[Script Info]
Title: 
[Events]
Format: Layer, Start, End, Style, Name, MarginL, MarginR, MarginV, Effect, Text
Dialogue: 0,0:00:15.62,0:00:17.13,Default,,0000,0000,0000,,Good afternoon.
Dialogue: 0,0:00:17.56,0:00:18.75,Default,,0000,0000,0000,,My name is Marvin Chun,
Dialogue: 0,0:00:18.75,0:00:20.92,Default,,0000,0000,0000,,and I am a researcher,
Dialogue: 0,0:00:20.92,0:00:24.23,Default,,0000,0000,0000,,and I teach psychology at Yale.
Dialogue: 0,0:00:24.52,0:00:25.99,Default,,0000,0000,0000,,I'm very proud of what I do,
Dialogue: 0,0:00:25.99,0:00:29.77,Default,,0000,0000,0000,,and I really believe in the value\Nof psychological science.
Dialogue: 0,0:00:31.01,0:00:33.55,Default,,0000,0000,0000,,However, to start with a footnote,
Dialogue: 0,0:00:33.55,0:00:36.60,Default,,0000,0000,0000,,I will admit that I feel\Nself-conscious sometimes
Dialogue: 0,0:00:36.60,0:00:38.75,Default,,0000,0000,0000,,telling people that I'm a psychologist,
Dialogue: 0,0:00:38.75,0:00:41.48,Default,,0000,0000,0000,,especially when you're meeting\Na stranger on the airplane
Dialogue: 0,0:00:41.48,0:00:43.77,Default,,0000,0000,0000,,or at a cocktail party.
Dialogue: 0,0:00:43.77,0:00:44.87,Default,,0000,0000,0000,,And the reason for this
Dialogue: 0,0:00:44.87,0:00:48.09,Default,,0000,0000,0000,,is because there's\Na very typical question that you get
Dialogue: 0,0:00:48.09,0:00:51.09,Default,,0000,0000,0000,,when you tell people\Nthat you're a psychologist.
Dialogue: 0,0:00:51.09,0:00:53.31,Default,,0000,0000,0000,,What do you think that question might be?
Dialogue: 0,0:00:53.31,0:00:54.74,Default,,0000,0000,0000,,(Audience) What am I thinking?
Dialogue: 0,0:00:54.74,0:00:56.17,Default,,0000,0000,0000,,"What am I thinking?" exactly.
Dialogue: 0,0:00:56.17,0:00:57.27,Default,,0000,0000,0000,,"Can you read my mind?"
Dialogue: 0,0:00:57.27,0:01:00.57,Default,,0000,0000,0000,,is a question that a lot\Nof psychologists cringe at
Dialogue: 0,0:01:00.57,0:01:02.30,Default,,0000,0000,0000,,whenever they hear it
Dialogue: 0,0:01:02.30,0:01:04.64,Default,,0000,0000,0000,,because it really reflects\Na misconception about -
Dialogue: 0,0:01:04.64,0:01:08.75,Default,,0000,0000,0000,,a kind of a Freudian misconception\Nabout what we do for a living.
Dialogue: 0,0:01:08.75,0:01:11.85,Default,,0000,0000,0000,,So you know, I've developed\Na lot of answers over the years:
Dialogue: 0,0:01:11.85,0:01:13.14,Default,,0000,0000,0000,,"If I could read your mind,
Dialogue: 0,0:01:13.14,0:01:15.33,Default,,0000,0000,0000,,I'd be a whole lot richer\Nthan I am right now."
Dialogue: 0,0:01:16.72,0:01:18.81,Default,,0000,0000,0000,,The other answer I came up with is,
Dialogue: 0,0:01:18.81,0:01:20.81,Default,,0000,0000,0000,,"Oh, I'm actually a neuroscientist,"
Dialogue: 0,0:01:20.81,0:01:22.72,Default,,0000,0000,0000,,and then that really makes people quiet.
Dialogue: 0,0:01:22.72,0:01:24.21,Default,,0000,0000,0000,,(Laughter)
Dialogue: 0,0:01:24.75,0:01:26.32,Default,,0000,0000,0000,,But the best answer of course is,
Dialogue: 0,0:01:26.32,0:01:28.14,Default,,0000,0000,0000,,"Yeah, of course I can read your mind,
Dialogue: 0,0:01:28.14,0:01:30.28,Default,,0000,0000,0000,,and you really should be\Nashamed of yourself."
Dialogue: 0,0:01:30.28,0:01:31.91,Default,,0000,0000,0000,,(Laughter)
Dialogue: 0,0:01:35.57,0:01:39.70,Default,,0000,0000,0000,,But all that is motivation\Nfor what I'll share with you today,
Dialogue: 0,0:01:39.70,0:01:43.90,Default,,0000,0000,0000,,which is I've been a psychologist\Nfor about 20 years now,
Dialogue: 0,0:01:43.90,0:01:49.43,Default,,0000,0000,0000,,and we've actually reached this point\Nwhere I can actually answer that question
Dialogue: 0,0:01:49.43,0:01:52.72,Default,,0000,0000,0000,,by saying, "Yes,\Nif I can put you in a scanner,
Dialogue: 0,0:01:52.72,0:01:54.95,Default,,0000,0000,0000,,then we can read your mind."
Dialogue: 0,0:01:54.95,0:01:59.16,Default,,0000,0000,0000,,And what I'll share with you today\Nis to what extent, how far are we
Dialogue: 0,0:01:59.16,0:02:02.67,Default,,0000,0000,0000,,at being able to read\Nother people's minds.
Dialogue: 0,0:02:04.08,0:02:06.26,Default,,0000,0000,0000,,Of course this is the stuff\Nof science fiction.
Dialogue: 0,0:02:06.26,0:02:08.69,Default,,0000,0000,0000,,Even a movie as recent as "Divergent"
Dialogue: 0,0:02:08.69,0:02:10.38,Default,,0000,0000,0000,,has these many sequences
Dialogue: 0,0:02:10.38,0:02:12.90,Default,,0000,0000,0000,,in which they're reading out\Nthe brain activity
Dialogue: 0,0:02:12.90,0:02:14.23,Default,,0000,0000,0000,,to see what she's dreaming
Dialogue: 0,0:02:14.23,0:02:17.25,Default,,0000,0000,0000,,under the influence of drugs.
Dialogue: 0,0:02:18.08,0:02:20.67,Default,,0000,0000,0000,,And you know, again,\Nthese technologies do exist.
Dialogue: 0,0:02:20.99,0:02:23.70,Default,,0000,0000,0000,,In modern day, in the real word,\Nthey look more like this.
Dialogue: 0,0:02:23.70,0:02:27.10,Default,,0000,0000,0000,,These are just standard\NMRI machines and scanners
Dialogue: 0,0:02:27.10,0:02:29.12,Default,,0000,0000,0000,,that are slightly modified
Dialogue: 0,0:02:29.12,0:02:33.61,Default,,0000,0000,0000,,so that they can allow you\Nto infer and read out brain activity.
Dialogue: 0,0:02:35.65,0:02:39.15,Default,,0000,0000,0000,,These methods have been around\Nfor about 25 years,
Dialogue: 0,0:02:39.34,0:02:42.08,Default,,0000,0000,0000,,and in the first decade -\NI'd say the 1990s -
Dialogue: 0,0:02:42.08,0:02:44.18,Default,,0000,0000,0000,,a lot of effort was devoted
Dialogue: 0,0:02:44.18,0:02:48.18,Default,,0000,0000,0000,,to mapping out what the different\Nparts of the brain do.
Dialogue: 0,0:02:48.46,0:02:53.27,Default,,0000,0000,0000,,One of the most successful forms\Nof this is exemplified here.
Dialogue: 0,0:02:53.27,0:02:55.89,Default,,0000,0000,0000,,Imagine you're a subject\Nlying down in that scanner;
Dialogue: 0,0:02:55.89,0:02:57.23,Default,,0000,0000,0000,,there's a computer projector
Dialogue: 0,0:02:57.23,0:02:59.45,Default,,0000,0000,0000,,that allows you\Nto look at a series of images
Dialogue: 0,0:02:59.45,0:03:01.64,Default,,0000,0000,0000,,while your brain is being scanned.
Dialogue: 0,0:03:02.22,0:03:03.87,Default,,0000,0000,0000,,For some periods of time,
Dialogue: 0,0:03:03.87,0:03:07.30,Default,,0000,0000,0000,,you are going to be seeing\Nsequences of scene images
Dialogue: 0,0:03:07.60,0:03:10.84,Default,,0000,0000,0000,,alternating with other\Nsequences of face images,
Dialogue: 0,0:03:11.14,0:03:14.52,Default,,0000,0000,0000,,and they'll just go back and forth\Nwhile your brain is being scanned.
