Gabe Newell on brain-computer interface technology(tvnz.co.nz)
tvnz.co.nz
Gabe Newell on brain-computer interface technology
https://www.tvnz.co.nz/one-news/new-zealand/gabe-newell-says-brain-computer-interface-tech-allow-video-games-far-beyond-human-meat-peripherals-can-comprehend
87 comments
What is the timeframe you imagine for us to see the first consumer BCI-based products?
Gabe seems to talk a lot about "inserting" data (like feelings) into the brain, instead of reading it. Is the technology really there already? Can we reliably read data from the brain (i.e. using it as input for a digital system)? And regarding inserting, what is the coolest thing you've done that you can share with us?
Gabe seems to talk a lot about "inserting" data (like feelings) into the brain, instead of reading it. Is the technology really there already? Can we reliably read data from the brain (i.e. using it as input for a digital system)? And regarding inserting, what is the coolest thing you've done that you can share with us?
Right now, OpenBCI makes products that only handle the "read" side of the equation.
As far as "writing" back into the brain, the coolest thing I've seen was the "BrainNet" project from University of Washington which used transcranial magnetic stimulation (TMS)
https://www.washington.edu/news/2019/07/01/play-a-video-game...
The science and tech is advancing very fast, but I think it's not accurate enough to be in everyday use yet as a controller for devices. 90% accuracy sounds great in a paper, but imagine if your mouse clicks or keystrokes didn't register 1 out of 10 times.
What feels way more likely is that we'll see biometric data being collected by more consumer tech devices (cellphones, laptops, headphones) and used as one of many inputs to improve software applications and operating systems. Could EMG or EEG data be used to improve iOS autocorrect and reduce fat finger mistakes? That's a mundane application for crazy tech, but it's the kind of thing that I think will be a necessary intermediate step in us learning how to use these types of signals in everyday ways.
As far as "writing" back into the brain, the coolest thing I've seen was the "BrainNet" project from University of Washington which used transcranial magnetic stimulation (TMS)
https://www.washington.edu/news/2019/07/01/play-a-video-game...
The science and tech is advancing very fast, but I think it's not accurate enough to be in everyday use yet as a controller for devices. 90% accuracy sounds great in a paper, but imagine if your mouse clicks or keystrokes didn't register 1 out of 10 times.
What feels way more likely is that we'll see biometric data being collected by more consumer tech devices (cellphones, laptops, headphones) and used as one of many inputs to improve software applications and operating systems. Could EMG or EEG data be used to improve iOS autocorrect and reduce fat finger mistakes? That's a mundane application for crazy tech, but it's the kind of thing that I think will be a necessary intermediate step in us learning how to use these types of signals in everyday ways.
Considering I own Neurosity's Notion: https://neurosity.co/, which is a consumer BCI based product, I'd say a couple years ago.
Do you like it? Where is the science backing up it's ability to keep you focused?
I don't know much about the tech, but I'm curious if neural networks and/or machine learning is used to process data and find connections, correlations, or...cool stuff like that? I just think machine learning is neat man, and bio tech is also very neat. The two together....super neat...haha.
I know the example they put out, a horror game that responds to your fears by reading the data you produce. Is the tech actually at a point were a dedicated team can accurately and consistently identify bio markers correlated with a fear response? Or diffrentiate between fear, anger, happiness, etc?
Has there been any research into using this tech effectively to treat anxiety/depression? Really interested in this focus once "writing" information is possible, especially with more severe mental health issues that aren't related to physical brain deterioration. But it seems (per your comments) that that's still a ways a way, haha. Would be really cool to be able to see a psychatrist and have them treat my anxiety with a few bip bops from an industrial headset, or go home with a prescription program to run once a day, that just sounds like science fiction.
I know the example they put out, a horror game that responds to your fears by reading the data you produce. Is the tech actually at a point were a dedicated team can accurately and consistently identify bio markers correlated with a fear response? Or diffrentiate between fear, anger, happiness, etc?
Has there been any research into using this tech effectively to treat anxiety/depression? Really interested in this focus once "writing" information is possible, especially with more severe mental health issues that aren't related to physical brain deterioration. But it seems (per your comments) that that's still a ways a way, haha. Would be really cool to be able to see a psychatrist and have them treat my anxiety with a few bip bops from an industrial headset, or go home with a prescription program to run once a day, that just sounds like science fiction.
I often wonder about the parallels between reverse engineering games using memory inspection software like 'cheat engine' to trying to reverse engineer the brain using a BCI.
For example if you want to find the memory address for your guns ammo you search for a start value in memory say '30', get all addresses that match that value, fire the gun and then find which of those addresses now have the value 29. Continue the search until you narrow down the memory address to just one. Then you can use that address to query the ammo for a 3rd party program that alerts you that you're low on ammo or even write to the memory address to give yourself more ammo..
Obviously the brain isn't as discrete but I feel like if I could play around with a BCI I could find fun signals for when I'm thinking about 'apples' vs 'oranges' and slowly build up an interface.
Have you been able to use a BCI to detect when you're thinking about something specific?
