0:00:00.857,0:00:03.230
I'm going to talk about[br]the strategizing brain.

0:00:03.254,0:00:05.640
We're going to use an unusual[br]combination of tools

0:00:05.664,0:00:07.254
from game theory and neuroscience

0:00:07.278,0:00:10.548
to understand how people interact socially[br]when value is on the line.

0:00:10.572,0:00:14.044
So game theory is a branch of,[br]originally, applied mathematics,

0:00:14.068,0:00:17.496
used mostly in economics and political[br]science, a little bit in biology,

0:00:17.520,0:00:20.363
that gives us a mathematical[br]taxonomy of social life,

0:00:20.387,0:00:22.706
and it predicts what people[br]are likely to do

0:00:22.730,0:00:24.046
and believe others will do

0:00:24.070,0:00:27.043
in cases where everyone's actions[br]affect everyone else.

0:00:27.067,0:00:30.752
That's a lot of things: competition,[br]cooperation, bargaining,

0:00:30.776,0:00:33.128
games like hide-and-seek and poker.

0:00:33.954,0:00:35.887
Here's a simple game to get us started.

0:00:35.911,0:00:38.375
Everyone chooses a number[br]from zero to 100.

0:00:38.399,0:00:40.859
We're going to compute[br]the average of those numbers,

0:00:40.883,0:00:44.922
and whoever's closest to two-thirds[br]of the average wins a fixed prize.

0:00:44.946,0:00:47.692
So you want to be a little bit[br]below the average number

0:00:47.716,0:00:48.868
but not too far below,

0:00:48.892,0:00:52.513
and everyone else wants to be a little bit[br]below the average number as well.

0:00:52.537,0:00:54.084
Think about what you might pick.

0:00:54.108,0:00:55.269
As you're thinking,

0:00:55.293,0:00:58.347
this is a toy model of something like[br]selling in the stock market

0:00:58.371,0:00:59.989
during a rising market:

0:01:00.013,0:01:03.123
You don't want to sell too early,[br]because you miss out on profits,

0:01:03.147,0:01:06.195
but you don't want to wait too late,[br]to when everyone else sells,

0:01:06.219,0:01:07.444
triggering a crash.

0:01:07.468,0:01:11.048
You want to be a little bit ahead[br]of the competition, but not too far ahead.

0:01:11.072,0:01:14.157
OK, here's two theories[br]about how people might think about this,

0:01:14.181,0:01:15.381
then we'll see some data.

0:01:15.405,0:01:17.001
Some of these will sound familiar

0:01:17.025,0:01:19.084
because you probably[br]are thinking that way.

0:01:19.108,0:01:20.760
I'm using my brain theory to see.

0:01:20.784,0:01:24.198
A lot of people say, "I really don't know[br]what people are going to pick,

0:01:24.222,0:01:27.749
so I think the average will be 50" --[br]they're not being strategic at all --

0:01:27.773,0:01:29.835
and "I'll pick two-thirds[br]of 50, that's 33."

0:01:29.859,0:01:31.013
That's a start.

0:01:31.037,0:01:33.426
Other people, who are a little[br]more sophisticated,

0:01:33.450,0:01:34.691
using more working memory,

0:01:34.715,0:01:36.344
say, "I think people will pick 33,

0:01:36.368,0:01:38.579
because they're going[br]to pick a response to 50,

0:01:38.603,0:01:40.901
and so I'll pick 22,[br]which is two-thirds of 33."

0:01:40.925,0:01:43.492
They're doing one extra step[br]of thinking, two steps.

0:01:43.913,0:01:45.117
That's better.

0:01:45.141,0:01:47.868
Of course, in principle,[br]you could do three, four or more,

0:01:47.892,0:01:49.838
but it starts to get very difficult.

0:01:49.862,0:01:51.793
Just like in language and other domains,

0:01:51.817,0:01:54.882
we know that it's hard for people[br]to parse very complex sentences

0:01:54.906,0:01:56.197
with a recursive structure.

