A skeptic’s take on Neuralink and other consumer neurotech
statnews.com
statnews.com
I think this is misreading Google Glass. People didn't like or want to use Google Glass because there was no "killer app" for them. There was no useful app for Glass at all so far as I am aware. Had Google Glass enabled something new and useful people would have worn it regardless of how different it looked, and because people found use in wearing it we would get used to it.
Headphones are an example of this. Headphones would probably look goofy if you weren't used to seeing them. People wear headphones anyway because headphones deliver something useful to the wearer - private audio.
If I had to wear a silly hat, but in exchange could carry around a meaningful brain-computer interface, I absolutely would, and, assuming "meaningful" in a broad sense, I'm certain huge numbers of people would as well.
Right now people are struggling to make the tech work at all. Worrying about the form factor with which it will be delivered is premature - it can't really do anything yet. Once it's useful it will need a good form factor, but the use will drive adoption, not the other way around.
Social proof will do that. Doesn't matter how weird something looks, if the appearance becomes associated with wealth in the public consciousness, people will start thinking it looks cool and coming up with reasons why.
I think the other reason why people didn't want to adopt Google Glass, is the stories about other people objecting to people wearing them–primarily because of privacy concerns, that the devices were capable of recording video, and that while a phone camera can record video too, it is somewhat more obvious if someone is doing it with a phone camera than with Google Glass. If a product is going to be a social hassle, if other people and businesses are going to object to your use of it, a lot of people will decide the negative social consequences aren't worth it.
Something like Google Glass specifically targeted to people with disabilities would be more acceptable, because someone using it can always say that it is for their disability, at which point most people would be more forgiving, plus if any business tries to ban it they run the legal and PR risk of being accused of disability discrimination.
Although honestly i thought it was a bullshit argument. You can buy tiny disguised cameras. If someone wanted to creepily record you, they could easily do so and you would never know. They certainly wouldn't use something as ostentatious as google glass.
People don't care about things they can't see -- because they can't see them.
But if something is seen, it probably shouldn't have cause to make people worry.
Rio was there first with a better product. Apple had better marketing though and the rest was history.
Rio worked with anything and had no DRM. Any Rio was superior to all early iPods.
Apple Marketing was pretty overwhelming, but I'm afraid the Rio line was already doomed by the time the iPod began to dominate.
In fact at the peak, the iPod's market share in music players was about the same as what Android has in smartphones today.
Right or wrong, constantly pointing a camera at others gives off an air of “I could recording you right now”, and that makes people uncomfortable.
Yeah, there’s a difference between someone having the ability to take pics in their pocket and having a camera pointed at you all the time.
Is it more of a reaction than a logical thing? Maybe.
People have shown time and again they will adopt something they love, no matter how grotesquely bulky, like brick cellphones
Eventually using it to backup and data mine memories will likely happen, as in Black Mirror.
Using it to filter, annotate, and block parts of reality is also likely. Again, Black Mirror touches on this.
Some have posited the idea of gamifying reality, for all sorts of purposes.
I've often thought realtime logic-checking would be useful. Fact-checking would be more subject to biases in the software and data, but could also be useful.
Eventually the idea is probably that neural prosthetics will be developed that act as entire new brain subsystems.
Additionally, an implant is pretty damn invasive. So some cute demos or niche consumer applications are not going to cut it.
IMO the two possible paths here are either this will be an R&D money sink for our entire lifetimes (maybe funded by Elon for fun, maybe some government grants, etc.) or they will primarily operate as a medical device company.
The implants they are developing are definitely a tech improvement, so they could realistically improve existing DBS treatments. The company could also work on their own novel DBS paradigms for treatment of specific diseases. A lot of DBS (especially for psychiatric diseases like OCD) is currently rudimentary AND we have a poor understanding of why it even sort of works. But it does clearly work for at least some subset of patients, so there are real avenues for continued development.
This is a step further, but say we could interface directly with auditory areas to have access to a massive internal library of different sounds. You could play songs on demand, verbatim recall anything that was said to you in the past, etc.
Sounds pretty dope at first, but this kind of thing would likely be developed before future researchers fully understand the entire human brain. So that means they would not really understand how the brain would interface with it, especially across long time scales and a heterogeneous population.