For example if you want to find the memory address for your guns ammo you search for a start value in memory say '30', get all addresses that match that value, fire the gun and then find which of those addresses now have the value 29. Continue the search until you narrow down the memory address to just one. Then you can use that address to query the ammo for a 3rd party program that alerts you that you're low on ammo or even write to the memory address to give yourself more ammo..
Obviously the brain isn't as discrete but I feel like if I could play around with a BCI I could find fun signals for when I'm thinking about 'apples' vs 'oranges' and slowly build up an interface.
Have you been able to use a BCI to detect when you're thinking about something specific?
There's a huge chunk of neuroscience devoted to questions like this.
Several groups have shown that they can "decode" a remembered image from brain activity. This is comparatively easy when the images are simple and there are only a few possibilities, but can generalize to larger sets of images and even (sort of) never-before-seen ones. Sensory and motor information is relatively accessible; I don't know that anyone's making great progress decoding thoughts like "I should be home by 8pm".
If this is up your alley, you may want to check out work done by Frank Tong's lab at Princeton (e.g. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1808230/) or Jack Gallant's group at UC Berkley (e.g, https://www.sciencedirect.com/science/article/pii/S105381191...), among others.
Several groups have shown that they can "decode" a remembered image from brain activity. This is comparatively easy when the images are simple and there are only a few possibilities, but can generalize to larger sets of images and even (sort of) never-before-seen ones. Sensory and motor information is relatively accessible; I don't know that anyone's making great progress decoding thoughts like "I should be home by 8pm".
If this is up your alley, you may want to check out work done by Frank Tong's lab at Princeton (e.g. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1808230/) or Jack Gallant's group at UC Berkley (e.g, https://www.sciencedirect.com/science/article/pii/S105381191...), among others.
I thought about buying an OpenBCI kit to try something like this, but I think there's just not enough sensor resolution to get anything meaningful in the way you're suggesting. The strongest signals by far come from muscle movement, too, so that tends to be the basis of interfaces.
(FWIW: absolutely not an expert)
(FWIW: absolutely not an expert)
In the area of BCI, what is the most "Science-Fictiony" thing you have seen made into a reality?
I referenced this in my other reply, but this is one of the ones that blew my mind:
https://www.washington.edu/news/2019/07/01/play-a-video-game...
https://www.washington.edu/news/2019/07/01/play-a-video-game...
All the talk of implants makes me wonder what's possible without a direct physical link to the brain. Like using an EEG to measure brain signals and transform them into instructions, images, etc. Here's some research where they use the EEG to train a model that reconstructs what people are seeing in their visual cortex [0]
All the reconstructions are super low resolution, but I wonder how much better they can get? I'd be much more open to brain-computer interfaces if it didn't involve drilling through my skull.
[0] https://scitechdaily.com/image-reconstruction-from-human-bra...
All the reconstructions are super low resolution, but I wonder how much better they can get? I'd be much more open to brain-computer interfaces if it didn't involve drilling through my skull.
[0] https://scitechdaily.com/image-reconstruction-from-human-bra...
Most if not all of the very interesting signals are in the 100hz - 10Khz range. These signals are also very low voltage so. Add in that the skulls acts as a low pass filter and you end up in a situation where you just will not be able to recover the information needed to decode what the brain is doing at the required resolution.
You can use ECoG which is like EEG but placed on under dura-mater (a protective bag around the brain) but this is also highly invasive. Also the spatial resolution is pretty bad. You can get a hundred or so sensors in a few square centimeters.
You can use ECoG which is like EEG but placed on under dura-mater (a protective bag around the brain) but this is also highly invasive. Also the spatial resolution is pretty bad. You can get a hundred or so sensors in a few square centimeters.
What signals do you mean? Brain waves measured in EEG typically range from 0.5 Hz and above.
The signals you pick up between 0.5 Hz and 40Hz are mostly likely "population" level signals or put another way the average activity of millions of neurons spread across large swaths of the brain. A good analogy re: trying to decode the brain from EEG would be akin to trying to decode what is going on in a soccer match but only using the noise of the crowds.
How are signals up to 10Khz generated? I thought the absolute refractory period of a neuronal action potential limits the rate at which they can fire.
From start to finish, an individual action potential lasts about 1ms, followed by a few millisecond refractory period. There is probably not much biologically-relevant activity happening at faster timescales.
However, one often wants to "sort" spikes so that those arising from individual neurons are grouped together. To do so, you need to sample fast enough to capture the shape of the action potential, not just the fact that it happened. This usually requires sampling rates in the 10s of kHz. (~30 kHz would be good).
However, one often wants to "sort" spikes so that those arising from individual neurons are grouped together. To do so, you need to sample fast enough to capture the shape of the action potential, not just the fact that it happened. This usually requires sampling rates in the 10s of kHz. (~30 kHz would be good).
So, to be clear, the signal of interest does not go above 1kHz.
Obviously if you want to sample a signal you have to use a higher frqeuency.
Obviously if you want to sample a signal you have to use a higher frqeuency.
Eh, it depends on what you're calling "the signal of interest."