0:01:56.221,0:01:58.370
This is called the cognitive[br]hierarchy theory,

0:01:58.394,0:02:00.643
something I've worked on[br]and a few other people,

0:02:00.667,0:02:02.437
and it indicates a kind of hierarchy,

0:02:02.461,0:02:05.900
along with some assumptions about[br]how many people stop at different steps

0:02:05.924,0:02:07.931
and how the steps of thinking are affected

0:02:07.955,0:02:10.399
by lots of interesting variables[br]and variant people,

0:02:10.423,0:02:11.623
as we'll see in a minute.

0:02:11.647,0:02:14.785
A very different theory, a much[br]more popular one and an older one,

0:02:14.809,0:02:17.288
due largely to John Nash[br]of "A Beautiful Mind" fame,

0:02:17.312,0:02:19.397
is what's called "equilibrium analysis."

0:02:19.421,0:02:22.234
So if you've ever taken[br]a game theory course at any level,

0:02:22.258,0:02:24.035
you'll have learned a bit about this.

0:02:24.059,0:02:25.887
An equilibrium is a mathematical state

0:02:25.911,0:02:29.211
in which everybody has figured out[br]exactly what everyone else will do.

0:02:29.235,0:02:30.579
It is a very useful concept,

0:02:30.603,0:02:32.657
but behaviorally,[br]it may not exactly explain

0:02:32.681,0:02:35.955
what people do the first time they play[br]these types of economic games

0:02:35.979,0:02:37.900
or in situations in the outside world.

0:02:37.924,0:02:40.725
In this case, the equilibrium[br]makes a very bold prediction,

0:02:40.749,0:02:43.458
which is: everyone wants[br]to be below everyone else,

0:02:43.482,0:02:45.183
therefore, they'll play zero.

0:02:45.723,0:02:46.880
Let's see what happens.

0:02:46.904,0:02:49.011
This experiment's been done[br]many, many times.

0:02:49.035,0:02:51.237
Some of the earliest ones[br]were done in the '90s

0:02:51.261,0:02:53.067
by me and Rosemarie Nagel and others.

0:02:53.091,0:02:55.611
This is a beautiful data set[br]of 9,000 people

0:02:55.635,0:02:58.856
who wrote in to three newspapers[br]and magazines that had a contest.

0:02:58.880,0:03:00.923
The contest said, send in your numbers,

0:03:00.947,0:03:04.281
and whoever is close to two-thirds[br]of the average will win a big prize.

0:03:04.305,0:03:08.038
As you can see, there's so much data[br]here, you can see the spikes very visibly.

0:03:08.062,0:03:10.776
There's a spike at 33 --[br]those are people doing one step.

0:03:10.800,0:03:13.019
There is another spike visible at 22.

0:03:13.043,0:03:16.059
Notice, by the way, most people[br]pick numbers right around there;

0:03:16.083,0:03:18.241
they don't necessarily[br]pick exactly 33 and 22.

0:03:18.265,0:03:20.181
There's something a bit noisy around it.

0:03:20.205,0:03:22.173
But you can see those spikes on that end.

0:03:22.197,0:03:23.682
There's another group of people

0:03:23.706,0:03:26.193
who seem to have a firm grip[br]on equilibrium analysis,

0:03:26.217,0:03:27.953
because they're picking zero or one.

0:03:27.977,0:03:29.624
But they lose, right?

0:03:29.648,0:03:33.032
Because picking a number that low[br]is actually a bad choice

0:03:33.056,0:03:35.795
if other people aren't doing[br]equilibrium analysis as well.

0:03:35.819,0:03:37.494
So they're smart, but poor.

0:03:37.518,0:03:39.582
(Laughter)

0:03:39.606,0:03:42.073
Where are these things[br]happening in the brain?

0:03:42.097,0:03:45.790
One study by Coricelli and Nagel[br]gives a really sharp, interesting answer.

