[00:00.340 --> 00:07.320]  So we have an OS problem, so I'm going to have to show it in this view.
[00:08.120 --> 00:29.780]  The thing about this talk that I want to emphasize is a lot of the work I'm going to show here is experimental, both in the artistic and the scientific, but it's going to give you an overview of what's coming down the pike in the next 10 years for research and both covert and overt about the brain.
[00:29.780 --> 00:35.460]  And a lot of people aren't aware of this, not because it's hidden, but because it's not their focus.
[00:35.900 --> 00:42.700]  So I'm going to merge both the art world and the science world and the surveillance world in this.
[00:42.820 --> 01:03.380]  And I'm going to finish, if there's enough time, with a one-minute clip of the world's first brain opera that I just premiered in Hong Kong, in which a performer launched video, a sonic environment, and a libretto just with their brain waves, interacting with the audience lifetime,
[01:03.840 --> 01:06.660]  as their brain waves were on view.
[01:07.780 --> 01:09.900]  So I'll show you that, too.
[01:10.800 --> 01:11.360]  Okay?
[01:16.230 --> 01:21.890]  So, the question is, is the brain the next hackable driver?
[01:22.230 --> 01:24.850]  And I'm sorry about this view, but it's the one I have to use.
[01:24.950 --> 01:27.090]  And the answer is, yes, it is.
[01:28.170 --> 01:32.850]  Most people think of this as that kind of EEG setup.
[01:33.150 --> 01:38.870]  But what's happened now is we've gotten BCIs, brain computer interfaces.
[01:39.410 --> 01:49.290]  And they basically take the signals from your brain, and you can control different things with them, prosthetic limbs, so on and so forth.
[01:49.450 --> 01:50.990]  That's the medical uses of them.
[01:50.990 --> 01:55.110]  Now there are consumer grade BCIs.
[01:55.530 --> 01:59.190]  And some of them are open-source, some of them are not.
[01:59.530 --> 02:05.430]  And they range in price from a little over $100 to $1,000.
[02:06.390 --> 02:08.610]  And you can buy them on the open market.
[02:09.490 --> 02:16.210]  Most people think of brain computer interfaces as fMRIs, also.
[02:16.570 --> 02:18.630]  You know, you go in, you get an MRI.
[02:19.290 --> 02:25.790]  And I'd like to show this one because it shows the brain effect differences between musicians and non-musicians.
[02:26.230 --> 02:27.550]  That there are differences.
[02:27.790 --> 02:29.330]  So that's what most people think of.
[02:29.330 --> 02:34.390]  However, BCIs began in the 1970s.
[02:35.010 --> 02:38.530]  And right at the start, DARPA was involved with BCIs.
[02:39.390 --> 02:40.950]  Right from the get-go.
[02:41.890 --> 02:44.370]  And they started at UCLA.
[02:46.230 --> 02:54.710]  And then there was early work mostly done on monkeys, rats, and so on and so forth.
[02:54.710 --> 03:00.090]  And we know that that led to other things.
[03:00.270 --> 03:06.110]  So I'm going to give you a five-second crash course in how an EEG works.
[03:06.110 --> 03:14.550]  So your brain puts out electrical signals, which is basically the difference in ionic charge with sodium and potassium.
[03:15.270 --> 03:22.170]  And if you can see in this graphic, the electrode translates the signal or the charge.
[03:23.910 --> 03:26.810]  How does the brain turn thoughts into action?
[03:26.810 --> 03:29.690]  Well, your neurons get a charge.
[03:29.870 --> 03:31.390]  They go to the next one.
[03:31.390 --> 03:35.270]  And it works for about 250th of a second.
[03:35.470 --> 03:40.330]  I'm not going into the science of it because that's not what this is about.
[03:42.310 --> 03:45.430]  And I'm going fast because I have a lot of things to show you.
[03:47.930 --> 03:56.830]  The skull in the EEG or an external EEG as opposed to an internal EEG blocks a lot of the signals that come into your brain.
[03:56.830 --> 04:00.810]  And the signal has to be smooth through an algorithm.
[04:01.070 --> 04:06.130]  And the algorithms are varied and have different degrees of efficacy.
[04:07.730 --> 04:10.450]  There's more machine learning now and algorithms.
[04:10.830 --> 04:13.890]  This is what's known as the basic 10-20 system.
[04:14.110 --> 04:19.270]  And that's what you saw that one of the first visuals with lots of lines coming out of the head.
[04:19.270 --> 04:22.990]  That was a 10-20 system that's used in most medical labs.
[04:24.930 --> 04:27.590]  However, this is really how it works.
[04:28.150 --> 04:33.050]  And a lot of you will love this because it's about signal acquisition.
[04:33.930 --> 04:35.810]  You digitize the signal.
[04:35.870 --> 04:37.210]  You process it.
[04:37.210 --> 04:38.370]  You translate it.
[04:38.450 --> 04:39.530]  You give it a command.
[04:39.870 --> 04:42.310]  And this, you have the BCI application.
[04:42.590 --> 04:46.290]  And I think that's pretty true of a lot of signal processing.
[04:46.290 --> 04:51.490]  But in this case, what you get is you get things like...
[04:52.550 --> 04:54.070]  You can change it.
[04:54.210 --> 05:07.330]  And you can turn it into lights, or colors, or video, or music, or tactile, or thermal, or in the future, scent, or odors, or this kind of thing, all from your brain.
[05:08.390 --> 05:10.230]  You can really do that.
[05:10.610 --> 05:14.250]  The thing is what kind of signal you're measuring is very important.
[05:14.250 --> 05:21.010]  Now, in terms of the actual physiology of your brain, you have your brain, your skull, the skin.
[05:21.590 --> 05:25.950]  There's invasive and non-invasive types of information.
[05:25.950 --> 05:30.190]  And as you know, it becomes digitized just into zeros and ones.
[05:30.550 --> 05:33.130]  But this is really good for people who want to...
[05:33.130 --> 05:41.070]  There are people who can now control wheelchairs and prosthetic limbs with this by thinking that they want to do something.
[05:41.070 --> 05:53.850]  And there's been a lot of work, especially with people with ALS, or Lou Gehring's disease, who at this point can't even blink their eyes, yet they can think a thought and something will happen.
[05:54.090 --> 05:54.330]  Wow.
[05:54.330 --> 06:00.970]  However, but in the art world, which I'm part of, BCIs go way back.
[06:00.970 --> 06:03.950]  They go back to 1965.
[06:03.970 --> 06:22.990]  And I'm going to show you a very short clip of Alvin Lussier's 1965 Music for Solo Performer, in which it was the first brainwave piece with an electrode that he attached to his scalp and he made drums play.
[06:23.930 --> 06:30.430]  And this was before there were really BCIs for the most part.
[06:33.750 --> 06:36.110]  So, let's just see if we get sound.
[06:37.870 --> 06:41.910]  And I'm going to let you see him and then I'm going to fast forward it.
[06:45.620 --> 06:49.480]  So that's Alvin Lussier with his little brain band.
[06:50.500 --> 06:54.260]  And he looks... he's concentrating and I'm going to fast forward this.