Forget any sort of malicious action, and still God only knows the potential side effects. It could massively fuck with ability to learn over time, cause terrible rumination issues, maybe it could trigger psychosis in high risk individuals, maybe it would fuck up your dreams.
Like seriously, for anybody that is trying to cut back phone usage, just imagine if you could check your feed purely by thinking about it. How freaking hard would it be to control yourself then? And that is an obvious immediate consequence type thing, not a deep consideration of potential problems.
Anyway, this mostly reminded me of that Black Mirror episode about the parent watching the kid via a brain implant, which I thought was too heavy handed. There are a ton of ways brain implants could be scary without using the super obvious surveillance angle. Honestly just the "blurring out of bad things" feature would have made for a cooler episode IMO.
We'd definitely have to regulate the shit out of tech then. Going to example of phones - the root problem there isn't that phones are addictive, it's that every aspect of them is controlled by commercial interests that benefit from making them addictive and user-hostile. If we were to create BCI the way smartphones where - where the Googles and Apples of BCI will release a consumption-oriented platform you don't own ("because security!"), that serves primarily to access the Facebooks and Twitters of BCI, in the way the platforms tell you to (again, "because security!") - then it's game over for humanity.
If, however, we allow control over devices and their operations to be held by the user (to the extent it's unprecedented in this industry), and regulate social media like tobacco, alcohol and gambling, then we can reduce this problem to something comparable with alcoholism. Not ideal, but much closer to optimally balancing individual freedom with preventing fully predictable individual failure modes.
Are you envisioning playback on a computer screen?
I'm not sure Neuralink would disagree. Their first applications are medical, not consumer mind-reading. They're open about their bigger vision being farther away.
> Neuroscience is far from understanding how the mind works — much less having the ability to decode it.
We don't necessarily have to understand everything about the brain to communicate with it. And we'll learn a lot in the process.
> a neurotech device would have to add significant value for a consumer to get one implanted in her skull.
Again, I don't think Neuralink would disagree. I guess the disagreement is in whether an implant could provide that significant value. And nobody really knows yet, of course. A lot depends on how well the brain can adapt to a new form of I/O.
> Consumer brain-recording devices have been on the market for roughly 15 years
Yeah but the technology is totally different and quite lame honestly. Existing devices have fundamental limitations that are not shared by an actual brain implant. Not a strong argument IMO.
> Helmets and other headgear face an uphill battle to adoption
The external hardware for an implant can get pretty small. Much less obtrusive than the given example of Google Glass.
> It may be more taxing to control a device with a BCI than without it
Pure speculation at this point.
Honestly it is wise to be skeptical. But mostly I am skeptical of the timeline, not the eventual usefulness of the product. Neuralink is likely too early, but one day useful consumer brain implants will be possible.
For example, maybe computers get so smart our need to interact with them diminishes and we interact with them rather passively.
Maybe the opposite happens and we need even tighter coupling leading to a genetic engineering solution or a nanobot pill solution that crosses the blood brain barrier.
The problem is everything is pure speculation at this point. I do hope they accomplish some of their goals but any business potential seems a total guess.
You first.
As it happens, neuroscientists have become quite proficient in reading out the activity of large numbers of neurons in recent times. We do whole-brain activity imaging (of fruit flies) routinely in the lab that I work in (albeit not at single-cell resolution). The problem is that these techniques are incredibly invasive. They involve genetically encoded fluorescent protein sensors, surgery and lasers so bright that imaging time is quite limited (because the organism gets cooked).
That approach is never going to work in humans. The genetic engineering alone is enough of a problem, let alone the optical bench (not exactly portable) and the hundreds of thousands of dollars of equipment you need.
And the whole reason that approach became so popular in the first place is because of how limiting electrophysiology is. You can get at best a couple of thousands of neurons with ephys. Certainly not the routine whole-brain imaging that is happening right now in neuro labs around the world.
So there's pretty good reason to be skeptical of neuralink. I don't think you'll ever meet a working neuroscientist who thinks this has any chance of working. What GP is expressing is entirely baseless. Of course neuralink would disagree, they have a profit motive and snake-oil to sell. Ask someone who actually works in the field and you're going to hear a different story.
If what neuralink was doing was actually worth pursuing, HHMI Janelia would be investing in it. They were doing ephys on dragonflies in flight way back on 2014 for gods sake [0]. They've got the chops, and that they're not doing it speaks volumes.
