Musk has stuck a Neuralink device into a person’s head. His announcement isn’t particularly informative.
The first human received an implant from @Neuralink
yesterday and is recovering well.
Initial results show promising neuron spike detection.
…promising neuron spike detection
? Do you realize how meaningless that is? I’ve jammed steel pins into a cockroach’s butt and gotten “promising neuron spike detection”. I’ve stuck sharpened tungsten wires into a zebrafish’s hindbrain and gotten “promising neuron spike detection”. This is a trivial accomplishment. Living brains are big sparking balls of continuous electrical activity, it’d be stunning if you put a wire in one and couldn’t get some measure of current.
This being Elon Musk, he continues with promises for the future rather than giving any details on what his company has actually accomplished.
In follow-up tweets sent in between arguing about video games and bantering with far-right influencers, the businessman said the first Neuralink product was called Telepathy.
”It enables control of your phone or computer, and through them almost any device, just by thinking,” he wrote. “Initial users will be those who have lost the use of their limbs. Imagine if Stephen Hawking could communicate faster than a speed typist or auctioneer. That is the goal.”
Musk has a long history of bold promises but a spottier record of fulfilling them. In 2016, he wrongly predicted that within two years it would be possible for a Tesla to drive autonomously from New York to Los Angeles. That year he said his SpaceX rocket company would fly to Mars in 2018 – it still has not.
Don’t forget the Hyperloop!
So can we expect this patient to be making calls to the press with his mind? Musk is not going to say, but I will boldly predict that…no, they will not. I will further guarantee that it will do none of the things Tesla PR promises.
I’ve seen people joking on other forums that the chip will make you watch adverts or you will lose the use of your eyes under his Tees & Cees.
No doubt that concert tour would be cheaper than Taylor’s, but considerably less value for the dollar.
oops, wrong thread!
There is an aspect of self-control or deliberation, which may be lost on Muskrat, where mere impulses are often quashed before being broadcast. Consult Dennett on his free will arguments.
Anyway in the typical dynamic of impulse, deliberation, and action much of what is covert never becomes overt behavior. Having some cyber-interface connected to your so-called smartphone could result in untold exhibitions of impulsiveness. Even in the best case that Neuralink achieves some success in bridging thought and online action this should be a warning as to consequences: https://www.youtube.com/watch?si=hWvzwVD4AjNCMbAG&v=cw4C8YNySpk
Imagine if we all had ability to instantly broadcast all our most private thoughts into the comments section here. I mean it can still get bad already, but for most people there is prefrontal executive function and a threshold to cross before the fingers engage.
I have enough of a time scrolling through news apps and what-not and my clumsy thumb accidentally launching ads. Impulsive thoughts could be disastrous. Plus much of what transpires is subliminal or nonconcious so there could be stuff that seems alien making it through.
And that’s just the outbound stuff. What about incoming from the internet bombarding your brain. Who creates that firewall and based on what knowledge of such low level neurotransmission?
Plus what differentiates meaningfulness and noise? And Musk’s output is bad enough. Imagine him creating and using a real life Shitter.
The Elon Musk “Brainchips Across America” tour? Yes on both counts. ;)
Sorry PZ but you are absolutely misunderstanding Musk, as if his genius is beyond your level. What “promising neuron spike detection” means is that the subject did not die which is an amazing accomplishment. For a minute, I deeply thought about what Elon has achieved here and it literally moved me one minute into the future. Amazing.
@lotharloo #6
yet
You’re actually, taking this at face value, PZ, rather than seeing it for the stock-goosing lie it (almost) certainly is? And you call yourself a skeptic!
The illustration suggests that Musk can upload his consciousness. I’d be much more pleased if he could download a conscience.
Frankly speaking, I think you can tell everything you need to know about Neuralink’s current state by the fact that, contra his own hype about it, he’s not getting the prototype installed.
Compare/contrast the fact he routinely brags about using a Tesla (of varying types) as his daily driver.
Cure depression? Upload your consciousness?
This is Elon Musk we’re talking about. More likely, you’ll upload depression and cure your consciousness.
In other Elon news, do a happy dance because DE says he can’t personally collect an extra $50B from Tesla.
I have a friend who has a form of epileptic seizures. He had a vagus nerve implant a year or so ago to control them. So, the idea of using electromagnetic devices for neurological stimulation in humans ain’t that radical. However, my friends vagus nerve implant hasn’t worked that well…it affects his voice…and has been off most of the time. He saw a new neurologists recently to adjust the timing, but she couldn’t figure that out. She’s going to have to talk to the manufacturer’s tech support. So there’s a lot more to do on this front.
Incidentally, there’s a fascinating article in the current issue of Scientific American titled “Minds Everywhere” on research that suggests that “brain” is bigger than neurons in our skull.
Hemidactylus @ 4, only one clumsy thumb? I’ve got 10 of them on some days!