A spike train, the output of a neuron, can reasonably be represented as a binary signal (1=spike, 0=no spike) at 1 kHz. In fact, I'd say this is nearly standard.
The neuron's membrane potential, a combination of its current inputs, intrinsic properties, and recent history, changes on faster timescales. Whatever processing the brain does probably does not use this information itself, but one probably wants access to it anyway for experimental and practical reasons.
The signal that an extracellular electrode picks up isn't that 1 kHz spike train. It's many of them mixed together and you'll need finer features to tell them apart. In other words, you can't just invoke Nyquist, sample at 2 kHz, and call it a day.
A spike train, the output of a neuron, can reasonably be represented as a binary signal (1=spike, 0=no spike) at 1 kHz. In fact, I'd say this is nearly standard.
The neuron's membrane potential, a combination of its current inputs, intrinsic properties, and recent history, changes on faster timescales. Whatever processing the brain does probably does not use this information itself, but one probably wants access to it anyway for experimental and practical reasons.
The signal that an extracellular electrode picks up isn't that 1 kHz spike train. It's many of them mixed together and you'll need finer features to tell them apart. In other words, you can't just invoke Nyquist, sample at 2 kHz, and call it a day.
Quite a lot simpler to seperate the mixed signals using a modern multi-electrode array.
I've written multi-tetrode cluster analysis SW. It is very much not simple. Deconvolving all the inputs is terribly difficult and computationally expensive. Generally we had undergrads as classifiers for the clustering ML, which never really worked all that well, TBH. The higher the sampling frequency, the better the spatiotemporal resolution, but the more data to work through. When you do get a cluster, the overlaps are very close together and it really is hard to say that they are 2 neurons. It takes a lot of data/time to be certain.
One thing to remember is that noise is a huge problem with all this. Not just the 'normal' electronic sources (how do you ground a brain well? It's intentionally evolved to be nothing but loops!), but the neuronal too. Neurons will often fire just because (or not), they will jiggle with animal movement, they will move with heartbeat, and they will die off or move on their own, the electrode dances about, the brain attacks the electrodes, some portion of the electrode snaps off, etc. It is a very hard thing to manage well.
Sorry, bad 'memories' of those projects.
One thing to remember is that noise is a huge problem with all this. Not just the 'normal' electronic sources (how do you ground a brain well? It's intentionally evolved to be nothing but loops!), but the neuronal too. Neurons will often fire just because (or not), they will jiggle with animal movement, they will move with heartbeat, and they will die off or move on their own, the electrode dances about, the brain attacks the electrodes, some portion of the electrode snaps off, etc. It is a very hard thing to manage well.
Sorry, bad 'memories' of those projects.
If only this were true.
(By "modern" I assume you mean densely-spaced [~10um] multi-pad linear silicon probes, the NeuroPixel being the most celebrated example among the non-specialist public.)
What many pads close together can give you is the potential for the waveform from a spike nearby to register on more than one pad.
In theory this should make spike-sorting easier, because you can distinguish two neurons whose spikes might have the same waveform on pad #1 but different waveforms on pad #2.
In practice, spike sorting improves, but not by as much as you'd think.
Part of this is down to physical factors (which are improving): probe geometry, pad shape, pad material, pad impedance, and so on.
Another part is down to software (which is also improving). Single-channel spike-sorting is by now probably close to as good as it's going to get given the information content of its input, and the algorithms and software to perform it are well-understood and stable.
Algorithmic approaches to multi-channel spike sorting, however, are the subject of active research with multiple promising avenues of progress, and software to perform it is ... well, charitably, let's call it "rough-and-ready." (It's nearly all lab-grown software, which means it's written by enthusiastic amateur programmers [among whom I'd count myself, no shade intended here] who soon move on to new projects because of the structure of academic science. This means the software is buggy, poorly documented, inconsistently supported, and constantly evolving.)
Now, using dense silicon arrays does markedly increase the rate at which I can record well-isolated neurons, but a significant part of this increase is just having more pads in the target brain region - many of the neurons I get from these multi-channel spike-sorting programs only show significant power on one or two pads.
And all of this doesn't even touch on some of the significant challenges that come with using multi-electrode arrays, including higher initial inflammation, later gliosis, and data-collection and storage (a 64-channel array [NB: the NeuroPixel 1 can record from 384 pads] producing 16-bit ints at 20kHz [just about the minimum sampling rate for decent spike-sorting] generates a bit less than 10GB/hour - a volume that real big data people might laugh at, but it sure isn't small!).
And finally, all of this is done off-line. On-line spike-sorting is harder.
So I'm sorry to say that just having a modern multi-electrode array absolutely does not make things "quite a lot simpler."
(By "modern" I assume you mean densely-spaced [~10um] multi-pad linear silicon probes, the NeuroPixel being the most celebrated example among the non-specialist public.)
What many pads close together can give you is the potential for the waveform from a spike nearby to register on more than one pad.
In theory this should make spike-sorting easier, because you can distinguish two neurons whose spikes might have the same waveform on pad #1 but different waveforms on pad #2.
In practice, spike sorting improves, but not by as much as you'd think.
