 3 Character names.
Re: (Score:2)
It worked for Robocop.
Re: (Score:2)
Re: (Score:2)
You mean 8 dot 3
Ban Banning. (Score:2)
This is all well and good for you sitting in your chair in your mom's basement with functioning arms, legs, and senses. Get off your privileged little molehill and think bigger than that.
You want to ban this, eliminating it as a possible escape for people locked in to useless bodies. You think you should get to decide that the person trapped in a chair, shitting into diapers, and wholly dependent on others cannot have any hope o
Re: (Score:2)
Re: (Score:2)
You can't possibly be that thick.
What about people that can't lift their fingers? Or open their mouths?
Oh, you were just considering yourself and not the people that this tech could make a measurable difference in quality of life. Selfish asshole.
Will this ever work? (Score:2)
Not your daddy's MindMouse!(TM)
Anybody remember that thing from like 2000? They had a great website but they disappeared when they didn't get enough funding.
Re: (Score:1)
Why didn't they use the Borg icon I haven't seen in a while on /.
Re: Will this ever work? (Score:3)
They managed to get rid of the Bill Gates icon though?
Re: (Score:2)
They had a great website but they disappeared when they didn't get enough funding.
That sums up many, many startups. The website generally is their best feature.
Re: (Score:3)
>we do no tknow what our pets are trying to tell us.
Well then perhaps you should try paying attention. Our pets have fairly rudimentary communication skills, much of it instinctual and body language. It's not that hard to learn to understand the basics, and the more effort you put into understanding them, the more effort they'll put into getting their point across. They tend to understand us pretty well, especially if we speak simply, and they only have a brain capable of developing a vocabulary of ma
Re: (Score:2)
>we do no tknow what our pets are trying to tell us.
Well then perhaps you should try paying attention. Our pets have fairly rudimentary communication skills, much of it instinctual and body language.
Spotted the person who's never owned a cat. Cats are like 83 year old grandmas. They might be literally dying, but they'll just soldier on and be like, "Oh, I wouldn't want to be a bother."
My cat was sleeping a little more than normal, but other than that doing all the normal cat things. And peeing blood. That's potentially life-threatening, but he didn't seem overly concerned about it. We're really fortunate to have caught him doing it so we could take care of the issue. Who knows how long it was going on,
Re: (Score:2)
I don't know about cats but dogs are simple. They just listen to their body and act accordingly. They don't try to interpret their sensation or power through.
- Feel tired, they sleep
- Not hungry, they don't eat
- Some movement cause pain, they avoid doing that movement
- Some weird thing is happening but it doesn't impair them, they don't bother
They only complain when something causes real acute pain.
In nature, there is no animal doctor that will diagnose trouble and provide treatment. So no need to signal an
Re: (Score:2)
Actually I've had a cat most of my life. My dad had a cat with some urinary track problem and it made a point of jumping up an peeing blood in the bathroom sink while he was on the toilet. They're probably not going to do something like that though unless you've already established decent communication and they think you're going to be able to fix it somehow - as GuB-42 said, the concept of a doctor is fairly alien to animals. If they have a problem they groom, eat helpful foods if instinct suggests them
Re: (Score:2)
Animals often don't want to signal that they're hurt, in case it makes them a target. That wouldn't change with mind-reading, a hurt cat wouldn't want to mind-read to someone that they're vulnerable.
Inspired by Iain Banks (Score:1)
Re: (Score:3)
If you want to be self-sustaining, you have to have access to resources. There are no physical resources for an orbital habitat - just energy. Now, of course you can move material between different orbits - say, from asteroids. But this involves huge dV changes for every kg you send, which at least with tech in the near to mid term, isn't exactly like hopping down to the local corner store. And unlike sending things to Earth - where they can aerobrake into orbit or even direct capture - you can't realist
Re: (Score:2)
Well, for starters we should probably build orbital habitats where the resource are, shouldn't we? That is to say, in orbit around the sun, rather than Earth, where the moon keeps everything swept relatively clean. That is to say - perhaps around Earth's L4 and L5 points, where we have at least some asteroids (though likely not enough for a good-sized habitat.), and primarily in or near the asteroid belt, and the clouds of them at Jupiters L4 and L5 (though sunlight is getting dim out that far).
As for ae
Re: (Score:2)
Outside of a gravity well, delta v for moving around orbits isn't that high. An orbital habitat that shares orbits with asteroids would have easy access to a lot of stuff. Low energy manoeuvres can get stuff around the asteroid belt with little hassle even if they take a long time, sort of like a freight train.
Re: (Score:2)
Your premise is false.
