![Biotech Biotech](http://a.fsdn.com/sd/topics/biotech_64.png)
![AI AI](http://a.fsdn.com/sd/topics/ai_64.png)
ChatGPT Has Been Integrated Into a Brain Implant (cnet.com) 34
CNET visits a leading-edge company making an implantable brain-computer-interface that's "experimenting with ChatGPT integration..."
We previously covered Synchron's unique approach to implanting its brain-computer-interface (BCI) without the need for open brain surgery. Now the company has integrated OpenAI's ChatGPT into its software, something it says is a world's first for a BCI company...
Typing out messages word by word with the help of a BCI is still time consuming. The addition of AI is seen as a way to make communication faster and easier by taking in the relevant context, like what was last said in a conversation, and anticipating answers a person might want to respond with, providing them with a menu of possible options. Now, instead of typing out each word, answers can be filled in with a single "click." There's a refresh button in case none of the AI answers are right... [ALS patient Mark, one of 10 people in the world testing Synchron's brain implant in a clinical trial] has noticed the AI getting better at providing answers that are more in line with things he might say. "Every once in a while it'll drop an f-bomb, which I tend to do occasionally," he says with a laugh.
Synchron CEO Tom Oxley tells me the company has been experimenting with different AI models for about a year, but the release of OpenAI's ChatGPT-4o in May raised some interesting new possibilities. The "o" in ChatGPT-4o stands for "omni," representative of the fact that this latest version is capable of taking in text, audio and visual inputs all at once to inform its outputs... Oxley envisions the future of BCIs as... having large language models like ChatGPT take in relevant context in the form of text, audio and visuals to provide relevant prompts that users can select with their BCI... Synchron's BCI is expected to cost between $50,000 and $100,000, comparable with the cost of other implanted medical devices like cardiac pacemakers or cochlear implants.
CNET has also released a video — titled "What It's Like Using a Brain Implant With ChatGPT."
Typing out messages word by word with the help of a BCI is still time consuming. The addition of AI is seen as a way to make communication faster and easier by taking in the relevant context, like what was last said in a conversation, and anticipating answers a person might want to respond with, providing them with a menu of possible options. Now, instead of typing out each word, answers can be filled in with a single "click." There's a refresh button in case none of the AI answers are right... [ALS patient Mark, one of 10 people in the world testing Synchron's brain implant in a clinical trial] has noticed the AI getting better at providing answers that are more in line with things he might say. "Every once in a while it'll drop an f-bomb, which I tend to do occasionally," he says with a laugh.
Synchron CEO Tom Oxley tells me the company has been experimenting with different AI models for about a year, but the release of OpenAI's ChatGPT-4o in May raised some interesting new possibilities. The "o" in ChatGPT-4o stands for "omni," representative of the fact that this latest version is capable of taking in text, audio and visual inputs all at once to inform its outputs... Oxley envisions the future of BCIs as... having large language models like ChatGPT take in relevant context in the form of text, audio and visuals to provide relevant prompts that users can select with their BCI... Synchron's BCI is expected to cost between $50,000 and $100,000, comparable with the cost of other implanted medical devices like cardiac pacemakers or cochlear implants.
CNET has also released a video — titled "What It's Like Using a Brain Implant With ChatGPT."
I hope he never needs comma-separated lists (Score:5, Funny)
ChatGPT always ignores that request.
Re: (Score:2)
do you ask nicely?
Re: (Score:2)
'do you ask nicely?'
I do, I always have to ask at least TWICE though.
Re: (Score:2)
if you use openai's interface you might try the "customize chatgpt" option (from the top right menu). these instructions are supposed to apply to all subsequent requests.
usability improved considerably for me just by adding "short and to the point" to "How would you like ChatGPT to respond?".
Re: (Score:2)
Thanks, I'll try that.