Dialogue: 0,0:03:14.72,0:03:17.30,Default,,0000,0000,0000,,And the key motivation here\Nfor the researchers
Dialogue: 0,0:03:17.30,0:03:19.66,Default,,0000,0000,0000,,is to compare what's\Nhappening in the brain
Dialogue: 0,0:03:19.66,0:03:22.85,Default,,0000,0000,0000,,when you're looking at scenes\Nversus when you're looking at faces
Dialogue: 0,0:03:22.85,0:03:27.49,Default,,0000,0000,0000,,to see if anything different happens\Nfor these two categories of stimuli.
Dialogue: 0,0:03:27.49,0:03:30.31,Default,,0000,0000,0000,,And I'd say even within 10 minutes\Nof scanning your brain,
Dialogue: 0,0:03:30.31,0:03:34.44,Default,,0000,0000,0000,,you can get a very reliable pattern\Nof results that looks like this,
Dialogue: 0,0:03:34.44,0:03:39.62,Default,,0000,0000,0000,,where there is a region of the brain\Nthat's specialized for processing scenes,
Dialogue: 0,0:03:39.62,0:03:41.72,Default,,0000,0000,0000,,shown here in the warm colors,
Dialogue: 0,0:03:42.16,0:03:45.18,Default,,0000,0000,0000,,and there is also\Na separate region of the brain,
Dialogue: 0,0:03:45.18,0:03:47.17,Default,,0000,0000,0000,,shown here in blue colors,
Dialogue: 0,0:03:47.17,0:03:51.14,Default,,0000,0000,0000,,that is more selectively active\Nwhen people are viewing faces.
Dialogue: 0,0:03:51.35,0:03:53.70,Default,,0000,0000,0000,,There's this disassociation here.
Dialogue: 0,0:03:54.26,0:03:56.64,Default,,0000,0000,0000,,And this allows you to see
Dialogue: 0,0:03:56.64,0:03:59.26,Default,,0000,0000,0000,,where different functions\Nreside in the brain.
Dialogue: 0,0:03:59.26,0:04:01.90,Default,,0000,0000,0000,,And in the case\Nof scene and face perception,
Dialogue: 0,0:04:01.90,0:04:05.80,Default,,0000,0000,0000,,these two forms of vision\Nare so important in our everyday lives -
Dialogue: 0,0:04:05.80,0:04:08.41,Default,,0000,0000,0000,,or course scene processing\Nimportant for navigation,
Dialogue: 0,0:04:08.41,0:04:11.78,Default,,0000,0000,0000,,face processing important\Nfor social interactions -
Dialogue: 0,0:04:11.78,0:04:14.61,Default,,0000,0000,0000,,that our brain has\Nspecialized areas devoted to them.
Dialogue: 0,0:04:15.66,0:04:18.43,Default,,0000,0000,0000,,And this alone, I think,\Nis pretty amazing.
Dialogue: 0,0:04:18.90,0:04:22.14,Default,,0000,0000,0000,,A lot of this work\Nstarted coming out in the mid-1990s,
Dialogue: 0,0:04:22.40,0:04:24.97,Default,,0000,0000,0000,,and it really complimented and reinforced
Dialogue: 0,0:04:24.97,0:04:28.31,Default,,0000,0000,0000,,a lot of the patient work\Nthat's been around for dozens of years.
Dialogue: 0,0:04:29.04,0:04:32.52,Default,,0000,0000,0000,,But really, the methodologies of fMRI
Dialogue: 0,0:04:32.52,0:04:37.19,Default,,0000,0000,0000,,have really gone far beyond\Nthese very basic mapping studies.
Dialogue: 0,0:04:37.19,0:04:39.71,Default,,0000,0000,0000,,You know, in fact,\Nwhen you look at a study like this,
Dialogue: 0,0:04:39.71,0:04:41.48,Default,,0000,0000,0000,,you may say, "Oh, that's really cool,
Dialogue: 0,0:04:41.48,0:04:43.82,Default,,0000,0000,0000,,but what does this have to do\Nabout everyday life?
Dialogue: 0,0:04:43.82,0:04:49.02,Default,,0000,0000,0000,,How is this going to help me understand\Nwhy I feel this way or don't feel this way
Dialogue: 0,0:04:49.02,0:04:52.29,Default,,0000,0000,0000,,or why I like a certain person\Nor don't like certain other people?
Dialogue: 0,0:04:52.29,0:04:55.00,Default,,0000,0000,0000,,Will this tell me what kind\Nof career I should take?"
Dialogue: 0,0:04:55.00,0:04:56.82,Default,,0000,0000,0000,,You know, you can't really get too far
Dialogue: 0,0:04:56.82,0:04:59.54,Default,,0000,0000,0000,,with this scene and face\Narea mapping alone.
Dialogue: 0,0:05:00.96,0:05:04.57,Default,,0000,0000,0000,,So it wasn't until this study\Ncame up around 2000 -
Dialogue: 0,0:05:04.79,0:05:06.94,Default,,0000,0000,0000,,And most of these studies,\NI should emphasize,
Dialogue: 0,0:05:06.94,0:05:08.71,Default,,0000,0000,0000,,are from other labs around the field.
Dialogue: 0,0:05:08.71,0:05:09.76,Default,,0000,0000,0000,,Any studies from Yale,
Dialogue: 0,0:05:09.76,0:05:12.17,Default,,0000,0000,0000,,I'll have the mark of "Yale"\Non the corner.
Dialogue: 0,0:05:12.17,0:05:18.07,Default,,0000,0000,0000,,But this particular study was done at MIT,\Nby Nancy Kanwisher and colleagues.
Dialogue: 0,0:05:18.07,0:05:20.92,Default,,0000,0000,0000,,They actually had subjects\Nlook at nothing.
Dialogue: 0,0:05:21.41,0:05:25.17,Default,,0000,0000,0000,,They had subjects\Nclose their eyes in the scanner
Dialogue: 0,0:05:25.82,0:05:28.01,Default,,0000,0000,0000,,while they gave them instructions,
Dialogue: 0,0:05:28.01,0:05:31.88,Default,,0000,0000,0000,,and the instructions\Nwere one of two types of tasks.
Dialogue: 0,0:05:31.88,0:05:35.92,Default,,0000,0000,0000,,One was to imagine\Na bunch of faces that they knew.
Dialogue: 0,0:05:35.92,0:05:38.64,Default,,0000,0000,0000,,So you might be able to imagine\Nyour friends and family,
Dialogue: 0,0:05:38.64,0:05:40.71,Default,,0000,0000,0000,,cycle through them one at a time
Dialogue: 0,0:05:41.06,0:05:42.68,Default,,0000,0000,0000,,while your brain is being scanned.
Dialogue: 0,0:05:42.68,0:05:44.92,Default,,0000,0000,0000,,That would be the face\Nimagery instruction.
Dialogue: 0,0:05:44.92,0:05:47.63,Default,,0000,0000,0000,,And then the other instruction\Nthat they gave the subjects
Dialogue: 0,0:05:47.63,0:05:49.63,Default,,0000,0000,0000,,was of course a scene imagery instruction,
Dialogue: 0,0:05:49.63,0:05:53.58,Default,,0000,0000,0000,,so imagine after this talk,\Nyou walk home or take your car home,
Dialogue: 0,0:05:53.58,0:05:56.22,Default,,0000,0000,0000,,get into your home,\Nnavigate throughout your house,
Dialogue: 0,0:05:56.22,0:05:58.08,Default,,0000,0000,0000,,change into something more comfortable,
Dialogue: 0,0:05:58.08,0:05:59.98,Default,,0000,0000,0000,,go to the kitchen, get something to eat.
Dialogue: 0,0:05:59.98,0:06:04.47,Default,,0000,0000,0000,,You have this ability to visualize\Nplaces and scenes in your mind,
Dialogue: 0,0:06:04.47,0:06:08.06,Default,,0000,0000,0000,,and the questions is, "What happens\Nwhen you're visualizing scenes,
Dialogue: 0,0:06:08.06,0:06:10.60,Default,,0000,0000,0000,,and what happens\Nwhen you're visualizing faces?
Dialogue: 0,0:06:10.91,0:06:13.01,Default,,0000,0000,0000,,What are the neural correlates of that?"
Dialogue: 0,0:06:13.21,0:06:18.14,Default,,0000,0000,0000,,And the hypothesis was, well, maybe\Nimagining faces activates the face area
Dialogue: 0,0:06:18.14,0:06:21.47,Default,,0000,0000,0000,,and maybe imagining scenes\Nwould activate the scene area,
Dialogue: 0,0:06:21.47,0:06:24.23,Default,,0000,0000,0000,,and that's in fact indeed what they found.