Part of this is down to physical factors (which are improving): probe geometry, pad shape, pad material, pad impedance, and so on.
Another part is down to software (which is also improving). Single-channel spike-sorting is by now probably close to as good as it's going to get given the information content of its input, and the algorithms and software to perform it are well-understood and stable.
Algorithmic approaches to multi-channel spike sorting, however, are the subject of active research with multiple promising avenues of progress, and software to perform it is ... well, charitably, let's call it "rough-and-ready." (It's nearly all lab-grown software, which means it's written by enthusiastic amateur programmers [among whom I'd count myself, no shade intended here] who soon move on to new projects because of the structure of academic science. This means the software is buggy, poorly documented, inconsistently supported, and constantly evolving.)
Now, using dense silicon arrays does markedly increase the rate at which I can record well-isolated neurons, but a significant part of this increase is just having more pads in the target brain region - many of the neurons I get from these multi-channel spike-sorting programs only show significant power on one or two pads.
And all of this doesn't even touch on some of the significant challenges that come with using multi-electrode arrays, including higher initial inflammation, later gliosis, and data-collection and storage (a 64-channel array [NB: the NeuroPixel 1 can record from 384 pads] producing 16-bit ints at 20kHz [just about the minimum sampling rate for decent spike-sorting] generates a bit less than 10GB/hour - a volume that real big data people might laugh at, but it sure isn't small!).
And finally, all of this is done off-line. On-line spike-sorting is harder.
So I'm sorry to say that just having a modern multi-electrode array absolutely does not make things "quite a lot simpler."
Tangentially related, but I wanted to post an article and a paper.
I remember hearing way back around 2010 that Valve was messing around with BCI's. Valve has as recently as 2019 talked about EEG's and gaming [0]. Also recall reading about similar concepts over a decade ago. [1]
[0] https://venturebeat.com/2019/03/24/valve-psychologist-explor...
[1] https://vpa.sabanciuniv.edu/phpBB2/vpa_views.php?s=4&serial=...
I remember hearing way back around 2010 that Valve was messing around with BCI's. Valve has as recently as 2019 talked about EEG's and gaming [0]. Also recall reading about similar concepts over a decade ago. [1]
[0] https://venturebeat.com/2019/03/24/valve-psychologist-explor...
[1] https://vpa.sabanciuniv.edu/phpBB2/vpa_views.php?s=4&serial=...
I have an OpenBCI headset that I have been messing with. Signal is definitely noisy, but they've been able to get a degree of mind control working, like flying little helicopters.
Doesn't reconstruction like this rely on the decoding mechanism being trained on the original images? This doesn't scale to reconstructing any visual image, only to differentiating between those the decoder has already seen.
It doubt very much that it reconstructs from the visual cortext. That is a non-trivial task with fMRI which has a much higher SNR than EEG.
It doubt very much that it reconstructs from the visual cortext. That is a non-trivial task with fMRI which has a much higher SNR than EEG.
Here's their preprint.
It's basically using a pretrained VGG network as a latent space, which lets them synthesize images.
https://www.biorxiv.org/content/10.1101/787101v3
It's basically using a pretrained VGG network as a latent space, which lets them synthesize images.
https://www.biorxiv.org/content/10.1101/787101v3
This scares me a bit.
Let's say you have certain nero indicators which indicate person X may do Y .
Do you we then minority report lock up people.
At the same time , it's going to be amazing for people in a coma or unable to communicate. We could have a fantastic amazing world, where we effectively live forever via some type of matrix like interface.
Let's say you have certain nero indicators which indicate person X may do Y .
Do you we then minority report lock up people.
At the same time , it's going to be amazing for people in a coma or unable to communicate. We could have a fantastic amazing world, where we effectively live forever via some type of matrix like interface.
Worse yet, while the government may be banned from looking at this data intrusively. We're likely to freely give it away to a few large tech companies. Without limits on their use, their power will only continue to grow.
To me the primary issue of our day at this moment, is we have no users bill of rights. We are forced to use the services of large tech companies to participate in modern life yet there are few restrictions on their powers.
It's bad enough when it's our written thoughts. But the idea that it can be my unwritten thoughts too... that scares the heck out of me.
To me the primary issue of our day at this moment, is we have no users bill of rights. We are forced to use the services of large tech companies to participate in modern life yet there are few restrictions on their powers.
It's bad enough when it's our written thoughts. But the idea that it can be my unwritten thoughts too... that scares the heck out of me.
You have to just be careful who gets your data.
The less you use social media the happier you'll generally be. No matter what you say the context around it can change and your life destroyed with it.
Breaking out of the social media matrix has been one of my best decisions, I was able to really live in 2019. Made tons of friends as well as a few partners. All of this happened in this place called reality.
The less you use social media the happier you'll generally be. No matter what you say the context around it can change and your life destroyed with it.
Breaking out of the social media matrix has been one of my best decisions, I was able to really live in 2019. Made tons of friends as well as a few partners. All of this happened in this place called reality.
Imagine targeting advertising based on a certain state of mind yikes
This is why I stopped using Spotify. With all the places it shares data with, it could easily "randomly" play an inspiring song right when I'm on the fence about buying something.
