Funnily enough, FDA forbids Elon Musk's Neuralink human experiments Despite years of proclamations by Elon Musk that human tests were just around the corner, it turns out his Neuralink brain-implant startup has already asked the US Food and Drug Administration for permission to conduct human tests – and been rejected. For those who can't recall, Neuralink is trying to make a chip that can be …
He's trying for approval of an experimental BRAIN implant.
His management style and corporate culture would make it hard to get and artificial toenail implant approved. He's trying to get the FDA to approve an experimental medical implant with no known therapeutic application. In the brain. I mention that because along with things like your heart and lungs they tend to be very cautious about brain implants. They also tend to focus on ones that treat actual and often life threatening medical conditions.
While I think the field of research is valid and important, human trials on a brain machine interface aren't where you should move fast and break things.
While I think the field of research is valid and important, human trials on a brain machine interface aren't where you should move fast and break things
I think the issue could be easily resolved by the FDA simply saying "You first!" and making Elon patient zero. I kinda went through some of this process many years ago when trying to develop a simpler interface to control prosthetics. Had a willing volunteer in a mate who lost a forearm in a bike accident, a willing neurosurgeon and reasonable cunning plan. All we wanted to do was attach sensors so we could record nerve impulses, and then try to proxy them. Assuming we could make any sense of the data. But I also learned that we've evolved to be really good at rejecting foreign objects, or just encapsulating them in fibrous tissue that then becomes rather difficult to remove. Plus there are a lot of nerves involved, and condensing a couple of trolleys worth of tin into something that could fit inside a prosthetic cuff was a bit of a wicked problem for the future. But first, the data. In researching that though, I did find examples of where it had been tried before, including one where an implant had been attached to a volunteer's optic nerve, and a screen showing a very low-res image of some simple patterns. Think that was from the late '60s, and technology has moved on a bit, so better DSPs to crunch signals.. But still a wicked problem to solve, even if some progress has been made.
https://www.hopkinsmedicine.org/health/wellness-and-prevention/the-brain-gut-connection
"Scientists call this little brain the enteric nervous system (ENS). And it’s not so little. The ENS is two thin layers of more than 100 million nerve cells lining your gastrointestinal tract from esophagus to rectum."
That could be the answer for Musk. Go for a "pluggable" ENS interface. The FDA are more likely to approve a non invasive interface.
Worst case, descope like Boring Co/Hyperloop, and just call it a Butt Plug. There's a ready alternative market for it
descope like Boring Co/Hyperloop, and just call it a Butt Plug
I think you mean Auditor of the Month Trophy. The original sold for $60,000 at auction.
Putting batteries into the body is technically very difficult and risky - I can see a way to avoid this problem and power the device without batteries, just a decent 1F capacitor, kept charged via a rework of my Nexus phone wireless charger. I think it's potentially an interesting testing/evaluation environment so eliminating the batteries to avoid internal risks would be worthwhile, but it would require an internal low-current design.
A 1F capacitor stores a potentially-deadly amount of power, and they are physically-large.
As a kid, I was given a 100,000 uF@25 VDC capactor by an old man, who demonstrated the potential hazard by charging it up a bit, then shorting the terminals with a screwdriver. The resultant arc melted/evaporated a substantial chunk of metal from both the positive terminal and from the screwdriver shaft. He gave me the screwdriver, too. I think he meant it as a "reminder" of the hazard.
Depends on the chemistry/technology that Neuralink want to use - are they looking at alternatives to the existing technology used for example in cardiac pacemakers since the '70's?
https://en.wikipedia.org/wiki/Artificial_cardiac_pacemaker
"Lithium-iodide or lithium anode cells increased the life of pacemakers from one year to as long as eleven years, and has become the standard for pacemaker designs.
As of 2021 enough brainy "investors" apparently thought it was close enough to be ready for market that it garnered $363M.
Personally, I would say it is obviously basic research only, without a chance of marketability.
Nobody want to up money for fabs, but unproductive sci-fi brain implants and money losing Twitter are OK.
Go figure.
As of 2021 enough brainy "investors" apparently thought it was close enough to be ready for market that it garnered $363M.
That's the power of marketing. Musk is the real-life Tony Stark, the smartest guy in the room, and the chap who told the world the Hypeloop was really simple. Just like an air hockey table. In a vacuum tube. Then he gave Las Vegas a bunch of Teslas travelling slowly through a storm drain. Plus plenty of investors have probably read SF, and dream of the possible, especially if they're also seeking yield, or just have too much money.
Personally, I would say it is obviously basic research only, without a chance of marketability.
There is a potentially huge market for a real-life neural lace, even if all it was doing was providing you a 4K graphics overlay in your visual field.. But it's kinda hard, especially when our brains have billions of neurons and we don't really understand how they work. Something I did when I was researching this stuff was to try and learn braille. Simple enough because it's a standard language, so just observe nerve impulses as a fingertip passes over some raised bumps that form a known pattern. In reality, really not simple, especially as our fingertips also generate additional data like tenperature, texture etc etc. Especially if you're 'just' trying to find out where those nerve impulses terminate in the brain, interpret them and act on them.. somehow. It was one of those cases where it'd be a lot simpler to maybe take the position sensor used in castrated mice, build that into a finger cover and run the scan data through a bump-to-speech convertor. Then from a prosthetics POV, have that as an 'FSD' feature in a hand or limb prosthetic that had enough fine motor control to allow someone to read braille.
It's one of those things engineers learn to dread.. That things like this are 'technically feasible', if you can somehow figure out a way to overcome a huge number of challenges. But if you can, then there's potentially a huge market. Assuming your product actually works, and doesn't kill your patient in the many ways a brain implant could.
"Neuralink insiders who spoke about the animal investigation last year reportedly called the surgeries "hack jobs," and in-house scientists expressed frustration at having to perform surgeries so quickly and in volume."
Naaaa, probably not a good idea with anything medical but especially not brain implants.
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2025