[0] https://www.janelia.org/news/anthony-leonardo-tracking-drago...
If their first goal is to restore some functionality to severely paralyzed people with less big/bulky and doesn't leave a hole (infection point) in your skull, what won't work about that? Is it the body rejecting the implant? Is it the wires breaking down or migrating out of where they're placed? What is it?
Now you go tell the paraplegics that this work is pointless...
People should always be skeptical of timelines when it comes to Musk's ventures. He is famous for short deadlines, high expectations, and a competitive environment.
There's a TED talk somewhere out there with SpaceX's CEO where this comes up a bunch, and she plays on it very well.
Very true! However, we're still learning a lot about the brain. For instance, astrocytes/glia are still quite understudied despite being ~1/2 of the cells in your brain (maybe). Mostly, this is because they are not very electrically reactive. But they are known to be a pathway for NMDAr recycling as well as surrounding the synapse in certain parts of the brain. This issue of astrocytes is one of thousands that we're still trying to get a grip on. Essentially, we have okay-ish knowledge of large-scale phenomena (~1cm3) and small-scale (synapse, nucleus) behavior, but the middle ground (10's of neurons) is quite hard. Unfortunately, that seems to be were all the magic happens.
> A lot depends on how well the brain can adapt to a new form of I/O.
Typically, it does this via scarring. Making sure the leads are engineered to deal with scarring is the first hurdle. Then you'd need to stimulate only a select few neurons, a feat that is quite difficult (dynamic positioning, heartbeat, discharges, feedback, etc). Not necessarily insurmountable problems, but we're not really sure how to do that compactly with wires. One way is just use light and optogenetics and CLARITY, but then you have to genetically modify a brain and wash it out in situ, which is downright crazy hard.
Sometimes “obviously” inferior solutions are actually local maximums, and finding the global maximum is extremely difficult.
Why in the world would these supposed Gatekeepers gatekeep away a technology that would provide a massive competitive advantage? Wouldn't they instead do exactly what they've been doing for decades and just buy up this hypothetical new technology? That sure seems to be what Apple is doing; there's an entire wiki page dedicated to Apple acquisitions and what Apple products were derived from it. Why would they suddenly flip into suppressing something better rather than just repeating their standard practice?
Perhaps the more plausible answer is that nobody has come up with an interface that beats the smart phone for the general population?
What competition? The disabled are lucky to get a single workable solution.
For regular people, it seems that voice is a somewhat important interface these days. "Ok google remind me to buy toilet paper" and so on. And honestly… isn't that already better than implanted magic in some ways? People (who can) naturally communicate with other people using voice, not telepathy.
At SoundMind (https://SoundMind.co), we're building a headband which monitors your sleep state and uses sound (auditory stimulation) to improve your sleep.
Commercial tCDS is a real thing for treatment of depression. There are multiple consumer tCDS devices on the market as well, I won't link to them, because I haven't seen any that I don't believe are snake oil. But the commercial scale devices will get shrunk down to consumer usable size eventually.
This is just the beginning. Consumer neurotech shouldn't mean implanted, just as sight correction doesn't mean lasik (just wear glasses).
I think this skeptic is perhaps too focused on the technology, rather than the use case.
I'm not sure if we'll ever willingly get an neural implant, but let's not throw all of neurotech to mean that small segment of the market.
There is so much hype (“it does everything: makes you stronger, faster, happier, smarter, and better at video games and jazz”) and anti-hype (“it does nothing and it’s all placebo or peripheral stimulation”)[0].
I know about as much about it as anyone and I think that it a) has tons of promise and b) should not be anywhere near anyone but an research participant for the foreseeable future.
However, the simplicity of the device makes people think its effects will be equally simple. This isn't true.
We don't fully understand what the electric fields do to the brain, and the simple story of "turning the brain up/down" almost certainly doesn't hold up. In most cases, we also don't understand the problem to be solved (e.g., what in the brain is causing your depression--or your bad golf game). Taken together, it's a nightmare: poking at an unknown cause with an unknown effect.
My colleagues and I have been looking at this in monkeys, which we can measure everything from behavioral effects down to the activity of single neurons. It's very exciting and has, I think, a lot of potential, but it blows my mind that people are already selling products. The crazy thing is that the field has been through this hype cycle before in the 1960s-1970s....