Dialogue: 0,0:06:24.74,0:06:27.40,Default,,0000,0000,0000,,First, you bring in a bunch\Nof subjects into the scanner,
Dialogue: 0,0:06:27.40,0:06:29.37,Default,,0000,0000,0000,,show them faces, show them scenes,
Dialogue: 0,0:06:29.37,0:06:31.22,Default,,0000,0000,0000,,you map out the face area on top,
Dialogue: 0,0:06:31.22,0:06:34.16,Default,,0000,0000,0000,,you map out the place area\Nin the second row.
Dialogue: 0,0:06:34.16,0:06:35.80,Default,,0000,0000,0000,,And then the same subjects,
Dialogue: 0,0:06:36.85,0:06:38.56,Default,,0000,0000,0000,,at a different time in the scanner,
Dialogue: 0,0:06:38.56,0:06:41.95,Default,,0000,0000,0000,,have them close their eyes,\Ndo the instructions I just had you do,
Dialogue: 0,0:06:41.95,0:06:43.42,Default,,0000,0000,0000,,and you compare.
Dialogue: 0,0:06:43.42,0:06:44.71,Default,,0000,0000,0000,,When you have face imagery,
Dialogue: 0,0:06:44.71,0:06:47.91,Default,,0000,0000,0000,,you can see in the second row\Nthat the face area is active,
Dialogue: 0,0:06:47.91,0:06:49.81,Default,,0000,0000,0000,,even though nothing's on the screen.
Dialogue: 0,0:06:49.81,0:06:53.61,Default,,0000,0000,0000,,And in the fourth row here,\Nthe second row of the bottom pair,
Dialogue: 0,0:06:53.78,0:06:57.48,Default,,0000,0000,0000,,you see that the place area is active\Nwhen you're imagining scenes;
Dialogue: 0,0:06:57.48,0:06:58.62,Default,,0000,0000,0000,,nothing's on the screen,
Dialogue: 0,0:06:58.62,0:07:02.66,Default,,0000,0000,0000,,but simply when you're imagining scenes,\Nit will activate the place area.
Dialogue: 0,0:07:03.45,0:07:06.78,Default,,0000,0000,0000,,In fact, even just\Nby looking at fMRI data alone
Dialogue: 0,0:07:06.78,0:07:10.31,Default,,0000,0000,0000,,and looking at the relative activity\Nfor the face area and place area,
Dialogue: 0,0:07:10.31,0:07:12.48,Default,,0000,0000,0000,,you can guess over 80% of the time,
Dialogue: 0,0:07:12.48,0:07:13.97,Default,,0000,0000,0000,,even in this first study,
Dialogue: 0,0:07:13.97,0:07:16.02,Default,,0000,0000,0000,,what subjects were imagining.
Dialogue: 0,0:07:16.02,0:07:17.02,Default,,0000,0000,0000,,Okay?
Dialogue: 0,0:07:17.02,0:07:19.95,Default,,0000,0000,0000,,And so, in this sense, in my mind,
Dialogue: 0,0:07:19.95,0:07:23.29,Default,,0000,0000,0000,,I believe that this is the first study\Nthat started using brain imaging
Dialogue: 0,0:07:23.29,0:07:26.25,Default,,0000,0000,0000,,to actually read out\Nwhat people were thinking.
Dialogue: 0,0:07:26.25,0:07:29.30,Default,,0000,0000,0000,,You can take a naïve person,\Nlook at this graph, you know,
Dialogue: 0,0:07:29.30,0:07:32.39,Default,,0000,0000,0000,,and ask which bar is higher,\Nwhich line is higher,
Dialogue: 0,0:07:32.39,0:07:35.89,Default,,0000,0000,0000,,and they can guess better than chance,\Nway better than chance,
Dialogue: 0,0:07:35.89,0:07:37.76,Default,,0000,0000,0000,,what people were thinking.
Dialogue: 0,0:07:38.62,0:07:40.67,Default,,0000,0000,0000,,Now, you may say, "Okay,\Nthat's really cool,
Dialogue: 0,0:07:40.67,0:07:44.61,Default,,0000,0000,0000,,but there're a whole lot more\Ninteresting things to do with a scanner
Dialogue: 0,0:07:44.61,0:07:46.84,Default,,0000,0000,0000,,or with what we want to know
Dialogue: 0,0:07:46.84,0:07:49.77,Default,,0000,0000,0000,,than whether people\Nare thinking of faces or scenes."
Dialogue: 0,0:07:49.77,0:07:51.07,Default,,0000,0000,0000,,And so the rest of my talk
Dialogue: 0,0:07:51.07,0:07:55.13,Default,,0000,0000,0000,,will be devoted to the next 15 years\Nof work since this study
Dialogue: 0,0:07:55.13,0:07:58.58,Default,,0000,0000,0000,,that have really advanced\Nour ability to read out minds.
Dialogue: 0,0:08:00.20,0:08:02.37,Default,,0000,0000,0000,,This is one study\Nthat was published in 2010
Dialogue: 0,0:08:02.37,0:08:04.62,Default,,0000,0000,0000,,in the New England Journal of Medicine
Dialogue: 0,0:08:04.94,0:08:06.03,Default,,0000,0000,0000,,by another group.
Dialogue: 0,0:08:06.03,0:08:10.58,Default,,0000,0000,0000,,They were actually studying patients\Nwho were in a persistent vegetative state,
Dialogue: 0,0:08:10.58,0:08:13.93,Default,,0000,0000,0000,,who were locked in,\Nhad no ability to express themselves;
Dialogue: 0,0:08:13.93,0:08:17.36,Default,,0000,0000,0000,,it was not even clear if they were\Ncapable of voluntary thought,
Dialogue: 0,0:08:17.36,0:08:22.04,Default,,0000,0000,0000,,and yet because you can instruct people\Nto imagine one thing versus another -
Dialogue: 0,0:08:22.04,0:08:25.64,Default,,0000,0000,0000,,in this case, imagine walking\Nthrough your house, navigation,
Dialogue: 0,0:08:25.64,0:08:27.91,Default,,0000,0000,0000,,in the other case, imagine playing tennis
Dialogue: 0,0:08:27.91,0:08:30.24,Default,,0000,0000,0000,,because that activates\Na whole motor circuitry
Dialogue: 0,0:08:30.24,0:08:33.62,Default,,0000,0000,0000,,that's separate from the spacial\Nnavigation circuitry -
Dialogue: 0,0:08:33.89,0:08:37.48,Default,,0000,0000,0000,,you attach each of those imagery\Ninstruction to "yes" or "no,"
Dialogue: 0,0:08:37.48,0:08:39.62,Default,,0000,0000,0000,,and then you can ask them\Nfactual questions,
Dialogue: 0,0:08:39.62,0:08:44.39,Default,,0000,0000,0000,,such as "Is your father's name Alexander,\Nor is your father's name Thomas?"
Dialogue: 0,0:08:44.39,0:08:47.34,Default,,0000,0000,0000,,And they were able to demonstrate\Nthat in this patient,
Dialogue: 0,0:08:47.34,0:08:49.78,Default,,0000,0000,0000,,who otherwise had no ability\Nto communicate,
Dialogue: 0,0:08:49.78,0:08:54.92,Default,,0000,0000,0000,,he was able to respond\Nto these factual questions accurately,
Dialogue: 0,0:08:54.92,0:08:59.17,Default,,0000,0000,0000,,and even in control subject\Nwho do not have problems.
Dialogue: 0,0:08:59.65,0:09:01.13,Default,,0000,0000,0000,,This method is so reliable
Dialogue: 0,0:09:01.13,0:09:05.86,Default,,0000,0000,0000,,that almost over 95%\Nof their responses were decodable
Dialogue: 0,0:09:05.86,0:09:08.88,Default,,0000,0000,0000,,just through brain imagining alone -\N"yes" versus "no."
Dialogue: 0,0:09:08.88,0:09:11.44,Default,,0000,0000,0000,,So you can imagine,\Nwith a game of 20 questions,
Dialogue: 0,0:09:11.44,0:09:14.54,Default,,0000,0000,0000,,you can really get insight\Ninto what people are thinking
Dialogue: 0,0:09:14.54,0:09:16.68,Default,,0000,0000,0000,,using these methodologies.
Dialogue: 0,0:09:17.82,0:09:19.29,Default,,0000,0000,0000,,But again, it's limited;
Dialogue: 0,0:09:19.29,0:09:20.72,Default,,0000,0000,0000,,we only have two states here -
Dialogue: 0,0:09:20.72,0:09:21.72,Default,,0000,0000,0000,,"yes," "no,"
Dialogue: 0,0:09:21.72,0:09:25.09,Default,,0000,0000,0000,,think tennis or think\Nwalking around your house.