[06:54.840 --> 06:57.320]  Because there's a lot of setup, they show everything.
[06:57.660 --> 07:00.120]  And he also has something on his hand as well.
[07:01.640 --> 07:02.860]  This is the piece.
[07:03.020 --> 07:04.960]  This is from 1965 again.
[07:07.140 --> 07:10.080]  And here, you should...
[07:11.500 --> 07:13.060]  Is our audio working?
[07:17.650 --> 07:19.430]  I'm not getting an audio signal.
[07:19.590 --> 07:20.570]  Can someone come up here?
[07:20.710 --> 07:21.950]  The audio has been disabled.
[07:23.410 --> 07:25.010]  Are there any tech people here?
[07:27.290 --> 07:27.810]  No?
[07:28.050 --> 07:30.530]  Well, they disabled my audio.
[07:31.090 --> 07:31.690]  Someone's coming.
[07:31.850 --> 07:32.830]  Okay, good.
[07:33.170 --> 07:35.890]  Because they disabled my audio when they set me up.
[07:36.050 --> 07:37.410]  And I don't know what they did.
[07:38.990 --> 07:40.530]  It may actually be fine on your computer.
[07:40.710 --> 07:41.930]  That means you can't control the language.
[07:42.070 --> 07:46.950]  I know, but I would like the person who set it up to, you know, come and fix it.
[07:48.890 --> 07:49.830]  Okay, they're coming.
[07:50.410 --> 07:50.890]  All right.
[07:52.990 --> 07:53.790]  I'm gonna...
[07:53.790 --> 07:54.730]  Oh, okay.
[07:54.850 --> 07:55.870]  You're gonna come up and...
[07:55.870 --> 07:57.070]  Can you fix my audio, please?
[08:01.150 --> 08:02.730]  Are you getting it on the live stream?
[08:05.910 --> 08:07.870]  Yeah, please give me my audio back.
[08:09.210 --> 08:10.690]  Yeah, there is audio, I did say.
[08:12.470 --> 08:13.750]  Oh, you didn't plug it in.
[08:14.950 --> 08:16.710]  Okay, we have lots of plugs up here.
[08:16.710 --> 08:17.550]  So I wasn't...
[08:17.550 --> 08:18.650]  I'm not concentrating.
[08:18.710 --> 08:19.470]  Bring it up on the console.
[08:19.990 --> 08:20.190]  Okay.
[08:20.630 --> 08:21.730]  Because look, I'm not getting...
[08:21.730 --> 08:22.210]  Oh, there we go.
[08:22.930 --> 08:24.230]  And let's hear the sound.
[08:25.190 --> 08:25.530]  Thank you.
[08:28.310 --> 08:34.930]  So Alvin Lussier's brain waves are giving out vibrations which go into speakers here.
[08:35.190 --> 08:38.070]  We're still not getting sound, although I'm on full sound.
[08:38.310 --> 08:39.150]  Hold on one more second.
[08:39.330 --> 08:39.730]  All right.
[08:43.120 --> 08:46.740]  It's good we're doing it on this piece because there's other pieces I need to show you.
[08:58.910 --> 09:00.710]  Yeah, well, it's a sound.
[09:00.910 --> 09:02.450]  I don't know if it's the correct sound.
[09:02.850 --> 09:03.430]  Let's hear it.
[09:04.430 --> 09:05.450]  Yeah, it's the correct sound.
[09:05.490 --> 09:06.050]  You hear this?
[09:07.510 --> 09:11.550]  This is his brain waves in 1965.
[09:15.050 --> 09:15.450]  Okay.
[09:19.080 --> 09:22.380]  Okay, I just wanted you to hear that because now we have that.
[09:22.380 --> 09:30.200]  In 1973, John Cage began using brain wave signals also with Nam Joon Pike.
[09:30.420 --> 09:34.380]  I'm just trying to give you a sense of how far back this goes in the art world.
[09:35.740 --> 09:46.180]  David Rosenboom just premiered a piece at the Whitney last year in which he worked with four brains and produced a whole concert with that.
[09:46.480 --> 09:50.920]  So this was from 1972, his diagram.
[09:50.920 --> 09:59.760]  And what you can see is it's a system for four brains and they go into and make sound.
[10:00.000 --> 10:05.080]  And he realized this last year at the New Whitney in the meat market.
[10:05.080 --> 10:10.320]  So, now I want to switch back to DARPA.
[10:10.700 --> 10:26.700]  In 1973, the first term brain-computer interface was published by a scientist, Vidal, and that was also partially funded by DARPA.
[10:28.780 --> 10:38.920]  What happened after that is a man named Lawrence Farwell really jumped on it in the late 1980s.
[10:38.920 --> 10:49.000]  And he started making something called the P300 Mimer signal, which is a signal in your brain of recognition.
[10:49.000 --> 10:57.180]  And what that means is when you recognize something, like you see your dog or something, your brain spikes with recognition.
[10:57.180 --> 10:59.200]  And there's no way you can avoid that.
[10:59.500 --> 11:09.100]  It's automatic and it takes one three hundredth of a second, which is why it's called a Mimer response.
[11:09.100 --> 11:20.200]  And that's something the FBI, the CIA, the Navy, and everyone else has been involved with, with Farwell since it first began.
[11:20.480 --> 11:34.080]  So, what I'm trying to do with this presentation is show you both the art world and DARPA were sort of running neck and neck with working with BCIs, although they were unbeknownst to each other.
[11:37.560 --> 11:41.920]  Farwell is still very active and you can Google him.
[11:42.240 --> 11:46.780]  He has huge contracts out for the P300 signal.
[11:47.180 --> 11:55.340]  And that signal has already been tested in side channel attacks for neuro gaming, which I will talk about a little later.
[11:56.760 --> 12:03.720]  The brain produces basically four types of brain waves, delta, theta, alpha, and beta.
[12:04.840 --> 12:13.120]  They have different frequencies and most people know alpha, which is meditation, and beta, which is interest.
[12:13.840 --> 12:18.300]  Delta and theta are more about sleep or very deep concentration.
[12:18.300 --> 12:24.700]  I bet both of you are either in beta or not in beta right now, which means interest or non-interest.
[12:26.340 --> 12:32.100]  Again, this is a little more explanation about the frequency ranges of these signals.
[12:32.260 --> 12:38.200]  And that's a lot of what the hertz levels is a lot of what these commercial BCIs pick up.
[12:39.820 --> 12:44.120]  There's OpenBCI, which is the open-source version, and it's in development.
[12:44.300 --> 12:47.120]  It keeps being development where you can read things.
[12:47.440 --> 12:49.820]  That's the early prototype of the headset.
[12:49.820 --> 12:52.020]  That's the early board.
[12:52.320 --> 12:54.340]  They've updated it since then.
[12:54.660 --> 12:55.880]  It's Bluetooth enabled.
[12:56.180 --> 12:57.180]  You can buy one.
[12:57.340 --> 12:59.000]  You can use it on your own brain.
[12:59.240 --> 13:00.300]  It's pretty raw stuff.