We've been speaking with neurologists here in Australia who are giving us guidance, but if you have any thoughts, I'd be very keen to hear them. Either here, or via email. My address is in my profile.
Thanks
We did some work on DARPA's RAM REPLAY program, and there was a consortium trying that for "tagging" memories: they'd pair something to be memorized with a sound during (awake) study sessions, and then play the sound to a sleeping participant to reactivate that memory during sleep.
Stephen Simons was the lead from Teledyne; there were also folks from UCR and Endonovo Therapeutics, I think. It sorta seemed to work (here's the paper: https://www.frontiersin.org/articles/10.3389/fnhum.2018.0002...) , though I remember they ended up needing to build a fancy system to optimize the cue volume so it was loud but not so loud that it woke the participants up.
They worked, more or less. Watching someone play a video game, the ears would show "attentive" while they played a level, and dropped back to rest when they finished. At an anime convention, I saw a group of girls all wearing them. Someone called out the name of one, and her ears popped up. Only hers, not the others.
They were a bit too bulky (4 AAA batteries), the ears were too big, and the forehead sensor wouldn't stay in place. Someone should try that again in a more compact form.
More fun than the quack medical applications.
The mind is not a joystick.
Throughput is not the measurement of a good BCI.
BCI as a term should be abolished.
Someone is going to make something so seemingly simple, but unbelievably meaningful, to rule this space.
There are many ways to get high fidelity 'joystick' level control out of a brain, especially if given one-on-one training.
The problems for Neuralink will be the same with any implant - immune responses and infections - and in a sensitive area no less.
I was at a conference where people were debating the merits of non-invasive vs. invasive neural interfaces. A surgeon argued that people would obviously never accept non-invasive ones...ON COSMETIC GROUNDS: EEG caps (and similar) look dorky.
Many people do go to great pains to hide disabilities, but I was (and still am) really skeptical that this totally trumps fears around highly-invasive neurosurgery!
Hypothetically let's say I'm 90, I'd love to be able to jack into the matrix to play with my grandkids.
Even though it's an incredibly effective treatment, surgeons are pretty reluctant to do DBS implants on older patients with Parkinson's Disease because of the risks of the surgery itself. That's with a pretty clear medical case too.
You haters are literally lying lmao.
For example, as of this writing, you could have an experimental procedure that results in light sensitive neurons, that when excited by a laser, stop your breathing and remove the sensation of asphyxiation to let you die a painless, playful death.
(Optogenetics, table stakes)
So I think it’s massively important to discuss what should and shouldn’t be done with Neurotech, otherwise with the rate the “throughput” argument is going... aka all of these joystick developments... your life value will be pegged to how many blockchains your mind can mine... enjoy the matrix!
(and it's "peddling" as in "selling", not "pedaling" like riding a bike)
The video tries to argue that Elon Musk is trying to steal credit for founding the company [2]. He doesn't actually quote Elon Musk when he accuses him of stealing credit. Why does he circle text on Wikipedia rather than quote Elon Musk? If you go to the Wikipedia article and check the citations for that claim you will discover Elon Musk does not state the thing which the liar in the video states that he claims. Why does he lie about what Elon Musk actually said? Apply common sense and skepticism to that question and you will be enlightened. I'll give you a hint: he is working backward from a narrative, not forward from the facts.
Alternatively, want to do a long bet for $10k that in 10 years Musk will still be a millionaire despite the contents of the video you linked? Void of bet in the event that some other cause leads to misfortune.
[1]: https://youtu.be/BvmA_gQ-95c?t=155 [2]: https://youtu.be/BvmA_gQ-95c?t=185
[2] https://www.spiegel.de/wirtschaft/elon-musk-tesla-gruenheide...
> 2,291 days since Elon Musk tweeted tat Tesla was developing a charger that automatically connects to cars. (12/31/2014) > "We are actually working on a charger that automatically moves out from the wall & connects like a solid metal snake. For realz." > https://elonmusk.today/#snake-charger
Here is the metal snake charger it claims doesn't exist:
I hope you don't own any Tesla shares because that will be a rude awakening in a not to distant future.
The information source that you are using isn't high quality. This is obvious, but I'll explain why in more detail.