Re: (Score:2)
That's unfair - I should probably elaborate to make my point. Here's [nasa.gov] a list of dV requirements for Hohmann transfers to asteroids. Yes, you can use lower energy transfers that take longer, but they not only significantly increase transfer times, but there's a limit to how much lower you can realistically get. The lowest energy NEO out there takes 3,76km/s to reach from Earth. The average NEO - remember, these are Near-Earth Objects - takes about 7km/s to reach from Earth.
Nothing is - or can be - too easy
Excellent! (Score:5, Funny)
I will be frank, this is like sci-fi WTF?? (Score:5, Interesting)
First, hello Slashdot. It's been quite some time since I last logged in. Greets and all that jazz.
This flexible polymer with seeking type robot just oozed a comment. Elon is definitely out to see some future happen in his lifetime. I am a fan, just because he's perfectly willing to take the big risks and see what can be done. Works his ass off too.
Not for everyone, but while we've got him. Thanks! Let's hope for some goodness.
In my mind, there are two things:
One, any kind of two way interface. Yeah, there will be risks, but it's like having an extra channel. I suspect just getting it in there and figuring out how to setup the feedback loop needed for a person to be aware of it and explore is the hard part. Once they do? Look out. Pretty soon, talking to someone may not be any assurance you are talking just to them. They may be consulting all sorts of information resources, even just acting. Crazy times ahead. File all the obvious enhancements under this heading. Speculate away!
The other thing is fixing broken people. Seems like amazing possibilities here. That's exciting, if nothing else.
Re: (Score:2)
Welcome back to Slashdot, PotatoHead. I see you've noticed that the AC trolls are still here ;)
Re: (Score:2)
Welcome back to Slashdot, PotatoHead. I see you've noticed that the AC trolls are still here ;)
While the Agile/Scrum, Gang of Four, Ruby on Rails and other fads have predictably faded after being overhyped, trolls, like the poor, will always be with us.
In some cases, it's because they have no other options. APK hasn't had a job or even a "jerb" in 25 years - what else CAN he do? Small towns don't come with much in the way of employment opportunities, and his history of contempt for women (he claims that estrogen makes women's brains unsuitable for tech work), what else is a middle aged incel going
Re: (Score:2)
1. OMG, a product had a delay. Stop the presses!
2. Your link is actually to your own post. Which is in turn two links which dont' say what you claim they say. I wrote "they expect to be level 4 by the end of the year." Since when are "I" and "They" synonyms? Odd, though, that you think that something that's not supposed to happen until the end of the year is an example of a person being wrong regardless.
3. Once again, you continue to link to yourself rather than me. And had an AC step in and correct you
Re: (Score:2)
Pretty soon, talking to someone may not be any assurance you are talking just to them. They may be consulting all sorts of information resources, even just acting. Crazy times ahead.
While this may all seem like sci-fi WTF right now, I lived over 20 years before cell phones became common and I hardly remember a world where you couldn't find nearly any information with a 2 minute search on a device you carry everywhere. In another 30 years my kids might not remember a world where you couldn't find nearly any information with a 200 microsecond search performed by your thoughts and aided by an AI which makes today's Google seem quaint.
Re: (Score:2)
Plus my first computer only had 64k and my current one has 16GB. Thus we will be living on Mars soon.
Re: (Score:2)
>>we stand on the verge of SEVERAL massive revolutions in nuclear power
I want to believe.
Re: (Score:2)
Seems reasonable, especially when you consider that the MOSFET [wikipedia.org] was invented only sixty years ago. Not even one human lifespan.
Unfortunately technology seems to develop faster that society can properly cope with it. If technology can be weaponized, it will be.
Re: (Score:1)
Yeah thanks! Not only does he himself create the idea of electric cars, space(!) rockets(!!) that can be reused(!!!), electronic money, brain-computer interfaces but he also protects 30 year old children against demonic pedophiles!!!
So he have money to spend (mostly the money of others), likes to take risks and then push the idea he himself is a creative genius. Nice if something useful comes out of it but no reason to idolize a pretty disgusting human being.
natural born cyborgs (Score:1)
Re: (Score:1)
Re: (Score:2)
Re: (Score:2)
You must be new here. This has been a major gripe from the community even back when CmdrTaco was in charge.
Re: (Score:2)
It does some non-ascii characters, but not many. Doesn't do degree signs. I can't type placenames from where I live because it doesn't do the letter 'thorn'. It's just a big annoyance.
Re: (Score:2)
Re: (Score:2)
Livestream Info (Score:5, Informative)
I watched the whole livestream. The current version uses ~1000 electrodes, and they're expecting a version that uses 10k by the time they start human trials; it was aspirational that they'd have FDA approval and human trials starting late next year, so 2Q20 sounds like Elon-time.
Each electrode supposedly produces 200Mbps of data, which is filtered and compressed; although I suspect the speaker misspoke given he said it was 20kHz sampling at 10-bit resolution, which would be 200kilobits/sec. The electrodes can read or write data; that is, they can also trigger nearby neurons.
