Latest / Elon Musk Thinking / Elon Musk Neuralink Update!

Transcript

Elon Musk Neuralink Update! · Elon Musk Thinking

0:00
Speaker 1: Welcome to Neuralink Live Update. We're gonna tell you about
0:05
Speaker 1: the progress of the first patient with the neuralink olland
0:10
Speaker 1: and sort of your recap of the progress there, then
0:15
Speaker 1: talk about what changes we were making for the second patient,
0:19
Speaker 1: which we're hoping to do an implant in the next
0:23
Speaker 1: week or so. And this is for our first product,
0:26
Speaker 1: which is called Levathy, which enables you to control a
0:29
Speaker 1: computer or a phone just by thinking. So let's in fact,
0:33
Speaker 1: so we'll start off with just some introductions. Do you start? Hi?
0:36
Speaker 2: Everyone, My name is DJ sull I'm an electrical engineer
0:39
Speaker 2: and a chip designer by training. I led the design
0:43
Speaker 2: of first several generations of the Neuralink implant. Currently I
0:48
Speaker 2: was on the founding team and currently a president.
0:51
Speaker 1: I'm Matthew McDougal.
0:52
Speaker 3: I'm a practicing neurosurgeon and a head of neurosurgery at Neuralink.
0:57
Speaker 4: Yeah go ahead, yeah, head of burning the first as applications.
1:02
Speaker 5: And I'm plussed some software engineer at Nerlink trying to
1:04
Speaker 5: figure out how to turn brain activity into the stuff
1:07
Speaker 5: in the work.
1:08
Speaker 1: All right, thank you, Well, so let's see, so we'll
1:10
Speaker 1: just get going into the presentation. So our first product
1:16
Speaker 1: is something like I said, we call telepathy, which enables
1:20
Speaker 1: the person with the neural link implant to control their
1:24
Speaker 1: phone or computer just by thinking. And once you can
1:26
Speaker 1: control your phone and computer, you can essentially control almost
1:29
Speaker 1: anything just and literally just by thinking. So there's no
1:33
Speaker 1: eye tracking or anything. It is purely purely your thoughts.
1:38
Speaker 1: So this is really quite quite a profound device that
1:44
Speaker 1: can help a lot of people who have lost the
1:46
Speaker 1: connection between their brain and body. So imagine people like
1:52
Speaker 1: Stephen Hawking who you know. You imagine if if he
1:55
Speaker 1: could communicate at the same speed as someone who had
1:58
Speaker 1: still had the connection to their brain and body. So
2:01
Speaker 1: it's it's really uh something that can help millions of
2:04
Speaker 1: people around the world, and it's a it's part of
2:07
Speaker 1: our overall goal of enabling a very high bandwidth connection
2:13
Speaker 1: between the brain and and your and and the rest
2:16
Speaker 1: of the world and your computers. The long term goal,
2:20
Speaker 1: which sounds a little esoteric, is to mitigate the the
2:23
Speaker 1: risk of of the civil civilizational risk of AI by
2:29
Speaker 1: having a sort of closer symbiosis between human intelligence and
2:35
Speaker 1: digital intelligence. But but that that that will take many years. Uh.
2:39
Speaker 1: Along the way, We're we're going to help solve a
2:42
Speaker 1: lot of brain injury or spinal injury issues. Uh so
2:48
Speaker 1: uh and with without first product sypathy, that's that's going
2:52
Speaker 1: to be really quite profound. Uh that there is also
2:56
Speaker 1: potential long term for bridging the gaps. So if there
3:01
Speaker 1: are damaged or severed neurons, being able to span the
3:05
Speaker 1: gap between the brain's motor cortex to the spine to
3:08
Speaker 1: enable someone to use their body again, I think that
3:11
Speaker 1: would be very exciting, and it's you know that that
3:16
Speaker 1: is something that is possible in the long term and
3:21
Speaker 1: they're not. Our second product, which we've demonstrated to work
3:26
Speaker 1: with monkeys, is blind sight, which would enable someone who
3:30
Speaker 1: is completely blind, lost both eyes or completely lost that
3:34
Speaker 1: are optic nerve to be able to see. So that's
3:39
Speaker 1: that's something that we hope to demonstrate in the future.
3:47
Speaker 1: So this just gives you a sense of what the
3:49
Speaker 1: device is. A way to think about the neuralink device
3:52
Speaker 1: is kind of like a fitbit or an Apple watch
3:54
Speaker 1: with tiny wires or electrodes. Those those tiny wires are
4:00
Speaker 1: implanted in the in the brain and they are read
4:04
Speaker 1: and write electrical signals. So a lot of people think
4:09
Speaker 1: the brain is this incredibly mysterious thing. It's it is
4:12
Speaker 1: mysterious a lot of ways, but but it is. Actually
4:15
Speaker 1: it does operate with like electrical signals, so if you
4:17
Speaker 1: can read and write those electrical signals, you can interface
4:20
Speaker 1: with the brain. And the device is sized so that
4:26
Speaker 1: it is the same size as the as the piece
4:29
Speaker 1: of skull that is removed. So if it's like a
4:32
Speaker 1: few centimeters diameter of skull skull that's removed, we replace
4:36
Speaker 1: that with the device after implanting the tiny wires with
4:40
Speaker 1: the surgical robot, and that enables read write capability to
4:45
Speaker 1: the neurons completely wirelessly. Yeah, yes, exactly, it's completely wirelessly.
4:50
Speaker 1: So like I could I could have a neuralank right now.
4:54
Speaker 1: You wouldn't know. And it charges inductively so you could
4:58
Speaker 1: just uh it's they have electromagnetic ad that that you
5:04
Speaker 1: charge the device with. So yeah, it's like an Apple
5:07
Speaker 1: Watch exactly.
5:08
Speaker 2: So uh, except that it's actually a much harder technical
5:12
Speaker 2: challenge to solve that there's limit as to how much
5:16
Speaker 2: get the print issue or whereas in for phones and.
5:19
Speaker 6: Really care if you're sitting on the table.
5:23
Speaker 1: Yeah. Also it's got to go through skin and stuff
5:25
Speaker 1: as well. And okay, so it is a tougher challenge
5:28
Speaker 1: to to charge and to have vivemand with communications, given
5:32
Speaker 1: that has got to go through skin and hair and
5:35
Speaker 1: stuff we have we have solved. So yeah.
5:41
Speaker 4: Yeah, So our first step with the telepathy is basically
5:44
Speaker 4: to unlock digital independence for people with policies and to
5:50
Speaker 4: allow them to control the computer just with their mind
5:53
Speaker 4: without moving their body. And our goal is to provide
5:57
Speaker 4: them the same level of control, functionality, and reliability that
6:01
Speaker 4: I have when I'm using a computer, even better than
6:04
Speaker 4: the level of control I have.
6:09
Speaker 5: Just to be clear, this guy he's controlling us with
6:11
Speaker 5: his brain. So he's not like you can't see his
6:13
Speaker 5: hands in this video, but he's not using a mouse
6:15
Speaker 5: and keyboard, just thinking about how to move the cursor
6:18
Speaker 5: and playing civilizations.
6:19
Speaker 1: No eye trackers from meat, just thinking.
6:26
Speaker 6: Just a couple of days a cursor move here.
6:28
Speaker 1: Yeah.
6:29
Speaker 5: Yeah, it's like the last standard two nights ago.
6:31
Speaker 2: Yes, yeah, I think I think the way he also
6:33
Speaker 2: described that is he's using them.
6:35
Speaker 4: Yeah, there's many more videos on his on the platform.
6:40
Speaker 4: Definitely checked them out. Yeah, so he can he's string
6:43
Speaker 4: with that live and also can talk and like move
6:45
Speaker 4: his head with that problem multi time.
6:48
Speaker 5: Yeah, you guys, like, if you join his livestream, you
6:49
Speaker 5: can ask him questions. You'll he'll tell you all about
6:52
Speaker 5: what it's like to move.
6:54
Speaker 2: Also, I think I haven't played Civilization myself, but I
6:58
Speaker 2: think this is actually not easy.
6:59
Speaker 6: About this is expert Emperor mode.
7:01
Speaker 5: Emperor if you have played stiff empro mote is like
7:03
Speaker 5: the highest difficulty level. Just the point is that this
7:05
Speaker 5: is a kind of demanding task. And while wife streaming
7:07
Speaker 5: playing the hardest motive team and he's able to do
7:10
Speaker 5: that while.