First off, you are right. Obviously the site intends the meaning that you gave.
Making up a different claim other than the one he said is stupid to do. With just the claim that 2 = 1 you can prove that you are the pope in two steps. The set of you and the pope is two people. 2 = 1, therefore the set of you and the pope is actually one people. Therefore you are the pope. This childish game of choosing what you believe rather than believing what is true is pointless.
Everything is provable, so the entire worldview is justifiable and anyone who can make you angry at something has control of your opinion and you can prove whatever you set out to prove.
For that reason, it is important to understand reality as it is rather than reality as you think it is. You don't get to make up whatever meaning you like for words if you want to be rational. If you are going to assume a meaning, it is wise to choose the most charitable possible reading, because otherwise you aren't grounded in evidence. This isn't just the principal of loving your neighbor as yourself - though it is that too. It goes deeper in a mathematical sense.
If there are two readings, one of which is supported by evidence, but the other isn't, but you choose to read the meaning which isn't supported by the evidence than you are making a calculated decision to use the reading which has less evidence supporting it. Your beliefs aren't justified by evidence. So they aren't rooted in a wellspring of truth.
The site does choose to be ungrounded from the truth. It doesn't choose the most charitable reading of his statement. It restates instead something he didn't say. Something which has less evidence backing it than another reading of what he said which was absolutely true with strong evidence backing the interpretation.
We know that there is a more charitable reading, because he claimed they were working on something and that thing which they claimed to have worked on is shown to exist. Working on something doesn't imply that it becomes massively adopted. If it did, that would imply that most people in all of reality have never worked a day in their life. Someone who spent all of their life in hard labor, according to your definition for what work is, would not have worked even once - they, after all, didn't get mass adoption of whatever it was they worked on.
So the twisting of the words isn't right. It requires distortion. It requires reaching. It isn't as supported by the evidence. It is wrong.
Yet the site chose to use that wrong interpretation. It gave you a source of information about Elon Musk, but it gave you one which wasn't rooted in the evidence. It wasn't written for your reason, but for your emotion. It is intended to provoke emotion, which is why it reads like it does. It isn't intended to be read rationally, because wisdom despises lies.
In forming your opinions atop this information source, you're not thinking based on the evidence, but based on emotion. It doesn't matter how clever you are - and I'm sure you're clever. The information you're using is faulty. It may lead you to correct conclusions despite itself, but it doesn't have a strong logical justification for doing so.
Therefore, the information source is quite obviously one you should disavow and avoid if your goal is to make good and profitable decisions.
Keep in mind, this doesn't mean you are wrong. You could be right - it just won't be because of the information source you used, which is quit soundly refuted.
Afterward, think on the emotional connotation of the restatement. Ask yourself questions like why they chose to restate things so as to produce a falsehood. Then ask yourself about Elon's text in that arrangement. There are lessons about good communication within it and it directly corresponds with mathematics.
Why does the information source choose to turn something which has deep mathematical truth justifying it were the claims to be true with something which is obviously non-casual?
It isn't just wrong. It is obviously wrong. Everyone knows cause and effect.
They do it anyway. They get it wrong, despite it being obviously wrong.
Why? The answer of course is because doing so works. They hooked you, didn't they?
Or at least, it seems to work. Truly, it doesn't, and the fact that it doesn't is probably best explained by someone other than me: read Plato. In particular read the dialogue in which the professor of rhetoric and Colt talk with Socartese about what it is that rhetoric is. He wrote about sites like the one you are using thousands of years ago.
Asking me to take a bet that doesn't match with my stated predictions isn't as reasonable. Why should I bet on something that I never claimed would happen? That said - I'm already invested in long positions in various AI & robotics companies. You can bet against my actual beliefs by taking positions which would benefit you if those companies aren't the future.
Teeth implants, knee implants they generate additional risk of low level inflammation in the body over time. Even generally unproblematic silicon breast implants that just hang there have scar tissue and can cause immune system reactions.
Brain surgery is many times riskier. If Neuralink could create way to circumvent all these problems it would be huge breakthrough even without the BCI.
Those things are still in the basic research stage, and it's been really hard to make them work properly. Neuralink is not even leading in research here, but i don't think they expect to make consumer grade anytime soon either. OTOH, all sorts of multiunnit recordings and optogenetics stimulation is possible today - if you're a mouse.