0:03:45.814,0:03:49.626
They had people play this game[br]while they were being scanned in an fMRI,

0:03:49.650,0:03:50.807
and two conditions:

0:03:50.831,0:03:52.217
in some trials, they're told,

0:03:52.241,0:03:54.838
"You're playing another person[br]who's playing right now.

0:03:54.862,0:03:57.865
We'll match up your behavior[br]at the end and pay you if you win."

0:03:57.889,0:04:00.617
In other trials, they're told,[br]"You're playing a computer,

0:04:00.641,0:04:02.165
they're just choosing randomly."

0:04:02.189,0:04:04.351
So what you see here[br]is a subtraction of areas

0:04:04.375,0:04:07.334
in which there's more brain activity[br]when you're playing people

0:04:07.358,0:04:08.936
compared to playing the computer.

0:04:08.960,0:04:11.496
And you see activity[br]in some regions we've seen today,

0:04:11.520,0:04:13.769
medial prefrontal cortex,[br]dorsomedial, up here,

0:04:13.793,0:04:16.185
ventromedial prefrontal cortex,[br]anterior cingulate,

0:04:16.209,0:04:19.224
an area that's involved[br]in lots of types of conflict resolution,

0:04:19.248,0:04:20.984
like if you're playing "Simon Says,"

0:04:21.008,0:04:24.181
and also the right and left[br]temporoparietal junction.

0:04:24.205,0:04:27.080
And these are all areas[br]which are fairly reliably known to be

0:04:27.104,0:04:29.355
part of what's called[br]a "theory of mind" circuit

0:04:29.379,0:04:30.905
or "mentalizing circuit."

0:04:30.929,0:04:34.436
That is, it's a circuit that's used[br]to imagine what other people might do.

0:04:34.460,0:04:38.367
These were some of the first studies[br]to see this tied in to game theory.

0:04:38.778,0:04:41.024
What happens with these[br]one- and two-step types?

0:04:41.048,0:04:43.299
So, we classify people[br]by what they picked,

0:04:43.323,0:04:46.853
and then we look at the difference[br]between playing humans versus computers,

0:04:46.877,0:04:48.942
which brain areas[br]are differentially active.

0:04:48.966,0:04:50.934
On the top, you see the one-step players.

0:04:50.958,0:04:52.343
There's almost no difference.

0:04:52.367,0:04:55.244
The reason is, they're treating[br]other people like a computer,

0:04:55.268,0:04:56.419
and the brain is too.

0:04:56.443,0:04:59.466
The bottom players, you see[br]all the activity in dorsomedial PFC.

0:04:59.490,0:05:02.497
So we know the two-step players[br]are doing something differently.

0:05:02.521,0:05:04.522
Now, what can we do with this information?

0:05:04.546,0:05:06.987
You might be able to look[br]at brain activity and say,

0:05:07.011,0:05:10.654
"This person will be a good poker player,"[br]or "This person's socially naive."

0:05:10.678,0:05:14.213
We might also be able to study things[br]like development of adolescent brains

0:05:14.237,0:05:16.674
once we have an idea[br]of where this circuitry exists.

0:05:16.698,0:05:17.850
OK. Get ready.

0:05:17.874,0:05:19.974
I'm saving you some brain activity,

0:05:19.998,0:05:22.735
because you don't need to use[br]your hair detector cells.

0:05:22.759,0:05:26.021
You should use those cells[br]to think carefully about this game.

0:05:26.045,0:05:27.558
This is a bargaining game.

0:05:27.582,0:05:30.599
Two players who are being[br]scanned using EEG electrodes

0:05:30.623,0:05:33.401
are going to bargain[br]over one to six dollars.

0:05:33.425,0:05:36.108
If they can do it in 10 seconds,[br]they'll earn that money.

0:05:36.132,0:05:39.281
If 10 seconds go by and they haven't[br]made a deal, they get nothing.

0:05:39.305,0:05:40.928
That's kind of a mistake together.

