Neuroscientists Weigh In On Elon Musk's Mysterious 'Neural Lace' Company (ieee.org)

the_newsbeagle writes: Elon Musk has set out to change the world with SpaceX's reusable rockets and Tesla's electric cars, and now he plans to change your brain. His new company, Neuralink, will reportedly build delicate brain implants called "neural lace" to help people with neuropsychiatric disorders and to give healthy people strange new mental abilities. But the news announcements about the company contained scant details about what kind of hardware Neuralink might actually build, and what engineering challenges the company will have to overcome in pursuit of miniaturized and safe brain implants. Here, five neuroscience experts describe those challenges, and give hints on what to expect from Musk's neural dust. One of the neuroscientists is Mary Lou Jepsen, founder of the Openwater startup, which is looking for ways to develop a noninvasive BCI for imaging and telepathy. Jepsen was also "an engineering executive at Facebook working on its Oculus virtual reality gear; before that she spent three years at Google X, running advanced projects on display technology," reports IEEE Spectrum. She says that Neuralink will likely face many medical hurdles, even if their process doesn't require splitting open patients' skulls. "The approach as I understand it (not much is published) involves implanting silicon particles (so called "neural lace") into the bloodstream. One concern is that implanting anything in the body can cause unintended consequences," says Jepsen. "For example, even red blood cells can clog capillaries in the brain when the red blood cells are made more stiff by diseases like malaria. This clogging can reduce or even cut off the flow of oxygen to the parts of the brain. Indeed, clogging of cerebral capillaries has been shown to be a major cause of Alzheimer's progression. Back to neural lace: One concern I would have is whether the silicon particles could lead to any clogging."

Comments

Cereberal Network Variability

[rtb61]

The problem with any direct connection to the brain is variability with neurons networks, entirely unique. Whilst broad cerebral manipulations are certainly doable, fine work is not, think waking and sleeping, happiness and sadness, really broad and dangerous brush strokes because you have no real idea the consequences on fine neuron networking detail. Then there is the extreme danger of hacking, whether corporate or government or just the idiot kid down the street with the wrong software. Even something as relatively safe as a sleep inducer can become very dangerous when triggered at the wrong times ie driving a car.

Easier to tap nerve bundles than the actual brain ie eyes, ears, spinal cord and they are also quite dangerous ie stimulation of data input versus extremes of pain stimulation.

For actual brain input, it would really have to be grown and the user and the system aligned, think useful brain tumour, with the tumour creating the communications links but the tumour also the idea representing the real risk involved.

The biggest problems are; Would the government hack it if they could, yes. Would Corporation hack it if they could, yes. Would individuals hack it if they could, yes. The temptation to hack if for total control, is far to great for it to be done, apart from very limited medical actions, sleeping being the obvious target for simple manipulation and a real warning about the risk ie putting someone to sleep when they are driving a car.

Re:Cereberal Network Variability

[Gravis Zero]

The problem with any direct connection to the brain is variability with neurons networks, entirely unique.

Actually, this is far less of a problem than you think it is. We've already created BCIs that allows basic actions to be made like manipulating a robotic arm without hooking to the circuitry for arms. How did we manage such a thing, well, the person with the BCI implant learned how to do it through practice. There are regions that typical brains have in common and if we tap into those parts, we can learn to truly interface minds with machines.

It's not easy feat and there is a lot of work to be done and breakthroughs to be made but it is possible.

Then there is the extreme danger of hacking

I believe the endgame here is implanted technology. If you connect that shit to the internet, you're implant has failed because you're still dumb a shit. ;)

Re:Cereberal Network Variability

[interkin3tic]

Not "entirely unique" in all areas. Variability is a concern, however it's not like each brain's layout is completely scrambled from one to another. Deep brain stimulation has been around for a while and is showing promise based on targeting the right areas.

There's also nothing that says that this thing would need to be hardwired up in a certain way and if two were inverted, fuck it, it's gone. I assume there would be a training period in which the software (like electronic software) would learn what was hooked up to what.

For that matter, there's plenty of training and learning on the wetware side that could work.