I wonder why consumer-grade doesnt pursue the more feasible goal of reading from a motor nerve. I wish i could train myself typing by thinking of finger movements.
Being skeptical about the accuracy of these technologies/feasibility of surgery is valid, but this article reads as pretty cynical rather than just skeptical, haha.
>> Would I invest my money in it?
Hell no
>> Would I have neurosurgery for no good reason?
Hell no*10
To me this is clearly a longshot product, but I'm very happy that some private citizen is willing to invest his own money into this. If someone stumbles on how to transform thoughts into movement signals or how to encode digital images into optic nerve signals I'm all for it
If I was about to bet on the future I assume, that aspirations of hybridisation/creolisation/amalgamation of human&machine will happen on the biological register; that this development will go the other way round: from semiconductors to biotechnology. E.g. the brain is a fabulous architecture; damn effective and very low energy consumption at the same time.
Wich makes me think: A substantially useful AI that can immerse with neurological coginition (and not just be a fancier interface that safes you from carrying a calculator around), will probably need A LOT of computational power. Damn these things can't even steer a hand through useful movements while im typing a way in language (which is its own insane technology [1]) with quite some musicality.
What I mean is: Let's assume you could really supply the population with these interfaces and they really work. Who is going to be supplied with her/his own infrastructure of supercomputers running that AI that will merge with her/his cognition? I don't think that it would be possible to supply even a marginal amount of the population with access to these entitys. So this amounts to some dream of the (misunderstood version) of the Übermensch, as a caressing of the narcissitic hurt that the being of tekhne inflicts to some.
[1] It has been argued that language speaks. Not only do subjects speak language, but language also speaks subjects. (that's spoken through Heidegger). Thus there is something like a life of lanugage (it makes sense to think about if it is alife). Think of language as an AI.
Now, for people with spinal injuries, etc. then this totally seems like a good thing to be investigating.
Most people cannot think faster than they type or speak, at least coherently. The rapid information transfer with a computer via "thoughts" ignores the fact that a concept of a "car", for example, exists as different neural activity patterns, distributed over different parts of the brain for different people and changes with time. So that would imply the BCI would need to catalog how different all possible concepts are cataloged within an individual, to achieve such "direct thought communication", which would require whole-brain scale recording, something definitely NOT achievable with ephys (as a thought experiment, you'd have to introduce wires whose total volume become significant compare to brain volume).
Further, the act of speaking and typing is also a way to articulate and organize your (abstract) thoughts. If one skips that stage, would the output be meaningful at all?
The limiting communication bandwidth idea is a very fancy marketing message easy for people to hype up on, in typical Elon fashion.
Likely going to sound quite pretentious - but I definitely can and have been able to do so for as long as I can remember. It's honestly quite annoying as I grow older. I like to think it may directly correlate with the fact I was able to type 160+ WPM on a shitty membrane keyboard at school in the 5th grade quite easily.
But yes, as you said and what I've come to believe... is that most can't.
I am also like you, in the sense that I can think orders of magnitude faster than I can type or speak.
It is bizarre to me that others wouldn't be able to. But now that I think about it, yeah, I think you two might be right.
I agree with you that language helps formalize and structure thoughts in an intelligible manner, and I do also have reserves on the hypothesis that one can think much faster than it speaks.
Though, is it completely wrong to assume that a neurolink interface would act like a language?
I mean, you mentioned that it would not be possible to probe the whole brain, which makes sense, but I could imagine that to communicate easily with a neurolink interface, you would somehow need to convert your thoughts to _something_ that is probed and intelligible to the interface, just like you do when you articulate sounds to speak.
For instance, if I just vaguely think of moving my hand, that's a pretty blurry and noisy thought. But when I focus, I can almost physically feel that I am at the very edge of moving it, my sensations in the hand increase, my muscles don't move but I feel like they are ready to at any time.
Surely, thoughts that are the result of a specific focus like this could be more easily probed. In fact, it could also be that your brain would need to "learn" how to properly focus thoughts that work with the interface.
That makes for a much less appealing interface though, because it's not "effortless" to articulate thoughts in a specific manner that an interface can probe.
There are a lot of things that we ignore or don't register consciously. It is possible to decode items in one's working memory in controlled situations (see for example: https://pubmed.ncbi.nlm.nih.gov/28155212/).