7:11
Speaker 6: Moving on talking engaging with you know, the audience.
7:14
Speaker 3: Playing One of the other games he likes to play
7:17
Speaker 3: a lot is chess. I think it gets lost sometimes
7:20
Speaker 3: that he's actually playing speed chess against me, which requires
7:25
Speaker 3: an incredibly high fidelity degree of control and speed of
7:29
Speaker 3: control in order to be able to win.
7:35
Speaker 4: So also, another cool stuff about about our device is
7:38
Speaker 4: that is it anywhere anytime, also on a plane during
7:41
Speaker 4: a flight, while sating really cool memes of care. And
7:47
Speaker 4: also our device unlocks things that previously were impossible for participants.
7:53
Speaker 4: For example, we're able to connect him to his living consoles,
7:57
Speaker 4: which they marcott with friends and family, and it was
8:01
Speaker 4: lovely to see them playing together after years that he
8:04
Speaker 4: couldn't do it since his injury.
8:07
Speaker 5: I know, if you're sitting over from this guy playing
8:11
Speaker 5: look over, he's going a cap me love, no hands,
8:15
Speaker 5: no movement.
8:17
Speaker 1: Even away over.
8:18
Speaker 5: Yeah, it's strange, strange time.
8:21
Speaker 6: Yeah.
8:22
Speaker 4: And he loves using the device and using independently daily
8:25
Speaker 4: to watch videos and read play games using the baser.
8:32
Speaker 4: And the key metrics that week are is to make
8:35
Speaker 4: sure our device is actually useful to basically the amount
8:38
Speaker 4: of hours and use the device daily and weekly, and
8:41
Speaker 4: we track it weekly since the since the surgery, and
8:45
Speaker 4: on weeks that he's not too busy and not traveling,
8:48
Speaker 4: he can even reach seventy hours of using the.
8:51
Speaker 1: Device a week.
8:52
Speaker 4: This is amazing and he would of course love to
8:55
Speaker 4: use it more, but need to run result sessions, he
8:59
Speaker 4: needs to sleep sometimes and also of course to charge
9:02
Speaker 4: the device once in a while hopefully will improve that
9:04
Speaker 4: over time.
9:06
Speaker 5: Maybe not obvious to people who are watching this like
9:08
Speaker 5: it's a normal math book. He's controlling. This isn't like
9:10
Speaker 5: some limited addition thing. But there's only a few options,
9:13
Speaker 5: Like you can just do anything that you can do
9:14
Speaker 5: on a mapbook pro that same one I have on
9:16
Speaker 5: my desk. Actually it's the exact same one.
9:19
Speaker 4: Yeah, And another interesting point is that on the first
9:25
Speaker 4: day he used bciurn Control, was able to break the
9:29
Speaker 4: previous world record for curser control just by using the
9:34
Speaker 4: brain and recently he even doubled it and was able
9:37
Speaker 4: to outperform about ten percent of our engineer at New Link.
9:42
Speaker 4: And you can be sure that we are very good
9:43
Speaker 4: in this game and very quick. And if you want
9:46
Speaker 4: to check out how well how well you can do it,
9:48
Speaker 4: you can do it on our website. And it's very
9:51
Speaker 4: addictive games. Yeah, it's a very simple game. You just
9:54
Speaker 4: have to click on the square.
9:55
Speaker 1: But it's actually even though it sounds silly, it's it's
10:01
Speaker 1: quite a Yeah, it can be quite a It sounds
10:04
Speaker 1: like it could be quite addictive, and it's especially if
10:07
Speaker 1: you get a low score and you think there's no
10:08
Speaker 1: way I got it. So, I mean, anyone who wants
10:11
Speaker 1: to try this, I recommend going to the neuralink dot
10:13
Speaker 1: com its website and seeing seeing if you can beat
10:16
Speaker 1: Nolan's record, and it's that you will find that's actually
10:19
Speaker 1: quite difficult to do so. And this is really with
10:23
Speaker 1: version one of the device and with only a small
10:25
Speaker 1: percentage of the electrodes that are working. So this is
10:30
Speaker 1: this is really just the beginning, but even the beginning
10:33
Speaker 1: is twice as good as the world record. This is
10:36
Speaker 1: important to emphasize the you know, the media has a
10:40
Speaker 1: habit of saying that the glasses ten percent empty, but
10:45
Speaker 1: actually it's ninety percent full. So I think it's really
10:48
Speaker 1: quite accomplishment of the Neuralink team to have achieved with
10:52
Speaker 1: with the first patient, first device, twice the world record
10:58
Speaker 1: for the range computer madwidth. That's really an astonishing, an
11:05
Speaker 1: amazingly great outcome, and it's only going to get better
11:09
Speaker 1: from here. So the potential is to ultimately get I
11:12
Speaker 1: think to megabit level. So that's that's probably the long
11:18
Speaker 1: term goal of improving the bandwidth of the brain computer interface.
11:22
Speaker 1: If you think about like how low the bandwidth normally
11:26
Speaker 1: is between a human and advice, it's the average bandwidth
11:30
Speaker 1: is extremely low. It's it's i say, less than one
11:33
Speaker 1: bit per second over the course of a day. If
11:36
Speaker 1: there are eighty six thousand, four hundred seconds in a day,
11:38
Speaker 1: you're outputting less than that number of bets to any
11:42
Speaker 1: any given device except perhaps very rare circumstances. So this
11:50
Speaker 1: is it's actually quite important for for AI. You know,
11:57
Speaker 1: basic for human AI sebiosis is just being able to
12:03
Speaker 1: communicate at a speed. Yeah, I can follow.
12:11
Speaker 3: So yeah, just to emphasize again, he's performing at this
12:16
Speaker 3: extremely high level with about fifteen percent of his channels functional,
12:22
Speaker 3: and so we want to mitigate any of the problems
12:25
Speaker 3: that led to that situation. So, you know, the brain
12:28
Speaker 3: is a fascinating organ. I'll share with you some of
12:31
Speaker 3: the secrets about the brain. During any typical brain surgery,
12:36
Speaker 3: a small amount of air is introduced into the skull.
12:40
Speaker 3: That's because neurosurgeons like to have as much room as
12:42
Speaker 3: possible around the brain, and so there's this little known
12:46
Speaker 3: control mechanism of allowing the CO two concentration in the
12:53
Speaker 3: blood to rise a bit, which allows the brain to
12:57
Speaker 3: either expand or contract, depending on where you you target
13:00
Speaker 3: that CO two. But typically neurosurgeons will have the brain
13:04
Speaker 3: shrink by lowering CO two. What we're going to do
13:07
Speaker 3: in our future surgeries is keep the CO two concentration
13:12
Speaker 3: actually quite normal, maybe even slightly elevated, and I'll allow
13:15
Speaker 3: the brain to stay its normal size and shape during surgery.
13:19
Speaker 3: That should eliminate this air pocket that we saw in
13:21
Speaker 3: the first participant that air pocket we think may have
13:24
Speaker 3: contributed to eating up some of the thread slack as
13:29
Speaker 3: the air bubble migrated to be under the implant push
13:32
Speaker 3: the brain away from the implant, and so that's easy
13:35
Speaker 3: enough to fix. Another consideration that we want to focus
13:41
Speaker 3: on for our upcoming participants is that the brain. Think
13:47
Speaker 3: of it like a really complex folded onion. It's layer
13:51
Speaker 3: upon layer of sheets of neurons all over the surface
13:54
Speaker 3: of the brain, folded into this odd looking shape.
14:00
Speaker 1: The folds of.
14:01
Speaker 3: The brain traveled down deep into the brain, and along
14:05
Speaker 3: with it go those onion layers of neurons. And if
14:08
Speaker 3: we insert very close to one of the folds where
14:12
Speaker 3: there may be very useful information encoded in neurons, we
14:15
Speaker 3: may end up traveling with our threads parallel to some
14:19
Speaker 3: of the layers of neurons that were most interested in
14:22
Speaker 3: avoiding them Entirely. To avoid that possibility, we're going to
14:25
Speaker 3: insert in our future participants more close to the middle
14:30
Speaker 3: of the apex of the folds, ensuring that we're crossing
14:35
Speaker 3: the layers of interest layer five of the cortex.
14:37
Speaker 2: I also think that it's important to highlight here those
14:42
Speaker 2: tiny wires that Elon mentioned.