Without real technical details, it's hard to say "This can't possibly work because of X thing that is already known." On top of that, Elon Musk is not known for proposing things that are going to run headfirst into problems one could identify just from a wiki page. I'd assume it's technically feasible.

Then there is the extreme danger of hacking, whether corporate or government or just the idiot kid down the street with the wrong software.

Really? You're going to poo-poo an awesome technology (again, assuming Musk hasn't just totally gone bonkers) based on "brain hacking?" After just saying "we really can't know where stuff will connect with the brain?" Some idiot kid down the street is going to know the wiring of your brain and the backdoors into it better than any neuroscientist does? Because neuroscientists don't know that level of detail. This won't be life-and-death stuff like breathing: your body does that on it's own.

If it had the potential to do something like "press here to cause a seizure," then Musk is going to have a hard time getting that to market or past his lawyers. He can't even get his car sold in some states, he's not going to railroad this thing despite obvious safety concerns. Again, he's not an idiot. If there were any chance of someone putting someone else to sleep with it, one would assume he would put that into the "Neat but stupid" idea pile.

Speaking of, I'm sure there are some fascinating but fatally flawed ideas that have come from Musk's team. They should publish a book of them.

Re:Cereberal Network Variability

[pz]

Uh, no. If what you asserted were the case, then things like cortical visual prostheses would not be a possibility. I suggest you look them up. While they are still under development, they most certainly do intend on creating fine-grained control over neural activity. Same for cortical somatosensory prostheses. On the flip side, we can definitely read-out fine-grained information about neural activity, such as is used for motor system prostheses, ranging from limb prostheses to vocal chord prostheses.

On both the read-out (decoding) side, and the driving (encoding) side, we have the ability to receive and transmit information on an individual basis. Yes, there is a lot of variability, but that is part and parcel of the challenge. Just as individual variation in foot size and shape does not preclude the creation of shoes because there is an underlying structure, so individual variation in brain morphology and wiring is unlikely to preclude creation of brain/machine interfaces because again there is an underlying structure.

Any time you hear someone say it is impossible to do something, it's likely they are just not thinking in advanced enough terms to overcome whatever barrier they perceive. I myself am guilty of such mistaken proclamations.

Re:Cereberal Network Variability

[Kiuas]

Easier to tap nerve bundles than the actual brain

You're right, but there are many conditions for which the brain is the problem. I've have cerebral palsy (more specifically spastic diplegia) as a result of a brain injury caused by oxygen loss during a premature birth. Through various surgeries addressing orthopedic and muscular issues I've reached a point wherein I can now even stand still without any external support. Walking is possible with canes or as little support as one finger to hold onto with both of my hands. The core of the issue is that the part of the motor cortex that processes incoming information from the balance organs and muscles about the posture of the body is partially dead, so the brain is unable to regulate balance accordingly. Outside stem cells or other such theoretical ways of regenerating nerve tissue, implants are the only thing that may one day solve this.

The biggest problems are; Would the government hack it if they could, yes. Would Corporation hack it if they could, yes.

A hypothetical implant for something like my case of CP would be just a chip that sits on top of the brain and does the calculations that the dead part of the cortex would normally do. It doesn't have to be connected to the external world, and should I ever get to see a day where such an implant is a reality, I certainly would not accept one that was. I'm 26 now and realistically speaking I don't expect to see this tech becoming widespread during my lifetime, nor will I volunteer as a test subject because I've reached a point wherein I can live by myself, work and drive a vehicle, so my quality of life is pretty much as close to normal as is currently feasible for people with CP, thanks in large part to the medical expertise of the university of Helsinki hospital, so my disability does not bother me nearly enough for me to desire to try experimental high risk treatments. But having seen the already very promising results that for example deep brain stimulation has had for people with Parkinson's etc I do think this kind of solution is far more feasible after some more decades of advances than most people currently think.

Security and Robustness

[sinij]

I don't think anyone, well maybe aviation software developers, are capable of programming such devices. It must be efficient, secure, fault tolerant... the list goes on. By contrast, today commercial software is expected to fail and security updates are routine that nobody minds. This would be disastrous when applied to such tech. Just imagine someone figuring out how to crash your visual cortex, inducing seizures and such.

