Daniel Wolpert: The real reason for brains

343,833 views ใƒป 2011-11-03

TED


ืื ื ืœื—ืฅ ืคืขืžื™ื™ื ืขืœ ื”ื›ืชื•ื‘ื™ื•ืช ื‘ืื ื’ืœื™ืช ืœืžื˜ื” ื›ื“ื™ ืœื”ืคืขื™ืœ ืืช ื”ืกืจื˜ื•ืŸ.

ืžืชืจื’ื: Yubal Masalker ืžื‘ืงืจ: Ido Dekkers
00:15
I'm a neuroscientist.
0
15260
2000
ืื ื™ ื—ื•ืงืจ ืžืขืจื›ื•ืช ืขืฆื‘ื™ื.
00:17
And in neuroscience,
1
17260
2000
ื‘ืžื“ืขื™ ืขืฆื‘,
00:19
we have to deal with many difficult questions about the brain.
2
19260
3000
ืขืœื™ื ื• ืœื”ืชืžื•ื“ื“ ืขื ื”ืจื‘ื” ืฉืืœื•ืช ืงืฉื•ืช ื‘ื ื•ื’ืข ืœืžื•ื—.
00:22
But I want to start with the easiest question
3
22260
2000
ืื‘ืœ ื‘ืจืฆื•ื ื™ ืœื”ืชื—ื™ืœ ืขื ื”ืฉืืœื” ื”ื›ื™ ืงืœื”
00:24
and the question you really should have all asked yourselves at some point in your life,
4
24260
3000
ื•ื”ืฉืืœื” ืฉื›ื•ืœื ื• ื”ื™ื™ื ื• ืฆืจื™ื›ื™ื ืœืฉืื•ืœ ืืช ืขืฆืžื ื• ืžืชื™ ืฉื”ื•ื,
00:27
because it's a fundamental question
5
27260
2000
ื›ื™ ื–ื• ืฉืืœืช ื™ืกื•ื“
00:29
if we want to understand brain function.
6
29260
2000
ืื ื‘ืจืฆื•ื ื ื• ืœื”ื‘ื™ืŸ ืืช ืคืขื•ืœืช ื”ืžื•ื—.
00:31
And that is, why do we and other animals
7
31260
2000
ื”ืฉืืœื” ื”ื™ื, ืžื“ื•ืข ืœื ื• ื•ืœื—ื™ื•ืช
00:33
have brains?
8
33260
2000
ื™ืฉ ืžื•ื—ื•ืช?
00:35
Not all species on our planet have brains,
9
35260
3000
ืœื ืœื›ืœ ื”ืžื™ื ื™ื ื™ืฉ ืžื•ื—ื•ืช,
00:38
so if we want to know what the brain is for,
10
38260
2000
ื•ืœื›ืŸ ืื ื‘ืจืฆื•ื ื ื• ืœื“ืขืช ืœืžื” ืฆืจื™ืš ืžื•ื—,
00:40
let's think about why we evolved one.
11
40260
2000
ื™ืฉ ืœื—ืฉื•ื‘ ืขืœ ื”ืกื™ื‘ื” ืœื”ืชืคืชื—ื•ืชื•.
00:42
Now you may reason that we have one
12
42260
2000
ืืคืฉืจ ืœื•ืžืจ ืฉื”ื•ื ืงื™ื™ื
00:44
to perceive the world or to think,
13
44260
2000
ื›ื“ื™ ืœืชืคื•ืก ืืช ื”ืขื•ืœื ืื• ื›ื“ื™ ืœื—ืฉื•ื‘,
00:46
and that's completely wrong.
14
46260
2000
ื•ื–ื” ืœื’ืžืจื™ ืœื ื ื›ื•ืŸ.
00:48
If you think about this question for any length of time,
15
48260
3000
ืื ื—ื•ืฉื‘ื™ื ืขืœ ืฉืืœื” ื–ื• ืžืกืคื™ืง ื–ืžืŸ,
00:51
it's blindingly obvious why we have a brain.
16
51260
2000
ื–ื” ื‘ืจื•ืจ ืžืืœื™ื• ืžื“ื•ืข ื™ืฉ ืœื ื• ืžื•ื—.
00:53
We have a brain for one reason and one reason only,
17
53260
3000
ื™ืฉ ืœื ื• ืžื•ื— ื‘ื’ืœืœ ืกื™ื‘ื” ืื—ืช ื•ื™ื—ื™ื“ื”,
00:56
and that's to produce adaptable and complex movements.
18
56260
3000
ื•ื”ื™ื ืœื™ืฆื•ืจ ืชื ื•ืขื•ืช ืกื’ื™ืœื•ืช ื•ืžื•ืจื›ื‘ื•ืช.