14:44
Speaker 6: They're fraction of a human hair.
14:46
Speaker 2: They're very flexible intentionally so because brains constantly moving and
14:50
Speaker 2: you want the electros to be moving with the brain
14:53
Speaker 2: causing less of the scarring. And it's actually impossible for
14:59
Speaker 2: human eurosurgeon, however talented matthew Is to actually maneuver them
15:03
Speaker 2: by so we have a surgical robot that we built
15:05
Speaker 2: that can actually precisely target them in any three dimensional
15:08
Speaker 2: space x, y as well as z with micron level
15:11
Speaker 2: precision while avoiding bascoature so that you don't disrupt and
15:16
Speaker 2: cause immune response from happening. So we actually have the
15:19
Speaker 2: technology to be able to place them exactly where we
15:21
Speaker 2: want them in three.
15:23
Speaker 3: Yeah, it's truly amazing to see the surface of the
15:25
Speaker 3: brain after the robot had inserted all the electrodes on
15:29
Speaker 3: the first participant, without a drop of blood in sight.
15:33
Speaker 3: It's really quite an achievement.
15:36
Speaker 1: Yes, something that probably most people don't realize is that
15:39
Speaker 1: the brain appears to be sort of somewhat undifferentiated. So
15:42
Speaker 1: if you look at the cortex, it looks like a
15:45
Speaker 1: whole bunch of folds that were you know, maybe like that.
15:49
Speaker 1: It's not obviously just looking at to take a picture
15:51
Speaker 1: of the brain. That that it's the brain is highly differentiated.
15:55
Speaker 1: That there's you pretty much know exactly where the part
16:00
Speaker 1: of the brain is that controls your right hand and
16:03
Speaker 1: your left hand and your leg and that kind of
16:05
Speaker 1: thing or vision. It's it's actually quite precisely located. It's
16:11
Speaker 1: not some people like might might think look the brain like, oh,
16:15
Speaker 1: could be it could be anywhere, but actually we it
16:19
Speaker 1: it's it's your brain is highly differentiated, even though it
16:23
Speaker 1: doesn't look it. It's.
16:25
Speaker 2: Yeah, do you want to describe how we actually were,
16:28
Speaker 2: like how we identify word to drill the quen So.
16:31
Speaker 3: We can we can put a patient that is considering
16:34
Speaker 3: this implant into an f MRI so a functional magnetic
16:39
Speaker 3: resonance imaging machine and ask them to imagine hand movements
16:43
Speaker 3: that you know, because of the spinal cord injury don't happen,
16:46
Speaker 3: but just imagining those hand movements causes these areas of
16:49
Speaker 3: the brain to light up in the fMRI scanner. And
16:52
Speaker 3: so we have a pretty good idea based in in fact,
16:57
Speaker 3: for each individual participant which part their brain is going to,
17:02
Speaker 3: you know, respond to imagined movements of the hand. And
17:06
Speaker 3: so we can map those imagined movements, much as we
17:08
Speaker 3: all do when moving a mouse to controlling a cursor
17:12
Speaker 3: on a screen, even without the use of a mouse.
17:17
Speaker 1: Yeah, but I think it's kind of an important point that, like,
17:19
Speaker 1: it's not like your the part of your brain that
17:21
Speaker 1: controls your hand might be anywhere in the cortex. It's
17:24
Speaker 1: this is not the case. It's going to be in
17:25
Speaker 1: a very specific region, and it's going to be extremely
17:28
Speaker 1: common across people. Precision is key too.
17:31
Speaker 5: Yeah, the left handed right handed in my mind too, Like,
17:37
Speaker 5: if you're right handed, you want.
17:39
Speaker 6: The fight on the left side.
17:41
Speaker 5: Let the lotto side to the hand gets your commet.
17:43
Speaker 1: Yeah, the left side of your brain controls right right.
17:45
Speaker 6: Yeah, everything's crop.
17:47
Speaker 3: Another of the risk mitigations we're looking at in the
17:50
Speaker 3: future is that, you know, the implant has a certain size.
17:54
Speaker 3: The depth of the bottom of the implant. It's actually
17:58
Speaker 3: dinner than the average human skull. And so what we
18:02
Speaker 3: want to be able to do is control the size
18:05
Speaker 3: of the gap under the implant, give the threads that
18:09
Speaker 3: travel from the implant into the brain as much slack
18:13
Speaker 3: as possible. We didn't do this in the first participant
18:17
Speaker 3: because we didn't want to manipulate any of their tissue
18:21
Speaker 3: that we didn't absolutely have to in upcoming implants.
18:24
Speaker 1: Our plan is to sort.
18:26
Speaker 3: Of sculpt the surface of the skull very intentionally to
18:31
Speaker 3: minimize the gap under the implant, such that the bottom
18:34
Speaker 3: of the implant travels perfectly flush with the normal contour
18:38
Speaker 3: of the inner side of the skull. That will put
18:41
Speaker 3: the implant closer to the brain. It will eliminate some
18:44
Speaker 3: of the tension on the threads, and we think it
18:46
Speaker 3: will reduce some of the tendency of threads to retract.
18:50
Speaker 6: And we actually built the tool to do.
18:53
Speaker 1: Yeah, this is actually this is a very important detail.
18:57
Speaker 1: You really want the inner contraur of the skull to
19:02
Speaker 1: be blushed so that the implant there's there's no the
19:06
Speaker 1: brain doesn't want to pucker up into the into the gap.
19:08
Speaker 1: That's really quite a big deal. So, like like minimizing
19:14
Speaker 1: the air procket and the implant being flushed with the
19:19
Speaker 1: inside contour of the skull are two very important improvements.
19:23
Speaker 2: The additional benefit here is that, uh, you know, you
19:26
Speaker 2: do see some amount of stick up what we call
19:28
Speaker 2: stick ups, so you might have bumped in the head,
19:31
Speaker 2: but this actually eliminates it even further.
19:33
Speaker 1: Yeah, yeah, I mean it's like, really our goal is
19:36
Speaker 1: that that if you run your hand over the top
19:40
Speaker 1: of the skull, you don't feel any any momp, you
19:41
Speaker 1: don't feel any any device, and that even if someone
19:45
Speaker 1: was bold, you wouln't really even notice it. And uh
19:48
Speaker 1: and then from the inner inner contraur of the skull
19:51
Speaker 1: that the brain from physical standpoint, doesn't really notice that
19:54
Speaker 1: there's a divot in the skull because there's no dibit. Okay.
20:06
Speaker 3: Another aspect of the human brain that you know, obviously
20:11
Speaker 3: differs from any of the animals that we tested in
20:14
Speaker 3: is that the human brain is a lot bigger. And
20:16
Speaker 3: so you may not realize that that means the human
20:19
Speaker 3: brain moves quite a bit more than any of these
20:24
Speaker 3: other smaller brained creatures. And so when we open the skull,
20:29
Speaker 3: we see the brain travel toward and away from the
20:33
Speaker 3: robot about three millimeters in total, as the heart beats
20:35
Speaker 3: and the breathing takes place. And so that movement, you know,
20:41
Speaker 3: it adds a small challenge for the robot in precisely
20:45
Speaker 3: choosing a depth to insert each thread. It's not an
20:48
Speaker 3: enormous challenge, and we've already upgraded the robot's capabilities to
20:52
Speaker 3: be able to even more precisely target depth in even
20:57
Speaker 3: a very rapidly moving brain with a high amplitude of movement.
21:05
Speaker 3: You may think the most obvious mitigation for threads that
21:07
Speaker 3: pulled out of the brain is to insert them deeper.
21:10
Speaker 3: We think so too, and so we're going to broaden
21:14
Speaker 3: the range of depths at which we insert threads. So
21:19
Speaker 3: for the very first participant, we had an enormous amount
21:22
Speaker 3: of data from our animal work, and we had very
21:25
Speaker 3: highly optimized our insertion depth to maximize the crossing of
21:31
Speaker 3: the layers of interest in the cortex with the electrodes
21:34
Speaker 3: that we're recording from. Now that we know retraction is
21:38
Speaker 3: a possibility, we're going to insert at a variety of
21:43
Speaker 3: depths that even in several cases of different amounts of
21:46
Speaker 3: retracting threads, we're going to have electrodes at the proper
21:50
Speaker 3: depth and with the deepest threads be able to track
21:55
Speaker 3: how much retraction has occurred across the surface of the
21:57
Speaker 3: brain from each thread. And so we're going to you know,
22:01
Speaker 3: both have more threads in the right layer and have
22:05
Speaker 3: better data on how much retraction has occurred.