Dialogue: 0,0:09:25.70,0:09:29.91,Default,,0000,0000,0000,,But fortunately, there're tremendously\Nbrilliant scientists all around the world
Dialogue: 0,0:09:29.91,0:09:35.64,Default,,0000,0000,0000,,who are constantly refining the methods\Nto decode fMRI activity
Dialogue: 0,0:09:35.64,0:09:38.57,Default,,0000,0000,0000,,so that we can understand the mind.
Dialogue: 0,0:09:39.76,0:09:43.86,Default,,0000,0000,0000,,And I'd like to use this slide\Nfrom Nature magazine
Dialogue: 0,0:09:43.86,0:09:47.46,Default,,0000,0000,0000,,to motivate the next few studies\NI'll share with you.
Dialogue: 0,0:09:47.92,0:09:53.09,Default,,0000,0000,0000,,One huge limitation of brain imaging,\Nat least in the first 15 years,
Dialogue: 0,0:09:53.09,0:09:56.72,Default,,0000,0000,0000,,is that there are not\Nmany specialized areas in the brain.
Dialogue: 0,0:09:56.72,0:09:59.24,Default,,0000,0000,0000,,The face area is one.\NPlace area is another.
Dialogue: 0,0:09:59.24,0:10:03.69,Default,,0000,0000,0000,,These are very, very important visual\Nfunctions that we carry out every day,
Dialogue: 0,0:10:03.69,0:10:08.38,Default,,0000,0000,0000,,but there is no "shoe" area in the brain,\Nthere is no "cat" area in the brain.
Dialogue: 0,0:10:08.38,0:10:12.31,Default,,0000,0000,0000,,You don't see separate blobs\Nor patches of activity
Dialogue: 0,0:10:12.31,0:10:14.05,Default,,0000,0000,0000,,for these two different categories,
Dialogue: 0,0:10:14.05,0:10:16.99,Default,,0000,0000,0000,,and indeed, for most categories\Nthat we encounter
Dialogue: 0,0:10:16.99,0:10:18.74,Default,,0000,0000,0000,,and are able to discriminate,
Dialogue: 0,0:10:18.74,0:10:21.02,Default,,0000,0000,0000,,they don't have separate brain regions.
Dialogue: 0,0:10:21.02,0:10:23.47,Default,,0000,0000,0000,,And because there are\Nno separate brain regions,
Dialogue: 0,0:10:23.47,0:10:24.48,Default,,0000,0000,0000,,the question then is,
Dialogue: 0,0:10:24.48,0:10:27.05,Default,,0000,0000,0000,,"How do we study them,\Nand how do we discriminate them,
Dialogue: 0,0:10:27.05,0:10:31.13,Default,,0000,0000,0000,,how do we get at these more fine\Ndetailed differences that matter so much
Dialogue: 0,0:10:31.13,0:10:33.23,Default,,0000,0000,0000,,if we really want to understand
Dialogue: 0,0:10:33.23,0:10:35.90,Default,,0000,0000,0000,,how the mind works\Nand how we can read it out
Dialogue: 0,0:10:35.90,0:10:37.66,Default,,0000,0000,0000,,using fMRI?
Dialogue: 0,0:10:38.49,0:10:44.46,Default,,0000,0000,0000,,Fortunately, some neuroscientists\Ncollaborated with computer vision people
Dialogue: 0,0:10:44.46,0:10:48.58,Default,,0000,0000,0000,,and with electrical engineers\Nand computer scientists and statisticians
Dialogue: 0,0:10:48.58,0:10:51.54,Default,,0000,0000,0000,,to use very refined mathematical methods
Dialogue: 0,0:10:51.54,0:10:56.18,Default,,0000,0000,0000,,for pulling out and decoding\Nmore subtle patterns of activity
Dialogue: 0,0:10:56.18,0:10:58.56,Default,,0000,0000,0000,,that are unique to shoe processing
Dialogue: 0,0:10:58.56,0:11:03.01,Default,,0000,0000,0000,,and that are unique\Nto the processing of cat stimuli.
Dialogue: 0,0:11:03.01,0:11:05.79,Default,,0000,0000,0000,,So, for instance,\Neven though the shoes and cats,
Dialogue: 0,0:11:05.79,0:11:07.79,Default,,0000,0000,0000,,you show a whole bunch\Nof them to subjects,
Dialogue: 0,0:11:07.79,0:11:11.15,Default,,0000,0000,0000,,it'll activate same part of cortex,\Nthe same part of the brain,
Dialogue: 0,0:11:11.15,0:11:13.45,Default,,0000,0000,0000,,and so that if you average\Nactivity in that part,
Dialogue: 0,0:11:13.45,0:11:17.18,Default,,0000,0000,0000,,you won't see any differences\Nbetween the shoes and the cats.
Dialogue: 0,0:11:17.18,0:11:19.18,Default,,0000,0000,0000,,As you can see here on the right,
Dialogue: 0,0:11:19.40,0:11:22.98,Default,,0000,0000,0000,,the shoes are still associated\Nwith different fine patterns,
Dialogue: 0,0:11:22.98,0:11:24.87,Default,,0000,0000,0000,,or micropatterns of activity
Dialogue: 0,0:11:24.87,0:11:28.85,Default,,0000,0000,0000,,that are different from the patterns\Nof activity elicited by cats.
Dialogue: 0,0:11:29.18,0:11:33.72,Default,,0000,0000,0000,,These little boxes over here,\Nthese cells are referred to voxels;
Dialogue: 0,0:11:34.11,0:11:36.14,Default,,0000,0000,0000,,the way fMRI collects data from the brain
Dialogue: 0,0:11:36.14,0:11:39.05,Default,,0000,0000,0000,,is it kind of chops up your brain\Ninto tiny little cubes,
Dialogue: 0,0:11:39.05,0:11:41.16,Default,,0000,0000,0000,,and each cube is a unit of measurement -
Dialogue: 0,0:11:41.16,0:11:44.94,Default,,0000,0000,0000,,it gives you numerical data\Nthat you can use to analyze.
Dialogue: 0,0:11:44.94,0:11:47.55,Default,,0000,0000,0000,,And if you imagine\Nthat the brightness of these squares
Dialogue: 0,0:11:47.55,0:11:50.18,Default,,0000,0000,0000,,corresponds to a different\Nlevel of activity,
Dialogue: 0,0:11:50.18,0:11:53.15,Default,,0000,0000,0000,,quantitatively measurable\Ndifferent level of activity,
Dialogue: 0,0:11:53.15,0:11:55.11,Default,,0000,0000,0000,,and you can imagine over a spacial region
Dialogue: 0,0:11:55.11,0:11:59.70,Default,,0000,0000,0000,,that the pattern of these activities\Nmay be different for shoes versus cats,
Dialogue: 0,0:11:59.98,0:12:03.65,Default,,0000,0000,0000,,then all you need to do\Nis to work with a computer algorithm
Dialogue: 0,0:12:03.65,0:12:06.45,Default,,0000,0000,0000,,to learn these subtle\Ndifferences in patterns,
Dialogue: 0,0:12:06.45,0:12:09.80,Default,,0000,0000,0000,,so we have what we might call a "decoder."
Dialogue: 0,0:12:10.12,0:12:12.23,Default,,0000,0000,0000,,And you train up a computer decoder
Dialogue: 0,0:12:12.23,0:12:15.62,Default,,0000,0000,0000,,to learn the differences\Nbetween shoe patterns and cat patterns,
Dialogue: 0,0:12:15.92,0:12:18.44,Default,,0000,0000,0000,,and then what you have\Nis here in the testing phase:
Dialogue: 0,0:12:18.44,0:12:20.73,Default,,0000,0000,0000,,after you train up\Nyour computer algorithm,
Dialogue: 0,0:12:20.73,0:12:22.68,Default,,0000,0000,0000,,you can show a new shoe,
Dialogue: 0,0:12:23.32,0:12:26.05,Default,,0000,0000,0000,,it will elicit a certain pattern\Nof fMRI activity,
Dialogue: 0,0:12:26.05,0:12:27.60,Default,,0000,0000,0000,,as you can see over here,
Dialogue: 0,0:12:27.60,0:12:30.35,Default,,0000,0000,0000,,and the computer\Nwill tell you its best guess
Dialogue: 0,0:12:30.35,0:12:32.98,Default,,0000,0000,0000,,as to whether this pattern over here
Dialogue: 0,0:12:32.98,0:12:36.84,Default,,0000,0000,0000,,is closer to a shoe\Nor closer to that for a cat.