The wireless version will use an induction coil, with the implanted chip only being activated when a combination battery/computation device is placed on the skin which gives power inductively, and also computes transmitted data, although Bluetooth will also be utilized for phone connectivity. Something that was repeatedly emphasized is that this N1 device (their first commercial version) will be implantable in an outpatient procedure in an office, comparable to Lasik surgery, with the robot working automatically and programmed by someone trained to use it rather than a neurosurgeon using e.g. a Da Vinci surgical robot; and that the device is designed to be used continuously everywhere, like a prosthetic, not just for a session at a clinic.
Elon also let slip that they've already done trials with monkeys, allowing one to control a computer with its mind. They're apparently going to start trials with quadriplegics, to control mobile devices & computers. Later they'll try applications to help those with Parkinson's, dementia, depression, and other neurological disorders. One possibility that was raised is the ability to simulate the sensation of touch (haptics), which could be combined with reading motor intent to allow for virtual limbs. Controlling computer avatars was specifically mentioned, which I reckon could eventually lead to full-dive VR systems. Artificial visual sensations were also mentioned, although this seemed further out; this could be used to help blind people see, although it could also replace screens if computers could directly tap into our visual systems. Speech can apparently be synthesized from motor intent, allowing for the mute to speak (or for people to 'speak' without making a sound.)
The sci-fi applications were mostly glossed over, as this presentation was a broad overview of everything. The implication is that they're creating a platform which will be capable of more and more things as time goes on and the tech and neuroscience improve. It's pretty amazing what they've achieved in just over 2 years, and I'm pretty excited over what they might be enabling two more years from now.
Re: (Score:2)
I have no idea why this was the thought process going through my mind as I read your comment (good comment by the way):
Everyone will be implanted with a version of this at birth. The version I am speaking of has the ability to "read your mind" and "paralyze" you when someone (police) with a particular device is nearby and presses a button.
Got a bar fight with a bunch of bikers going on? No problem. Send one officer in with the "device" and he pushes the button and everyone falls to the floor no longer in co
WHOA, I know KUNG FU (Score:2)
Just teak them so they experience ecstasy while picking litter up from the side of the road. Or charging into battle against the enemies of the state, or beating up hippies or whatever.
Re: (Score:2)
Steady on Charlie Brooker.
Sorry, but: Do. Not. Want. (Score:2, Interesting)
Re: (Score:2)
Re: (Score:2)
Re: (Score:2)
Just because you're stuck in your wageslave job where all you do is 'do' and not 'achieve' doesn't give you license to hate on people who actually 'achieve'.
Re: (Score:2)
Re: (Score:2)
Technically you can implant an electrode array and do things like see pictures the person is imagining. Presumably you could make that work better and use it as an interrogation device.
But drugs would probably be easier.
Cyberpunk (Score:2)
Cool! I'm finally going to get to LARP my Netrunner!
So long as I am never forced to it (Score:1)
All I keep thinking about is (Score:2)
Re: (Score:2)
Re: (Score:2)
Wirelessly? What could go wrong?!? (Score:2)
>but ultimately the goal is to create a system than can work wirelessly.
Yeah, because that's what I've always wanted, a direct neural interface that can be wirelessly hacked by any asshole that wants to fuck with my brain.
Re: (Score:2)
Ah, you prefer a socket?
Re: (Score:2)
Surface contact probably. I prefer my brain to be easily airgapped.
Usb-c? (Score:1)
I've got a picture of an early prototype (Score:2)
Research potential (Score:2)
There is also potential for abuse. For example, if we can understand exact data path and processing algorithm for visual cortex, it might be possible to develop exploits (and weaponize them). Imagine inducing seizures by simply showing a short clip, something like rapid flashing and epileptics, but
Elysium....here we come! (Score:2)
The photo of the implant looks almost exactly like the implant that was used in Elysium. Just need the orbiting paradise built and then the rest of the world can go to pot...
i have seen this ... (Score:1)
Re: (Score:2)
Allows diplomatic victory, so how could it be wrong? ;-)
Braindead (Score:1)
I hope musk volunteers himself and the operation runs as smoothly as a Tesla delivery deadline.
Wireless interface? Great idea! (Score:2)
As far as I am aware, there is no secure wireless tech on the planet. Sure, you can put a VPN over wireless or ssh or the like, but I doubt they have even thought so far. Well, I guess, "Brain Hacking" will become a thing.
Ian M. Banks (Score:1)
It will fail (Score:2)
Terminal man (Score:1)
Re: (Score:1)
Slashdot is a place for smart people. what are you doing here?
Re: (Score:2)
But they won't ask if you're left-handed. Many studies exclude lefties, because the brains are wired a bit different, but the resulting treatments and drugs are somehow still expected to work on them.