Neural Lace

[PopeRatzo]

I'm more of a neural leather guy.

"Neural lace" is a term invented by Iain M. Banks

[realmolo]

He was the author of "The Culture" series of sci-fi novels. Which is *easily* the greatest science-fiction series ever written. It's everything you ever wanted from sci-fi. If you haven't read them, do so immediately.

Start with "The Player of Games", or even better, "Use of Weapons" if you are an "andvanced" reader.

Re:Elon Musk is . . .

[Hylandr]

I assume he read about something like this in a scifi novel when he was a kid and now that he's rich he's throwing money at it

As a fan for the progression of technology I hail him as a hero for sacrificing his funds to make the world a better place. He deserves all the fame in the world. It's not like he's selfishly using his riches to be Batman, he's contributing his gobs of money for the betterment of Humanity.

What have you done to help your fellow man today, other than to post as an anonymous coward to belittle the selfless ambitions of a man using his billions for the greater good?

Re: Fuck that.

[Anonymous Coward]

I suffer from extreme chronic pain among other issues. Just like pain medication, smoking weed only treats the symptoms but doesn't cure the true on going issues. I would be more than willing to let Musk's researchers drop neural lace or slap a chip in my melon if I thought it was a cure or a more permanent solution for my constant pain. I have been to many different doctors try to find a way to lessen my pain or find a cure. I've had multiple injections using cocktails of meds or just regular steroids to get a cure. I'm heading to the Cleveland Clinic next to go through a 4 week program to help cure or at least control my pain. This is kind of the last chance to try something to help me. So if Musk's doctors can implant a chip or whatever to help me then I'd let them when they have a useful product.

Scary Stuff

[JimSadler]

Such a technology could be applied without consent in many cases. For example convicts might be implanted in such a way that they could no longer commit crimes. Some of the mentally ill would fight like crazy any attempt to actually cure their mental illness. For some of these folks the illness and the person are the same thing. Remove the illness and their universe would simply cease to exists. For example a person with paranoia may feel that he is such a mess in life as that mysterious enemy has forced him into doing things. Blast that fantasy away and he will be forced to admit he is the one that failed, did wrong etc.. That could leave a man with no paranoia but an overwhelming urge to commit suicide. Changing people might have terrible consequences. A priest once told me that when dealing with some nasty, angry and bitter old folks that one had to be careful not to take away that bitterness as sometimes it was all that person had.

Re:Scary Stuff

[Baron_Yam]

I think he had the wrong perspective. The elderly tend to get 'cranky' for three reasons:

One, there's a portion of the brain governing impulse control that actually atrophies with age.

Two, their world has often shrunk drastically as their mobility is impaired, friends have died off, and younger family is busy and can't visit constantly. They're left with much less to think about, so little things become disproportionately important to them.

Three, they're often losing control and independence, and sometimes their faculties. You try remaining friendly when you're treated like a baby because your body and maybe a bit of your own mind are betraying you.

Nothing much can be done about the first (unless it's a symptom of the other two instead of another cause)... but there's a lot that can be done about items 2 & 3 that we don't do because it's easier to simply warehouse the elderly and give them minimal care while waiting for them to die so we can say we miss them.

Re:Elon Musk is . . .

[Hylandr]

So what if he is? Why should I have a problem with someone trying to turn a profit while improving our quality of life?

Are you mental?

If you had a million dollars, would you spend it all on tents for the homeless and be broke or would you invest it so every year you could buy 100,000 worth of tents and training for the homeless every year?

Even if he didn't buy tents. It's his money, he earned it through sweat and or intelligence. Don't be so jealous.

LIkely to be safer than oral/inhaled/absorb drugs.

[BlueCoder]

Imagine if you could artificial control your own mood without taking drugs. You could fall asleep when you want to and wake up quickly at a certain time. You could be happy, social, and less anxious when you want. You could turn on or off sexual desire. You could have the mental focus of a robot and feel bored. How about turning off pain in whole or selectively. Turn off anxiety and fear. These and a million other uses are guaranteed. How much would any of those be worth if you could control them? It's a guaranteed future for he that invests.

And we haven't even gotten to the point of discussing illegal wire-heads and simply do it to get a better drug free high.