So I suppose one possible approach would be to hold things in your working memory and have the Neuralink actively decode those. But in that case, as you've pointed out, requires a lot of attention and most likely would rely more heavily on contextual AI to narrow the decoder search space.
The basic premise of my skepticism with the "we need higher bandwidth I/O" statement still stays though -- the bottleneck of human communication is how quickly coherent thoughts can be formed in a way someone else can understand it (output), and how quickly information can be understood (input). For the latter example, you can scan an entire page in less than second. With immense speed reading practice, you can get a "gist" of a page in around that time. But understanding the details in that time is not possible.
It appears to be "neuralink and others are promising full mind reading and this whole thing is going to fail"
It is similar to the "self-driving argument". People complain and complain about Elon Musk's full-self-driving, yet telsa owners are out on crowded freeways, their cars happily driving them down the center of the lane and judiciously following the car in front. Yes, it probably can't get drive through at mcdonalds, but big whoop.
(also this person has no concept of what "the public" will do - honestly they would probably pay $5k to have alexa beep in their mind when someone's at the door)
Personally I would love a ctrl-labs keyboard, but now with facebook owning them, you know there will be a login requirement.
What exactly do you gain if you still have to sit in the driver’s seat and can’t look away for more than a moment?
I’d say that’s only a 1% improvement.
Not having to look at all, you could read or work. That’s another 10%.
Not having to even be present opens up a whole world of innovation and societal benefits.
Who would be dumb enough to implant something from facebook into their brain?!
Costly high-risk surgery that might kill or lobotomise you is not very appealing compared to glaring at a screen and moving your fingers for the purpose of sharing meaningless drivel with the world.
The hardware in that system is from Blackrock Microsystems. You can call them up and buy it right now (if you’ve got the cash...It’s not cheap!) It’s currently intended for research, so you’ll need ethics approval and the decoder isn’t included (but that’s the fun part anyway).
Ripple Neuro sells similar gear (better, even, maybe) and Neuralynx and some smaller companies also have competing products. It’s not in every corner store, but you could have a similar setup this month if you really wanted to.
There will be some interesting BCI tech in the next few years but as the author notes, it's going to be very specific and limited, and probably quite expensive.
So what? Stop trying? Give up? What's the point of this statement? Why is the timeline you personally find realistic relevant to anyone else but you?
I just don't get this attitude. It's such a sour grapes type of attitude. Either help out, or get out.
It’s called “skepticism”. You read enough history, you realize the charlatans greatly outnumber the geniuses, and you do your due-diligence before you invest, or form some weird cult.
Just comparing the mission statement of NeuraLink (from their homepage) with the wild claims of their most prominent founder (Musk) should make you stop and think for a second.
The actual goals are quite humble and achievable in the near term. What Musk says on Joe Rogen's podcasts and other public forums, however, is more fiction than science.
Unfortunately most publications focus on the latter and ignore the former and way too many people fall for this and seriously believe that highly invasive and risky brain surgery will become routine in just a couple of years.
We're talking about the same guy who said there could be "1 million full self-driving robo taxis on the road by 2020" during the April 2019 Tesla investor day [0]. The same guy, mind you, who's company two years later, in 2021 told regulatory bodies that their technology is only SAE level-2... [1]
The technical press back then was very sceptical about the claim as well (and it turns out they were right to do so). Autonomous cars are on a whole different level than BCI, though, and there's no breakthrough science required to achieve autonomous cars.
So in light of this I feel that scepticism has its place w.r.t. BCI and outrageous claims (e.g. streaming music directly into the brain or curing neurological diseases) about the near-term abilities and development stage of the technology.
[0] https://www.extremetech.com/extreme/290029-tesla-well-have-f...
[1] https://www.thedrive.com/tech/39647/tesla-admits-current-ful...
Secondly, nothing that was shown (apart from being far from new, exciting or ground-breaking in any way) was independently verified so we just have to take their word for it...
I am not impressed at all.
[1] https://github.com/monkalynn813/Mindwave_controlling_cursor
Leaving aside that this person was probably in this field before Neuralink and the others, in matters such as these, offering an expert opinion on a market filled with snake oil certainly is helping out. Otherwise we rely on people whose first priority is not truth but profit.