22:07
Speaker 5: If you're a pc I NERD. You might know that
22:09
Speaker 5: being able to control individual z depth per thread it
22:12
Speaker 5: is not something that most neural interface devices offer. Most
22:16
Speaker 5: neural interface devices are kind of a static, fixed rigid
22:19
Speaker 5: array that you push in and all the lectures are
22:21
Speaker 5: on depth.
22:22
Speaker 1: Right.
22:22
Speaker 5: To be able to do this is actually pretty pretty
22:24
Speaker 5: novel part of the robots.
22:26
Speaker 3: Yeah, the historical approach is to actually pound in a
22:30
Speaker 3: sort of bed of nails with an air hammer into.
22:31
Speaker 1: The Look it looks crazy that that is, Yeah, just
22:36
Speaker 1: with a with a aumuntic hammer. That's the that's what
22:39
Speaker 1: this is. It sounds come somewhat by braic. This is
22:41
Speaker 1: not what we do, but this is the what's been
22:44
Speaker 1: done before. Is that's literally just hammering and what looks
22:47
Speaker 1: like a better nails with the brain, which actually works.
22:50
Speaker 1: It's astonishing that.
22:53
Speaker 5: Some people like man like DBS props you're just sticking
22:55
Speaker 5: in by hand.
22:58
Speaker 3: But those are several several or is a magnitude more
23:01
Speaker 3: volume of brain tissue that you're destroying.
23:03
Speaker 6: Compared to what we're doing.
23:04
Speaker 1: But that deep brain simulation stuff does actually work and
23:06
Speaker 1: it actually helps people a lot. Yeah, Yeah, that's a
23:10
Speaker 1: great product. Yeah, I mean I think we'll be able
23:14
Speaker 1: to do a much more finessed version of that down
23:16
Speaker 1: the road. So I mean it's really difficult. Like the
23:23
Speaker 1: neuraling device is something that really absolutely minimizes damage to
23:26
Speaker 1: the brain, absolutely minimizes the load on the patient, and
23:31
Speaker 1: the goal is to allow someone to live a completely
23:33
Speaker 1: normal life that you won't even notice that someone even
23:37
Speaker 1: has the device. So, like I said, restoring the ability
23:42
Speaker 1: to control your computer and phone, that's a telepathy. And
23:46
Speaker 1: then next device being able to allow people to see
23:49
Speaker 1: that could not see before. In fact, you could allow
23:51
Speaker 1: people to see kind of like Doorty of the Forge
23:53
Speaker 1: and Star Trek in any whatever. Yeah, infrared ultraviolet radar
24:00
Speaker 1: are so so I think another way of saying is
24:07
Speaker 1: that we want to give people superpowers. So it's not
24:10
Speaker 1: just that we're restoring your prior brain functionality, but that
24:14
Speaker 1: you actually have functionality far greater than a normal human.
24:18
Speaker 1: That's a super big deal.
24:20
Speaker 2: And I also think you know, oftentimes the questions that
24:22
Speaker 2: we get a lot is why do you have to
24:23
Speaker 2: actually go into the brain, What if you place it
24:26
Speaker 2: on the surface or most types of score. Basically, the
24:30
Speaker 2: long story short, the physics of how it works, you
24:32
Speaker 2: really need to get the sensors, which are these facing
24:36
Speaker 2: in the brain next to the source, which neuron as
24:40
Speaker 2: close to it as possible. Otherwise, what you get is
24:42
Speaker 2: you get a population respond and not be able to
24:45
Speaker 2: kind of do the level of controls that we believe of.
24:48
Speaker 1: Yeah, I mean maybe a good sort of analogy would
24:52
Speaker 1: be like if you're trying to understand what goes on
24:54
Speaker 1: in a factory, you kind of need to go into
24:56
Speaker 1: the factory. You can't just put a death scope on
24:57
Speaker 1: the wall and try to figure out what's going like
25:00
Speaker 1: anything on the outside of the Trying to read things
25:03
Speaker 1: from the outside is like putting a statoscope on the
25:06
Speaker 1: world of a factory trying to understand what's going in
25:08
Speaker 1: the factory. It's not going to be effective. You've got
25:12
Speaker 1: to be the threads. You gotta be in there. So
25:17
Speaker 1: but I just want to be emphasized again, like the
25:19
Speaker 1: goal is to give people superpowers, not just to restore
25:23
Speaker 1: prior functionality. So I think that's very exciting and I
25:28
Speaker 1: think that should give hope to a lot of people
25:31
Speaker 1: in the world that the future is going to be
25:33
Speaker 1: exciting and inspiring and the technology is going to give
25:37
Speaker 1: them superpowers. I mean, that's that's amazing. Oh yeah, I
25:50
Speaker 1: guess yeah, and could can you multitest? But that Yeah,
25:54
Speaker 1: In fact, if you look at Nolan's streaming and you
25:57
Speaker 1: can just check out Nolan's streams on on the x platform,
26:02
Speaker 1: he's multitasking all the time. So he's playing video games
26:04
Speaker 1: while talking and uh the podcast podcast. Yeah, yeah exactly.
26:11
Speaker 1: So it's truly just like if you're using your hands
26:14
Speaker 1: and you can be you know, playing a video game
26:17
Speaker 1: while talking.
26:18
Speaker 5: So and we don't take our word for it, they
26:19
Speaker 5: just go watch. I mean, yeah, he's out there on
26:21
Speaker 5: the internet doing the thing.
26:22
Speaker 1: Yeah exactly. So can you do keyboards shortcuts or is
26:27
Speaker 1: it just the mouse?
26:28
Speaker 5: Yeah, that's what you're working on right now.
26:31
Speaker 4: Sure, so Godley is walking with the mouse. But we
26:33
Speaker 4: are also exploring recording more dimensions out from the newer
26:36
Speaker 4: activity multiple clicks, so to shortcuts or just able to
26:43
Speaker 4: draw more games games with an xbook controller. But also
26:49
Speaker 4: in the future we're expend we plan to expand to
26:51
Speaker 4: the code text, not just the mouse control, but also
26:54
Speaker 4: allow what participant to type much faster.
27:00
Speaker 5: And yeah, actually so maybe going back to the discussion
27:03
Speaker 5: of thread attraction, you know, one of the very exciting
27:05
Speaker 5: parts to me about this story is that we're able
27:07
Speaker 5: to do so much with fifteen percent of channels, and
27:09
Speaker 5: you have more channels, what that actually offers you is
27:11
Speaker 5: not just faster and mouse control, because in the motor
27:14
Speaker 5: cortext neurons don't all represent the same thing. So if
27:17
Speaker 5: you're trying to understand, like, you know, what an individual
27:19
Speaker 5: finger is trying to do, you might or might not
27:22
Speaker 5: have an electro next to it. And the more channels
27:24
Speaker 5: you have in the bringing in, the higher likelihood you
27:25
Speaker 5: have representation or decodability of all fingers on the hand.
27:29
Speaker 5: And so if you're trying to do something like I'll
27:31
Speaker 5: put texts at a fast rate is something that matters
27:33
Speaker 5: a lot for people who are completely locked in, who
27:35
Speaker 5: cannot speak at all, who are trying to, you know,
27:37
Speaker 5: just say I love you to a loved one in
27:38
Speaker 5: their family, or ask for a glass of water or
27:40
Speaker 5: scratch or whatever, you know, being able to type it
27:42
Speaker 5: a faster ate.
27:42
Speaker 6: It's extremely important.
27:44
Speaker 5: And the more fingers you have access to, the higher
27:46
Speaker 5: probability you can do that efficient.
27:48
Speaker 6: And so yeah, you know, I'm.
27:50
Speaker 5: Super excited about how high the ceiling is that we
27:53
Speaker 5: can get to as we resolve this tut attraction too.
27:56
Speaker 1: Yeah, I mean we're kindly at approximately ten ten us
27:59
Speaker 1: per second PA. Great, but ultimately we want to get
28:02
Speaker 1: to a megabit, and I think ultimately whole brain interface,
28:07
Speaker 1: I think, you know, many years from now. I think
28:10
Speaker 1: gigabit level as possible. So that's pretty soonishing now, you know,
28:18
Speaker 1: with there is still version of one of our device.