[13:00.500 --> 13:02.980]  It doesn't have a lot of interpretation at this point.
[13:04.460 --> 13:08.520]  It's both dry point and paste, depending on what version you get.
[13:09.200 --> 13:16.940]  Commercial versions that you can buy are the Neurosky, which is the entry level one that gets alpha and beta.
[13:16.940 --> 13:18.180]  It's a little over $100.
[13:19.260 --> 13:25.500]  The Muse headset, which is very good for meditation, and it's a little easier to use.
[13:25.940 --> 13:32.840]  The Emotive, which I used for my brain opera, and has 14 points, and it's a little more robust.
[13:33.040 --> 13:34.680]  It's also a little more expensive.
[13:35.580 --> 13:43.740]  And it takes eight seconds to calibrate specific contours and folds of any individual brain.
[13:44.080 --> 13:45.760]  And the Emotive can do that.
[13:46.280 --> 13:53.800]  So I want to show you one or two things that have been done with the Emotive as a segue before we get into the heavy surveillance stuff.
[13:55.160 --> 14:01.760]  Yehudi Duenas, who's an artist, created the Ascent, which is basically you strap on a harness.
[14:02.000 --> 14:04.740]  It premiered in Sunset Park four years ago.
[14:05.060 --> 14:09.040]  And with your own concentration, you lift up in the harness.
[14:09.360 --> 14:11.900]  And if you lose concentration, you lower.
[14:12.100 --> 14:15.080]  But if you keep concentration, you lift up 30 feet.
[14:15.080 --> 14:21.080]  And bells go off, and confetti comes down.
[14:21.360 --> 14:22.720]  So it's sort of like a game.
[14:22.880 --> 14:26.160]  And I'm going to show you a quick one-minute clip of that.
[14:38.580 --> 14:40.260]  This was in Sunset Park.
[15:41.680 --> 15:53.720]  Okay, so Yehudi now wants to make this into a sort of traveling roadshow in which between eight and 24 people are meditating and levitating at the same time.
[15:53.720 --> 15:59.720]  You see, it kind of looks like something out of a sci-fi movie where the spirits are ascending to heaven.
[15:59.980 --> 16:01.840]  And that's his drawings about it.
[16:02.340 --> 16:06.540]  So that shows you a little bit of, you know, the fun part of it.
[16:07.360 --> 16:16.800]  Also, there was an installation in Toronto called My Virtual Dream with more than 500 adults wearing Muse wireless headsets.
[16:17.040 --> 16:22.380]  And they played a collective neurofeedback computer game.
[16:22.380 --> 16:24.300]  And that was in 2013.
[16:24.300 --> 16:25.660]  That was in a dome.
[16:26.000 --> 16:28.620]  There's also the Cyborg Foundation.
[16:29.620 --> 16:33.460]  Some of you may or may not know who Neil Harbison is.
[16:33.780 --> 16:34.840]  He wears...
[16:34.840 --> 16:38.300]  He was born without being able to see color.
[16:38.300 --> 16:48.120]  So he wears a device outside his head where he picks up 360 types of colors and turns them into sound waves inside his brain.
[16:49.060 --> 16:49.640]  Yeah.
[16:50.340 --> 17:00.600]  And here, you can see how he makes the visual to the chromatic, the sounds.
[17:00.860 --> 17:06.920]  And at first, he used to do things like meet celebrities and make pictures of them, how they sound.
[17:07.480 --> 17:11.420]  This is him also showing the color wheel of the sound.
[17:11.580 --> 17:14.300]  And that's Moon Rebus, also part of the Cyborg Foundation.
[17:15.380 --> 17:16.980]  And they're showing the...
[17:16.980 --> 17:19.220]  You can see at the top, it's a sound wave.
[17:19.440 --> 17:20.700]  And at the bottom, it's color.
[17:21.480 --> 17:24.180]  And that's sort of what happens to the brain.
[17:24.180 --> 17:32.220]  So, what Neil did is he did the world's first skull transmitted painting.
[17:32.620 --> 17:33.820]  Yeah, that's right.
[17:33.940 --> 17:43.100]  A skull transmitted painting in which a painting was done in Times Square, sent via Skype into a room.
[17:43.440 --> 17:46.800]  And what happened is it was uploaded into his brain.
[17:48.160 --> 17:51.640]  And I'm going to show you a little bit of that very quickly.
[17:53.280 --> 17:55.560]  I think you'll all find...
[17:55.560 --> 18:03.800]  So, basically we feel that all these sensors that we are developing with the Cyborg Foundation are very, very natural because they're related to sensors that already exist.
[18:04.240 --> 18:08.220]  It might have an antenna makes me feel closer to insects that already have antennas.
[18:08.380 --> 18:13.320]  Also, hearing through bone conduction makes me feel closer to dolphins that also hear through bone conduction.
[18:13.560 --> 18:17.540]  And perceiving infrared and ultraviolet makes me feel closer to insects that also perceive these colors.
[18:17.620 --> 18:32.960]  The second performance will be something we've never done before, Which is, I'll be connected to a white canvas that is now in Times Square, then passerbys will paint different colored lines on the canvas, and then this will be connected to my head.
[18:33.080 --> 18:41.740]  Because since a couple of years ago, I have Internet connection to my head, so I can now receive colors from other parts of the world.
[18:42.060 --> 18:48.200]  So there's like now five different people in the world that have connections that can send colors from different continents.
[18:48.200 --> 18:54.580]  So now I'm not only relying on the colors that I have in front, I can also sense colors that are way far away from here.
[18:54.740 --> 19:00.600]  So we're gonna test creating exactly the same painting at the same time, one in Times Square, one here.
[19:00.780 --> 19:09.840]  We'll be connecting through Skype so you can see the process live there, and I'll be here facing the other way so I can sense and paint exactly the same painting.
[19:09.980 --> 19:12.660]  It might not work, but as we said, it's experimental.
[19:12.860 --> 19:13.840]  We've never tried it before.
[19:14.500 --> 19:27.500]  They'll be picking up, they have been given specific instructions, so they have to paint a color line, and then I'll be receiving the color line in my head, and then I'll be painting exactly the same color line.
[19:27.640 --> 19:34.320]  So it will be seven people that will pick up any color, and then we'll be doing the lines, and it should be the same.
[19:34.320 --> 19:37.720]  So that was in Times Square, and it was projected on a wall.
[19:37.880 --> 19:39.140]  This is a video image.
[19:39.320 --> 19:41.860]  And they would zoom in to, like, the color green.
[19:43.200 --> 19:58.820]  And he, Neil, would hear the sine wave of the color green coming into his head, and then he would match the sine wave of the color green to the tube of paint without ever seeing it.
[19:58.820 --> 20:02.480]  He would just hear it through the Internet hookup in his head.
[20:02.820 --> 20:10.220]  So what he's doing is, he's putting his eye board to hear the sine waves on the different colors that he's experiencing.
[20:12.820 --> 20:14.940]  So this is all done wirelessly.
[20:20.440 --> 20:22.640]  And as you can see, he never turns around.
[20:35.740 --> 20:40.760]  So that's what's going on in Times Square, and then you can see how it's being reflected on his.