0:05:40.952,0:05:43.568
The twist is that one player, on the left,

0:05:43.592,0:05:45.941
is informed about how much[br]on each trial there is.

0:05:45.965,0:05:48.683
They play lots of trials[br]with different amounts each time.

0:05:48.707,0:05:50.804
In this case, they know[br]there's four dollars.

0:05:50.828,0:05:54.428
The uninformed player doesn't know,[br]but they know the informed player knows.

0:05:54.452,0:05:56.647
So the uninformed player's[br]challenge is to say,

0:05:56.671,0:05:57.822
"Is this guy being fair,

0:05:57.846,0:05:59.775
or are they giving me a very low offer

0:05:59.799,0:06:03.572
in order to get me to think there's only[br]one or two dollars available to split?"

0:06:03.596,0:06:06.315
in which case they might reject it[br]and not come to a deal.

0:06:06.339,0:06:09.392
So there's some tension here[br]between trying to get the most money

0:06:09.416,0:06:12.091
but trying to goad the other player[br]into giving you more.

0:06:12.115,0:06:14.593
And the way they bargain[br]is to point on a number line

0:06:14.617,0:06:16.296
that goes from zero to six dollars.

0:06:16.320,0:06:19.154
They're bargaining over how much[br]the uninformed player gets,

0:06:19.178,0:06:21.179
and the informed player will get the rest.

0:06:21.203,0:06:23.347
So this is like[br]a management-labor negotiation

0:06:23.371,0:06:25.101
in which the workers don't know

0:06:25.125,0:06:28.331
how much profits[br]the privately held company has,

0:06:28.355,0:06:30.790
and they want to maybe[br]hold out for more money,

0:06:30.814,0:06:33.210
but the company might want[br]to create the impression

0:06:33.234,0:06:36.194
that there's very little to split:[br]"I'm giving the most I can."

0:06:36.218,0:06:39.622
First, some behavior: a bunch[br]of the subject pairs play face-to-face.

0:06:39.646,0:06:42.086
We have other data[br]where they play across computers.

0:06:42.110,0:06:44.684
That's an interesting difference,[br]as you might imagine.

0:06:44.708,0:06:46.482
But a bunch of the face-to-face pairs

0:06:46.506,0:06:49.233
agree to divide the money[br]evenly every single time.

0:06:49.257,0:06:51.919
Boring. It's just not[br]interesting neurally.

0:06:52.308,0:06:54.532
It's good for them --[br]they make a lot of money.

0:06:54.556,0:06:56.096
But we're interested in:

0:06:56.120,0:06:59.873
Can we say something about when[br]disagreements occur versus don't occur?

0:06:59.897,0:07:02.659
So this is the other group[br]of subjects, who often disagree.

0:07:02.683,0:07:06.160
They bicker and disagree[br]and end up with less money.

0:07:06.184,0:07:09.147
They might be eligible to be[br]on "Real Housewives," the TV show.

0:07:09.171,0:07:10.259
(Laughter)

0:07:10.283,0:07:11.966
You see on the left,

0:07:11.990,0:07:14.624
when the amount to divide[br]is one, two or three dollars,

0:07:14.648,0:07:16.270
they disagree about half the time;

0:07:16.294,0:07:18.645
when it's four, five, six,[br]they agree quite often.

0:07:18.669,0:07:20.868
This turns out to be[br]something that's predicted

0:07:20.892,0:07:22.853
by a very complicated type of game theory

0:07:22.877,0:07:25.984
you should come to graduate school[br]at CalTech and learn about.

0:07:26.008,0:07:28.396
It's a little too complicated[br]to explain right now,

0:07:28.420,0:07:31.063
but the theory tells you[br]that this shape should occur.

0:07:31.087,0:07:33.149
Your intuition might tell you that, too.

0:07:33.173,0:07:36.040
Now I'm going to show you[br]the results from the EEG recording.

0:07:36.064,0:07:37.215
Very complicated.