28:21
Speaker 1: As we mentioned, it's version one with only fifteen percent
28:23
Speaker 1: of the threads working. The current device has sixty four
28:28
Speaker 1: threads with sixteen electrodes on each thread. Our next device
28:32
Speaker 1: has one hundred and twenty eight threads with eight electrodes
28:36
Speaker 1: per thread. Because as we get more confident about how
28:41
Speaker 1: we're exactly to place the electrode the thread, you need
28:45
Speaker 1: fewer electrodes per thread. So we can, essentially, with the
28:48
Speaker 1: current device without substantial changes, potentially double the band if
28:54
Speaker 1: we are accurate with the placement of the threads, and
28:58
Speaker 1: then our next generation device will have maybe even more channels.
29:02
Speaker 1: Yeah yeah, so next vice amy four Yeah, three thousand channels,
29:08
Speaker 1: So this will just keep getting better and better, really
29:12
Speaker 1: moving up, I think, and for his magnitude in factors
29:15
Speaker 1: of ten, basically in what kind of ad withth So
29:18
Speaker 1: I think it won't be it won't be all that
29:22
Speaker 1: long before someone with a neural link device can communicate
29:26
Speaker 1: faster than someone who is has a fully functional body.
29:29
Speaker 1: And yeah, so I think, you know, faster than the
29:34
Speaker 1: fastest speed typist or auctioneer.
29:37
Speaker 6: The esports tournaments are going to be like.
29:38
Speaker 1: You literally won't be able to speak faster than someone
29:41
Speaker 1: can communicate with a neural link to lepathy device.
29:45
Speaker 4: It may it may be a very interesting part of this. Basically,
29:48
Speaker 4: we currently connect to stand up inputs to the computer
29:52
Speaker 4: through hous and keyboards. Very soon as we will have
29:55
Speaker 4: a part of bendwith we need to think about new
29:58
Speaker 4: ways to actually the interface for the devices.
30:02
Speaker 1: Is this something that makes said, yeah, no, that's that's
30:07
Speaker 1: a good point because the the current import devices are
30:11
Speaker 1: centered around human hands. So it's like we've got these
30:15
Speaker 1: you know, little meat sticks that we move and there
30:20
Speaker 1: is a certain rate at which you can move your
30:22
Speaker 1: little move your fingers and and so we've got like
30:24
Speaker 1: the mouse and the keyboard and or the joystick control
30:27
Speaker 1: you know, like Xbox control or something like that. But
30:29
Speaker 1: you really don't need that. You can actually, you know,
30:34
Speaker 1: you don't, you don't need since you're no longer if
30:36
Speaker 1: you're if you're not trying to use your hands, you
30:38
Speaker 1: don't you actually don't need those conventional uh control mechanisms.
30:45
Speaker 1: And so that's why I like, ultimately I think you'll
30:47
Speaker 1: be able to do conceptual telepathy, like where you can
30:52
Speaker 1: communicate entire concepts uncompressed to someone else within your link
30:58
Speaker 1: or to the computer.
31:00
Speaker 5: Even today, we have some problems here where like, you know,
31:02
Speaker 5: if you don't feel the mouse clicking under your finger,
31:04
Speaker 5: how do you know it actually happened, Because you know,
31:06
Speaker 5: you're you're seeing it on the screen, but you don't
31:07
Speaker 5: actually feel the mouse click. You don't have appropriate sective
31:09
Speaker 5: feedback of you know, the keys under your fingertips or
31:11
Speaker 5: the track pad on your So there's all sorts of
31:14
Speaker 5: interesting UAX challenges. How to actually give the user some
31:17
Speaker 5: sense of what their decoder is actually doing, what the
31:20
Speaker 5: EARTHNK is actually doing.
31:21
Speaker 6: How we're gonna trying to use.
31:26
Speaker 1: So wireless.
31:30
Speaker 5: Yeah, it's bluetooth, just a Bluetooth connection, just like how
31:33
Speaker 5: your normal Apple mouse or like Apple Magic keyboard connects
31:37
Speaker 5: to your computer, same exact thing in fact, in yeah,
31:41
Speaker 5: we can basically have this exposed as an HID interface
31:43
Speaker 5: if we want. HID is just the name of the
31:45
Speaker 5: protocol for like sending bits from a mouse into a computer. Yeah,
31:49
Speaker 5: I can plug into basically anything.
31:51
Speaker 1: Yeah.
31:52
Speaker 2: Yeah, I mean I think we chose that interface because
31:54
Speaker 2: it's ubiquitous. Yeah, basically any devices are have Bluetooth capabilities.
32:00
Speaker 2: Our long term goal is to actually have our own protocol.
32:02
Speaker 2: You know that is safe and secure, but for now,
32:05
Speaker 2: you know, chosen it for interoperability.
32:09
Speaker 3: So the question is kind of neuralink should prepare the
32:11
Speaker 3: paralysis in the long term. You know, we can't do
32:14
Speaker 3: that right now. We have done sort of preliminary work
32:18
Speaker 3: planting a second neuralink in the spinal cord, and we
32:22
Speaker 3: can restore naturalistic looking hand and leg movements in animal models.
32:31
Speaker 3: But this isn't something that is you know, don't don't
32:33
Speaker 3: hold your breath waiting for it. It's going to be
32:35
Speaker 3: a while. We've got a lot of work to do.
32:37
Speaker 3: But yes, there's no reason in theory that we can't
32:41
Speaker 3: repair paralysis.
32:43
Speaker 1: Yeah, I mean, essentially to to, I mean, there's there's
32:50
Speaker 1: no there's no physics barrier to fully solving paralysis. That
32:53
Speaker 1: is perhaps a way to say that you've got signals
32:57
Speaker 1: coming from your motor cortex. Uh that if they are
33:02
Speaker 1: transferred past the point where the nose are damaged. Essentially
33:07
Speaker 1: just it's basically a communications bridge to bridge the communications
33:11
Speaker 1: from the motor cortex past the point in the necrospine
33:16
Speaker 1: where the nose are damaged. And you should like it
33:19
Speaker 1: is physics. It is possible from a physics standpoint to
33:22
Speaker 1: restore full body functionality. For physics standpoints, it's a very
33:26
Speaker 1: hard technical problem, but there is nothing that it prevents
33:30
Speaker 1: it happening from a physics standpoint. So in terms of
33:38
Speaker 1: next phases roll out, well, we really want to make
33:41
Speaker 1: sure that we make as much progress as possible between
33:45
Speaker 1: each neuralink patient. So this is we're only just moving
33:48
Speaker 1: now to our second neuralink patient, but we hope to
33:52
Speaker 1: have you know, if things go well, high single digits
33:56
Speaker 1: this year, and I don't know, maybe this is somewhat
34:02
Speaker 1: dependent on regulatory approval and how much technical progress we make,
34:06
Speaker 1: but within a few years hopefully thousands.
34:10
Speaker 2: Yeah, And I think one thing that is important to
34:12
Speaker 2: highlight is that you know, it's not that we built
34:15
Speaker 2: only one device and one surgery.
34:17
Speaker 6: We've built bunderds of surgery.
34:18
Speaker 2: We've built thousands and thousands of devices even for just
34:22
Speaker 2: the ability to unearth any sort of low frequency failure mode.
34:26
Speaker 2: So we have already been investing very heavily in infrastructure
34:30
Speaker 2: to be able to scale this thing on the device
34:32
Speaker 2: manufacturing side as well as on the surgery side of things.
34:35
Speaker 2: We want to be able to help as many people
34:37
Speaker 2: as quickly as well. Yeah, we go through obviously the
34:39
Speaker 2: appropriate hurdles, right that are regulatory.
34:42
Speaker 6: Challenges and proving out the device with one part.
34:46
Speaker 1: Yeah, and the device implantation really needs to become almost entirely,
34:51
Speaker 1: if not entirely automatic, in the same way that's a
34:54
Speaker 1: laser eye surgery is done. You know, you don't have
34:58
Speaker 1: an ophthalmologist with a a cutter by hand, that that
35:01
Speaker 1: would be crazy. But the uphomologist oversees the lasic machine
35:07
Speaker 1: and make sure that the settings are correct, and then
35:10
Speaker 1: the machine does everything and restores your eyesight. It's really
35:13
Speaker 1: remarkable how many people have had their eyesight restored with laser.