[21:13.800 --> 21:16.880]  The world's first skull transmission painting.
[21:17.220 --> 21:19.280]  He still can't see the colors there.
[21:20.300 --> 21:25.540]  So what's actually going to happen now is the canvas is going to come to the red door from Times Square.
[21:25.880 --> 21:26.540]  It's on its way.
[21:26.880 --> 21:29.540]  And we'll be able to actually compare them side by side.
[21:32.260 --> 21:37.580]  He was picking up a lot of the infrared in Times Square, because he can also pick up infrared.
[21:38.380 --> 21:42.140]  So he messed orange and red a little bit in this picture.
[21:45.240 --> 21:55.180]  Eventually, somewhere else in the world, someone could be creating a painting, and it's going straight to your brain, and you're like in synergy and in one with it.
[21:55.560 --> 21:59.680]  This is Anita Marunta, who was the director of our Times Square operation.
[22:03.400 --> 22:04.280]  All right.
[22:06.060 --> 22:08.800]  So one has more orange and more red.
[22:08.900 --> 22:10.220]  Just ignore the second one.
[22:14.100 --> 22:19.200]  Okay, so I just wanted you to see that, because that's the fun stuff.
[22:19.480 --> 22:21.000]  That's the art stuff, right?
[22:21.240 --> 22:23.420]  So now we're going to get a little more serious.
[22:27.060 --> 22:32.360]  And remember, what was it, HOPE, with Ed Snowden and Daniel Ellsberg?
[22:32.700 --> 22:43.240]  Well, we all know that if you think what's been done with network technology has been something, wait till you see what's coming with the brain.
[22:46.020 --> 22:55.060]  So most governments in the world right now have ten-year or more brain initiatives to map every neuron in the human brain.
[22:56.580 --> 23:06.640]  That includes the U.S. government, the European Union, the Australian government, the Japanese government, the Israeli government, the Canadian government, and a little bit the Chinese government.
[23:07.430 --> 23:12.060]  Probably the amounts with all those governments is in the trillions of dollars.
[23:12.660 --> 23:16.950]  In terms of the United States government, there's the Obama Brain Initiative.
[23:16.950 --> 23:22.950]  And if you follow the money, DARPA is getting half the money for that.
[23:23.590 --> 23:30.030]  The National Institute of Health is getting a quarter of it, and the National Science Foundation is getting the rest.
[23:30.350 --> 23:39.190]  And there's so much money that DARPA just went and combined a lot of other agencies and created a new agency, or merged agency, I would say.
[23:40.730 --> 23:48.250]  And I'll show you that in a minute, but they have now a special section of DARPA called the Biological Technologies Office.
[23:49.250 --> 23:51.370]  Just... and they're very user-friendly now.
[23:51.670 --> 23:54.850]  They want you to reach out and touch somebody.
[23:55.430 --> 23:57.250]  They solicit proposals.
[23:57.250 --> 24:07.070]  So here, it's like they send themselves, says, Hi, we want to give you opportunities, and we want to crowdsource, and we want to love you.
[24:07.250 --> 24:11.970]  And it's very shocking how out there they are with it.
[24:12.050 --> 24:13.530]  And you can go on the site and look.
[24:14.870 --> 24:16.430]  Here is iARPA.
[24:16.570 --> 24:22.730]  And that's the new wing that came out, not just of DARPA, of many different agencies.
[24:22.730 --> 24:27.730]  It's called the Intelligence Advanced Research Projects Activity.
[24:28.010 --> 24:34.290]  And you can go on their site, and you can see just how brain-specific it is.
[24:34.430 --> 24:38.010]  If you look, it says, iARPA joins White House Brain Initiative.
[24:38.170 --> 24:39.730]  They've got press releases.
[24:40.150 --> 24:42.690]  They've got, you know, outreach.
[24:42.690 --> 24:44.810]  They want to go into schools.
[24:45.090 --> 24:47.390]  They want to have kids hack the brain.
[24:47.750 --> 24:50.250]  I mean, they're really out there with it.
[24:50.250 --> 24:53.230]  Which is both wonderful and a little strange.
[24:54.690 --> 24:57.110]  Here they have seedling information.
[24:57.430 --> 24:59.410]  Submit your ideas to us.
[24:59.570 --> 25:02.790]  Here's what we seeded before, you know.
[25:03.110 --> 25:08.370]  And you can, all of you, if you're interested, you can go on and see what they're looking at.
[25:09.050 --> 25:19.910]  Now, why I find this particularly strange is because I want to show you just the state of brain fingerprinting at this moment.
[25:21.110 --> 25:22.410]  Fingerprinting your brain.
[25:29.300 --> 25:32.760]  Now that you've been practicing this more, Mavi, can you see it when it starts to entrain?
[25:33.080 --> 25:38.940]  Brain print is a technique for identifying people with their brain activity.
[25:39.280 --> 25:41.780]  And can you just adjust him a little, side to side?
[25:41.920 --> 25:51.380]  We record people's brain activity while they look at a series of images that are designed to elicit very unique responses from person to person.
[25:51.380 --> 26:01.220]  So you might see a picture of a sushi, and a picture of a pizza, and a picture of Adam Sandler, and a picture of Kim Kardashian, and the word conundrum, and a whole bunch of things rapidly flashing at you.
[26:01.400 --> 26:02.540]  Take a reading, 25.
[26:03.000 --> 26:03.660]  Where did you put it?
[26:03.900 --> 26:04.520]  Right there.
[26:05.020 --> 26:17.000]  So when you take hundreds of those, where every person is going to feel differently about each individual one, then you can be really accurate in identifying which person it was that looked at them just from their brain activity.
[26:18.140 --> 26:22.320]  The key idea of brain printing is we want to establish a new biometric.
[26:22.620 --> 26:28.140]  The concept of biometric, you may be more familiar with like a fingerprint, iris recognition, face recognition.
[26:28.480 --> 26:35.620]  Those biometrics, they are good, they are very accurate in identifying individuals, but they have several key limitations.
[26:35.980 --> 26:36.960]  So, look at that.
[26:37.480 --> 26:43.920]  Using EEG or brain activity as a way of biometrics, you know, has been proposed before.
[26:43.920 --> 26:48.620]  But the difference is previously people more focused on active thinking.
[26:48.940 --> 26:54.880]  We are focusing on the more challenging, but more unique non-volitional, intuitive brain response.
[26:58.820 --> 27:01.760]  Non-volitional also is that P300 signal.
[27:01.760 --> 27:08.380]  We want to identify and recognize the individual person based on their inside thinking, okay, inside brain activity.
[27:08.580 --> 27:09.960]  That is non-visible.
[27:10.300 --> 27:13.440]  That means even the user cannot be aware of that.
[27:14.460 --> 27:20.200]  My collaboration with Zheng Peng Jin is kind of unique because he's an engineer and I'm a psychologist.
[27:20.900 --> 27:24.400]  And most of the past attempts at brain biometrics have been made just by engineers.
[27:24.660 --> 27:31.360]  And that collaboration has been really fruitful because it crosses cybersecurity and biometrics and cognitive neuroscience and psychology.