00:59
There is no other reason to have a brain.
19
59260
2000
ืื™ืŸ ืกื™ื‘ื” ืื—ืจืช ืœืงื™ื•ื ื”ืžื•ื—.
01:01
Think about it.
20
61260
2000
ืชื—ืฉื‘ื• ืขืœ ื–ื”.
01:03
Movement is the only way you have
21
63260
2000
ืชื ื•ืขื” ื”ื™ื ื”ื“ืจืš ื”ื™ื—ื™ื“ื”
01:05
of affecting the world around you.
22
65260
2000
ืœื”ืฉืคื™ืข ืขืœ ื”ืขื•ืœื ืกื‘ื™ื‘ื ื•.
01:07
Now that's not quite true. There's one other way, and that's through sweating.
23
67260
3000
ื–ื” ืœื ืœื’ืžืจื™ ื ื›ื•ืŸ. ื™ืฉ ื“ืจืš ื ื•ืกืคืช ื•ื”ื™ื ื‘ืืžืฆืขื•ืช ื”ื–ืขื”.
01:10
But apart from that,
24
70260
2000
ืื‘ืœ ืžืœื‘ื“ ื–ืืช,
01:12
everything else goes through contractions of muscles.
25
72260
2000
ื›ืœ ื“ื‘ืจ ืื—ืจ ืžืชื‘ืฆืข ื‘ืืžืฆืขื•ืช ืชื ื•ืขื•ืช ืฉืจื™ืจื™ื.
01:14
So think about communication --
26
74260
2000
ืชื—ืฉื‘ื• ืขืœ ืชืงืฉื•ืจืช --
01:16
speech, gestures, writing, sign language --
27
76260
3000
ื“ื™ื‘ื•ืจ, ืฉืคืช-ื’ื•ืฃ, ื›ืชื™ื‘ื”, ืฉืคืช ืกื™ืžื ื™ื --
01:19
they're all mediated through contractions of your muscles.
28
79260
3000
ื”ื›ืœ ืžืชื‘ืฆืข ื‘ืืžืฆืขื•ืช ืชื ื•ืขื•ืช ืฉืœ ืฉืจื™ืจื™ื.
01:22
So it's really important to remember
29
82260
2000
ืœื›ืŸ ื–ื” ื—ืฉื•ื‘ ืžืื•ื“ ืœื–ื›ื•ืจ
01:24
that sensory, memory and cognitive processes are all important,
30
84260
4000
ืฉืœืžืจื•ืช ืฉืชื”ืœื™ื›ื™ื ืชื—ื•ืฉืชื™ื™ื ื•ืชืคื™ืกืชื™ื™ื ื”ื ื—ืฉื•ื‘ื™ื,
01:28
but they're only important
31
88260
2000
ื”ื ื—ืฉื•ื‘ื™ื ืจืง ื‘ืฉื‘ื™ืœ
01:30
to either drive or suppress future movements.
32
90260
2000
ืœื™ืฆื•ืจ ืื• ืœื”ืคืกื™ืง ืชื ื•ืขื•ืช ืขืชื™ื“ื™ื•ืช.
01:32
There can be no evolutionary advantage
33
92260
2000
ืื™ืŸ ืฉื•ื ื™ืชืจื•ืŸ ืื‘ื•ืœื•ืฆื™ื•ื ื™
01:34
to laying down memories of childhood
34
94260
2000
ื‘ื–ื™ื›ืจื•ื ื•ืช ื™ืœื“ื•ืช
01:36
or perceiving the color of a rose
35
96260
2000
ืื• ื‘ืงืœื™ื˜ืช ืฆื‘ืข ืฉืœ ืคืจื—
01:38
if it doesn't affect the way you're going to move later in life.
36
98260
3000
ืื ืื™ืŸ ืœื”ืŸ ื”ืฉืคืขื” ืขืœ ื”ื“ืจืš ื‘ื” ืื ื• ืขื•ืžื“ื™ื ืœื ื•ืข ื‘ืขืชื™ื“.
01:41
Now for those who don't believe this argument,
37
101260
2000
ืขื‘ื•ืจ ืืœื” ืฉืื™ื ื ืžืืžื™ื ื™ื ืœื˜ื™ืขื•ืŸ ื–ื”,
01:43
we have trees and grass on our planet without the brain,
38
103260
2000
ืขืฆื™ื ื•ื“ืฉื ืื™ืŸ ืœื”ื ืžื•ื—,
01:45
but the clinching evidence is this animal here --
39
105260
2000
ืื‘ืœ ื”ื”ื•ื›ื—ื” ื”ื—ื•ืชื›ืช ื”ื™ื ื”ื™ืฆื•ืจ ื”ื–ื” --
01:47
the humble sea squirt.