0:07:37.239,0:07:39.631
The right brain schematic[br]is the uninformed person,

0:07:39.655,0:07:41.055
and the left is the informed.

0:07:41.079,0:07:43.825
Remember that we scanned[br]both brains at the same time,

0:07:43.849,0:07:46.118
so we can ask about time-synced activity

0:07:46.142,0:07:49.158
in similar or different[br]areas simultaneously,

0:07:49.182,0:07:51.447
just like if you wanted[br]to study a conversation,

0:07:51.471,0:07:54.049
and you were scanning two people[br]talking to each other.

0:07:54.073,0:07:56.331
You'd expect common[br]activity in language regions

0:07:56.355,0:07:58.316
when they're listening and communicating.

0:07:58.340,0:08:02.171
So the arrows connect regions[br]that are active at the same time.

0:08:02.195,0:08:03.517
The direction of the arrows

0:08:03.541,0:08:06.307
flows from the region[br]that's active first in time,

0:08:06.331,0:08:10.126
and the arrowhead goes[br]to the region that's active later.

0:08:10.150,0:08:12.197
So in this case, if you look carefully,

0:08:12.221,0:08:14.244
most of the arrows[br]flow from right to left.

0:08:14.268,0:08:17.552
That is, it looks[br]as if the uninformed brain activity

0:08:17.576,0:08:19.187
is happening first,

0:08:19.211,0:08:23.063
and then it's followed[br]by activity in the informed brain.

0:08:23.087,0:08:26.538
And by the way, these are trials[br]where their deals were made.

0:08:26.562,0:08:28.319
This is from the first two seconds.

0:08:28.343,0:08:31.499
We haven't finished analyzing this data,[br]so we're still peeking in,

0:08:31.523,0:08:34.931
but the hope is that we can say something[br]in the first couple of seconds

0:08:34.955,0:08:36.918
about whether they'll make a deal or not,

0:08:36.942,0:08:39.947
which could be very useful in thinking[br]about avoiding litigation

0:08:39.971,0:08:41.835
and ugly divorces and things like that.

0:08:41.859,0:08:45.936
Those are all cases in which a lot[br]of value is lost by delay and strikes.

0:08:46.630,0:08:48.794
Here's the case where[br]the disagreements occur.

0:08:48.818,0:08:51.212
You can see it looks different[br]than the one before.

0:08:51.236,0:08:52.577
There's a lot more arrows.

0:08:52.601,0:08:55.252
That means that the brains[br]are synced up more closely

0:08:55.276,0:08:56.896
in terms of simultaneous activity,

0:08:56.920,0:08:59.123
and the arrows flow clearly[br]from left to right.

0:08:59.147,0:09:01.435
That is, the informed brain[br]seems to be deciding,

0:09:01.459,0:09:03.651
"We're probably not going[br]to make a deal here."

0:09:03.675,0:09:06.418
And then later, there's activity[br]in the uninformed brain.

0:09:06.799,0:09:09.203
Next, I'm going to introduce you[br]to some relatives.

0:09:09.227,0:09:11.388
They're hairy, smelly, fast and strong.

0:09:11.412,0:09:13.906
You might be thinking back[br]to your last Thanksgiving.

0:09:13.930,0:09:14.946
(Laughter)

0:09:14.970,0:09:17.446
Maybe, if you had a chimpanzee with you.

0:09:17.470,0:09:21.476
Charles Darwin and I and you broke[br]off from the family tree from chimpanzees

0:09:21.500,0:09:22.900
about five million years ago.

0:09:22.924,0:09:24.735
They're still our closest genetic kin.

0:09:24.759,0:09:26.478
We share 98.8 percent of the genes.

0:09:26.502,0:09:29.463
We share more genes with them[br]than zebras do with horses.

0:09:29.487,0:09:31.397
And we're also their closest cousin.

0:09:31.421,0:09:34.042
They have more genetic relation[br]to us than to gorillas.