[27:31.680 --> 27:37.900]  And I think it's why our brain print protocol is more accurate and faster than any of the existing protocols.
[27:37.900 --> 27:40.100]  You can do that in the brain, too.
[27:40.320 --> 27:46.040]  My graduate student, Maria Ruiz Blondet, has shown that we can achieve 100% identification accuracy.
[27:46.200 --> 27:47.160]  And that was a huge breakthrough.
[27:47.520 --> 27:48.580]  100% accuracy.
[27:48.580 --> 27:54.220]  This idea is based on the scientific rationale, okay, everyone is supposed to have the unique memory and knowledge.
[27:54.480 --> 27:55.000]  We're unique.
[27:55.180 --> 27:56.000]  Our brains are unique.
[27:56.140 --> 27:56.940]  Our thoughts are unique.
[27:57.060 --> 27:57.980]  Our feelings are unique.
[27:57.980 --> 27:59.960]  It's really, like, quantifiable.
[28:00.380 --> 28:02.740]  You are not the same as any other person.
[28:03.040 --> 28:12.580]  We definitely see the applications of brain print as being used for access to very high security locations, like the Pentagon.
[28:13.280 --> 28:19.270]  So we think the brain print will definitely represent the very promising next generation biometrics.
[28:20.040 --> 28:21.940]  Have a seat, young research assistant.
[28:23.940 --> 28:29.280]  Okay, so next generation biometrics, 100% accurate, your brain print.
[28:29.880 --> 28:33.380]  Think about the implications of that, okay?
[28:34.240 --> 28:35.880]  So there's more.
[28:36.240 --> 28:38.160]  That's just the tip of the iceberg.
[28:40.020 --> 28:45.120]  If you look at this picture, there's a picture of a newscaster's face.
[28:46.140 --> 28:56.340]  And there is a clip reconstructed from different MRI parts of the brain, which is now called the semantic brain lighting up.
[28:57.180 --> 29:01.700]  This has also been used for dream reconstruction, mostly in Japan.
[29:02.700 --> 29:05.820]  And I'm going to show you just a little bit of what that's about.
[29:06.040 --> 29:14.220]  Now, if you look at this, and in the next slide we'll explain this more, this is a map of what's called the semantic brain.
[29:14.220 --> 29:22.560]  What research is finding out is your brain is not right and left, but that it's very coded for specific things.
[29:22.840 --> 29:26.860]  And I'm going to let this explain it much better than I can.
[29:27.000 --> 29:29.200]  This is the actual researcher.
[29:30.500 --> 29:36.880]  Hi, my name is Alex Huth, and I'm going to tell you about a study I did here at UC Berkeley, along with Shinji Nishimoto, Ann Vu, and Jack Gallant.
[29:37.600 --> 29:42.480]  When you look around the world, your brain is busily recognizing and categorizing everything that you see.
[29:43.000 --> 29:46.340]  So we presume that different categories of things are represented differently in the brain.
[29:46.620 --> 29:49.300]  But that doesn't mean that they're represented in completely different ways.
[29:49.980 --> 29:51.840]  Take cars and motorcycles for example.
[29:52.160 --> 29:53.240]  They share a lot of features.
[29:53.460 --> 29:56.080]  They both have wheels, they both have headlights, they both go on roads.
[29:56.360 --> 29:57.860]  They fulfill pretty much the same functions.
[29:58.800 --> 30:01.820]  So you might think that they're represented pretty similarly in the brain.
[30:02.500 --> 30:06.320]  But how can we extend this idea to all the thousands of categories that we can see?
[30:06.840 --> 30:14.220]  One way would be to use a semantic space, where nearby categories are semantically similar, but distant categories are semantically very different.
[30:14.840 --> 30:16.820]  So what does the brain's semantic space look like?
[30:17.200 --> 30:21.480]  Which of the thousands of categories that we can see are represented similarly in the brain, and which aren't?
[30:22.300 --> 30:33.860]  To answer these questions, we put five human subjects in an MRI machine, and recorded their brain activity while we showed them two hours of movie trailers, where we had labeled all the objects and actions that appeared during each second of the movies.
[30:34.600 --> 30:39.020]  A functional MRI measures blood flow in little rectangular areas called voxels.
[30:39.680 --> 30:45.360]  Our voxels are about 2 by 2 by 4 millimeters, and it takes about 30,000 of them to cover the entire cortex.
[30:46.380 --> 30:52.220]  So our next step was to determine how each category of object and action affected the blood flow in each voxel.
[30:52.920 --> 30:55.280]  We did this using regularized regression analysis.
[30:56.180 --> 31:03.820]  What we're left with is a model that tells us how each of the 30,000 voxels responds to each of the 1,705 categories in the movies.
[31:04.820 --> 31:11.020]  Then, we used principal components analysis, or PCA, to recover a semantic space from each subject's brain.
[31:12.200 --> 31:20.160]  PCA gives us a low-dimensional space, where categories that are represented similarly in the brain are close together, but categories that are represented very differently tend to be far apart.
[31:21.460 --> 31:26.180]  Interestingly, we found that the first four dimensions seem to be shared across our subjects.
[31:26.520 --> 31:36.460]  Now, of course, our true shared semantic space has more than four dimensions, but we're limited here by the size of our stimulus set and the resolution of fMRI.
[31:38.360 --> 31:40.560]  So, what do the shared semantic dimensions look like?
[31:40.860 --> 31:48.880]  We're going to visualize them here using a graphical structure drawn from WordNet, which is a semantic taxonomy painstakingly hand-constructed by a team of linguists.
[31:48.880 --> 31:50.940]  So, your brain categorizes things in locations in the cortex.
[31:51.060 --> 31:57.640]  Each category is shown as a node, and the links between categories show is-a relationships, like an athlete-is-a-human.
[31:58.620 --> 32:02.580]  Verb categories, like communication verbs and movement verbs, are shown as separate trees.
[32:03.580 --> 32:05.740]  Here we're showing the first shared semantic dimension.
[32:06.260 --> 32:12.760]  It seems to distinguish things that move, like people, animals, and vehicles, from things that don't, like buildings in the sky.
[32:13.400 --> 32:18.500]  This is not really surprising, since we know that bright, fast things just tend to elicit more activity in the visual cortex.
[32:19.520 --> 32:26.120]  The next three dimensions distinguish categories associated with people, categories associated with civilization, and biological categories.
[32:26.280 --> 32:27.620]  So, your brain is more organized than you think.
[32:27.620 --> 32:30.860]  But instead of looking at each dimension separately, we can visualize all three at the same time.
[32:32.360 --> 32:35.140]  Now we've colored each node to show where it lies in each dimension.
[32:35.680 --> 32:41.780]  The red component of the color is set by the second dimension, the green component is set by the third, and the blue component is set by the fourth.
[32:42.580 --> 32:55.220]  So for example, a category that's high on the second and third dimensions, but low on the So if you see a mammal and it lights up, that can be detected in an fMRI and it can begin to be reconstructed, which is what they did.
[32:55.360 --> 33:00.560]  Here we see that people and communication verbs are represented similarly, and animals are also not so different.