Dialogue: 0,0:12:36.84,0:12:37.85,Default,,0000,0000,0000,,And it will allow you
Dialogue: 0,0:12:37.85,0:12:40.27,Default,,0000,0000,0000,,to guess something\Nthat's a lot more refined
Dialogue: 0,0:12:40.27,0:12:42.06,Default,,0000,0000,0000,,and that is not possible
Dialogue: 0,0:12:42.06,0:12:45.62,Default,,0000,0000,0000,,when you don't have areas\Nlike the face area or the place area.
Dialogue: 0,0:12:48.36,0:12:52.06,Default,,0000,0000,0000,,And now people are really taking off\Nwith these types of methodologies.
Dialogue: 0,0:12:52.06,0:12:56.63,Default,,0000,0000,0000,,Those started coming out in 2005,\Nso a little than a decade ago.
Dialogue: 0,0:12:56.63,0:12:59.91,Default,,0000,0000,0000,,This is a study published\Nin the New England Journal of Medicine,
Dialogue: 0,0:12:59.91,0:13:03.05,Default,,0000,0000,0000,,year 2013, literally last year,
Dialogue: 0,0:13:04.24,0:13:07.32,Default,,0000,0000,0000,,a very important clinical\Napplication of studying
Dialogue: 0,0:13:07.32,0:13:09.96,Default,,0000,0000,0000,,or trying to find a way\Nto objectively measure pain,
Dialogue: 0,0:13:09.96,0:13:13.33,Default,,0000,0000,0000,,which is a very subjective state, we know.
Dialogue: 0,0:13:13.71,0:13:15.05,Default,,0000,0000,0000,,You can use fMRI,
Dialogue: 0,0:13:15.05,0:13:17.40,Default,,0000,0000,0000,,and these researchers have developed ways
Dialogue: 0,0:13:17.40,0:13:20.88,Default,,0000,0000,0000,,of mapping and predicting\Nthe amount of heat
Dialogue: 0,0:13:20.88,0:13:23.18,Default,,0000,0000,0000,,and the amount of pain people experience
Dialogue: 0,0:13:23.18,0:13:27.77,Default,,0000,0000,0000,,as you increase the temperature\Nof a noxious stimulus on their skin,
Dialogue: 0,0:13:28.03,0:13:32.33,Default,,0000,0000,0000,,and the fMRI decoder\Nwill kind of give you an estimate
Dialogue: 0,0:13:32.33,0:13:35.26,Default,,0000,0000,0000,,of how much heat\Nand maybe even how much pain
Dialogue: 0,0:13:35.26,0:13:37.62,Default,,0000,0000,0000,,someone is experiencing,
Dialogue: 0,0:13:37.62,0:13:41.33,Default,,0000,0000,0000,,to the extent that they can predict,\Njust based on fMRI activity alone,
Dialogue: 0,0:13:41.33,0:13:45.25,Default,,0000,0000,0000,,how much pain a person is experiencing\Nover 93% of the time,
Dialogue: 0,0:13:45.25,0:13:47.53,Default,,0000,0000,0000,,or with 93% accuracy.
Dialogue: 0,0:13:49.73,0:13:52.86,Default,,0000,0000,0000,,This is an astonishing version\Nthat I'm sure many of you have seen,
Dialogue: 0,0:13:52.86,0:13:53.86,Default,,0000,0000,0000,,published in 2011,
Dialogue: 0,0:13:53.86,0:13:57.65,Default,,0000,0000,0000,,from Jack Gallant's group\Nover in UC, Berkley.
Dialogue: 0,0:13:57.99,0:14:02.56,Default,,0000,0000,0000,,They showed people a bunch\Nof these videos from YouTube,
Dialogue: 0,0:14:02.56,0:14:04.29,Default,,0000,0000,0000,,here shown here on the left,
Dialogue: 0,0:14:04.29,0:14:07.53,Default,,0000,0000,0000,,and then they trained up\Na classifier, a decoder,
Dialogue: 0,0:14:07.53,0:14:13.09,Default,,0000,0000,0000,,to learn the mapping\Nbetween the videos and fMRI activity;
Dialogue: 0,0:14:13.50,0:14:16.88,Default,,0000,0000,0000,,and then when they showed\Npeople novel videos
Dialogue: 0,0:14:16.88,0:14:19.85,Default,,0000,0000,0000,,that elicited certain patterns\Nof fMRI activity,
Dialogue: 0,0:14:19.85,0:14:22.58,Default,,0000,0000,0000,,they were able to use\Nthat fMRI activity alone
Dialogue: 0,0:14:22.58,0:14:25.51,Default,,0000,0000,0000,,to guess what videos people were seeing.
Dialogue: 0,0:14:25.51,0:14:28.27,Default,,0000,0000,0000,,And they pulled the guesses\Noff of YouTube as well
Dialogue: 0,0:14:28.27,0:14:29.27,Default,,0000,0000,0000,,and averaged them,
Dialogue: 0,0:14:29.27,0:14:31.06,Default,,0000,0000,0000,,and that's why it's a little grainy,
Dialogue: 0,0:14:31.06,0:14:33.76,Default,,0000,0000,0000,,but what you'll see here -\Nand this is all over YouTube -
Dialogue: 0,0:14:34.08,0:14:36.13,Default,,0000,0000,0000,,is what people saw
Dialogue: 0,0:14:36.13,0:14:39.02,Default,,0000,0000,0000,,and, on the right, what the computer\Nis guessing people saw
Dialogue: 0,0:14:39.02,0:14:41.35,Default,,0000,0000,0000,,based on fMRI activity alone.
Dialogue: 0,0:15:13.14,0:15:16.43,Default,,0000,0000,0000,,So it's pretty astonishing;\Nthis literally is reading out the mind.
Dialogue: 0,0:15:16.43,0:15:19.57,Default,,0000,0000,0000,,I mean, you have complicated models,\Nyou have to train it and such,
Dialogue: 0,0:15:19.57,0:15:22.44,Default,,0000,0000,0000,,but still, this is based\Non fMRI activity alone
Dialogue: 0,0:15:22.44,0:15:24.86,Default,,0000,0000,0000,,that they're able to make\Nthese guesses over here.
Dialogue: 0,0:15:25.08,0:15:29.85,Default,,0000,0000,0000,,I had a student at Yale University.\NHis name is Alan Cowen.
Dialogue: 0,0:15:30.05,0:15:32.19,Default,,0000,0000,0000,,He had a very strong\Nmathematics background.
Dialogue: 0,0:15:32.19,0:15:35.22,Default,,0000,0000,0000,,He came to my lab,\Nand he said, "I really love this stuff."
Dialogue: 0,0:15:35.53,0:15:38.34,Default,,0000,0000,0000,,He loved this stuff so much\Nthat he wanted to do it himself.
Dialogue: 0,0:15:38.34,0:15:42.71,Default,,0000,0000,0000,,I actually did not have the ability\Nto do this in the lab,
Dialogue: 0,0:15:42.71,0:15:46.01,Default,,0000,0000,0000,,but he developed it\Ntogether with my post doc Brice Kuhl,
Dialogue: 0,0:15:46.01,0:15:49.62,Default,,0000,0000,0000,,who's now an assistant professor\Nat New York University.
Dialogue: 0,0:15:49.62,0:15:51.11,Default,,0000,0000,0000,,They wanted to ask the question
Dialogue: 0,0:15:51.11,0:15:56.24,Default,,0000,0000,0000,,of "Can we limit this type\Nof analysis to faces?"
Dialogue: 0,0:15:57.05,0:15:58.05,Default,,0000,0000,0000,,As you saw before,
Dialogue: 0,0:15:58.05,0:16:00.82,Default,,0000,0000,0000,,you can kind of see what categories\Npeople were looking at,
Dialogue: 0,0:16:00.82,0:16:04.20,Default,,0000,0000,0000,,but certainly the algorithm\Nwas not able to give you guesses
Dialogue: 0,0:16:04.20,0:16:06.26,Default,,0000,0000,0000,,as to which person you were looking at
Dialogue: 0,0:16:06.26,0:16:08.80,Default,,0000,0000,0000,,or what kind of animal\Nyou were looking at.
Dialogue: 0,0:16:08.80,0:16:11.38,Default,,0000,0000,0000,,You're really getting\Nmore categorical guesses
Dialogue: 0,0:16:11.38,0:16:12.70,Default,,0000,0000,0000,,in the prior example.
Dialogue: 0,0:16:12.70,0:16:14.75,Default,,0000,0000,0000,,And so Alan and Brice,
Dialogue: 0,0:16:14.75,0:16:19.46,Default,,0000,0000,0000,,they thought that, well, if we focused\Nour algorithms on faces alone -
Dialogue: 0,0:16:19.46,0:16:22.64,Default,,0000,0000,0000,,you know, the faces\Nbeing so important for everyday life,
Dialogue: 0,0:16:22.64,0:16:23.64,Default,,0000,0000,0000,,and given the fact
Dialogue: 0,0:16:23.64,0:16:27.48,Default,,0000,0000,0000,,that there are specialized mechanisms\Nin the brain for face processing -
Dialogue: 0,0:16:27.48,0:16:29.77,Default,,0000,0000,0000,,maybe if we do something\Nthat's that specialized,
Dialogue: 0,0:16:29.77,0:16:33.65,Default,,0000,0000,0000,,we might be able to individually\Nguess individual faces.