35:17
Speaker 1: And I think there's an one could smile, it's they
35:20
Speaker 1: keep making it better. We need to have something similar
35:22
Speaker 1: for a neuralink implantation, so that you basically sit down
35:26
Speaker 1: and whatever the whatever kind of upgrades or you know,
35:30
Speaker 1: brain fixes are needed. That's reviewed by medical expert. Obviously,
35:36
Speaker 1: want to make sure that that is reviewed correctly, but
35:40
Speaker 1: it really needs to be automatic. So you sit down
35:43
Speaker 1: and within ten minutes you have a neuralin device installed,
35:48
Speaker 1: very very fast. I mean it's very sort of cyberpunk,
35:51
Speaker 1: you know diosex if you played those.
35:53
Speaker 4: Games, and we'll annually start to interface with adult devices
35:59
Speaker 4: like wheelchair. It's a great question. We're currently focusing on
36:04
Speaker 4: john in computers unlock independence in the virtual world. Of course,
36:10
Speaker 4: our plan is as you mentioned earlier, body arm and
36:13
Speaker 4: a wheelchair to unlock independence in the physical world because,
36:16
Speaker 4: of course at the additional risk if you make your computer.
36:19
Speaker 4: They shouts to that, but we're welcome with the FDA
36:23
Speaker 4: to allow us to quite so, what.
36:26
Speaker 7: Seems like if the wheelchair has an app, well, the
36:30
Speaker 7: wheelchair just needs to have an airpace, So if the
36:33
Speaker 7: wheelchair has a ge tooth interface, you could just go
36:36
Speaker 7: tooth interface to the wheelchair and.
36:40
Speaker 1: That's probably something we should do. We're working on pretty soon.
36:44
Speaker 5: Paperwork it safely you don't drive off.
36:47
Speaker 1: Well, I think, well we can limit the speed it
36:51
Speaker 1: doesn't go careating often the disaster. But you know, so
36:55
Speaker 1: just make it go slowly at first. But yeah, so
36:59
Speaker 1: being to sort of it, really, the Nework device just
37:03
Speaker 1: should work generally for anything that's got a bluetooth their.
37:06
Speaker 2: Face, including potentially an optimism.
37:10
Speaker 1: Yes, yeah, you yes, you could communicate with optimists. Uh, yep,
37:18
Speaker 1: absolutely optimists. Well we also also be able to talk
37:23
Speaker 1: to optimists, but like, but you could just yeah, instead
37:28
Speaker 1: of talking, just you could just be in it directly.
37:30
Speaker 1: Or if if someone has lost the use of speech,
37:33
Speaker 1: then then they can still communicate to an optimist. They
37:36
Speaker 1: can humunicate telepathically to optimists or Bluetooth and and and
37:43
Speaker 1: so even if someone has completely lost the ability to speak,
37:47
Speaker 1: they could still control optimists or their computer or phone.
37:51
Speaker 5: We also like, if you have an optimist and you
37:52
Speaker 5: have a NERD, you can just directly map the signal
37:55
Speaker 5: to control of the physical arm of the robot. And
37:57
Speaker 5: that's a very meaningful thing. Like if you're you know,
37:59
Speaker 5: folks that have buncod injury, one of the biggest requests
38:01
Speaker 5: to be able to scratch yourself. It's something that quite
38:04
Speaker 5: annoying actually, And if you have a scratch on your face,
38:07
Speaker 5: you can't fall asleep until you scratch it. You know,
38:10
Speaker 5: it's very convenient to be able to move something physically
38:12
Speaker 5: towards you to scratch. Similar things like eating food. You know,
38:15
Speaker 5: if you need somebody to feed, you very hard to
38:17
Speaker 5: have dinner with friends in a way that is, you know,
38:20
Speaker 5: sort of a normal social experience. And so if you
38:22
Speaker 5: can feed yourself, pick up a fork and actually eat
38:25
Speaker 5: a piece of chicken on your own, yeah, that's a
38:28
Speaker 5: big deal. It prevents and saves a lot of interactions
38:32
Speaker 5: with caretakers and other people in your life that you
38:34
Speaker 5: rely on to take care of. It really increases you.
38:38
Speaker 1: I think an exciting possibility long term also is to
38:41
Speaker 1: say if you take parts of the Optimus Optimius humanoid
38:44
Speaker 1: robot and you combine that with a neural link. Let's
38:46
Speaker 1: say somebody has lost their arms or legs, Well, we
38:50
Speaker 1: could actually attach an Optimist arm or Optimist legs and
38:54
Speaker 1: do a neural link implant so that the motor commands
38:58
Speaker 1: from your brain that goat would goat to your biological arms,
39:01
Speaker 1: now go to your robot arms or robot legs, and
39:04
Speaker 1: again you'd have basically cybernetic superpowers.
39:07
Speaker 5: Actually, so the lenksy from the neuralink to your hand
39:10
Speaker 5: would probably be something faster than it is just to
39:13
Speaker 5: go to.
39:13
Speaker 6: Your physical hand.
39:14
Speaker 5: So you can imagine, like if you're a piano player
39:16
Speaker 5: or I don't know, anything that requires extremely fast hand movements,
39:20
Speaker 5: that you could actually have a pretty imbalanced right hand
39:22
Speaker 5: robotic arm control versus left hand physical hand control.
39:25
Speaker 1: That's one of them. Yeah, Like I said, it's just
39:28
Speaker 1: kind of a cyberpunk dio sex in the future where
39:31
Speaker 1: you have cybernetic upgrades that are actually better than your
39:35
Speaker 1: biological limbs. And it's certainly that we'll have a much
39:45
Speaker 1: you know as particularly as we span to a large
39:48
Speaker 1: number of of of customers or patients for neuralink, the
39:54
Speaker 1: understanding of the brain will improve dramatically because really there
40:00
Speaker 1: isn't a fine, very fine grained understanding of the brain
40:03
Speaker 1: today because it's just the sensors are not good enough.
40:05
Speaker 1: You've got fMRI, which is pretty good, but it's still
40:09
Speaker 1: not as good as actually having I bound with electroids
40:12
Speaker 1: in the brain.
40:13
Speaker 3: Yeah, I think this is underappreciated as a research tool
40:16
Speaker 3: to move that whole effort forward of really knowing, you know,
40:20
Speaker 3: what the physical substance of human thought is. We don't
40:24
Speaker 3: know to the to the degree that we need to.
40:27
Speaker 3: So neuralink is actually a very powerful research tool.
40:32
Speaker 1: Yeah, I mean we I think we can ultimately understand
40:37
Speaker 1: and and fix it quite severe psychosis or like if
40:42
Speaker 1: if somebody's got like the if somebody's got like a.
40:47
Speaker 3: Like a delusion that they have a chip in their brain.
40:51
Speaker 1: I was wondering if you're gonna mention that one. We
40:53
Speaker 1: just want to be clear if there's only one person
40:55
Speaker 1: with a neural link chip in their brain. So for
40:58
Speaker 1: people out there who think we've put a chip on
41:00
Speaker 1: their brain, would like to assure you or what it's worth,
41:03
Speaker 1: you probably won't believe us, but we did not put
41:06
Speaker 1: each ep in your brain. Okay, So there's actually a
41:12
Speaker 1: remarkable number of people who think we have put a
41:14
Speaker 1: trip in their brain, but we have not. But in
41:17
Speaker 1: the future, if you would like us to put a
41:19
Speaker 1: trip in your brain, which will perhaps help with the
41:22
Speaker 1: issue of thinking that you have a chip in your brain,
41:25
Speaker 1: then we will be able to do so. So there
41:28
Speaker 1: are people that have severeria, schizophrenia, They've got basically things
41:32
Speaker 1: that their brain is malfunctioning in some way, and this
41:37
Speaker 1: is actually due to really like physical circuitry issues. You
41:41
Speaker 1: can think of the brain as like, really it's a
41:46
Speaker 1: biological computer and if if some of the circuits are crusted,
41:49
Speaker 1: it's gonna you know, it's going to crash. It's or
41:52
Speaker 1: it's gonna have issues that it's not work. But with
41:57
Speaker 1: a neuralin device, we can fix those issues. And you know,
42:01
Speaker 1: give someone who I think has to say serbisica, spraying
42:05
Speaker 1: or psychosis of some kind, allow them to live a
42:09
Speaker 1: normal life. I think that is one of the likely
42:11
Speaker 1: things in the future. So yeah, I mean, yeah, you
42:23
Speaker 1: can certainly imagine, like I'm sure people have like parents,
42:26
Speaker 1: grandparents who you know, have memory that's not working as
42:33
Speaker 1: well as it used to be. Sometimes they forget who
42:36
Speaker 1: who their grandchildren are or what day it is, and
42:39
Speaker 1: this is something that in your O like device could
42:41
Speaker 1: help fix.