40
107260
2000
ืจื›ื™ื›ืช-ื™ื.
01:49
Rudimentary animal, has a nervous system,
41
109260
3000
ื™ืฆื•ืจ ืœื ืžืคื•ืชื— ืฉื™ืฉ ืœื” ืžืขืจื›ืช ืขืฆื‘ื™ืช
01:52
swims around in the ocean in its juvenile life.
42
112260
2000
ื•ื”ื™ื ืฉื•ื—ื” ื‘ืื•ืงื™ื™ื ื•ืก ื‘ืฆืขื™ืจื•ืชื”.
01:54
And at some point of its life,
43
114260
2000
ื‘ืฉืœื‘ ื›ืœืฉื”ื• ื‘ื—ื™ื™ื”,
01:56
it implants on a rock.
44
116260
2000
ื”ื™ื ื ื“ื‘ืงืช ืœืกืœืข. ื•ื”ื“ื‘ืจ ื”ืจืืฉื•ืŸ
01:58
And the first thing it does in implanting on that rock, which it never leaves,
45
118260
3000
ืฉื”ื™ื ืขื•ืฉื” ืขื ื”ื™ื“ื‘ืงื•ืชื” ืœืกืœืข ืฉื”ื™ื ืœืขื•ืœื ืœื ืขื•ื–ื‘ืช,
02:01
is to digest its own brain and nervous system
46
121260
3000
ื–ื” ืœืขื›ืœ ืืช ื”ืžื•ื— ื•ืืช ื”ืžืขืจื›ืช ื”ืขืฆื‘ื™ืช ืฉืœ ืขืฆืžื”
02:04
for food.
47
124260
2000
ื‘ืชื•ืจ ืžื–ื•ืŸ.
02:06
So once you don't need to move,
48
126260
2000
ื›ืš, ื‘ืจื’ืข ืฉืœื ื–ืงื•ืงื™ื ืœืชื ื•ืขื”,
02:08
you don't need the luxury of that brain.
49
128260
3000
ืื™ืŸ ืฆื•ืจืš ื‘ืœื•ืงืกื•ืก ืฉืœ ื”ืžื•ื—.
02:11
And this animal is often taken
50
131260
2000
ื•ื™ืฆื•ืจ ื–ื” ืžืฉืžืฉ ืœืขื™ืชื™ื
02:13
as an analogy to what happens at universities
51
133260
2000
ื›ืื ืœื•ื’ื™ื” ืœืžื” ืฉืžืชืจื—ืฉ ื‘ืื•ื ื™ื‘ืจืกื™ื˜ืื•ืช
02:15
when professors get tenure,
52
135260
2000
ื›ืืฉืจ ืคืจื•ืคืกื•ืจื™ื ืžืงื‘ืœื™ื ืงื‘ื™ืขื•ืช,
02:17
but that's a different subject.
53
137260
2000
ืื‘ืœ ื–ื” ื ื•ืฉื ืื—ืจ.
02:19
(Applause)
54
139260
2000
(ืžื—ื™ืื•ืช ื›ืคื™ื™ื)
02:21
So I am a movement chauvinist.
55
141260
3000
ืื ื›ืš, ืื ื™ ืงื ืื™ ืœืชื ื•ืขื”. ืื ื™ ืžืืžื™ืŸ ืฉืชื ื•ืขื”
02:24
I believe movement is the most important function of the brain --
56
144260
2000
ื”ื™ื ื”ืชื™ืคืงื•ื“ ื”ื—ืฉื•ื‘ ื‘ื™ื•ืชืจ ืฉืœ ื”ืžื•ื— --
02:26
don't let anyone tell you that it's not true.
57
146260
2000
ื•ืืœ ืชื ื™ื—ื• ืœืืฃ ืื—ื“ ืœื•ืžืจ ืœื›ื ืฉืื™ืŸ ื–ื” ื ื›ื•ืŸ.
02:28
Now if movement is so important,
58
148260
2000
ืื‘ืœ ืื ืชื ื•ืขื” ื”ื™ื ื›ื” ื—ืฉื•ื‘ื”,
02:30
how well are we doing
59
150260
2000
ืขื“ ื›ืžื” ืื ื• ืžื‘ื™ื ื™ื ื›ื”ืœื›ื”
02:32
understanding how the brain controls movement?
60
152260
2000
ื›ื™ืฆื“ ื”ืžื•ื— ืฉื•ืœื˜ ื‘ืชื ื•ืขื”?
02:34
And the answer is we're doing extremely poorly; it's a very hard problem.