[33:01.560 --> 33:05.980]  Vehicles and buildings and movement verbs are also pretty similar to each other, but different from people and animals.
[33:07.220 --> 33:13.320]  This movie is showing the same data, but we have also positioned each node according to where it lies, on the second, third, and fourth dimensions.
[33:14.460 --> 33:17.460]  So now we have a pretty good idea of what the brain's semantic space looks like.
[33:18.060 --> 33:23.100]  We can use this semantic space to map out how different categories are represented in different parts of the brain.
[33:24.180 --> 33:28.740]  The first step is to take a high-resolution anatomical MRI of each subject's brain.
[33:29.520 --> 33:32.960]  Then we use this image to construct a 3D model of the cortical surface.
[33:33.560 --> 33:37.400]  Then we flatten out the cortical surface so that we can see the entire cortex at the same time.
[33:38.120 --> 33:42.400]  Now we can color in each voxel according to what part of the semantic space it's selected for.
[33:43.560 --> 33:45.380]  These colors are the same that we saw earlier.
[33:45.720 --> 33:50.660]  Green is for humans, yellow is for other animals, pink and violet are for vehicles, and dark blue is for buildings.
[33:50.680 --> 33:52.100]  And this can be used for your dreams also.
[33:52.360 --> 33:56.220]  What we see is a really complex pattern of semantic selectivity throughout higher visual cortex.
[33:56.820 --> 34:02.760]  These patterns are consistent with the well-known semantically selective brain areas, like the fusiform face area and peri-hypocampal place area.
[34:03.180 --> 34:06.100]  But we also see semantic selectivity in much of the surrounding cortex.
[34:08.100 --> 34:11.900]  Now if you find this as exciting as I do, you'll be itching to play with this data yourself.
[34:12.840 --> 34:14.040]  Well, you're in luck.
[34:14.260 --> 34:17.480]  We've made an online viewer where you can see the flattened and 3D brain maps.
[34:17.640 --> 34:20.940]  You can click on each voxel to see exactly which categories it represents.
[34:21.180 --> 34:24.860]  Or you can click on each category to see exactly how it's represented across the cortex.
[34:26.140 --> 34:30.400]  This viewer is available now at gallantlab.org slash semanticmovies.
[34:30.400 --> 34:31.740]  So that's it.
[34:31.960 --> 34:34.160]  If you want to learn more, read the paper at Neuron.
[34:34.380 --> 34:35.260]  And thanks for listening.
[34:36.000 --> 34:50.180]  Okay, so what this basically means is that your brain is organized into categories that at this point through an fMRI, but at another point through other means, can be tracked.
[34:50.180 --> 35:01.500]  So if you're someone and you're dreaming and you're in a situation where it can be tracked, people can tell what you're dreaming at this point very vaguely.
[35:02.220 --> 35:04.200]  It's going to take a lot more research.
[35:04.600 --> 35:13.380]  Or if you're thinking of something or being shown something, you cannot fake that you recognize it or don't recognize it.
[35:13.380 --> 35:15.260]  So that's the state.
[35:15.480 --> 35:22.080]  And now there's something called optogenetics in which false memories can be implanted.
[35:22.360 --> 35:25.320]  And they are being implanted into mice.
[35:25.920 --> 35:28.440]  Optogenetics is a science of light.
[35:29.000 --> 35:31.260]  And it's a very recent science.
[35:31.260 --> 35:33.040]  It only came out in 2011.
[35:34.000 --> 35:45.660]  And basically how it works is it takes light-sensitive protein from algae, puts it into DNA in specific neurons in the brain, and transplants it.
[35:45.840 --> 35:52.600]  Right now it turns on and off with special light that's blue turns it on and red turns it off.
[35:52.840 --> 36:03.780]  So what that means is you could have a mouse and you could have a happy mouse and then you shock it and it's in a box.
[36:03.780 --> 36:07.480]  And you optogenetically manipulate it.
[36:07.620 --> 36:11.880]  And you take it out of the box and put it in a grassy field.
[36:12.060 --> 36:22.360]  And the mouse is running around in the grassy field and you turn the blue light on it and the mouse goes into shock because it thinks it's back in the box getting the shock.
[36:22.520 --> 36:24.240]  But actually nothing's happening.
[36:24.440 --> 36:26.140]  It's running around in the grassy field.
[36:26.340 --> 36:31.320]  So a false memory has been implanted into the mouse and been turned on.
[36:32.460 --> 36:35.160]  This is in an experimental stage right now.
[36:35.360 --> 36:38.760]  It takes a lot of genetic manipulation, but it works.
[36:40.960 --> 36:47.140]  The other thing that's going on is DARPA is developing something called the cortical modem.
[36:47.280 --> 36:58.200]  And the cortical modem will be the size of a nickel and it will be implanted right into your brain, giving you images directly into the brain with no other interface.
[36:58.200 --> 37:03.320]  Right now it's being used in zebrafish because, again, it's another genetic situation.
[37:04.040 --> 37:11.420]  It's being used because they think that'll be a good way for soldiers to communicate from just brain to brain.
[37:11.580 --> 37:13.340]  And that's why it's being developed.
[37:13.860 --> 37:17.900]  Now, I found out about this and DARPA had a conference in New York last year.
[37:18.180 --> 37:21.100]  And they wanted people to reach out and do something.
[37:21.100 --> 37:23.720]  So I thought, okay, I'll reach out and do something.
[37:23.760 --> 37:24.600]  So I wrote to them.
[37:24.980 --> 37:26.560]  I said, hi, DARPA.
[37:26.720 --> 37:28.080]  I want to come to your conference.
[37:28.980 --> 37:31.680]  And they were like, yeah, sure.
[37:32.000 --> 37:33.700]  And here's the email I got.
[37:33.840 --> 37:36.020]  And they said, I will forward your email.
[37:36.200 --> 37:37.240]  Never heard from them again.
[37:38.300 --> 37:38.700]  Okay?
[37:38.900 --> 37:41.760]  But I did reach out just like they wanted you to, right?
[37:41.820 --> 37:43.880]  I said, I want to see this cortical modem.
[37:44.180 --> 37:44.660]  Uh-uh.
[37:45.100 --> 37:46.260]  No way, right?
[37:46.860 --> 37:47.420]  Okay.
[37:47.980 --> 37:51.260]  So then there's also cloud brain.
[37:51.580 --> 37:54.760]  So now your brain data is being uploaded to the cloud.
[37:54.960 --> 37:55.600]  Did you know that?
[37:55.740 --> 37:57.620]  Not your brain data, but brain data.
[37:57.820 --> 38:02.040]  And there will be more and more of this, especially as neuro gaming goes on.
[38:03.800 --> 38:05.640]  So what is that about?
[38:06.080 --> 38:10.600]  There was the experimental tech conference and expo.
[38:10.700 --> 38:11.700]  They're around.
[38:11.900 --> 38:13.220]  They happen every year now.
[38:13.220 --> 38:19.400]  And that's where you get the people who are working on cloud brain.
[38:19.840 --> 38:21.880]  And brain information in the cloud.