Dialogue: 0,0:16:35.14,0:16:37.15,Default,,0000,0000,0000,,As a summary, that's what they did.
Dialogue: 0,0:16:37.15,0:16:39.13,Default,,0000,0000,0000,,They showed people a bunch of faces,
Dialogue: 0,0:16:39.13,0:16:41.03,Default,,0000,0000,0000,,they trained up these algorithms,
Dialogue: 0,0:16:41.03,0:16:46.47,Default,,0000,0000,0000,,and then using fMRI activity alone,\Nthey were able to generate good guesses,
Dialogue: 0,0:16:46.80,0:16:48.79,Default,,0000,0000,0000,,above-chance guesses,
Dialogue: 0,0:16:48.79,0:16:51.68,Default,,0000,0000,0000,,as to which faces\Nan individual is looking at,
Dialogue: 0,0:16:51.68,0:16:53.57,Default,,0000,0000,0000,,again, based on fMRI activity alone,
Dialogue: 0,0:16:53.57,0:16:56.10,Default,,0000,0000,0000,,and that's what's shown here on the right.
Dialogue: 0,0:16:56.93,0:16:58.88,Default,,0000,0000,0000,,Just to give you a little bit more detail
Dialogue: 0,0:16:58.88,0:17:01.21,Default,,0000,0000,0000,,for those who kind of\Nlike that kind of stuff.
Dialogue: 0,0:17:01.21,0:17:03.46,Default,,0000,0000,0000,,It's really relying\Non a lot of computer vision.
Dialogue: 0,0:17:03.46,0:17:04.95,Default,,0000,0000,0000,,This is not voodoo magic.
Dialogue: 0,0:17:04.95,0:17:07.19,Default,,0000,0000,0000,,There's a lot of computer vision\Nthat allows you
Dialogue: 0,0:17:07.19,0:17:13.30,Default,,0000,0000,0000,,to mathematically summarize\Nfeatures of a whole array of faces.
Dialogue: 0,0:17:14.42,0:17:15.92,Default,,0000,0000,0000,,People call them "eigenfaces."
Dialogue: 0,0:17:15.92,0:17:18.58,Default,,0000,0000,0000,,It's based on principle\Ncomponent analysis.
Dialogue: 0,0:17:18.58,0:17:22.60,Default,,0000,0000,0000,,So you can decompose these faces\Nand describe a whole array of faces
Dialogue: 0,0:17:22.60,0:17:25.26,Default,,0000,0000,0000,,with these mathematical algorithms.
Dialogue: 0,0:17:25.26,0:17:28.85,Default,,0000,0000,0000,,And basically what\NAlan and Brice did in my lab
Dialogue: 0,0:17:28.85,0:17:34.90,Default,,0000,0000,0000,,was to map those mathematical\Ncomponents to brain activity,
Dialogue: 0,0:17:34.90,0:17:39.07,Default,,0000,0000,0000,,and then once you train up a database,\Na statistic library, for doing so,
Dialogue: 0,0:17:39.07,0:17:42.34,Default,,0000,0000,0000,,then you can take a novel face\Nthat's shown up here on the right,
Dialogue: 0,0:17:42.98,0:17:46.22,Default,,0000,0000,0000,,record the fMRI activity\Nto those novel faces,
Dialogue: 0,0:17:46.22,0:17:49.69,Default,,0000,0000,0000,,and then make your best guess\Nas to which face people were looking at.
Dialogue: 0,0:17:49.69,0:17:53.10,Default,,0000,0000,0000,,Right now, we are about 65, 70% correct,
Dialogue: 0,0:17:53.51,0:17:58.98,Default,,0000,0000,0000,,and we and others are working very hard\Nto improve that performance.
Dialogue: 0,0:18:00.30,0:18:01.87,Default,,0000,0000,0000,,And this is just another example:
Dialogue: 0,0:18:01.87,0:18:03.92,Default,,0000,0000,0000,,a whole bunch of originals\Non the left here,
Dialogue: 0,0:18:03.92,0:18:07.44,Default,,0000,0000,0000,,and then reconstructions\Nand guesses on the right.
Dialogue: 0,0:18:07.69,0:18:11.44,Default,,0000,0000,0000,,This study actually just came out,\Njust got published this past week,
Dialogue: 0,0:18:11.44,0:18:13.00,Default,,0000,0000,0000,,so I'm pretty excited
Dialogue: 0,0:18:13.00,0:18:16.22,Default,,0000,0000,0000,,that you might even see\Nsome media coverage of the work,
Dialogue: 0,0:18:16.22,0:18:19.05,Default,,0000,0000,0000,,and again, I really want to credit\NAlan Cowen and Brice Kuhl
Dialogue: 0,0:18:19.05,0:18:20.72,Default,,0000,0000,0000,,for making this possible.
Dialogue: 0,0:18:21.57,0:18:22.57,Default,,0000,0000,0000,,Now, you may wonder,
Dialogue: 0,0:18:22.57,0:18:26.56,Default,,0000,0000,0000,,going back, kind of, to the more\Ncategorical reading out of brains -
Dialogue: 0,0:18:27.18,0:18:28.48,Default,,0000,0000,0000,,That work came out in 2011.
Dialogue: 0,0:18:28.48,0:18:31.01,Default,,0000,0000,0000,,I've been talking about it a lot\Nin classes and stuff,
Dialogue: 0,0:18:31.01,0:18:33.23,Default,,0000,0000,0000,,and a really typical\Nquestion that you get -
Dialogue: 0,0:18:33.51,0:18:35.18,Default,,0000,0000,0000,,and this question I actually like -
Dialogue: 0,0:18:35.18,0:18:37.70,Default,,0000,0000,0000,,is, "Well, if you can read out\Nwhat people are seeing,
Dialogue: 0,0:18:37.70,0:18:39.75,Default,,0000,0000,0000,,can you read out what they're dreaming?"
Dialogue: 0,0:18:39.75,0:18:42.50,Default,,0000,0000,0000,,because that's the natural next step.
Dialogue: 0,0:18:42.50,0:18:46.07,Default,,0000,0000,0000,,And this is work done\Nby Horikawa et al. in Japan,
Dialogue: 0,0:18:46.07,0:18:47.78,Default,,0000,0000,0000,,published in Science last year.
Dialogue: 0,0:18:47.78,0:18:51.48,Default,,0000,0000,0000,,They actually took a lot of the work\Nthat Jack Gallant did,
Dialogue: 0,0:18:51.48,0:18:54.44,Default,,0000,0000,0000,,modified it so they can decode\Nwhat people were dreaming
Dialogue: 0,0:18:54.44,0:18:57.06,Default,,0000,0000,0000,,while they were sleeping\Ninside the scanner.
Dialogue: 0,0:18:57.06,0:18:58.49,Default,,0000,0000,0000,,And what we have here on the -
Dialogue: 0,0:18:58.49,0:18:59.49,Default,,0000,0000,0000,,Ignore the words,
Dialogue: 0,0:18:59.49,0:19:03.10,Default,,0000,0000,0000,,these are just ways to try to categorize\Nwhat people were dreaming about.
Dialogue: 0,0:19:03.10,0:19:05.07,Default,,0000,0000,0000,,Just focus on the top left.
Dialogue: 0,0:19:05.42,0:19:09.71,Default,,0000,0000,0000,,That's the computer\Nalgorithm's guess, or decoder's guess
Dialogue: 0,0:19:09.71,0:19:11.72,Default,,0000,0000,0000,,as to what people were dreaming about.
Dialogue: 0,0:19:11.72,0:19:14.18,Default,,0000,0000,0000,,You can imagine\Nit's going to be a lot more messier
Dialogue: 0,0:19:14.18,0:19:16.45,Default,,0000,0000,0000,,than what you saw before for perception,
Dialogue: 0,0:19:16.45,0:19:17.88,Default,,0000,0000,0000,,but still, it's pretty amazing
Dialogue: 0,0:19:17.88,0:19:20.83,Default,,0000,0000,0000,,that you can decode dreams now\Nwhile people are sleeping,
Dialogue: 0,0:19:20.83,0:19:23.31,Default,,0000,0000,0000,,you know, while they are\Nin rapid eye movement sleep.
Dialogue: 0,0:19:23.31,0:19:26.12,Default,,0000,0000,0000,,So what we have here -\NI'll turn it on, it's a video.