0:09:34.066,0:09:36.805
So, how humans and chimpanzees[br]behave differently

0:09:36.829,0:09:38.923
might tell us a lot about brain evolution.

0:09:39.326,0:09:41.626
This is an amazing memory test

0:09:41.650,0:09:44.442
from [Kyoto], Japan,[br]the Primate Research Institute,

0:09:44.466,0:09:46.469
where they've done a lot of this research.

0:09:46.493,0:09:49.317
This goes back a ways.[br]They're interested in working memory.

0:09:49.341,0:09:51.057
The chimp will see, watch carefully,

0:09:51.081,0:09:54.665
they'll see 200 milliseconds' exposure --[br]that's fast, eight movie frames --

0:09:54.689,0:09:56.666
of numbers one, two, three, four, five.

0:09:56.690,0:09:58.935
Then they disappear[br]and are replaced by squares,

0:09:58.959,0:10:00.586
and they have to press the squares

0:10:00.610,0:10:02.810
that correspond to the numbers[br]from low to high

0:10:02.834,0:10:04.137
to get an apple reward.

0:10:04.161,0:10:05.658
Let's see how they can do it.

0:10:16.478,0:10:17.640
This is a young chimp.

0:10:17.664,0:10:20.581
The young ones are better[br]than the old ones, just like humans.

0:10:20.605,0:10:21.607
(Laughter)

0:10:21.631,0:10:23.109
And they're highly experienced,

0:10:23.133,0:10:25.456
they've done this thousands of times.

0:10:25.480,0:10:28.366
Obviously there's a big training[br]effect, as you can imagine.

0:10:28.390,0:10:29.402
(Laughter)

0:10:29.426,0:10:31.574
You can see they're very[br]blasé and effortless.

0:10:31.598,0:10:34.809
Not only can they do it very well,[br]they do it in a sort of lazy way.

0:10:34.833,0:10:35.837
(Laughter)

0:10:35.861,0:10:37.623
Who thinks you could beat the chimps?

0:10:37.647,0:10:38.707
(Laughter)

0:10:38.731,0:10:40.266
Wrong. (Laughter)

0:10:40.290,0:10:42.875
We can try. We'll try. Maybe we'll try.

0:10:42.899,0:10:46.893
OK, so the next part of the study[br]I'm going to go quickly through

0:10:46.917,0:10:49.893
is based on an idea of Tetsuro Matsuzawa.

0:10:49.917,0:10:53.037
He had a bold idea he called[br]the "cognitive trade-off hypothesis."

0:10:53.061,0:10:56.543
We know chimps are faster and stronger;[br]they're also obsessed with status.

0:10:56.567,0:10:59.248
His thought was, maybe[br]they've preserved brain activities

0:10:59.272,0:11:00.875
and practice them in development

0:11:00.899,0:11:04.714
that are really, really important to them[br]to negotiate status and to win,

0:11:04.738,0:11:07.730
which is something like strategic[br]thinking during competition.

0:11:07.754,0:11:09.290
So we're going to check that out

0:11:09.314,0:11:11.941
by having the chimps actually play a game

0:11:11.965,0:11:14.475
by touching two touch screens.

0:11:14.499,0:11:17.559
The chimps are interacting[br]with each other through the computers.

0:11:17.583,0:11:18.932
They'll press left or right.

0:11:18.956,0:11:22.434
One chimp is called a matcher;[br]they win if they press left-left,

0:11:22.458,0:11:25.603
like a seeker finding someone[br]in hide-and-seek, or right-right.

0:11:25.627,0:11:27.232
The mismatcher wants to mismatch;

0:11:27.256,0:11:29.955
they want to press[br]the opposite screen of the chimp.

0:11:29.979,0:11:32.451
And the rewards are apple cube rewards.

0:11:32.475,0:11:34.802
So here's how game theorists[br]look at these data.