61
154260
2000
ื”ืชืฉื•ื‘ื” ื”ื™ื ืฉื‘ืžื™ื“ื” ืžืื•ื“ ื“ืœื”; ื–ื• ื‘ืขื™ื” ืžืื•ื“ ืงืฉื”.
02:36
But we can look at how well we're doing
62
156260
2000
ืื‘ืœ ืืคืฉืจ ื’ื ืœื”ืชืจืฉื ืžื‘ื™ืฆื•ืขื™ื ื• ื”ื˜ื•ื‘ื™ื
02:38
by thinking about how well we're doing building machines
63
158260
2000
ืื ื—ื•ืฉื‘ื™ื ืขืœ ืื™ื›ื•ืช ื”ืžื›ื•ื ื•ืช ืฉืื ื• ื‘ื•ื ื™ื
02:40
which can do what humans can do.
64
160260
2000
ืืฉืจ ืžื‘ืฆืขื•ืช ืžื” ืฉื‘ื ื™-ืื“ื ืขื•ืฉื™ื.
02:42
Think about the game of chess.
65
162260
2000
ืชื—ืฉื‘ื• ืขืœ ืžืฉื—ืง ื”ืฉื—ืžื˜.
02:44
How well are we doing determining what piece to move where?
66
164260
3000
ืขื“ ื›ืžื” ื˜ื•ื‘ ื•ื ื›ื•ืŸ ืื ื• ืžื—ืœื™ื˜ื™ื ืœื”ื–ื™ื– ืืช ื”ื›ืœื™ื ื•ืœื”ื™ื›ืŸ?
02:47
If you pit Garry Kasparov here, when he's not in jail,
67
167260
3000
ืื ื ืฆื™ื‘ ืืช ื’ืืจื™ ืงืกืคืืจื•ื‘, ื›ืืฉืจ ืื™ื ื• ื‘ื›ืœื,
02:50
against IBM's Deep Blue,
68
170260
2000
ื›ื ื’ื“ ืžื—ืฉื‘ IBM "ื›ื—ื•ืœ ืขืžื•ืง",
02:52
well the answer is IBM's Deep Blue will occasionally win.
69
172260
3000
ื”ืชืฉื•ื‘ื” ืชื”ื™ื” ืฉื”ืžื—ืฉื‘ ื™ื ืฆื— ืœืขื™ืชื™ื.
02:55
And I think if IBM's Deep Blue played anyone in this room, it would win every time.
70
175260
3000
ื•ืื ื”ืžื—ืฉื‘ ื™ืฉื—ืง ื ื’ื“ ื›ืœ ืžื™ืฉื”ื• ื‘ืื•ืœื ื–ื”, ื”ืžื—ืฉื‘ ื™ื ืฆื— ืชืžื™ื“.
02:58
That problem is solved.
71
178260
2000
ื›ืืŸ ื”ืชืฉื•ื‘ื” ื‘ืจื•ืจื”.
03:00
What about the problem
72
180260
2000
ืžื” ืœื’ื‘ื™ ื”ื‘ืขื™ื” ืฉืœ
03:02
of picking up a chess piece,
73
182260
2000
ื”ืจืžืช ื›ืœื™-ืฉื—ืžื˜,
03:04
dexterously manipulating it and putting it back down on the board?
74
184260
3000
ืœืฉืœื•ื˜ ื‘ื• ื‘ืžื™ื•ืžื ื•ืช ื•ืœื”ื ื™ื—ื• ื‘ื—ื–ืจื” ืขืœ ื”ืœื•ื—?
03:07
If you put a five year-old child's dexterity against the best robots of today,
75
187260
3000
ืื ืฉืžื™ื ืืช ืžื™ื•ืžื ื•ืชื• ืฉืœ ื™ืœื“ ื‘ืŸ 5 ื›ื ื’ื“ ื”ืจื•ื‘ื•ื˜ื™ื ื”ื›ื™ ืžืฉื•ื›ืœืœื™ื ืฉืœ ื”ื™ื•ื,
03:10
the answer is simple:
76
190260
2000
ื”ืชืฉื•ื‘ื” ื‘ืจื•ืจื”:
03:12
the child wins easily.
77
192260
2000
ื”ื™ืœื“ ื™ื ืฆื— ื‘ืงืœื•ืช.
03:14
There's no competition at all.
78
194260
2000
ืื™ืŸ ื‘ื›ืœืœ ืขืœ ืžื” ืœื“ื‘ืจ.
03:16
Now why is that top problem so easy
79
196260
2000
ืื– ืžื“ื•ืข ื”ื‘ืขื™ื” ื”ืจืืฉื•ื ื” ื›ื” ืงืœื”
03:18
and the bottom problem so hard?