Re: (Score:2)
Did that and it completely ignored it. :-)
The worst tattoo (Score:2)
Hahahaha, oh god. Thereâ(TM)s gonna be actual early adopters for this, the poor sods. Theyâ(TM)re gonna have this thing in their head for decades, until they die and get buried with it. And itâ(TM)s gonna keep its value about as well as a Segway-mounted fax machine.
Re: (Score:2)
Hahahaha, oh god. Thereâ(TM)s gonna be actual early adopters for this, the poor sods.
https://en.wikipedia.org/wiki/... [wikipedia.org]
Re: (Score:2)
Time for the caffeine....
Re: (Score:2)
I'm not quite sure if they're doing this right from the description.
Markov Chain text prediction works well as next-word prediction, but struggles out past a couple words because it's only predicting words. Transformers-based architectures basically predict out full coherent thoughts. So they *could* be using this as a super-advanced for of autocomplete, with the model functioning akin to a speculative decoder and the user as the verifier. The model drafts a long chain of tokens, and wherever the user's th
Re: (Score:2)
Just to be clear, in case any of the terminology is unclear (I just realized, rereading this, that some might be):
Speculative decoding is an inference technique (inference = running a LLM, aka a Transformers model based on text) to greatly speed it up. In speculative decoding, you have two models: the main (heavyweight) model, which acts as a verifier, and an extremely lightweight model (ideally wide-but-shallow) for "drafting" speculative tokens. The lightweight model very rapidly generates a long sequenc
Great for hallucinations! (Score:2)
Now get them beamed directly into your brain!
That said, a specialized LLM may or may not be a good idea for this type of application. But a general one? Pure insanity.
File it under: F-That! (Score:2)
Imagine (Score:5, Insightful)
having a piece of e-waste stuck in your brain because the corporate mothership went out of business [windowscentral.com].
#DoNotWant (Score:2)
Was this sort of thing developed by people who watched Star Trek and were actually a bit confused over the concept that the Borg were the bad guys?
Re: (Score:2)
It's just fiction. (Score:2)
Star Trek is made-up fantasy. Here in realty, we are borg already. More-or-less.
You think of yourself as an individual being, but really you are a network of trillions of cells. If you are having trouble with this, consider this rhetorical question: which of your neurons is the real you?
Your sense of self is produced by a complex network of interconnected neurons, pulsing chemicals around to one-another all the time. All the rest of your mind is like that, too. Science has revealed it to be so. The "I
Re: (Score:1)
Ah yes, because you with your healthy ability to speak normally is the target audience for this.
God forbid we do things for the disabled that isn't applicable to healthy people. That's not profitable.
And still... (Score:2)
Sounds like Project Lazarus. (Score:2)
and when it hallucinates (Score:2)
No no this one is OMNI!!! (Score:1)
Oh yeah, they got OMNI in there.
What a bunch of crap.
AI. LLM. Quantum. Blockchain. Brain Interface. OMNI. Trumponomics.
Let the hedge funds line up to lose money on this made-up crap. It has all the buzzwords.
Re: (Score:1)
When you are so tech illiterate you think everything is a meaningless buzzword meant to confuse and beguile you.
The Terminal Man, by Michael Crichton. (Score:2)
The Saying-Is-Believing Effect (Score:2)
Research strongly supports the notion that what one says can influence what one believes, so on the balance of probability: biases inherent in the tool will be adopted by the user, given sufficient time.
One such study can be found here. [doi.org]
whatevergoesinhere (Score:2)
I mean, it doesn't put an AI in your head—it puts a client for an AI API in your head. If the AI service stops, or you block it, you don't have the AI "in your head" anymore.
Re: (Score:1)
What the fuck are you even talking about? The BCI is a control interface, the AI is on a smartphone or in the cloud. The AI stops working, they go back to using their other methods of communication.
LOL. If I ever have a brain implant... (Score:2)
Voices in your head (Score:2)
Can I just, can I just, can I just... make you buy this piece of crap you don't need?