Still, that firewall, well, long married people do learn that firewall technique of selective listening… ;)
Seriously though, we naturally do filter our senses, with most information being ignored. Something like what’s promised and be anticipated to arrive alongside commercial fusion power would just add to that workload.
But, I could see Muskrat successfully backing up his mind. How much storage capacity would one need to store the mental processes of a paramecium?
UnknownEric the Apostate @ 5, that’s twice I glanced at your comment and rather than see brainchips, initially registered brainships. In the latter case, Elon would most certainly travel in only the best of circles.
lotharloo @ 6, one will still detect neurological activity to some degree after pithing an animal.
I’ll simply suggest we wait to see what the long term survival is in his human subjects. His animal research did lose all of his partners, due to pretty much 1950’s non-survival rates and that his research basically was 1970’s state of the art – albeit with far better and smaller computers.
Who knows? He might just yet invent an implant to allow some deaf people to hear, perhaps by implanting electrodes inside of the cochlea and call that a new technology too.
Yeah, he’s only around 40 years behind everyone else, with infection rates right out of the 1920’s.
I find it much more interesting with the results of helping the blind “see” and other peer-reviewed research.
drew @ 11, unable to do the happy dance. I’m still laughing too hard.
robro @ 12, the first serious research on vagus nerve stimulation was in the 1930’s, meaningful in a modern context, research began in the mid ’80’s. De novo clearance was granted in 2018 for migraine treatment, but the first epilepsy treatment implant was in 1988.
But, we can ascertain how high the efficacy is by the tact that a non-invasive vagus nerve stimulation treatment was given emergency use authorization for COVID-19 in 2020. Pretty sure that was alongside snake oil and rattles…
@4, Hemidactylus,
Who wrote,
Hah, then there is the threshold to cross to hit post even after your fingers engage. I don’t post half the comments I write. Usually because the comment is too long, too boring, rambling, or completely opaque (What was I saying again? If I can’t figure it out how can I expect someone else to?).
I was wondering what the MAGA crowd and conspiracy theorists thought of their favorite billionaire troll’s announcement so I took a dive into the swamp for a quick look, and aside from jokes about Joe Biden needing one, they’re… conflicted. For example:
and
But it seems like Musk is smart to focus on the technology’s potential application (if it ever works) on helping the disabled and amputees, because it’s that point that comes up time and again in approving comments from his supporters. Even Alex Jones said that’s okay, but he’d “come after him with everything he’s got” if he tried to mass market it or promote it to schoolchildren (as if he had anything left to go after Musk with these days).
Not at all surprised they give Musk the type of benefit of the doubt they would never give to a more moderate or liberal billionaire, but Musk likely has an uphill battle to keep his paranoid supporters on his side should the tech show signs of progress, or maybe after his battle to keep Twitter afloat fails, he just won’t care anymore.
So, too early to build a Murderbot?
All this talk about “uploading consciousness” is getting way ahead of even plausible speculation, let alone reality. “Uploading consciousness,” or merely copying any sort of information to or from an organic brain, can’t even be considered a theoretical possibility unless and until we can answer this simple question: What is the “machine language” of an organic human brain? Do organic brains even HAVE anything that could be called a “machine language?”
Computers can communicate with each other, and transfer information in various forms to/from each other, because all the information is expressed and stored in various formats that the machines (via their operating systems) are able to process and manipulate. Any information not translated into any such machine-usable format, is not available to the machines.
So how, exactly, would one go about “uploading,” say, my memory of a movie I just watched — or one I watched ten years ago, or one I’ve watched several times over a period of several years, processing it differently each time — to an Internet server or some such? How would one be able to extract the bio-electrical signals that represent my memory of that movie, separate it from signals representing something else, and translate the former into a full and accurate document of what I’d seen (and how it made me feel, which would involve linking to other bits of memory, emotion, etc.)? Is any of this even possible at all?
Any uploading will be a copy, fellow meat computers. Your only hope is your descendants figuring out how to get your organic material and put it back together, if they like you enough to want to try.
Not much of this is actually new. Recording human neuronal electrical activity with indwelling electrodes is old hat – epilepsy centers have been implanting electrodes in brain to diagnose seizures for over 50 years. Closing the system for long-term use and adding neurostimulation to alter brain function is newer: the Neuropace RNS system records neuronal activity, analyzes it, and stimulates brain to stop seizures. It was FDA approved in 2014. Deep brain stimulation to treat Parkinson’s Disease was approved by FDA in 2015. Last year, a paraplegic patient was able to walk with assistance from a device that recorded brain activity and transmitted it to a spine-level stimulator (Lorach, 24May2023, Nature). So, to me the question is: what advance does Elon bring to this? All I see that’s new are unsupported grandiose claims about vague future uses, and also blatant disregard for proper animal safety in the preliminary experiments.
Not much of this is actually new.