80
198260
2000
ื•ื”ืฉื ื™ื” ื›ื” ืงืฉื”?
03:20
One reason is a very smart five year-old
81
200260
2000
ืกื™ื‘ื” ืื—ืช ื”ื™ื ืฉื‘ืŸ 5 ืฉื”ื•ื ืžืื•ื“ ืคื™ืงื—
03:22
could tell you the algorithm for that top problem --
82
202260
2000
ื™ื›ื•ืœ ืœืžืฆื•ื ืืช ื”ืืœื’ื•ืจื™ืชื ืœื‘ืขื™ื” ื”ืจืืฉื•ื ื” --
03:24
look at all possible moves to the end of the game
83
204260
2000
ื™ืกืชื›ืœ ืขืœ ื›ืœ ื”ืžื”ืœื›ื™ื ื”ืืคืฉืจื™ื™ื ื‘ืžืฉื—ืง
03:26
and choose the one that makes you win.
84
206260
2000
ื•ื™ื‘ื—ืจ ืืช ื–ื” ื”ื™ื›ื•ืœ ืœื ืฆื—.
03:28
So it's a very simple algorithm.
85
208260
2000
ืœื›ืŸ ื–ื” ืืœื’ื•ืจื™ืชื ืžืื•ื“ ืคืฉื•ื˜.
03:30
Now of course there are other moves,
86
210260
2000
ื‘ืจื•ืจ ืฉื™ืฉ ืžื”ืœื›ื™ื ืื—ืจื™ื,
03:32
but with vast computers we approximate
87
212260
2000
ืื‘ืœ ืขื ืžื—ืฉื‘ื™ื ืื ื• ืขื•ืฉื™ื ืงื™ืจื•ื‘ื™ื
03:34
and come close to the optimal solution.
88
214260
2000
ื•ื›ืš ืžืชืงืจื‘ื™ื ืœืคื™ืชืจื•ืŸ ื”ืžื™ื˜ื‘ื™.
03:36
When it comes to being dexterous,
89
216260
2000
ืื‘ืœ ื›ืืฉืจ ื–ื” ืžื’ื™ืข ืœืžื™ื•ืžื ื•ืช ื‘ืชื ื•ืขื”,
03:38
it's not even clear what the algorithm is you have to solve to be dexterous.
90
218260
2000
ื›ืœืœ ืœื ื‘ืจื•ืจ ืื™ื–ื” ืืœื’ื•ืจื™ืชื ืขืœื™ื ื• ืœืคืชื•ืจ ื›ื“ื™ ืœื”ื™ื•ืช ืžื™ื•ืžื ื™ื.
03:40
And we'll see you have to both perceive and act on the world,
91
220260
2000
ืื ื• ื ืจืื” ืฉืฆืจื™ืš ื’ื ืœืชืคื•ืก ืืช ื”ืกื‘ื™ื‘ื” ื•ื’ื ืœืคืขื•ืœ ืขืœื™ื”,
03:42
which has a lot of problems.
92
222260
2000
ืฉื–ื” ื™ื•ืฆืจ ื”ืžื•ืŸ ื‘ืขื™ื•ืช.
03:44
But let me show you cutting-edge robotics.
93
224260
2000
ืืจืื” ืœื›ื ืขื›ืฉื™ื• ืจื•ื‘ื•ื˜ื™ืงื” ืžืชืงื“ืžืช.
03:46
Now a lot of robotics is very impressive,
94
226260
2000
ื‘ื”ืจื‘ื” ืžืงืจื™ื ืจื•ื‘ื•ื˜ื™ืงื” ื–ื” ื“ื‘ืจ ืžืจืฉื™ื ื‘ื™ื•ืชืจ,
03:48
but manipulation robotics is really just in the dark ages.
95
228260
3000
ืื‘ืœ ืจื•ื‘ื•ื˜ื™ืงื” ืชืคืขื•ืœื™ืช ื”ื™ื ืขื“ื™ื™ืŸ ืคืจื™ืžื™ื˜ื™ื‘ื™ืช.
03:51
So this is the end of a Ph.D. project
96
231260
2000
ื–ื”ื• ื”ืชื•ืฆืจ ื”ืกื•ืคื™ ืฉืœ ืขื‘ื•ื“ืช ื“ื•ืงื˜ื•ืจื˜
03:53
from one of the best robotics institutes.
97
233260
2000
ื‘ืื—ื“ ืžืžื›ื•ื ื™ ื”ืจื•ื‘ื•ื˜ื™ืงื” ื”ืžืชืงื“ืžื™ื ื‘ื™ื•ืชืจ.