Just let the machine make you happy.
Yes, they will do. People are claiming for other people to be fired, disbarred or imprisoned for stupid tweets. Can you imagine the cancel brigade and the big social media companies in charge of this technology and its consequences?
We're already there on foreign soil with Palantir's monitoring software. You don't need to read people's brains to invasively spy on them and arrest them without traditional probable cause
I think the question is the maximum possible range. If it requires something that's a few centimeters (or less) away from the skull, then a government would either have to force you / make you agree to a scan, or would have to tap data generated by existing consumer devices (like they currently often do with phones/computers).
Is it theoretically plausible that this activity could be read from many feet away, or further?
Is it theoretically plausible that this activity could be read from many feet away, or further?
Eh, we'll have software for detecting and correcting criminal intent before the government gets its act together enough to start arresting people preemptively.
Aren't those things equivalent?
Heh. Remember back in the day when pacemakers were being hacked remotely because admin passwords weren't even changed from the default, on top of outdated wireless protocols? Oh right that's today!
Hardware makers often aren't good at software, not to mention software updates. And even that gets wonky, like when Microsoft update got hijacked.
Point is, if we can't get IoT 100% right, or even 90% right, how can we trust IoT with physical interfaces into our bodies? That's the problem. And then what happens if the company who made your implant goes out of business? What do you do when those updates stop? Look at cellphones, supporting a cellphone for 2 years is too much for most hardware makers, they rather never update it.
Point is, even if the tech is 100% possible, we're way too far from business setups that allow for this to happen.
Hardware makers often aren't good at software, not to mention software updates. And even that gets wonky, like when Microsoft update got hijacked.
Point is, if we can't get IoT 100% right, or even 90% right, how can we trust IoT with physical interfaces into our bodies? That's the problem. And then what happens if the company who made your implant goes out of business? What do you do when those updates stop? Look at cellphones, supporting a cellphone for 2 years is too much for most hardware makers, they rather never update it.
Point is, even if the tech is 100% possible, we're way too far from business setups that allow for this to happen.
These are all valid worries, but they all have the same fundamental solutions as current products.
> if we can't get IoT 100% right, or even 90% right, how can we trust IoT with physical interfaces into our bodies?
Trust is a risk-reward calculus for any product, whether it's BCI, the microcontrollers in your car, or the (hopefully uninfected) produce in your supermarket. Many folks will find the BCI value outweighs the risks for some feature set that matters to them.
> And then what happens if the company who made your implant goes out of business? What do you do when those updates stop?
Certain applications will have to be designed and evaluated with longevity or long-term support in mind. You don't need that when talking about cellphones. You do for today's pacemakers.
> if we can't get IoT 100% right, or even 90% right, how can we trust IoT with physical interfaces into our bodies?
Trust is a risk-reward calculus for any product, whether it's BCI, the microcontrollers in your car, or the (hopefully uninfected) produce in your supermarket. Many folks will find the BCI value outweighs the risks for some feature set that matters to them.
> And then what happens if the company who made your implant goes out of business? What do you do when those updates stop?
Certain applications will have to be designed and evaluated with longevity or long-term support in mind. You don't need that when talking about cellphones. You do for today's pacemakers.
Like everything else, it will be imperfect, there will be scandals, the occasional lawsuit, and we'll muddle through.
Software scandals don't end up with people dead or extorted at gun point. Yet.
Technically that has happened: https://www.bugsnag.com/blog/bug-day-race-condition-therac-2...
But you're right, software bugs don't generally result in lives lost, at least directly.
But you're right, software bugs don't generally result in lives lost, at least directly.
So, everyone needs to go back and watch the Ghost in the Shell: Stand Alone Complex series. In that world, pretty much everyone has a few enhancements, visual and audio overlay options are pretty much universal. Having your equipment hacked is a super rare, state-actor-level event. But, the story is focused on people who deal with events of that magnitude, so it happens all the time in the plotline.
> not to mention software updates.
Updates of any sort to the SW are generally very difficult to do. The certification process for any medical device is hard enough, let alone for implants, let alone for life-critical implants, let alone the recertification process for an already implanted life-critical device. This is why you'll see sonogram machines running XP still, completely disconnected from the larger internet [0].
The questions listed are mostly already considered and have many mitigation strategies per the regulatory agency in charge. There are many other questions that people like the FDA demand answers to.
One thing about the business side is that risk/reward ratio. In medical development, FDA authorization typically occurs at the 10-12 year mark for a product (though it varies widely). Meaning that your start-up only gets the go-ahead to make money after about a decade of investment. That said, once you get to sell, you have an effective monopoly on the market. Hence the costs of new drugs and devices being so insanely high; it's that risk/reward imbalance on the business side.
[0] Another reason why IoT medicine is very difficult, among many.
Updates of any sort to the SW are generally very difficult to do. The certification process for any medical device is hard enough, let alone for implants, let alone for life-critical implants, let alone the recertification process for an already implanted life-critical device. This is why you'll see sonogram machines running XP still, completely disconnected from the larger internet [0].