0:11:34.826,0:11:36.848
This is a graph of the percentage of times

0:11:36.872,0:11:39.078
the matcher picked right on the x-axis

0:11:39.102,0:11:41.256
and the percentage of times[br]they picked right

0:11:41.280,0:11:43.485
by the mismatcher on the y-axis.

0:11:43.509,0:11:46.838
So a point here is the behavior[br]by a pair of players,

0:11:46.862,0:11:49.058
one trying to match,[br]one trying to mismatch.

0:11:49.082,0:11:52.399
The NE square in the middle --[br]actually, NE, CH and QRE --

0:11:52.423,0:11:55.547
those are three different theories[br]of Nash equilibrium and others,

0:11:55.571,0:11:57.254
tells you what the theory predicts,

0:11:57.278,0:11:59.403
which is that they should match 50-50,

0:11:59.427,0:12:01.854
because if you play left[br]too much, for example,

0:12:01.878,0:12:04.844
I can exploit that if I'm the mismatcher[br]by then playing right.

0:12:04.868,0:12:07.840
And as you can see, the chimps --[br]each chimp is one triangle --

0:12:07.864,0:12:10.523
are circled around,[br]hovering around that prediction.

0:12:11.205,0:12:12.911
Now we move the payoffs.

0:12:12.935,0:12:16.422
We're going to make the left-left payoff[br]for the matcher a little higher.

0:12:16.446,0:12:17.941
Now they get three apple cubes.

0:12:17.965,0:12:21.240
Game theoretically, that should[br]make the mismatcher's behavior shift:

0:12:21.264,0:12:24.771
the mismatcher will think, "Oh, this guy's[br]going to go for the big reward,

0:12:24.795,0:12:27.323
so I'll go to the right,[br]make sure he doesn't get it."

0:12:27.347,0:12:29.375
And as you can see,[br]their behavior moves up

0:12:29.399,0:12:32.097
in the direction of this change[br]in the Nash equilibrium.

0:12:32.121,0:12:34.367
Finally, we changed[br]the payoffs one more time.

0:12:34.391,0:12:35.639
Now it's four apple cubes,

0:12:35.663,0:12:38.496
and their behavior again moves[br]towards the Nash equilibrium.

0:12:38.520,0:12:41.194
It's sprinkled around,[br]but if you average the chimps out,

0:12:41.218,0:12:42.792
they're really close, within .01.

0:12:42.816,0:12:45.444
They're actually closer[br]than any species we've observed.

0:12:45.468,0:12:48.566
What about humans? You think[br]you're smarter than a chimpanzee?

0:12:49.350,0:12:52.651
Here's two human groups in green and blue.

0:12:52.675,0:12:55.968
They're closer to 50-50; they're not[br]responding to payoffs as closely.

0:12:55.992,0:12:58.288
And also if you study[br]their learning in the game,

0:12:58.312,0:13:00.413
they aren't as sensitive[br]to previous rewards.

0:13:00.437,0:13:04.022
The chimps play better than the humans,[br]in terms of adhering to game theory.

0:13:04.046,0:13:07.247
And these are two different groups[br]of humans, from Japan and Africa;

0:13:07.271,0:13:08.611
they replicate quite nicely.

0:13:08.635,0:13:11.180
None of them are close[br]to where the chimps are.

0:13:11.670,0:13:12.964
So, some things we learned:

0:13:12.988,0:13:16.734
people seem to do a limited amount of[br]strategic thinking using theory of mind.

0:13:16.758,0:13:18.852
We have preliminary[br]evidence from bargaining

0:13:18.876,0:13:21.791
that early warning signs in the brain[br]might be used to predict

0:13:21.815,0:13:24.446
whether there'll be a bad[br]disagreement that costs money,

0:13:24.470,0:13:26.709
and chimps are "better"[br]competitors than humans,

0:13:26.733,0:13:27.975
as judged by game theory.

0:13:27.999,0:13:29.150
Thank you.

0:13:29.174,0:13:32.293
(Applause)