[38:22.220 --> 38:24.500]  And the people who are designing the systems.
[38:24.760 --> 38:27.140]  And I'm not going to go into all the systems that exist.
[38:27.380 --> 38:31.300]  But I will say that some are already in the startup phase.
[38:31.600 --> 38:34.480]  And are being well funded by venture capital.
[38:35.680 --> 38:38.400]  So that's... you can check out neuro gaming conferences.
[38:39.220 --> 38:39.800]  They exist.
[38:39.920 --> 38:41.180]  They're usually on the west coast.
[38:41.180 --> 38:45.680]  I think we're going into our third or fourth year of neuro gaming conferences.
[38:46.060 --> 38:50.060]  And that means your brain data is stored in the cloud or can be.
[38:50.220 --> 38:53.180]  And probably will be, especially for gaming.
[38:53.980 --> 38:59.380]  So that leads me to the final part of what I came here to talk about.
[38:59.380 --> 39:02.160]  Which is my own brain opera.
[39:02.160 --> 39:03.500]  And whoops.
[39:03.840 --> 39:05.100]  Consciousness is misspelled.
[39:05.300 --> 39:06.020]  Sorry about that.
[39:06.200 --> 39:06.480]  Anyway.
[39:07.400 --> 39:11.720]  Is there a place in human consciousness where surveillance cannot go?
[39:11.960 --> 39:16.840]  And I'm going to show you a very short excerpt from this.
[39:17.000 --> 39:21.280]  In which the main actor had four emotions register.
[39:21.280 --> 39:26.740]  Which were interest, excitement, meditation, and frustration.
[39:27.340 --> 39:34.140]  And these four emotions launched databanks of images that were...
[39:34.140 --> 39:38.700]  And sounds and words that were around those four emotions.
[39:38.700 --> 39:43.320]  In front of a live audience in a 360 degree theater.
[39:44.540 --> 39:46.280]  As she interacted.
[39:46.540 --> 39:47.960]  And I'm putting together the video.
[39:48.140 --> 39:51.060]  But I have a one minute clip I can show you right here.
[39:51.320 --> 39:51.820]  Very quickly.
[39:52.680 --> 39:56.040]  And I'll try and talk a little bit about what's going on.
[39:56.680 --> 39:58.160]  And you'll see bubbles.
[39:58.500 --> 40:01.300]  And the bubbles are different colors of brain states.
[40:01.460 --> 40:03.440]  Like you see those pink bubbles?
[40:03.920 --> 40:06.760]  Those pink bubbles are excitement.
[40:06.760 --> 40:08.540]  Yellow is interest.
[40:09.060 --> 40:10.400]  I colored them blue.
[40:10.820 --> 40:12.380]  Turk boys is meditation.
[40:12.640 --> 40:13.520]  And red is frustration.
[40:14.000 --> 40:15.140]  So here...
[40:15.140 --> 40:16.480]  I had to be very careful.
[40:16.720 --> 40:17.580]  I was hiding.
[40:17.980 --> 40:19.020]  All time.
[40:22.420 --> 40:25.420]  So here she's touching someone.
[40:25.980 --> 40:27.200]  Frustration is launching.
[40:29.700 --> 40:31.220]  That's the libretto.
[40:32.100 --> 40:33.420]  Sonic environment.
[40:35.400 --> 40:36.840]  We were betrayed.
[40:36.840 --> 40:38.480]  One by one.
[40:40.500 --> 40:42.520]  The life that I have is yours.
[40:52.160 --> 40:55.720]  Other members of the cell gave Noor up.
[40:56.260 --> 40:58.820]  Her code name was Madeline.
[40:59.280 --> 41:00.060]  Concentration counts.
[41:01.680 --> 41:02.760]  My father.
[41:03.320 --> 41:04.260]  My house a blessing.
[41:13.610 --> 41:18.590]  Okay, so in that clip, I was talking to the performer.
[41:18.590 --> 41:23.030]  But the back words like concentration camps.
[41:23.190 --> 41:23.970]  My house a blessing.
[41:24.210 --> 41:24.910]  That's the libretto.
[41:25.350 --> 41:26.750]  And her brain was triggering that.
[41:27.650 --> 41:29.550]  The images you were seeing.
[41:30.050 --> 41:32.690]  Her brain were triggering all those images.
[41:33.550 --> 41:35.090]  The dots you were seeing.
[41:35.530 --> 41:37.810]  Her brain was triggering all of that.
[41:38.370 --> 41:40.490]  Depending on what was going on in the story.
[41:40.490 --> 41:42.970]  And this performance took about 20 minutes.
[41:43.230 --> 41:46.030]  No one can wear a headset much longer than that.
[41:46.150 --> 41:49.610]  Without having a pain in the brain.
[41:50.050 --> 41:54.930]  And so to the best of my knowledge, it was the world's first brain opera.
[41:54.930 --> 41:59.870]  So where does that leave us in terms of what's going to happen?
[42:00.170 --> 42:04.590]  Well, there is one group that's taking responsibility for this.
[42:04.790 --> 42:11.150]  They're called CEREB, the Center for Responsible Brainwave Technologies.
[42:11.150 --> 42:12.550]  And there's their link.
[42:12.710 --> 42:21.150]  And they're beginning to put out publications about the ethics of brainwave technology and brainwave technology privacy guidelines.
[42:21.670 --> 42:26.150]  Let's hope people follow their instructions.
[42:26.150 --> 42:34.450]  But in terms of governments that don't adhere to this, it's an open question.
[42:35.090 --> 42:37.690]  So that's basically my presentation.
[42:38.070 --> 42:38.930]  And now, questions?
[42:47.320 --> 42:49.640]  Where are your slides posted?
[42:50.060 --> 42:51.100]  They're not.
[42:52.340 --> 42:53.320]  Will they be?
[42:53.980 --> 43:00.820]  I have to think about that because a lot of this is still in a position for me to publish about it.
[43:01.000 --> 43:04.060]  And I am publishing a lot about it in peer-reviewed journals.
[43:04.400 --> 43:08.480]  So that's why I'm not being too public about it yet.
[43:09.500 --> 43:10.280]  That's why.
[43:10.760 --> 43:15.520]  But this is all information that you can pretty much source or Google.
[43:15.820 --> 43:19.720]  There's not great secrets here, especially about DARPA and IARPA.
[43:24.570 --> 43:25.290]  Yeah.
[43:25.830 --> 43:28.970]  Which country's brain initiatives do you think are the most advanced?
[43:29.530 --> 43:36.130]  I personally think that the United States is going full tilt for brain initiatives.
[43:36.550 --> 43:43.890]  And I think that there's a very public face to IARPA and I think there's a very non-public face.
[43:43.890 --> 43:47.870]  And I can't penetrate it because I don't have security clearance.
[43:48.970 --> 43:49.610]  Yeah.
[43:49.810 --> 43:58.830]  How far are we from technology rather than having a helmet than doing a extensive radio training?
[43:58.830 --> 44:02.790]  That the military is designing...
[44:02.790 --> 44:09.910]  How far are we from, did I get this right, wearing a helmet instead of a physical helmet?