The questions listed are mostly already considered and have many mitigation strategies per the regulatory agency in charge. There are many other questions that people like the FDA demand answers to.
One thing about the business side is that risk/reward ratio. In medical development, FDA authorization typically occurs at the 10-12 year mark for a product (though it varies widely). Meaning that your start-up only gets the go-ahead to make money after about a decade of investment. That said, once you get to sell, you have an effective monopoly on the market. Hence the costs of new drugs and devices being so insanely high; it's that risk/reward imbalance on the business side.
[0] Another reason why IoT medicine is very difficult, among many.
The trick is I never expect anything to be 100% right, or even 90% right. Even if they promise to be 100% right.
Either accept that it can broke or always have a backup plan.
>then what happens if the company who made your implant goes out of business?
The company goes out of business doesn't always mean the product stop working anymore.
If the product stop working then I'll look for removal/replacement.
Either accept that it can broke or always have a backup plan.
>then what happens if the company who made your implant goes out of business?
The company goes out of business doesn't always mean the product stop working anymore.
If the product stop working then I'll look for removal/replacement.
And that's not even getting into the privacy aspect. I, for one, reserve no excitement whatsoever for a world where we can interface directly with computers. I'm just hoping that unlike smartphones, it remains viably optional.
Even worse, what if you have artificial eyes and one day the manufacturer blocks your vision functionality because you did something they disagree with politically? It used to be your account would get banned, now with cancel culture you can be banned from participating in media and society altogether, tomorrow maybe your ability to participate in your own body's functionality may be subject to the whims of the censors and the morality police
That seems stretch. People are getting accounts cancelled on online services they use, where the companies feel they may be morally or legally complicit in the message, or their employees, who decide they don't want to employ that person. I don't think most "cancelled" people are worried about their grocery stores not accepting them, or any number of other services where they aren't a platform used to further the behavior people are upset about.
If you got eyes that required the internet to function you made a poor and dangerous choice regardless of whether the public or some company wants to punish you for some perceived public behavior. If you got eyes that required some company on the internet's continued acceptance of you as a client, you made an even worse choice.
If you got eyes that required the internet to function you made a poor and dangerous choice regardless of whether the public or some company wants to punish you for some perceived public behavior. If you got eyes that required some company on the internet's continued acceptance of you as a client, you made an even worse choice.
Anyone who signs away to a corporation their right to see deserves whatever results from such an arrangement.
I don't recall where I first heard this, but I think it's true:
The theme of CES for the past decade has been making products worse through software.
The theme of CES for the past decade has been making products worse through software.
I'm not a huge anime person, but there's one I've watched that seems a bit relevant. In _Accel World_, they wear this thing on their neck called a "Neurolink" and basically are able to connect to things like the Internet or private networks (school, home, hospital) wirelessly with it. I believe it takes over signals coming to and from the brain. But they also have a wired connection mode so they can do that or do a special direct, wired connection with someone they are close to.
The ramifications of access like that are described surprisingly well in the show.
For instance, two close friends (they're in a relationship but young enough I'd just consider them close friends) use the direct wired connection feature because they feel comfortable enough doing so. But at some point one of them decides to use that private access to plant a virus of sorts. While the purpose is irrelevant here, it does show how brain connections, especially between people, can become increasingly vulnerable.
The main plot point of the anime though, is this game application that can be installed to the Neurolink, and basically it has the ability to also wipe away or hide memories.
While what Newell is talking about is still a ways away, it's interesting to consider exactly what we may need to be worried about when the time comes.
The ramifications of access like that are described surprisingly well in the show.
For instance, two close friends (they're in a relationship but young enough I'd just consider them close friends) use the direct wired connection feature because they feel comfortable enough doing so. But at some point one of them decides to use that private access to plant a virus of sorts. While the purpose is irrelevant here, it does show how brain connections, especially between people, can become increasingly vulnerable.
The main plot point of the anime though, is this game application that can be installed to the Neurolink, and basically it has the ability to also wipe away or hide memories.
While what Newell is talking about is still a ways away, it's interesting to consider exactly what we may need to be worried about when the time comes.
There's a better version of that in Ghost in the Shell SAC. You could have a cyber brain and connect to networks which could include websites, devices, other people, and even have a direct wired connection too.
I do a significant amount of breakthrough device fda/eu regulatory work in the BCI area. For those interested, its worth reading the FDA guidance: https://www.fda.gov/regulatory-information/search-fda-guidan....
Regarding security, I think people would be really freaked out by the scope of attacks. One infusion pump I worked on this year was succesfully attacked over 200 times.
That sounds interesting, is this a case of 200 vulnerabilities being found or 200 documented cases of attacks against (production?) devices?
big fan of his point that the technology is moving so fast that it doesn't make sense to lock-in and go mass market yet
This sounds like the spaceship problem. Whenever you build a spaceship to take people to another planet, it will be passed by a later ship which was built to go faster.
Reminds me of my games of Civilization. Should I spend few more turns to build additional engines to my spaceship, or should I launch it now and risk being overtaken by a faster ship.