03:55
And the student has trained this robot
98
235260
2000
ื”ืกื˜ื•ื“ื ื˜ ืœื™ืžื“ ืืช ื”ืจื•ื‘ื•ื˜
03:57
to pour this water into a glass.
99
237260
2000
ืœืฉืคื•ืš ืžื™ื ืœืชื•ืš ื›ื•ืก.
03:59
It's a hard problem because the water sloshes about, but it can do it.
100
239260
3000
ื–ื• ื‘ืขื™ื” ืงืฉื” ื›ื™ ื”ืžื™ื ืžืฉื›ืฉื›ื™ื, ืื‘ืœ ื”ื•ื ืžืฆืœื™ื—.
04:02
But it doesn't do it with anything like the agility of a human.
101
242260
3000
ืื‘ืœ ื”ื•ื ืื™ื ื• ืขื•ืฉื” ื–ืืช ื‘ืื•ืชื” ืžื™ื•ืžื ื•ืช ืฉืœ ืื“ื.
04:05
Now if you want this robot to do a different task,
102
245260
3000
ื›ืขืช ืื ืจื•ืฆื™ื ืฉืจื•ื‘ื•ื˜ ื–ื” ื™ื‘ืฆืข ืžืฉื™ืžื” ืื—ืจืช,
04:08
that's another three-year Ph.D. program.
103
248260
3000
ื–ื• ืขื‘ื•ื“ืช ื“ื•ืงื˜ื•ืจื˜ ืื—ืจืช ืฉืœ 3 ืฉื ื™ื.
04:11
There is no generalization at all
104
251260
2000
ืื™ืŸ ืžืฉื”ื• ืžืฉื•ืชืฃ
04:13
from one task to another in robotics.
105
253260
2000
ื‘ื™ืŸ ืฉืชื™ ืžืฉื™ืžื•ืช ื‘ืจื•ื‘ื•ื˜ื™ืงื”.
04:15
Now we can compare this
106
255260
2000
ื›ืขืช ื ืฉื•ื•ื” ื–ืืช
04:17
to cutting-edge human performance.
107
257260
2000
ืœื‘ื™ืฆื•ืขื™ ืื ื•ืฉ ื‘ื“ืจื’ื” ื”ื›ื™ ื’ื‘ื•ื”ื”.
04:19
So what I'm going to show you is Emily Fox
108
259260
2000
ืืชื ืชืจืื• ืืช ืืžื™ืœื™ ืคื•ืงืก
04:21
winning the world record for cup stacking.
109
261260
3000
ืฉืงื‘ืขื” ืฉื™ื ืขื•ืœื ื‘ืกื™ื“ื•ืจ ื›ื•ืกื•ืช.
04:24
Now the Americans in the audience will know all about cup stacking.
110
264260
2000
ื”ืืžืจื™ืงืื™ื ื‘ื™ืŸ ื”ืฆื•ืคื™ื ืคื” ื™ื•ื“ืขื™ื ื”ื›ืœ ืขืœ ืกื™ื“ื•ืจ ื›ื•ืกื•ืช.
04:26
It's a high school sport
111
266260
2000
ื–ื”ื• ืกืคื•ืจื˜ ืฉืœ ืชื™ื›ื•ื ื™ื
04:28
where you have 12 cups you have to stack and unstack
112
268260
2000
ืฉื‘ื• ืฆืจื™ืš ืœืกื“ืจ ื•ืœื”ืคืจื™ื“ 12 ื›ื•ืกื•ืช
04:30
against the clock in a prescribed order.
113
270260
2000
ื›ื ื’ื“ ื–ืžืŸ ื‘ืกื“ืจ ืžืกื•ื™ื™ื.
04:32
And this is her getting the world record in real time.
114
272260
3000
ื›ืืŸ ื–ื• ื”ื™ื ื‘ื–ืžืŸ ืงื‘ื™ืขืช ื”ืฉื™ื ื”ืขื•ืœืžื™.
04:39
(Laughter)
115
279260
8000
(ืฆื—ื•ืง)
04:47
(Applause)
116
287260
5000
(ืžื—ื™ืื•ืช ื›ืคื™ื™ื)
04:52
And she's pretty happy.
117
292260
2000
ื”ื™ื ืฉืžื—ื” ืœืžื“ื™.
04:54
We have no idea what is going on inside her brain when she does that,
118
294260
2000
ืื™ืŸ ืœื ื• ืžื•ืฉื’ ืขืœ ื”ืžืชื—ื•ืœืœ ื‘ืžื•ื—ื” ื›ืืฉืจ ื”ื™ื ืขืฉืชื” ื–ืืช,
04:56
and that's what we'd like to know.
119
296260
2000
ื•ื–ื” ืžื” ืฉื”ื™ื™ื ื• ืจื•ืฆื™ื ืœื“ืขืช.