42:43
Speaker 2: I mean, that's actually one of the personal reasons.
42:51
Speaker 6: In many way like forms of you're literally losing and
42:55
Speaker 6: then part of your right then.
42:57
Speaker 1: Which is just a very very.
43:00
Speaker 5: Go through it.
43:00
Speaker 1: Yeah, it's really just it's a glitch in the biological
43:05
Speaker 1: computer that is a fixable glitch. Like it's a short circuit.
43:10
Speaker 5: Essentially, how does the device charge and how long does
43:15
Speaker 5: the charge last. Yeah, So the current version that Nolan
43:18
Speaker 5: has lasts but four to five hours on a charge,
43:21
Speaker 5: and it takes about forty five minutes to charge. The
43:24
Speaker 5: thing we've learned from Nolan is that that's actually one
43:26
Speaker 5: of the main limitters for him using it more. It's
43:28
Speaker 5: actually pretty hard to use a product more than like
43:31
Speaker 5: seventy hours a week.
43:32
Speaker 1: But that's about what he has used it for in
43:34
Speaker 1: some weeks.
43:35
Speaker 6: Yeah, seventy hours a week.
43:36
Speaker 5: Yeah, I mean just for contexts, like you sleep roughly
43:38
Speaker 5: eight hours a night, so that's you know, we're doing
43:40
Speaker 5: better than the bed. Like the bed is fifty six
43:42
Speaker 5: hours a week if you used roughly, and so seventy
43:44
Speaker 5: hours a week uses I challenge you to think about
43:47
Speaker 5: products that you've actually used for that duration.
43:50
Speaker 1: But that's again some of these points are worth like
43:51
Speaker 1: emphasizing again, like the that Nolan, our first a Neuralink recipient,
43:57
Speaker 1: has used the neuralink device for seventy hours in a week,
44:01
Speaker 1: which is incredible.
44:02
Speaker 5: You probably won't enjoy that I'm sharing his computer use publicly,
44:05
Speaker 5: but I assure you for productive things only. But actually,
44:11
Speaker 5: so one of the things we've learned is that in
44:12
Speaker 5: the next version of advice, we really need to like
44:14
Speaker 5: double or increase that battery life, and so I think
44:16
Speaker 5: the next version is going to be.
44:17
Speaker 2: Double actually actually double without without increasing the charge.
44:22
Speaker 5: Correct in charging time, to double the battery of life,
44:25
Speaker 5: meaning you should get roughly eight hours of use, and.
44:27
Speaker 2: The goal is to actually get to all they use.
44:29
Speaker 2: So you can just charge, you know, maybe in your.
44:31
Speaker 1: Sleep, right, sleep and pillow exactly as soon as you've
44:34
Speaker 1: got like sixteen hours of usage, then you basically have
44:36
Speaker 1: twenty four hours of usage because you can charge while
44:38
Speaker 1: you're sleeping.
44:39
Speaker 5: One other thing that's important I think to call out
44:41
Speaker 5: your is if you're paralyzed, you can't you put the
44:43
Speaker 5: charger over your head yourself, And so it's important to
44:45
Speaker 5: think about like it's not just a duration of better use,
44:48
Speaker 5: but also can you recharge it yourself and dependently, So
44:51
Speaker 5: we spend a lot of time thinking about how to
44:52
Speaker 5: make that peaceable because then that means that you can
44:55
Speaker 5: this is what no one does. You can use the device,
44:57
Speaker 5: charge it, use the device, charge it, use the device
44:59
Speaker 5: with out needing anybody to come in and sort of
45:01
Speaker 5: help you with that, which is a big deal if
45:03
Speaker 5: you're trying to play set up until five am at
45:05
Speaker 5: night when your family.
45:07
Speaker 6: In the way in which he does that is that
45:09
Speaker 6: there is a charge of coil.
45:10
Speaker 2: That's a bigger you know about this big and we
45:13
Speaker 2: actually put it in the sleeve of a beanie or
45:18
Speaker 2: a beanie and then he wears it and then says
45:20
Speaker 2: with the voice command and.
45:22
Speaker 6: Charge charge your energize. That's the one you like.
45:31
Speaker 1: Howard writing work. So so uh yeah, the current device
45:38
Speaker 1: that Nolan has is reading, so it's trying to read
45:43
Speaker 1: his essentially like worst movement from from one one hand.
45:47
Speaker 1: That's also you know, worth pointing out like in the future,
45:51
Speaker 1: like and we're pretty cool to give Nolan the second
45:54
Speaker 1: implant that would allow the other hand to be used
45:56
Speaker 1: and also have higher uh obviously higher to ELECTORID accounts,
46:01
Speaker 1: so then you could play too, essentially play games two
46:03
Speaker 1: had it, because that's not only how you play games.
46:06
Speaker 1: And but then with with writing, it's really just it's
46:10
Speaker 1: an electrical impulse. Instead of like reading electrical impulses from
46:13
Speaker 1: the neuron to you issue an electrical impulse, which is
46:16
Speaker 1: obviously critical for visions. So vision is writing, which is
46:19
Speaker 1: just traggering electrical impulse in the vision part of the brain.
46:25
Speaker 1: And that like activates a pixel. So we actually do
46:29
Speaker 1: have this working in monkeys, and we had we've had
46:31
Speaker 1: it working with monkeys for a while now where you
46:34
Speaker 1: can sort of flash a pixel and then you watch
46:36
Speaker 1: where the monkey. Obviously the monkey is like whatever, monkey's
46:40
Speaker 1: a little surprised to see, like, hey, there's a flash
46:42
Speaker 1: here and a flash here, but it's it gets used
46:45
Speaker 1: to it after a while. But it just you can
46:47
Speaker 1: see that that the pixel is in the right location
46:50
Speaker 1: because the monkey's eyes will dark to that location.
46:52
Speaker 5: It's not on on the screen, like there's no pixel
46:54
Speaker 5: on the.
46:54
Speaker 1: Screen, there's no picks on the screen. On the brain. Yeah,
46:56
Speaker 1: just verify that that the that are triggering pixel in
46:59
Speaker 1: the right part of the brain. So you know, the
47:02
Speaker 1: initial resolution for vision will be relatively low, you know,
47:06
Speaker 1: sort of Atario graphics type of thing, but over time
47:09
Speaker 1: it could potentially be better than normal vision.
47:13
Speaker 2: And then I guess in terms of some additional applications
47:16
Speaker 2: for where writing to the brain can be useful, what
47:20
Speaker 2: are applications. As Bliss mentioned, there is feedback, there's appropriate
47:25
Speaker 2: sceptive feedback, there's a actile feedback.
47:27
Speaker 5: Especially robot arm if you're trying to grasp a cop right,
47:29
Speaker 5: you need to know about it.
47:31
Speaker 4: You want to an egg.
47:34
Speaker 2: It's very much a delicate balance of not just initiating
47:38
Speaker 2: the movement but getting the feedback and controlling it accordingly.
47:40
Speaker 2: So there is some metasensory cord text that's right adjacent
47:44
Speaker 2: to motor cortex that it could be better motor movements.
47:53
Speaker 1: So any changes in neural growth the devices and so
48:01
Speaker 1: did were don't see any any signs of neural damage,
48:06
Speaker 1: but I and I guess we we have seen some
48:09
Speaker 1: rebound on some of the electrodes, right.
48:15
Speaker 2: Correct, And then also I mean, I guess, I guess
48:17
Speaker 2: you know. Brain is very plastic, so it's not that plastic. Well,
48:22
Speaker 2: it does diminish quite a bit after each and why.
48:27
Speaker 3: Throughout throughout childhood, especially when you get to about twenty five,
48:34
Speaker 3: brain really done cooking. Yeah, but there are there is
48:43
Speaker 3: a little bit of damage done with each insertion, but
48:45
Speaker 3: it's a minuscule amount compared to anything else out there,
48:49
Speaker 3: and so it's an easy amount of damage recover from.
48:55
Speaker 3: And it's really only detectable on cutting pieces of the
49:00
Speaker 3: brain after after the animal is no longer alive, and
49:05
Speaker 3: looking at them under a microscope. You can't really tell
49:07
Speaker 3: during life that there's been any.