In real life its so much more complicated than that. Because choosing to "send ship now" will result in huge amounts of progress in overcoming obstacles you could not have known about until you actual began the project.
It seems like the answer is always send out a test ship as fast as possible, knowing that it will absolutely be overtaken by a ship with more engines AND the knowledge gain from the first ship.
It seems like the answer is always send out a test ship as fast as possible, knowing that it will absolutely be overtaken by a ship with more engines AND the knowledge gain from the first ship.
I don't think that's quite right - it's more that they're a lab and not a product manufacturer, so they can't decide to go mass-market.
Even if they decided that they want to go public with a virtual keyboard/mouse/controller and virtual heads-up display that you use by wearing an electrode net, the current team would not be able to make that pivot. End users won't be debugging Labview sketches, spending weeks training a neural network to recognize virtual keystrokes, or shaving spots on their scalp.
Personally, I expect that the V1 product here is approximately just that: a game controller. A few buttons, a pointer, maybe a little haptic feedback. It would be great for me if it supported text input and output faster than a keyboard and terminal, but I don't think that's likely to happen in an early version.
Even less likely is that we're going to jump straight to The Matrix. Valve needs to admit that, and be happy with a limited version, instead of letting it fizzle out in the lab.
Even if they decided that they want to go public with a virtual keyboard/mouse/controller and virtual heads-up display that you use by wearing an electrode net, the current team would not be able to make that pivot. End users won't be debugging Labview sketches, spending weeks training a neural network to recognize virtual keystrokes, or shaving spots on their scalp.
Personally, I expect that the V1 product here is approximately just that: a game controller. A few buttons, a pointer, maybe a little haptic feedback. It would be great for me if it supported text input and output faster than a keyboard and terminal, but I don't think that's likely to happen in an early version.
Even less likely is that we're going to jump straight to The Matrix. Valve needs to admit that, and be happy with a limited version, instead of letting it fizzle out in the lab.
One of the things I wish the most technologically is simply being able to record my thoughts. Sometimes in the middle of a stream of thought going back and putting it on paper, taking notes disrupts the process somewhat. Also obviously translating visual imagination and recollection is more difficult (more for someone that wasn't trained artistically). My memory is imperfect too and I've forgotten a few ideas I thought were amazing because I didn't write them down at the right time.
I, for one, welcome the mind-reading devices.
I'm sure it'll be limited by physical contact so I'm not much worried about some science fiction scenarios. However, the implications for police are indeed worrisome, enough it should warrant specific legislation as soon as possible (i.e. non-consent mind reading should be made a grave crime, and a war crime, as soon as the technology is available)
I, for one, welcome the mind-reading devices.
I'm sure it'll be limited by physical contact so I'm not much worried about some science fiction scenarios. However, the implications for police are indeed worrisome, enough it should warrant specific legislation as soon as possible (i.e. non-consent mind reading should be made a grave crime, and a war crime, as soon as the technology is available)
I thought that it was interesting that Gabe was using The Matrix as a comparison for BCIs since that movie was a warning about what this kind of tech could bring, but he did express a lot of concern around more real-world ramifications around what this technology could bring if it wasn't properly used or secured. What he was describing then reminded me more of the movie Strange Days where people would use neural connectors to record and play back experiences from others, but amplifying the signal too much would result in scrambled neurons.
Also reminds of "Better Than Life" from a cheesy, teen-sci-fi Shadowrun novel. The story starts out assuming Strange Days-style neural recording and playback is a commonplace consumer good. But, then quickly gets into highly edited experiences, generated experiences, and eventually into overwhelming, drug-like experiences.
> the experiences that you will have are things that you will be created and edited for you [..]
> sleep will now become an app that you will run where you say I need this much sleep, I need this much REM
I find this to go in a bizarre awkward direction. We as humans have self regulating capabilities that are to be tapped with some "self work". Is that a bad thing? Why are we trying to change that and create an app to control bodily functions? Our brain is perfectly capable of handling that and has been for millions of years in our ancestors
I find this to go in a bizarre awkward direction. We as humans have self regulating capabilities that are to be tapped with some "self work". Is that a bad thing? Why are we trying to change that and create an app to control bodily functions? Our brain is perfectly capable of handling that and has been for millions of years in our ancestors
Please tell me in great detail how I can go to sleep within 15 minuten of laying in bed. I'm very open to self work.
You need to learn to wind down. You can't expect to use your brain intensely and then shut off. But you can start by lowering the light, relaxing, stop using electronic devices, etc. (though listening to something unexciting does induce sleep). I also induce sleep by forced repeated yawing. Every couple of forced yawns trigger a real one, and with that some sleepiness. Another exercise I do is to imagine I lift my arms/legs up and let them fall by my side, helps me to relax my nervous system. I've done some meditation in the past and it helped as well but I don't actively do it before bed but you can absolutely try. Another thing that may or may not work for you is white/pink/brown noise. I prefer quiet if I could get it, sometimes that is impossible, e.g. having a window facing car traffic so in that case white noise can reduce the intensity of sudden sounds. So there is no formula but a bunch of habits really do help.
