Neurotech - humanity’s next frontier by Juan Benet | Devcon SEA

[Music] y okay perfect right hey everybody uh very excited to be here I'm going to move really fast uh I'm going to talk about Humanity's next Frontier which I believe to be uh neurotech uh I'm going to cover these three things in seven minutes uh so please pay attention to the slides not to me because your visual cortex can absorb way more information um that way so I'm going to leverage a lot of imagery uh to communicate a lot I'm going to talk about three things neurotech to continue Humanities Evolution neurotech as defensive acceleration and the rough road map for neurotech neurotech can be a a solution to mitigating and curing many diseas diseases and disorders of various sorts both neurological things and many other kind of um nerve damage type issues it can also lead to better self-regulation of our minds of our bodies and of our moods and um of our decisions so we by having the ability to upgrade our brains we'll be able to do everything we do uh hopefully better uh we can also do things like restoring Mobility restoring senses like sight and senses like hearing uh this was the realm of sci-fi in the 20th century it is happening today and many people are already the beneficiaries of these kinds of Technologies you will see this ramping up over the next 5 to 10 years now after repair uh we can start enhancing our senses um it's just an accidental fact of evolution U that we have the particular senses that we have or the particular capabilities that we have now that we have uh certain Technologies we can now start enhancing our capabilities and being able to see um a wider Spectrum being able to hear wider spectrum and so on we can also start using these systems to enhance our thought we can level up how we ourselves um think we can also improve the connectivity between ourselves so think of uh the ability to actually have something like telepathy this is this was science fiction in the 20th century it will become science fact in the 21st century uh the Corpus colossum is the way that the two loes of your brain communicate to each other they have I think hundreds of millions of connections something on the order of that or hundreds of millions of neurons going through maybe that connections there's many other it higher um but that's that's how you maintain a sense of self uh you can imagine that um type of connectivity being able to extend to other humans and eventually to the rest of humanity you could actually get um telepathy to scale uh you also have better connectivity to computers this means uh you can think of augmenting your um your current neocortex which is itself a brain on top of other brains right like our evolution in the brain is a a series of stacked brains that we evolved over uh hundreds of millions of years uh and you can think of now adding um a next layer which people commonly call the exocortex of being able to now have other devices um associated with your your brain then that could lead to connectivity with the internet more broadly being able to do things like recall the information that's in the internet being able to learn extremely quickly being able to actuate um other uh computers in the internet and so on but the most important thing and why uh we're here today in terms of DC day is connectivity to Ai and I'm going to make the case that this is absolutely critical from a um Humanity survival uh level scale uh look the the future is extremely different than the past uh normally Humanity doesn't like to think about this um we like our normal day-to-day experience and we believe this like wonderful fiction that the world was just always like this and always will be uh this is just not the case all of the evidence that we have about our past shows that um the entire universe was extremely different billions of years ago um the planet was very different uh billions and millions of years ago and our own species evolved um through various different paths and various different time frames even ourselves after achieving this level of biological scale uh changed a dramatic amount over the last um tens to hundreds of thousands of years ago or even millions of years ago so who you are today who uh humanity is today is radically different than the humanity of a thousand years ago 10,000 years ago 100,000 years ago so this is a totally different um um uh evolving landscape and so we now need to look ahead into 100 a thousand 10,000 years into the future and think about where we're headed so this is the hard part this is what I consider to be the most important graph uh for Humanity today it's a plot of the Computing scale uh over time uh and the horizontal lines are different um guesses as to what you would take computationally to be able to simulate a whole brain we're entering what I call the fund Zone where we're finally um for many models we should be able to um do actually model a human brain sometime in the in the next 10 to 20 years with that level of computational capacity um if you're much more conservative than that and think uh there's maybe many other uh lower level uh Atomic or or other or Quantum effects um it might take another 20 years but this is roughly roughly the timeline this also suggests if you can do this with the to the human brain which is you know um neurons are quite big we're talking about Micron scale