Dialogue: 0,0:19:26.12,0:19:29.27,Default,,0000,0000,0000,,I will mention that,\Na little stereotypically,
Dialogue: 0,0:19:30.25,0:19:31.65,Default,,0000,0000,0000,,this is a male subject,
Dialogue: 0,0:19:31.65,0:19:34.94,Default,,0000,0000,0000,,and male subjects tend to think\Nabout one thing when they're dreaming,
Dialogue: 0,0:19:34.94,0:19:35.94,Default,,0000,0000,0000,,(Laughter)
Dialogue: 0,0:19:35.94,0:19:39.80,Default,,0000,0000,0000,,but it is rated PG,\Nif you'll excuse me for that.
Dialogue: 0,0:19:42.77,0:19:45.22,Default,,0000,0000,0000,,And again, you know, it's not easy.
Dialogue: 0,0:19:45.22,0:19:47.37,Default,,0000,0000,0000,,It's, again, remarkable\Nthat they can do this,
Dialogue: 0,0:19:47.37,0:19:50.04,Default,,0000,0000,0000,,and it gets better\Nas it goes towards the end,
Dialogue: 0,0:19:50.04,0:19:53.56,Default,,0000,0000,0000,,right before they awakened the subject\Nto ask them what they were dreaming,
Dialogue: 0,0:19:53.56,0:19:55.70,Default,,0000,0000,0000,,but you can start seeing\Nbuildings and places,
Dialogue: 0,0:19:55.70,0:19:59.68,Default,,0000,0000,0000,,and here comes the dream,\Nthe real concrete stuff,
Dialogue: 0,0:19:59.68,0:20:00.68,Default,,0000,0000,0000,,right there.
Dialogue: 0,0:20:00.68,0:20:01.68,Default,,0000,0000,0000,,Okay.
Dialogue: 0,0:20:01.68,0:20:03.33,Default,,0000,0000,0000,,(Laughter)
Dialogue: 0,0:20:03.33,0:20:06.62,Default,,0000,0000,0000,,Then they wake up the subject and say,\N"What were you dreaming about?"
Dialogue: 0,0:20:06.62,0:20:09.72,Default,,0000,0000,0000,,and they will report something\Nthat's consistent with the decoder,
Dialogue: 0,0:20:09.72,0:20:11.64,Default,,0000,0000,0000,,as a way to access its validity.
Dialogue: 0,0:20:14.66,0:20:17.08,Default,,0000,0000,0000,,So, we can decode and read out the mind;
Dialogue: 0,0:20:17.08,0:20:19.57,Default,,0000,0000,0000,,the field, the kind\Nof neuroscience can do that.
Dialogue: 0,0:20:19.57,0:20:22.100,Default,,0000,0000,0000,,And the question is, "What are we\Ngoing to use this capability for?"
Dialogue: 0,0:20:23.62,0:20:26.38,Default,,0000,0000,0000,,Some of you may have seen this movie, \Nthe Minority Report.
Dialogue: 0,0:20:26.38,0:20:30.07,Default,,0000,0000,0000,,It's a fascinating, amazing movie.\NI encourage you to see it if you can.
Dialogue: 0,0:20:30.07,0:20:32.12,Default,,0000,0000,0000,,This movie is all\Nabout predicting behavior,
Dialogue: 0,0:20:32.12,0:20:33.85,Default,,0000,0000,0000,,you know, reading out brain activity
Dialogue: 0,0:20:33.85,0:20:36.66,Default,,0000,0000,0000,,to try to predict what's going\Nto happen in the future.
Dialogue: 0,0:20:36.66,0:20:41.51,Default,,0000,0000,0000,,And neuroscientists are working\Non this aspect of cognition as well.
Dialogue: 0,0:20:41.88,0:20:46.28,Default,,0000,0000,0000,,As one example, colleagues\Nover at the University of New Mexico,
Dialogue: 0,0:20:46.28,0:20:47.84,Default,,0000,0000,0000,,Kent Kiehl's group,
Dialogue: 0,0:20:49.04,0:20:53.09,Default,,0000,0000,0000,,scanned a whole bunch of prisoners\Nwho committed felony crimes.
Dialogue: 0,0:20:53.09,0:20:55.04,Default,,0000,0000,0000,,They were all released back into society,
Dialogue: 0,0:20:55.04,0:20:57.07,Default,,0000,0000,0000,,and the depended measure was:
Dialogue: 0,0:20:57.07,0:20:59.58,Default,,0000,0000,0000,,"How likely was it\Nfor someone to be rearrested?
Dialogue: 0,0:20:59.58,0:21:02.17,Default,,0000,0000,0000,,What was the likelihood of recidivism?"
Dialogue: 0,0:21:02.17,0:21:04.66,Default,,0000,0000,0000,,And they found that based on brain scans
Dialogue: 0,0:21:04.66,0:21:08.19,Default,,0000,0000,0000,,while people were\Ndoing tasks in the prison -
Dialogue: 0,0:21:08.19,0:21:10.71,Default,,0000,0000,0000,,in the scanner that was\Nlocated in the prison,
Dialogue: 0,0:21:10.71,0:21:12.64,Default,,0000,0000,0000,,they could predict reliably
Dialogue: 0,0:21:12.64,0:21:15.61,Default,,0000,0000,0000,,who was more likely\Nto come back to prison,
Dialogue: 0,0:21:15.61,0:21:17.47,Default,,0000,0000,0000,,be rearrested, re-commit a crime
Dialogue: 0,0:21:17.47,0:21:20.04,Default,,0000,0000,0000,,versus who was less likely to do so.
Dialogue: 0,0:21:20.38,0:21:24.66,Default,,0000,0000,0000,,In my own laboratory here at Yale,\NI worked with a undergraduate student.
Dialogue: 0,0:21:24.66,0:21:27.73,Default,,0000,0000,0000,,This is, again,\Nhis own idea, Harrison Korn,
Dialogue: 0,0:21:27.73,0:21:29.58,Default,,0000,0000,0000,,together with Michael Johnson.
Dialogue: 0,0:21:29.58,0:21:35.59,Default,,0000,0000,0000,,They wanted to see if they could measure\Nimplicit or unconscious prejudice
Dialogue: 0,0:21:36.04,0:21:38.77,Default,,0000,0000,0000,,while people were looking\Nat these vignettes
Dialogue: 0,0:21:39.21,0:21:42.05,Default,,0000,0000,0000,,which involve employment\Ndiscrimination cases.
Dialogue: 0,0:21:42.05,0:21:43.63,Default,,0000,0000,0000,,So you can read through that:
Dialogue: 0,0:21:43.63,0:21:45.90,Default,,0000,0000,0000,,"Rodney, a 19 year-old African American,
Dialogue: 0,0:21:45.90,0:21:49.29,Default,,0000,0000,0000,,applied for a job as a clerk\Nat a teen-apparel store.
Dialogue: 0,0:21:49.50,0:21:50.99,Default,,0000,0000,0000,,Had experience as a cashier,
Dialogue: 0,0:21:50.99,0:21:53.24,Default,,0000,0000,0000,,glowing recommendation\Nfrom a previous employer
Dialogue: 0,0:21:53.24,0:21:57.09,Default,,0000,0000,0000,,but was denied the job because he did not\Nmatch the company's look."
Dialogue: 0,0:21:57.09,0:21:58.09,Default,,0000,0000,0000,,Okay?
Dialogue: 0,0:21:58.09,0:21:59.81,Default,,0000,0000,0000,,And et cetera, et cetera, et cetera.
Dialogue: 0,0:21:59.81,0:22:00.82,Default,,0000,0000,0000,,And the question is,
Dialogue: 0,0:22:00.82,0:22:05.42,Default,,0000,0000,0000,,"In this hypothetical damage awards case,
Dialogue: 0,0:22:05.42,0:22:09.27,Default,,0000,0000,0000,,how much would you\Nas a subject award Rodney?"
Dialogue: 0,0:22:09.27,0:22:12.96,Default,,0000,0000,0000,,And you can imagine a range of responses,
Dialogue: 0,0:22:12.96,0:22:16.34,Default,,0000,0000,0000,,where you would award Rodney a lot\Nor award Rodney a little,
Dialogue: 0,0:22:16.34,0:22:18.80,Default,,0000,0000,0000,,and we had many other cases like this.
Dialogue: 0,0:22:18.80,0:22:19.80,Default,,0000,0000,0000,,And the question is,
Dialogue: 0,0:22:19.80,0:22:23.08,Default,,0000,0000,0000,,"Can we predict which subjects\Nare going to award a lot of damages
Dialogue: 0,0:22:23.08,0:22:26.17,Default,,0000,0000,0000,,and which subjects\Nare going to award few damages?"