>So what? Stop trying? Give up?
Her suggestion is to exercise skepticism and be realistic about what the actual applications and markets are.
Matthew MacDougall, Head of Neurosurgery at Neuralink and neurosurgeon at California Pacific Medical Center. He was previously working at Stanford where he worked in labs that implemented and designed brain-computer interfaces.
Vanessa Tolosa, Director of Neural Interfaces. She previously led a neurotechnology team at the Lawrence Livermore National Laboratory that worked with a wide variety of technology on technology prostheses that were used in clinical and academic settings.
I don't get this attitude. The author is an expert. Can't we learn from her?
I work in this field and I have such strong, mixed feelings about companies like these. On the one hand, more interest in the brain and neurotech is great. These are tough problems, and we need new ideas, new tools, and new approaches. The standard academic approach of throwing a few trainees at a problem for a few years each (mostly in isolation), might not be the best way to tackle a problem that ranges from biophysics to psychology (and everything in between: materials science, signal processing, etc).
On the other hand, I worry that excessive hype is going to blow the field up before it gets started. If Elon Musk says he can implant 3000 electrodes that will let you control your iPhone--by next year--why would anyone fund me to do the slower, slogging work that I think will be required to eventually make something like that possible? This isn't just a critique of industry; the same brand of hucksterism shows up in parts of academia too. Throwing cold water on people's hopes and dreams isn't fun, but I think it's important for the long-term health of neuroscience as a field.
Shorter term, companies should also think about who they're attracting with this hype. I'm on the job market (in a very low-key way), and one of my principle requirements is that I want to work somewhere that is serious about the science: I want to build something that will actually work, instead of burning a pile of VC cash chasing hype or, God forbid, Theranos. Many of my colleagues feel the same way.
There is nothing wrong with grounding a company's claims/goals against reality. Particularly relevant for any company lead by Elon Musk. His timelines are always wrong. He's been promising FSD is just a year away for the better part of the last decade.
SpaceX took nearly fifteen years to become the beast it is and Tesla took eight to increase deliveries from 2,600 to 500,000. Don't underestimate what these companies achieve in the long-term and we are always terrible to predict exponential growth (from pandemics, to EV, to space flights)
[1] https://www.nytimes.com/interactive/2018/science/spacex-falc...
[2]https://www.statista.com/chart/8547/teslas-vehicle-deliverie...
There isn't really a market for BCI except for cases of disability. In this case, he has to create the customer demand for a technology that doesn't exist. That's a really tough sell when even if it did work (which is very much in doubt), would require a literal brain surgery. This isn't like scaling up a factory or using modern compute to get a rocket to land itself. It's fundamental research with uncertainties on perhaps every angle.
Plenty of people get Lasik just to avoid the annoyance of contacts/glasses and eyes are technically part of the brain. Why wouldn't there be a market for BCI when the potential benefit is much greater?
I think your missing the point about what I'm highlighting needs to happen after the wires are inserted into some otherwise unsuspecting tissue that isn't specially adapted for this use case in any particular way.
Citation needed
Maybe machines in the future just won't need that much specification. Like we'll give high level goals to robots/computer and interaction will be much more passive. You'll just tell a program to do your taxes or tell a robot to clean the house and it'll get done.
It's possible that we can accomplish most of the benefits of BCI without the brain surgery part. By the time this technology is ready, you may be able to take nano-bot pills once a month and to uninstall, you just stop taking them. Or maybe we'll genetically engineer babies to be born with them.
It's so early that all we can do right now is speculate and there's a lot of reason to believe this one particular company won't succeed.
Applying similar patterns to Neuralink, the likelihood that a lot of people will jump at the opportunity initially is small. After that, it all depends on how exactly it improves lives (maybe even becomes necessary in some fields?) and how long they can absorb losses until the process is battle-tested enough.
As opposed to Lasik, Neuralink presents much greater risks through surgery and being an implant in the first place. My guess is the value proposition would need to be overwhelming for a lot of people to consider it and fairly strong for enough people to consider it to guarantee profitability.
I wouldn't call this widespread adoption. Maybe it's cost, maybe it's the risk of the surgery.
(1) https://www.webmd.com/eye-health/news/20180727/lasik-know-th...
The point is to remind us that all of the hype around this is just that: hype.