[44:10.470 --> 44:15.070]  Well, basically the military, that's a great question, is designing those helmets.
[44:15.490 --> 44:21.950]  And they're designing those helmets already for soldiers to have direct brain-to-brain communication.
[44:22.750 --> 44:27.950]  Which they feel will be much easier and much less open to misinterpretation.
[44:28.770 --> 44:30.270]  So that is in the design stage.
[44:30.850 --> 44:32.030]  You're right about that.
[44:33.010 --> 44:33.090]  Yeah.
[44:33.090 --> 44:40.350]  Do you have to wear a helmet to read someone's brain waves?
[44:41.490 --> 44:46.390]  That is really in the future if you don't have to wear a helmet to read someone's brain waves.
[44:46.550 --> 44:52.010]  But by the end of these 10-year initiatives by all these countries, I don't have an answer for that.
[44:52.450 --> 44:54.570]  The research is going very quickly.
[44:56.070 --> 44:56.570]  Yeah.
[44:58.930 --> 44:59.430]  Yeah.
[45:00.630 --> 45:04.830]  There's a great book by Annie Jacobson called The Pentagon's Brain.
[45:05.450 --> 45:09.190]  Annie Jacobson's The Pentagon's Brain, which was just mentioned.
[45:09.450 --> 45:13.270]  I reference that a lot in my written publications.
[45:13.730 --> 45:19.550]  Annie Jacobson just wrote a history of DARPA, which no one has ever done.
[45:20.150 --> 45:22.450]  And Annie mentions this initiative.
[45:22.450 --> 45:31.950]  And she mentions how terrified she is by this initiative because it's being run by the same people.
[45:32.690 --> 45:39.210]  Boeing, Lockheed, all of those people are now on the advisory board.
[45:39.410 --> 45:46.670]  So it's the people running it who sell the equipment to the military industrial complex.
[45:46.670 --> 45:48.850]  And it's a closed loop now.
[45:49.030 --> 45:51.510]  So that's why it's a little frightening.
[45:51.510 --> 45:54.410]  But Annie Jacobson's book is fantastic.
[45:54.750 --> 45:59.770]  She had a speculation in the book about soldiers having more invasive brain-machine interfaces.
[46:00.410 --> 46:01.690]  Do you share that speculation?
[46:01.970 --> 46:09.090]  The question is she had something about soldiers having more invasive brain-computer interfaces.
[46:10.210 --> 46:16.590]  I think that there will be, but it will be very covert and it will be very volunteer.
[46:16.890 --> 46:19.470]  And we won't really know about it.
[46:19.470 --> 46:23.210]  If that's the dark hole of DARPA-IARPA, that's going to be classified.
[46:23.630 --> 46:24.990]  But yeah, I agree with her on that.
[46:25.910 --> 46:26.230]  All right.
[46:26.950 --> 46:27.430]  Yeah.
[46:28.090 --> 46:34.630]  How far can a device be from the head to capture brain waves?
[46:38.290 --> 46:44.170]  Well, are you talking about MRI or EEG brain waves?
[46:47.530 --> 46:50.710]  EEG is you mount it on your skull.
[46:51.010 --> 46:55.970]  And at this point in an MRI, you have to go into a big machine.
[46:56.210 --> 46:59.120]  Although I'm sure those things will be miniaturized.
[47:00.700 --> 47:01.260]  Yeah.
[47:01.660 --> 47:08.960]  I should really include the Movies In Your Mind project from, I'm not saying I'm going to its presentation.
[47:09.600 --> 47:10.340]  Have you ever seen it?
[47:10.700 --> 47:11.920]  Movies In Your Mind?
[47:12.120 --> 47:14.180]  Not quite unless it's from Jack Gallant's lab.
[47:14.600 --> 47:15.500]  I think it is.
[47:15.760 --> 47:17.080]  Yes, I have.
[47:17.900 --> 47:18.200]  I have.
[47:21.000 --> 47:21.560]  Yeah.
[47:21.960 --> 47:34.680]  Well, the thing that the question was about is Jack Gallant's lab in Berkeley, where the Alexander Huth part was from, has a lot of this research about the dreams and the semantic brain.
[47:34.840 --> 47:36.700]  And that's part of the semantic brain research.
[47:37.060 --> 47:38.080]  So you're right.
[47:38.300 --> 47:42.000]  It's just there's only so much you can include in a one hour presentation.
[47:42.300 --> 47:43.040]  But thanks for that.
[47:44.420 --> 47:44.840]  Any...
[47:45.200 --> 47:52.940]  How feasible are tools like OpenBCI for prototyping and for amateur brain hackers to get started?
[47:53.880 --> 47:55.840]  OpenBCI is very feasible.
[47:56.160 --> 48:03.780]  But the thing is, right now it's raw data and it uses, in order to interpret it, you have to really use something called MATLAB.
[48:04.300 --> 48:09.120]  It's not that user friendly for the average user at this point.
[48:09.440 --> 48:13.780]  Although, if you can write really good code, you can do stuff with it, I think.
[48:13.960 --> 48:21.720]  It's more accurate, but it doesn't have smoothing algorithms, unlike the commercial grade ones.
[48:22.460 --> 48:23.520]  So, yeah.
[48:26.460 --> 48:31.500]  When they were showing the fMRI picture, it looked sort of basically with a 2D view.
[48:31.840 --> 48:36.620]  Didn't the 3D structure of the brain really play into how everything is coordinating into the construction?
[48:36.900 --> 48:43.800]  The question was, the brain looked like a 2D view in the fMRI and will the 3D view play in?
[48:43.880 --> 48:44.980]  Oh yeah, absolutely.
[48:45.320 --> 48:47.160]  It was just, that was just one video.
[48:47.340 --> 48:51.700]  Like I said, I could have spent a half hour just on the semantic brain.
[48:51.700 --> 49:00.120]  I'm just, you know, I'm throwing a lot of stuff out here very quickly because it's the HOPE conference and I want to get a lot of information out fast.
[49:00.480 --> 49:01.220]  But, yeah.
[49:01.640 --> 49:02.000]  Absolutely.
[49:03.040 --> 49:03.140]  Yeah.
[49:08.630 --> 49:12.430]  That's a question, does the brain fingerprint change over time?
[49:12.810 --> 49:23.490]  There's a whole critique of that called brain blobology by a neuroscientist called Sally Satel, who would ask that question.
[49:23.830 --> 49:32.830]  And that's a really great question which researchers haven't caught up with yet, but you better believe in the next 10 years they will ask those questions.
[49:33.410 --> 49:36.510]  So, that's a great question that has yet to be answered.
[49:38.810 --> 49:39.970]  Any more questions?
[49:41.330 --> 49:42.370]  Any more questions?
[49:42.810 --> 49:43.950]  Okay, that's it.
[49:44.170 --> 49:44.950]  Thanks so much.
[49:45.190 --> 49:45.430]  Thank you.
[49:47.440 --> 49:48.220]  Thank you.
[49:48.520 --> 49:48.840]  Thank you so much.
[49:48.840 --> 49:49.180]  Thank you.