7 ms·
Will This “Neural Lace” Brain Implant Help Us Compete with AI?
- AndrewKemendo 10y agoI don't anticipate a practical IO bci until after AGI is done. Why? We don't understand the brain's learning mechanism well enough yet and once we do that will be recipe for AGI which will by default see faster growth than individual to individual adoption of these bci. Not only that but a recent poll indicated that the majority of those polled would not get a bci to improve their cognitive ability. The best way to bridge the human/AGI gap in the beginning imo is with decision based integrated AR. It's not even really a bridge it's a symbiosis.
- goldfeld 10y agoWherein is the basis to assume AGI is even possible? Isn't a fundamental faith that consciousness and insight is replicable a same blind faith that there is a God, for example? How is one more arbitrary than the other?
- binalpatel 10y agoWe have proof that consciousness and insight are possible in nature - mainly us. We replicate it with each generation, and watch it grow as children age.
- tfgg 10y ago> consciousness and insight Being unmeasurable and undefined, and therefore irrelevant to measures of intelligence, "consciousness" is typically not considered part of AGI [1]. Not sure what you mean by "insight". [1] http://arxiv.org/abs/0712.3329 http://arxiv.org/abs/0712.3329
- mikeash 10y agoSeems like the exact opposite of faith in God to me. If you accept the supernatural, then there is the possibility that humanity has some "spark" which cannot be replicated artificially. But without the supernatural, humans are just matter organized a certain way, and as such demonstrate that intelligence can be constructed out of matter.
- Joof 10y agoIf it's all traditional physics, we can eventually just simulate the whole structure in the worst case where we have no idea how it works.
- goldfeld 10y agoBelieving "it's all traditional physics" has no basis, the only evidence from the past is that models keep getting overturned in counterintuitive directions, and current ones have been long at odds. Has the world raised a generation arrogant as to blindly accept we know it all, and trade away scientific inquiry for a little scientific rigour? Science and philosophy are names for different phases of a same process for reaching understanding. Shouldn't we at this confused point in time go back to the drawing board of philoshophy before piling on hopes in search of a holy grail of superhuman intelligence? What does human intelligence even mean and how is it limited, what are the constraints applicable? How does the perceptive, probing imagination and profound insight of someone like Einstein or Tesla work? Can machine learning create a theoretical physicist? We don't know.
- Joof 10y agoI'm just saying that it's very likely possible based on what we know. Not that we know how. Even if it's based on quantum physics, it only means that we can't simply copy a brain and will have to come up with our own system. I worked in an AI / Brain lab for some time. I see no reason it can't be done, but I don't expect it to happen for a while. If it isn't possible to build a brain, then how do we ourselves exist? Time and research will eventually be able to do it.
- duaneb 10y agoWhat's the end goal of this? How is this making humans happier? How is this improving quality of life? Competing with AI is an artificial competition. Quality of life is a) a real problem that isn't going away and b) is directly contrary to AR. Think emails are a problem now? It's not getting any better with inescapable interaction brought on by AR.
- AndrewKemendo 10y agoEnd goal is to merge with the AGI. AR gets people acclimated with human machine symbiosis as an optimal way to live.
- duaneb 10y agoI don't see how constant direct contact with a computer is going to improve anyone's life. Seems deeply psychologically unhealthy.
- AndrewKemendo 10y agoThink about it like decision support. Computers can make consistently better and more rational decisions, especially as they get feedback (reinforcement learning) giving you this data as you go through life should make things easier. It already has in most areas of life, it's just not persistent yet. Think more practical applications, like cost calculations, route planning, meal prep etc... not social networking and gossip.
- duaneb 10y agoThis makes sense, but why can't you just do that today with a phone?
- vorotato 10y agoThis seems like a conclusion made with absolutely no evidence or reasoning.
- dominotw 10y ago>We don't understand the brain's learning mechanism well enough yet and once we do that will be recipe for AGI Is there such a thing as 'ununderstandable' or do we have the ability to understand anything with enough effort and time.
- AndrewKemendo 10y agoIt's a spectrum in my estimation between formal proof and being able to manipulate actions in light of some understanding. We could "control" fire before we understood thermodynamics, similarly flight and aerodynamics. So there is a threshold for making breakthroughs that don't require quark level precise predictions. We aren't even at rubbing sticks together phase when it comes to understanding the mechanisms of intelligence.
- gnaritas 10y ago> We don't understand the brain's learning mechanism well enough yet and once we do that will be recipe for AGI I don't think we have to understand it well enough for AGI to interface with it because it will adapt to the inputs whether we fully understand them or not as I think we've already seen with hearing implants. We can utilize the brains extraordinary ability to find signal in noise without understanding how it does it.