My wife has sever insomnia and is having more success with this than she did with prescription sleep aids.
https://choosemuse.com/
Being able to run "sleep" like its an app on your phone and directly control your brain would be life changing for us. I really hope it happens in our lifetime.
https://choosemuse.com/
Being able to run "sleep" like its an app on your phone and directly control your brain would be life changing for us. I really hope it happens in our lifetime.
I think VR could use non-fresnel lenses and better ways to adapt your vision in the headset, ironically the DK1&2 had lenscups which now are found in zero headsets.
I keep my old glasses that are smaller to be able to fit them in my vive, but that makes development inconcievable: Remove terminal/monitor glasses and put on old glasses and then the headset every time I need to test something in VR. Nope I'm not doing that.
Also why are the media business still adding fake laughter and annoying background music to everything they touch?
I keep my old glasses that are smaller to be able to fit them in my vive, but that makes development inconcievable: Remove terminal/monitor glasses and put on old glasses and then the headset every time I need to test something in VR. Nope I'm not doing that.
Also why are the media business still adding fake laughter and annoying background music to everything they touch?
I found out last week that there are prescription inserts for headsets. I’ve yet to get mine, but as someone who wears glasses the concept sounds awesome. https://vroptician.com/prescription-lens-inserts/valve-index...
Sure, but on the DK 1&2 I made my own lens corrections in the cups, much cheaper and future proof (what I thought, since I can move the lenses to newer headsets) and then Valve and everyone else listened to the wrong lens guy and now everyone has to buy single serving solutions for the rest of existence + get aweful god rays!?
I went ahead and ordered them later that day. I had seen the custom options, but since I don't have a 3D printer/have awful eyesight, I wasn't sure it was worth the hassle of trial and error/poor fitting lenses. I am under the assumption that they will not have an effect on the visual quality, but I can update you if you are interested. I never tried the DK1/2, I've only really put time into the Vive and Index.
I think it would be neat if Valve made it so you can more easily pop the lenses in and out, allowing you to pick what you need prescription wise (and in your case fresnel vs non-fresnel). I am unsure how well that would work, since the focal distance between headsets could change.
I think it would be neat if Valve made it so you can more easily pop the lenses in and out, allowing you to pick what you need prescription wise (and in your case fresnel vs non-fresnel). I am unsure how well that would work, since the focal distance between headsets could change.
HP is getting into basic (compared to GabeN's ideas) work in this area with the Omnicept VR headset. It looks like eye tracking, pupillometry sensor, heart rate, and a face camera. Not direct neurotechnology but none the less using data to better understand and adapt to the user's response to VR.
https://www8.hp.com/us/en/vr/reverb-g2-vr-headset-omnicept-e...
https://www8.hp.com/us/en/vr/reverb-g2-vr-headset-omnicept-e...
The link below shows up on occasion and is a good reminder that not only do we need to keep up with physical/practical trust but philosophical as well.
https://marshallbrain.com/manna
(It explores implants and using them vs. being used by them.)
https://marshallbrain.com/manna
(It explores implants and using them vs. being used by them.)
This is the kind of technology that needs ironclad opt-in privacy.
Valve collects lots of data - that you can't turn off - and doesn't let you opt out.
(see if you can ask them to stop collecting time played in games or "achievements" or whatever)
Valve collects lots of data - that you can't turn off - and doesn't let you opt out.
(see if you can ask them to stop collecting time played in games or "achievements" or whatever)
this is a very appropriate thing for Valve to be working on considering their logo
The logo may not be a coincidence, but more an expression of Gabe Newell's life focus.
That looks like an awful over complicated way of performing a lobotomy.
Imagine getting a cookie popup every time you look at something.
https://borderlands.fandom.com/wiki/Adbug
"While technically hostile, Adbugs do not directly attack Vault Hunters. Instead, they follow their targeted Vault Hunter around and beam a semi-transparent advertisement at them, which is placed in a random position on the field of vision, obscuring part of the field of view."
"While technically hostile, Adbugs do not directly attack Vault Hunters. Instead, they follow their targeted Vault Hunter around and beam a semi-transparent advertisement at them, which is placed in a random position on the field of vision, obscuring part of the field of view."
It's pretty funny to me that the photo of Mr. Ambinder is now the best evidence I know that Valve is still working on TF2
Whoever added that background soundtrack should be shot.
Yeah I thought it was interfering with the interview too.
It will be horrible when you suddenly can't afford to subscribe your brain extensions. It will be like severe brain damage.
Imagine living only half-life.
Imagine living only half-life.
That's a recurring thing in Deus Ex: Human Revolution. Outside of the protagonist, most people need a costly antirejection drug for the augmentations that they're societally pressured to have. Because they're otherwise uncompetitive in their work.
Reduced to a life of just using a crowbar and not talking t anyone.
But all jokes aside, it could be brain damage, but drug addiction might be a better analogy. It could be something that takes time for neuroplasticity to repair or correct.
But all jokes aside, it could be brain damage, but drug addiction might be a better analogy. It could be something that takes time for neuroplasticity to repair or correct.
Happy to answer any questions people have about OpenBCI https://openbci.com/