Dialogue: 0,0:22:26.65,0:22:31.57,Default,,0000,0000,0000,,And Harrison found that if you scan people
Dialogue: 0,0:22:31.57,0:22:36.93,Default,,0000,0000,0000,,while they were looking\Nat white faces versus black faces,
Dialogue: 0,0:22:36.93,0:22:39.51,Default,,0000,0000,0000,,based on the brain response\Nto those faces,
Dialogue: 0,0:22:39.51,0:22:40.85,Default,,0000,0000,0000,,he could predict
Dialogue: 0,0:22:40.85,0:22:47.06,Default,,0000,0000,0000,,the amount of awards that were given\Nin this hypothetical case later on.
Dialogue: 0,0:22:47.47,0:22:50.92,Default,,0000,0000,0000,,And that's shown here\Nby these correlations on the left.
Dialogue: 0,0:22:51.40,0:22:54.47,Default,,0000,0000,0000,,As a final thing I want to share\Nof my lab right now,
Dialogue: 0,0:22:54.47,0:22:58.19,Default,,0000,0000,0000,,we're really interested in this issue\Nof "what does it mean to be in the zone?
Dialogue: 0,0:22:58.19,0:23:00.29,Default,,0000,0000,0000,,what does it mean\Nto be able to stay focused
Dialogue: 0,0:23:00.29,0:23:01.82,Default,,0000,0000,0000,,for a sustained amount of time?"
Dialogue: 0,0:23:01.82,0:23:03.01,Default,,0000,0000,0000,,which, you might imagine,
Dialogue: 0,0:23:03.01,0:23:05.59,Default,,0000,0000,0000,,is critical for almost anything\Nthat's important to us,
Dialogue: 0,0:23:05.59,0:23:08.66,Default,,0000,0000,0000,,whether that's athletic performance,\Nmusical performance,
Dialogue: 0,0:23:08.66,0:23:12.60,Default,,0000,0000,0000,,theatrical performance,\Ngiving a lecture, taking an exam.
Dialogue: 0,0:23:13.04,0:23:16.88,Default,,0000,0000,0000,,Almost all these domains\Nrequire sustained focus and attention.
Dialogue: 0,0:23:16.88,0:23:18.45,Default,,0000,0000,0000,,You all know that there are times
Dialogue: 0,0:23:18.45,0:23:20.59,Default,,0000,0000,0000,,when you're really in the zone\Nand can do well
Dialogue: 0,0:23:20.59,0:23:23.50,Default,,0000,0000,0000,,and there are other times\Nwhere even if you're fully prepared,
Dialogue: 0,0:23:23.50,0:23:24.50,Default,,0000,0000,0000,,you don't do as well,
Dialogue: 0,0:23:24.50,0:23:27.31,Default,,0000,0000,0000,,because you're not in the zone\Nor you're kind of distracted.
Dialogue: 0,0:23:27.31,0:23:29.88,Default,,0000,0000,0000,,So my lab is interested\Nin how we can characterize this
Dialogue: 0,0:23:29.88,0:23:31.76,Default,,0000,0000,0000,,and how we can predict that.
Dialogue: 0,0:23:32.23,0:23:35.01,Default,,0000,0000,0000,,And as work that's not even published yet,
Dialogue: 0,0:23:35.01,0:23:37.79,Default,,0000,0000,0000,,Emily Finn, a graduate student\Nin neuroscience program,
Dialogue: 0,0:23:37.79,0:23:40.82,Default,,0000,0000,0000,,and Monica Rosenberg,\Na graduate student in psychology,
Dialogue: 0,0:23:41.19,0:23:42.53,Default,,0000,0000,0000,,they're both working with me
Dialogue: 0,0:23:42.53,0:23:47.92,Default,,0000,0000,0000,,and Todd Constable over at the Magnetic\NResonance Research Center at Yale.
Dialogue: 0,0:23:48.92,0:23:53.20,Default,,0000,0000,0000,,They find that it's not easy to predict\Nwhat it means to be in the zone
Dialogue: 0,0:23:53.20,0:23:55.11,Default,,0000,0000,0000,,and what it means to be out of the zone,
Dialogue: 0,0:23:55.11,0:23:57.11,Default,,0000,0000,0000,,but if you start using measures
Dialogue: 0,0:23:57.11,0:24:01.64,Default,,0000,0000,0000,,that also look at how different brain\Nareas are connected with each other -
Dialogue: 0,0:24:01.64,0:24:03.98,Default,,0000,0000,0000,,something that we call\N"functional connectivity" -
Dialogue: 0,0:24:03.98,0:24:07.09,Default,,0000,0000,0000,,you can start getting at\Nwhat it means to be in the zone
Dialogue: 0,0:24:07.09,0:24:09.10,Default,,0000,0000,0000,,versus what it means\Nto be out of the zone.
Dialogue: 0,0:24:09.10,0:24:10.11,Default,,0000,0000,0000,,And what we have here
Dialogue: 0,0:24:10.11,0:24:14.75,Default,,0000,0000,0000,,are two types of networks \Nthat Emily and Monica have identified.
Dialogue: 0,0:24:14.75,0:24:18.71,Default,,0000,0000,0000,,There is a network of brain regions\Nand the connectivity between them
Dialogue: 0,0:24:18.71,0:24:20.70,Default,,0000,0000,0000,,that predicts good performance,
Dialogue: 0,0:24:20.70,0:24:23.26,Default,,0000,0000,0000,,and there is another complementary network
Dialogue: 0,0:24:23.26,0:24:27.73,Default,,0000,0000,0000,,that has different brain areas\Nand different types of connectivity
Dialogue: 0,0:24:27.73,0:24:30.65,Default,,0000,0000,0000,,that seems to be correlated\Nwith bad performance.
Dialogue: 0,0:24:30.65,0:24:33.58,Default,,0000,0000,0000,,So if you had an activity\Nin the blue network over here,
Dialogue: 0,0:24:33.58,0:24:34.98,Default,,0000,0000,0000,,people tend to do worse.
Dialogue: 0,0:24:34.98,0:24:38.20,Default,,0000,0000,0000,,If you have activity in the red network,\Npeople tend to do better.
Dialogue: 0,0:24:38.51,0:24:40.78,Default,,0000,0000,0000,,These networks are so robust
Dialogue: 0,0:24:40.78,0:24:44.94,Default,,0000,0000,0000,,that you can measure these networks\Neven when subjects are doing nothing.
Dialogue: 0,0:24:44.94,0:24:48.08,Default,,0000,0000,0000,,We put them in the scanner\Nbefore anything starts,
Dialogue: 0,0:24:48.08,0:24:50.72,Default,,0000,0000,0000,,have them close their eyes\Nand rest for 10 minutes -
Dialogue: 0,0:24:50.72,0:24:52.98,Default,,0000,0000,0000,,you call that a "resting state scan" -
Dialogue: 0,0:24:52.98,0:24:56.68,Default,,0000,0000,0000,,and based on their resting\Nstate activity alone,
Dialogue: 0,0:24:57.12,0:25:02.25,Default,,0000,0000,0000,,we were able to predict how well\Nthey were going to do over the next hour.
Dialogue: 0,0:25:02.80,0:25:05.96,Default,,0000,0000,0000,,Again, red network corresponds\Nwith better performance;
Dialogue: 0,0:25:05.96,0:25:08.70,Default,,0000,0000,0000,,blue network corresponds\Nwith worse performance.
Dialogue: 0,0:25:08.70,0:25:10.68,Default,,0000,0000,0000,,And just on a variety of tasks,
Dialogue: 0,0:25:10.68,0:25:14.53,Default,,0000,0000,0000,,you have very high\Npredictive power in these studies.
Dialogue: 0,0:25:14.86,0:25:20.57,Default,,0000,0000,0000,,The implications for studying ADHD\Nand many other types of domains,
Dialogue: 0,0:25:20.57,0:25:21.72,Default,,0000,0000,0000,,I think, are very large,
Dialogue: 0,0:25:21.72,0:25:24.02,Default,,0000,0000,0000,,so hopefully, this will get\Npublished soon.
Dialogue: 0,0:25:24.02,0:25:26.55,Default,,0000,0000,0000,,So in closing, I hope I've convinced you
Dialogue: 0,0:25:26.55,0:25:30.23,Default,,0000,0000,0000,,that I am no longer embarrassed to say\Nthat we can read your mind,
Dialogue: 0,0:25:30.23,0:25:31.92,Default,,0000,0000,0000,,if anyone asks me.
Dialogue: 0,0:25:32.17,0:25:35.09,Default,,0000,0000,0000,,And I thank you for your attention.
Dialogue: 0,0:25:35.09,0:25:37.99,Default,,0000,0000,0000,,(Applause)\N