- Symmetry 10y agoI'd agree that "I know kung fu" style memory alteration is probably off the table for the foreseeable future but we've already got people manipulating devices just via sensing caps. The implications of this for prosthetics seem like fairly simple extensions of known techniques
- AndrewKemendo 10y agoNot really. Non invasive bci through fmri is basically about mapping activation function in certain regions to some other computing function. The actual computer actuation is only as coupled with the activation as you can get from related measurement after giving the user a prompt. Oh and by the way it's not IO it's just one way. tDCS on the flip side is a very unknown, poorly targeted way to get some activation but only to certain levies and nobody really knows how or if it works consistently. The research jury is split so far. Invasive Deep brain stimulation is definitely better, however it's still solely used to create general simulation in order to prevent tremors and other maladies. The only thing I've seen recently that gives me some excitement from an input protective was work recently done by Ramirez and Liu on implanting memory into a mouse brain. That was a single study though and was very very narrow.
- cinquemb 10y ago>Not really. Non invasive bci through fmri is basically about mapping activation function in certain regions to some other computing function. The actual computer actuation is only as coupled with the activation as you can get from related measurement after giving the user a prompt. Oh and by the way it's not IO it's just one way. I think the better (spacial/temporal) we can understand how stimulus looks like from a electromagnetic perspective for a given brain, we can make some good estimates on what the em field would look like some time t, afterwords. I've done something like this with eeg using beamforming techniques with a sufficient number of leads and comparing brain states with pre and post fmri of same tasks while providing feedback visually or auditory every 1-2 seconds (which is a really long time). I also think this could actually get us pretty far (concerning decision making and task allocation) if we could record in environments not constrained to the lab and factor in the variables to sufficient degree that confound eeg typically by including gyroscopes/accelerometers/thermometers/drift in impedance (factoring out fluctuations in em field due to temperature/conductivity of scalp via sweat/head movements), or even environments that are pretty much stationary like gaming (also where the inputs are more observable by time-stamping screen with controller inputs to sensors on the brain). >Invasive Deep brain stimulation is definitely better, however it's still solely used to create general simulation in order to prevent tremors and other maladies. Saw Prof. Arnd Pralle, University at Buffalo[0], give a talk at MIT in May on his work on tetherless magnetothermal deep brain neuro-stimulation to modulate motor behavior in awake, unrestrained mice. Was pretty cool and got me thinking the same could be done with designed photonic thermal systems. I wish this field had as much money going into it as the folks in the compsci depts, esp with focusing on hardware/tooling for getting measurements vs all else going on in the neurofield. [0] http://www.acsu.buffalo.edu/~apralle/ http://www.acsu.buffalo.edu/~apralle/
- ythl 10y agoAI is still having trouble competing with us in terms of general intelligence and creative ability. I will be very surprised if AI can ever generate a movie, song, video game, or book completely on its own that can actually compete with human created entertainment.
- booleandilemma 10y agoYour comment reminds me of a scene in the movie iRobot, when Will Smith's character asks a robot if he can write a symphony or paint a masterpiece. The robot replies "Can you?"
- TrevorJ 10y agoHere's the state of things as far as AI-written films: https://www.youtube.com/watch?v=LY7x2Ihqjmc&ab_channel=ArsTechnicaVideos https://www.youtube.com/watch?v=LY7x2Ihqjmc&ab_channel=ArsTe...
- ythl 10y agoGenerating a screenplay from a bunch of existing screenplays is about as good as generating a poem from a bunch of existing poems. I think there was a story the other day showcasing AI-generated poems, and all of them pretty much sucked. They might be able to pass as "modern art", but they were a far cry from poems like "O Captain, My Captain" or "The Highwayman".
- readams 10y agoGenerating poems from a bunch of existing poems is pretty much what humans do also. Art is very derivative, and necessarily so.
- binalpatel 10y agoHumans do exactly that though - an artist isn't born in a vacuum.
- 10y ago
- pault 10y agoCan someone more knowledgeable explain the practical applications of this (beyond therapy)? Is it read-only or can it be used to stimulate different areas of the brain to create artificial sensory input? Can we have the matrix now?
- Joof 10y agoMy guess? No, they can't really explain. Measurements and stuff like deep brain stimulation maybe, but it's hard to say what the ramifications of this will be.
- internaut 10y agoNot sure about the Matrix but DARPA is a big fan of Ghost in the Shell's cyberbrains: http://www.darpa.mil/news-events/2015-01-19 http://www.darpa.mil/news-events/2015-01-19
- eli_gottlieb 10y agoMy God, that poor scientist had to indulge so much scifi stuff just to publicize his biocompatible materials-science work.
- JackFr 10y agoBetteridge's Law https://en.wikipedia.org/wiki/Betteridge%27s_law_of_headlines https://en.wikipedia.org/wiki/Betteridge%27s_law_of_headline...
- aminorex 10y agoThe notion of "competing" with AI seems fundamentally broken.