49:09
Speaker 5: Brain Another way to interpret this question, have the funded
49:13
Speaker 5: changes in neural growth after the device is inserted. One
49:16
Speaker 5: way to interpret that is like the user learning how
49:18
Speaker 5: to use the device, And I think on that side
49:19
Speaker 5: of things, there's been tremendous progress, and he's put in
49:23
Speaker 5: hundreds of hours trying to figure out the best way
49:25
Speaker 5: to use this device because he really thinks that you know,
49:26
Speaker 5: if you can figure this out and you can help
49:28
Speaker 5: share this knowledge. I mean, he's like on Friday night
49:31
Speaker 5: at eight pm, you know, he's starting a session of like,
49:34
Speaker 5: you know, figuring out himself how to how to push
49:36
Speaker 5: his own performance to the next level. And that's really
49:39
Speaker 5: a unique learning process because there's not many people in
49:41
Speaker 5: the world that had the experience of moving something with
49:43
Speaker 5: the brain, and so there's a lot of nuance to like, Okay,
49:46
Speaker 5: how exactly should I imagine or attempt to move my
49:48
Speaker 5: risk to get the thing to me.
49:51
Speaker 6: Yeah, he's really dialed that into it.
49:53
Speaker 2: Also, just a sheer number of hours that he us
49:56
Speaker 2: He's saying even in the past six months, right, Yeah. Anyways,
50:00
Speaker 2: Like I mean, he's using it in his travel, in
50:02
Speaker 2: his plain ride effectively PCI has left the lot.
50:05
Speaker 1: Yeah, yeah, yeah. I mean one of the questions is
50:07
Speaker 1: how close we can voting thoughts into text. I mean, right,
50:11
Speaker 1: right now, ask more about moving curse from the screen
50:12
Speaker 1: on a virtual keyboard. But long term, you should be
50:17
Speaker 1: able to really ask my entire words faster than anyone
50:20
Speaker 1: could possibly time.
50:22
Speaker 5: I'm able to type a little world today correctly, but
50:26
Speaker 5: we're still in the early days.
50:27
Speaker 2: Making that a pause experience it. I mean, the other
50:31
Speaker 2: things that we're looking at is sign language. Right at
50:33
Speaker 2: the end of the day, it is a movement of
50:35
Speaker 2: and into rights.
50:37
Speaker 1: Yeah, that's true.
50:40
Speaker 3: Was the brain trying to naturally push the threads out?
50:42
Speaker 3: I mean this is sort of a universal feature of
50:45
Speaker 3: any implant in the body. The body tries to reject it,
50:48
Speaker 3: and the goal of the surgeons and the technology team
50:51
Speaker 3: is to fight that. And so with artificial hips and
50:55
Speaker 3: with screws and the spine, we've done a really good
50:58
Speaker 3: job of finding biocompacts of AWL materials and techniques to
51:03
Speaker 3: fix those implants in the body. I mean, past a
51:07
Speaker 3: certain age, it's getting hard to find someone without some
51:09
Speaker 3: kind of implant, you know, knee, some kind of screws
51:13
Speaker 3: in their spine and so we've got this problem pretty
51:18
Speaker 3: well solved. So to answer your question, yes, the body
51:22
Speaker 3: is trying to get rid of any implant, but we
51:24
Speaker 3: can ensure that basically canned.
51:28
Speaker 2: It's also worth highlighting that the threads have not actually
51:31
Speaker 2: moved in the past five months. There's there's some still
51:36
Speaker 2: minor movements in terms of like some maybe maybe getting
51:38
Speaker 2: pushed in a little bit, pushed out a little bit,
51:40
Speaker 2: but it's it's more or less very stable and been
51:43
Speaker 2: stable for more And the reason for that is, you know,
51:46
Speaker 2: once you once you do a brain surgery, it takes
51:50
Speaker 2: some time for the tissues to come in and then
51:52
Speaker 2: and then you know, our tissue or the neo membrane
51:55
Speaker 2: to actually come in and then anchor the threads in place.
51:57
Speaker 2: And once that happens, it has been stable and seen
52:00
Speaker 2: much movement.
52:01
Speaker 3: That's where the world record performance starts to come in.
52:04
Speaker 6: Yeah, that was a couple of weeks ago.
52:08
Speaker 1: Yeah, the threads, like it is important that the threads
52:11
Speaker 1: be extremely tiny. If they're extremely tiny, then the brain
52:14
Speaker 1: does not The smaller they are, the less like the
52:17
Speaker 1: brain is to react to them. So that's why you
52:20
Speaker 1: want the thread to be extremely tiny and also to
52:22
Speaker 1: minimize any damage to your aunts.
52:27
Speaker 2: So on that note, we do plan to actually share
52:30
Speaker 2: some of the you know, the tissue response in detail
52:34
Speaker 2: and some of the later upcoming updates.
52:37
Speaker 1: Yeah, it is quite a challenging. It's challenging on many
52:41
Speaker 1: fronts to do something like this because you're trying to
52:45
Speaker 1: read and read and write electrical signals, but you need
52:48
Speaker 1: to have the threads themselves need to be electrically isolated
52:56
Speaker 1: and not subject to corrosion in the body. So like
52:59
Speaker 1: the you know, just middle by itself, it's so much
53:03
Speaker 1: sub subject to corrosion or being attacked. So it's it's
53:08
Speaker 1: it's good like in terms of the various coatings and things.
53:11
Speaker 1: To actually make this electrode work while not actually eroding
53:14
Speaker 1: its performance over time is very difficult. Human body is
53:17
Speaker 1: a very very harsh environment, very harsh envirotment.
53:19
Speaker 2: It's a bag of salt water with bad sensors, that's
53:23
Speaker 2: elevated temperature that is well regulated. I mean, I'm sure
53:26
Speaker 2: people have experienced dropping their electronic devices in a seawater
53:29
Speaker 2: and in an instant.
53:31
Speaker 1: Yeah, yeah, so we'll we're better sort of wrap this
53:35
Speaker 1: up soon and if there's like a few few questions.
53:40
Speaker 1: I guess, so a good question. So what about upgrades?
53:44
Speaker 1: So yeah, we we do think it's going to be
53:46
Speaker 1: important to be able to upgrade the device over time.
53:49
Speaker 1: Just like you wouldn't want like an iPhone one stuck
53:52
Speaker 1: in your brain forever. You know, if you've got an
53:56
Speaker 1: iPhone fifteen, you probably want the iPhone fifteen, not iPhone one.
54:00
Speaker 1: So I think people over time will be able to
54:05
Speaker 1: upgrade their their new links. We will take the neural
54:07
Speaker 1: link device out and put a new one in. And
54:11
Speaker 1: we we have done this with some of our animals
54:16
Speaker 1: and they're actually in one case we did it with
54:19
Speaker 1: we upgrade our device three times and and uh with
54:22
Speaker 1: a pig.
54:23
Speaker 6: We did with a monkey as well.
54:26
Speaker 1: Yes, and he's he's doing finely.
54:30
Speaker 6: Hit his I think his record the last.
54:32
Speaker 3: Yeah, with the with an upgrade, still beat him though.
54:36
Speaker 6: Still beat him.
54:37
Speaker 5: Yes, this is true. Humans are top of.
54:39
Speaker 1: The species board right now, pages like eight or.
54:41
Speaker 5: Something like eight point five bps. Okay, well that's a
54:45
Speaker 5: very high score. I'm not trying to put Patriot down.
54:47
Speaker 5: And also to train a monkey to do that, it's
54:48
Speaker 5: like a whole challenge on it. Something we have like
54:50
Speaker 5: the best animal care team world.
54:52
Speaker 1: Yeah, I just do want to have size. We we
54:54
Speaker 1: we we do our absolute best to take care of
54:56
Speaker 1: the animals and h when we had like a USDA
55:01
Speaker 1: inspector come through, she said that this was the nicest
55:05
Speaker 1: animal facility she has ever seen in her entire life.
55:08
Speaker 6: So breakfast.
55:13
Speaker 1: The monkey orders room service, Yes, yeah, we have monkey
55:17
Speaker 1: room service, which is rare. We're the only ones who
55:22
Speaker 1: offer monkey room service. So we really do everything we
55:25
Speaker 1: can to maximize well for the animals. So all right
55:29
Speaker 1: with that, thank you everyone for tuning in. I hope
55:32
Speaker 1: you found this interesting.