if you can do this in neurons you can certainly do it in artificial neurons Nets which are orders of magnitude smaller and faster meaning there's say an advantage for digital systems um that just operate dramatically faster and at smaller scale meaning that if you can do this to to a human brain in this time scale you can certainly create an AGI um much sooner than that order some magnitude uh sooner in terms of computing complexity so that means many years uh before we've seen AI seep into things we've seen the field accelerate um and now this is the main uh problem space uh we've seen capabilities um human capabilities fall to AI uh there's like this great graph from our world in data that shows how um we can build models and go very quickly from completely abysm performance to superhuman capability in a matter of a year or a few years um and so you know whenever you hear somebody say oh a AIS will never do that just pull up this graph and just start pondering at what point in time uh will that actually cease to be the case so just do not believe disbelieve um the the question of oh AIS are just never going to be able to do what humans can do just remember that you know we things that dinosaurs couldn't do um so just have some humility as a species um the top right uh graph is maybe one of the most optimistic Landscapes um a lot of the people deep in the AI field uh think that that's roughly what's going to happen in 2027 2028 we'll start getting a recursive self-improvement and then we'll go from AGI to ASI that's very um aggressive that I hold a large probability distribution there but I think it may take longer it may take quite a a lot longer so we have some window of time uh to upgrade ourselves uh to make matters more complicated we now have a whole bunch of Robotics appearing so these models even though you know they might not be as smart as us are going to be able to actuate uh bodies um like ours and bodies that are uh just different than ours in a bunch of regards so we have both humanid humanoids coming in the pipeline and then other things like drones and whatnot so we're really um headed for a very very difficult Choice um I like that story of comparing homerus to homos sapiens the neanthal and nans homoerectus was a hominid from um millions of years ago uh one to two million years ago that expanded Out of Africa the same way that we did and basically colonized a lot of the world uh but went extinct um millions of years later uh hundreds of thousands of years ago Homo sapiens neanthal and isans each had their own um uh expansion Out of Africa then eventually ended up merging and upgraded into the species that we have today so um most of us carry some amount of neth of and DNA um and we have this combined species with the bulk of it being Homo sapiens so this is the the trade of that we have we can go the way of homus or we can start merging um and improving ourselves and so really I think of it this way we have this digital Evolution pathway that we're building and either we're going to make a another digital species that will replace us um if we uh insist on remaining uh entirely biological and and refuse to enhance ourselves or we can take the path of um bootstrapping with neurotech and evolving uh ourselves in our future uh now I'll touch super briefly on the road map there's bcis and you'll see this um you're starting to see this develop um this will enhance over the next um 5 to 10 years whole bra emulation is the next Frontier after that uh I really think these two things are extremely critical to survive uh this is roughly the the road map that I expect in hurn emulation this might be too late by many accounts um and so if you want to dramatically decrease uh AI risk um I suggest you join me in accelerating the whole bra emulation road map and and also the you know BCI road map we could be doing something kind of like this potentially even tighter um and so that's both possible and and um and I think imperative uh we need breakthroughs in all kinds of tools and Technologies um and we really need a a um an R&D environment similar to the one that we just saw for AI development or for many other Technologies uh this is roughly the capital scale that we're talking about um which means you can usually um Orient it in terms of um you know philanthropy and Angel Investors initially then VC then tech companies because the the scaling the tens of billions of scale is just so expensive that just VC cannot afford this so either you start generating a cash flow to grow or you have to tap the tech companies themselves um so I think if you um want to defensively accelerate uh I strongly encourage you to support uh neurotech either by working directly in it by funding it um and by uh spreading the word uh I think a government isn't going to do it uh for a lot of reasons and I do think that we need a range of field building to happen um huge plug for for a lot of organizations that have been doing this um and especially um I think today we think that prizes and fast grants can greatly Accel accelerate the field you'll hear from other folks in neurotech about um a lot of this stuff in more detail um but specifically a way that the crypto Community can dramatically help here um is by supporting uh the field as a whole through prices and grants all right thank you very much thanks thank you Juan that was amazing