04:58
So in my group, what we try to do
120
298260
2000
ืœื›ืŸ ื‘ืฆื•ื•ืช ืฉืœื™, ืžื” ืฉืื ื• ืžื ืกื™ื ืœืขืฉื•ืช
05:00
is reverse engineer how humans control movement.
121
300260
3000
ื–ื” ืœื ืชื— ื”ื ื“ืกื™ืช ื›ื™ืฆื“ ืื ืฉื™ื ืฉื•ืœื˜ื™ื ื‘ืชื ื•ืขื”.
05:03
And it sounds like an easy problem.
122
303260
2000
ื–ื” ื ืฉืžืข ื›ืžื• ื‘ืขื™ื” ืคืฉื•ื˜ื”.
05:05
You send a command down, it causes muscles to contract.
123
305260
2000
ืคืงื•ื“ื” ื ืฉืœื—ืช, ื”ื™ื ื’ื•ืจืžืช ืœืฉืจื™ืจื™ื ืœื”ืชื›ื•ื•ืฅ.
05:07
Your arm or body moves,
124
307260
2000
ื”ื’ื•ืฃ ืื• ื”ื–ืจื•ืข ื ืขื™ื,
05:09
and you get sensory feedback from vision, from skin, from muscles and so on.
125
309260
3000
ืžืงื‘ืœื™ื ืžืฉื•ื‘ ื—ื•ืฉื™ ืžื”ืขื™ื ื™ื™ื, ืขื•ืจ, ืฉืจื™ืจื™ื ื•ื›ื•'.
05:12
The trouble is
126
312260
2000
ื”ื‘ืขื™ื” ื”ื™ื ืฉืื•ืชื•ืช ื”ืœืœื•
05:14
these signals are not the beautiful signals you want them to be.
127
314260
2000
ืื™ื ื ืื•ืชื•ืช ืื™ื“ืืœื™ื™ื ื›ืคื™ ืฉื”ื™ื™ื ื• ืจื•ืฆื™ื.
05:16
So one thing that makes controlling movement difficult
128
316260
2000
ืื—ื“ ื”ื“ื‘ืจื™ื ืฉื”ื•ืคื›ื™ื ืืช ื”ืฉืœื™ื˜ื”
05:18
is, for example, sensory feedback is extremely noisy.
129
318260
3000
ืขืœ ืชื ื•ืขื” ืœืงืฉื” ื”ื•ื ื”ืจืขืฉ ื”ื ืœื•ื•ื” ืœืžืฉื•ื‘.
05:21
Now by noise, I do not mean sound.
130
321260
3000
ื‘ืจืขืฉ, ืื™ื ื™ ืžืชื›ื•ื•ืŸ ืœืฆืœื™ืœื™ื.
05:24
We use it in the engineering and neuroscience sense
131
324260
2000
ื‘ื”ื ื“ืกื” ื•ืžื“ืขื™ ื”ืขืฆื‘ ื”ืžืฉืžืขื•ืช ื”ื™ื
05:26
meaning a random noise corrupting a signal.
132
326260
2000
ืจืขืฉ ืืงืจืื™ ืืฉืจ ื”ื•ืจืก ืืช ื”ืื•ืช.
05:28
So the old days before digital radio when you were tuning in your radio
133
328260
3000
ื–ื” ื›ืžื• ืคืขื, ืœืคื ื™ ื”ืจื“ื™ื• ื”ื“ื™ื’ื™ื˜ืœื™ ื›ืืฉืจ ื”ื™ื™ื ื• ืžื›ื•ื•ื ื™ื ืœืชื—ื ื”
05:31
and you heard "crrcckkk" on the station you wanted to hear,
134
331260
2000
ื•ืฉืžืขื ื• ืจืขืฉื™ื ืฆื•ืจืžื™ื ื‘ืชื—ื ื” ืฉืจืฆื™ื ื• ืœื”ืื–ื™ืŸ ืœื”,
05:33
that was the noise.
135
333260
2000
ืขืœ ืจืขืฉ ื›ื–ื” ืžื“ื•ื‘ืจ.
05:35
But more generally, this noise is something that corrupts the signal.
136
335260
3000
ื‘ืื•ืคืŸ ื™ื•ืชืจ ื›ืœืœื™, ืจืขืฉ ื›ื–ื” ื”ื•ืจืก ืืช ื”ืื•ืช.