To the idiots who make up #QElon’s personality cult, ALL of it is new — they never followed developments in this field until #QElon made them aware of what the #QElon-owned company was doing in it; and now that’s all they know.
(And, in fairness, I, too, have to be reminded of what other non-Muskrat companies have done in this field, because like so many other lay readers, I don’t follow this field of medicine all that closely either. This is how an obvious idiot like #QElon is able to make himself look like such a bold brave cutting-edge innovator without ever actually bringing anything real to the market.)
I’ve searched trashed offices in sci fi horror games for audio logs less suspicious than that.
Raging Bee @20:
Theoretically, you don’t have to understand the machine language of the human brain (or a computer) to replicate it. Not saying it would be easy (or even possible), but if you can record the operation of the brain with enough fidelity and replicate the operating environment accurately enough, it could be enough to bootstrap the duplicated brain by hitting play on the recording and letting it go from there.
As for uploading in general, a Ship of Theseus approach would seem to be far more plausible anyway — i.e. replacing the brain one a piece at a time (in some cases at the brain cell) with more robust artificial version, in situ, while you’re still alive.
Then, once your body gives up the ghost, your brain can be kept alive in an environment of your choice — connected to the cloud, or in an android body. (No doubt there would also be a market for “donor” human bodies, because there will always be people who will take the lowest road possible to survive.)
Aside from the increased practicality (going from the impossible to the next to impossible), there’s another important advantage to this approach. It would solve the problem of the continuity of self that making an uploaded copy would cause. There would always only be one you.
It’s fun to speculate, though at my age I have about a 50/50 chance of living another 22 years, so it’s all academic to me.
@8: Mu5k’s real problem is that the development cycle is so delayed that he was unable to implant the chips — and, of course, the (inevitably buggy) mind-control software not disclosed on the packaging — into one particular Delaware judge. But that leaves a critical question: Does getting one of these chips implanted mean thereafter going through life without one’s tinfoil hat? It won’t be alien mind-control rays — they’ll be coming not just from the CIA, but from Big Media…
All of which assumes, without any evidence, that (a) all the bugs have been worked out, (b) there are no features that are not actually bugs, (c) the documentation is accurate and hasn’t been approved by the marketing department, and (d) nobody has built any backdoors into the system for convenience of later debugging that will be later exploited by someone with less integrity than Mu5k. Like, say, his cobillionair-techbro friend whose company is now and always has been dead to me…
I am interested in how this got by an institutional review board.
If you get a Musk brain implant, the first thing you should do every morning is check to see if you suddenly have an X account, and take a look at how many likes there are for Musk’s tweets.
Theoretically, you don’t have to understand the machine language of the human brain (or a computer) to replicate it.
I suspect you’d have to understand both the machine language of the brain (the “software” operating system and stored information) and the specific neural pathways (the “hardware”) before you could read the contents of a brain and translate it with sufficient fidelity (which is to say, probably no less than 99.999%) onto any other medium that would process the information the same way as a human brain would.
Of course that’s all assuming organic brains even HAVE a “machine language.” They may not, and if they don’t, then there’s absolutely no way to copy or replicate any significant part of a person’s consciousness to any other storage or processing medium, let alone the whole of it. And writing any kind of information TO an organic brain by the same means becomes even less plausible.
Also, doesn’t the act of reading information from a hard disk or RAM chip involve shooting electrical impulses through the bits where the data is stored, and changing the actual charge on those bits and then changing it back again? How would a “mind upload” system do that to a person’s brain cells without destroying them? Computer chips are made for that sort of thing, organic brain cells aren’t. Best to stick with the one-cell-at-a-time replacement/upgrade idea instead.
robro @ #13, thank you for that reference! Here’s the Scientific American link for anyone interested:
“Brains Are Not Required When It Comes to Thinking and Solving Problems—Simple Cells Can Do It”:
@ wzrd1 :
Reminds me of Anne McCaffrey’s Ships novels starting with The Ship Who Sang :
https://en.wikipedia.org/wiki/The_Ship_Who_Sang
@ 7coffeepott :
Well, in fairness we all die evewntually. Question is whether this Muskian tech contributes tocausing that death and makes it occur sooner rather than later.
@ Hemidactylus : “Plus what differentiates meaningfulness and noise? And Musk’s output is bad enough. Imagine him creating and using a real life Shitter.”
Given “Shitter” is a Aussie slang for the dunny – okay, er, toilet – I’m pretty sure he’s used a shitter in real life. A Musk shitter could be different I guess though hate to think what he’d do with it as well as on it.
Meh. Phineas Gage stuck a metal piece into his brain two centuries earlier, so this is unimpressive.
Tacitus @ 25
The ‘bit by bit’ approach was explored by Stanislaw Lem’s SF story “Roly-poly” aka “Are you Alive Mr.Smith?” in the 1960s .