Elon Musk's Neuralink May Offer us Symbiosis With AI - Part 2

this is the place where robots meet look  you can see them here as slaves to logic and this man on the hill comes to  free them do you know who he is so i want to emphasize the the purpose of  neuralink like what do we what's our goal   our goal is to solve important spine and brain  problems with a seamlessly implanted device   so you want to have a device that you can  basically put in your head and feel and   look totally normal but it solves some important  problem in your brain or spine and the reality   is that almost everyone over time we'll develop  brain and spine problems these range from minor to   very severe but if you live long enough everyone's  going to basically have some kind of neurological   disorder from memory loss to brain damage but  the thing that's important to appreciate is that   an implantable device can actually solve these  problems i think a lot of people don't quite   realize that but all of your senses your sight  hearing feeling pain these are all electrical   signals sent by neurons to your brain and if  you can correct these signals you can solve   everything from memory loss blindness paralysis  depression insomnia extreme pain seizures   anxiety addiction strokes brain damage these can  all be solved with an implantable neural link in the short term neurolengths bmi may be used  to fix neurological problems and disorders   as elon musk has pointed out over time virtually  everyone who gets old will suffer at least one   if not multiple common neurological issues  such as memory loss hearing loss seizures   strokes and brain damage etc with  the development of neurolinks device   these problems may be a thing of the past better  yet the integration of neuralink's device with   the human brain may advertently solve the  artificial intelligence alignment problem   by achieving a symbiotic relationship between  humans and machines this is because there are many   cases where an ai and a biological intelligence  could benefit from each other's actions the ai   receiving data from the human brain and the human  brain receiving data from the ai the benefit of   this relationship would greatly outweigh the  costs to both humans and ai systems however it   is also very likely that ai systems and biological  intelligences will at some point be in conflict   i think the the first bit of advice would be to  really pay close attention to the development of   artificial intelligence we need to just be very  careful in how we adopt artificial intelligence   and to make sure that researchers don't get  carried away because sometimes what happens   is a scientist can get so engrossed in their work  they don't necessarily realize the ramifications   of what they're doing so i think it's important  for public safety that we you know governments   keep a close eye on artificial intelligence  and make sure that it does not represent a   danger to the public elon musk has commented  on the dangers of ai saying it is the greatest   risk we face as a civilization however in order  to prove that neurolink can solve this problem   two things will need to become clear how does  neuralink achieve symbiosis with the human brain   and what are the side effects and  potential drawbacks of this symbiosis   one possible downside would be that humans  would lose their sense of individuality   if we all interconnect our thoughts and ideas into  the cloud however if we can continue to retain our   individuality and use neuralink in a productive  way the benefits far outweigh the potential risks   worth mentioning is neurolenk's recent development  of a device which connects people's brainwaves to   computers to allow them to control machinery with  their thoughts alone the device is able to read   from and write to neural synapses in a person's  brain using very fine needle-like electrodes that   penetrate a person's cranium this technology  already in use on animals requires no evasive   surgery advanced versions of such devices inserted  into the brains of quadriplegics and those   suffering from brain trauma and other ailments  that prevent them from moving their physical   body could allow them to restore if not enhance  the physical capabilities of the human body [Music] we've simplified this to simply something  that is about the size of a large coin and it goes   in your skull replaces a piece of skull  and the wires then connect within a few   centimeters or about an inch away from the  device and this is sort of what it looks like   in a lot of ways it's kind of like a fitbit in  your skull with tiny wires current prototype   version 0.9 has about a thousand channels and  it's a 23 millimeters by eight millimeters in   terms of getting a link you need to have the  device you need to have a great robot that puts   in the electrodes and it does the surgery so you  want the surgery to be as automated as possible   and the only way you can achieve the level of  precision that's needed is with an advanced robot   the human race is advancing towards a great  merging with the technology it has created   and as our technology grows we will have to  confront the questions that arise from our merging   with it not just of how to create it but of how  to control it and what heights we might achieve   through this process for the first time there is  a real possibility of extending our individual   mental capabilities in a manner that goes  beyond physical and biological constraints   the goal of neuralink is to help achieve this  vision to come up with a neural vase that's   doing something beyond what we are doing right  now in a major way one of the main long-term   goals of neurolink is to help humans keep with  the upcoming rapid technological advancements   by significantly extending the natural  capabilities of the human brain we will possibly   create a global neural network of minds this will  allow us to either create or enhance a collective   consciousness an interconnected group of minds  working together to solve the problems of humanity   in the face of rapid technological advancements  thus far it seems the emerging solution for the   ai control problem is the best option for  humanity to make it into the next century   an additional argument for the merging option is  that the control problem might take a long time   to satisfactorily solve and some preliminary  work needs to be started as soon as possible   but also because of the possibility of a  certain intelligence explosion from subhuman to   superhuman ai in which case there might not be any  substantial or unambiguous warning before a super   intelligence arrives in addition it is possible  that insights gained from the control problem   could in the future end up suggesting that some  systems for artificial general intelligence   are more predictable and amenable to control than  other systems which in turn could helpfully nudge   early agi research towards the direction of the  more controllable ai proposed approaches to the   control problem include the creation of friendly  ai as nick bostrom states the best strategy to   assure ourselves that an ai remains friendly  may be to endow it with interest similar to ours   if this strategy is successful it will have the  additional benefits of directing the first super   intelligent system towards objectives that humans  value in other words the control problem may be   easier to solve if the super intelligence system  is designed so that it wants what humans want   which is also known as the value alignment problem  the thing that's going to be tricky here is that   it's going to be very tempting to use ai as a  weapon the on-ramp to serious ai the danger is   going to be more humans using  it against each other i think   most likely from a long-term existential  standpoint that's like the purpose of neural link   is to create a high bandwidth interface to  the brain such that we can be symbolic with ai [Music]   i think best case scenario we effectively merge  with ai where ais serves as a tertiary cognition   layer where we've got the limbic system primitive  brain essentially we've got the cortex if we do   have a third layer which is the ai extension of  yourself that is also symbiotic and there's enough   bandwidth between the cortex and the ai extension  of yourself such that the ai doesn't de facto   separate that could be a good outcome that could  be quite a positive outcome for the future [Music] an ai system that has already surpassed a human  level of intelligence might well deliberately hide   the moment at which it becomes super intelligent  in order to try to avoid any attempts to shut it   down however once an ai is fully matched and then  surpassed human intelligence it will be difficult   for it to hide it for long especially  if it is openly interacting with humans   and while it is possible that humanity will never  know what causes an ai system to suddenly become   super intelligent it is possible that  with foresight and an appropriate level   of caution this occurrence could be  predicted and even controlled [Music] thanks for watching did you like this  video then show your support by subscribing   ringing the bell and enabling notifications  to never miss videos like this [Music] [Music]

Channel: Science Time

Views: 113,153

Keywords: Elon Musk, Musk, Elon Musk on AI, Elon Musk AI, Neuralink, artificial intelligence, AI, artificial general intelligence, superintelligence, digital superintelligence, implantable brain machine interface, technology, science, science time

Id: awt5olRHtDo

Length: 10min 2sec (602 seconds)

Published: Sat Oct 17 2020