05:38
So for example, if you put your hand under a table
137
338260
2000
ืœื“ื•ื’ืžื, ืื ืฉืžื™ื ืืช ื”ื™ื“ ืžืชื—ืช ืœืฉื•ืœื—ืŸ
05:40
and try to localize it with your other hand,
138
340260
2000
ื•ืžื ืกื™ื ืœืืชืจ ืื•ืชื” ื‘ืขื–ืจืช ื”ื™ื“ ื”ืฉื ื™ื”,
05:42
you can be off by several centimeters
139
342260
2000
ืืคืฉืจ ืœืคืกืคืก ื‘ื›ืžื” ืกื ื˜ื™ืžื˜ืจื™ื
05:44
due to the noise in sensory feedback.
140
344260
2000
ื‘ื’ืœืœ ื”ืจืขืฉ ื‘ืžืฉื•ื‘ ืžืŸ ื”ื—ื•ืฉื™ื.
05:46
Similarly, when you put motor output on movement output,
141
346260
2000
ื‘ืื•ืคืŸ ื“ื•ืžื”, ื›ืืฉืจ ืžืขืจื‘ื‘ื™ื ืืช ืื•ืชื•ืช ื”ืžื ื•ืข ืขื ืื•ืชื•ืช ืฉืœ ืชื ื•ืขื”,
05:48
it's extremely noisy.
142
348260
2000
ื–ื” ืžืื•ื“ ืจื•ืขืฉ.
05:50
Forget about trying to hit the bull's eye in darts,
143
350260
2000
ืฉืœื ืœื“ื‘ืจ ืขืœ ืงืœื™ืขื” ืœืžื˜ืจื” ื‘ืืžืฆืขื•ืช ื–ืจื™ืงืช ื—ื™ืฆื™ื,
05:52
just aim for the same spot over and over again.
144
352260
2000
ืคืฉื•ื˜ ืœื›ื•ื•ืŸ ืœืื•ืชื” ื ืงื•ื“ื” ืฉื•ื‘ ื•ืฉื•ื‘.
05:54
You have a huge spread due to movement variability.
145
354260
3000
ื™ืฉ ืœื ื• ืคื™ื–ื•ืจ ืจื—ื‘ ืžืื•ื“ ื‘ื’ืœืœ ื”ืฉืชื ื•ืช ื”ืชื ื•ืขื”.
05:57
And more than that, the outside world, or task,
146
357260
2000
ื•ืืคื™ืœื• ืขื•ื“ ื™ื•ืชืจ, ื”ืขื•ืœื ืกื‘ื™ื‘ื ื•, ืื• ื”ืžืฉื™ืžื”,
05:59
is both ambiguous and variable.
147
359260
2000
ื”ื ืžืขื•ืจืคืœื™ื ื•ืžืฉืชื ื™ื.
06:01
The teapot could be full, it could be empty.
148
361260
2000
ืงื•ืžืงื•ื ื”ืชื” ื™ื›ื•ืœ ืœื”ื™ื•ืช ืžืœื ืื• ืจื™ืง.
06:03
It changes over time.
149
363260
2000
ื”ื•ื ื ืชื•ืŸ ืœืฉื™ื ื•ื™ื™ื ืœืื•ืจืš ื–ืžืŸ.
06:05
So we work in a whole sensory movement task soup of noise.
150
365260
4000
ื›ืš ืฉืื ื• ืคื•ืขืœื™ื ื‘ืžืจืง ืฉืœ ืจืขืฉื™ื ื”ื ื•ื‘ืขื™ื ืžืขื‘ื•ื“ืช ื”ื—ื•ืฉื™ื.
06:09
Now this noise is so great
151
369260
2000
ืจืขืฉ ื–ื” ื”ื•ื ื›ื” ืžื”ื•ืชื™
06:11
that society places a huge premium
152
371260
2000
ืฉื”ื—ื‘ืจื” ืžืขื ื™ืงื” ืชื’ืžื•ืœ ืขื ืง
06:13
on those of us who can reduce the consequences of noise.
153
373260
3000
ืœืžื™ ืžืื™ืชื ื• ื”ืžืกื•ื’ืœ ืœืฆืžืฆื ืืช ื”ืฉืคืขื•ืช ื”ืจืขืฉ.
06:16
So if you're lucky enough to be able to knock a small white ball
154
376260
3000
ื›ืš ืฉืื ื”ืชืžื–ืœ ืœืžื™ืฉื”ื• ื”ืžื–ืœ ื•ื”ื•ื ืžืกื•ื’ืœ ืœื”ื›ื•ืช ื‘ื›ื“ื•ืจ ืœื‘ืŸ ืงื˜ืŸ
06:19
into a hole several hundred yards away using a long metal stick,
155
379260
3000
ืœืชื•ืš ื’ื•ืžื—ื” ื‘ืžืจื—ืง ื›ืžื” ืžืื•ืช ืžื˜ืจื™ื ื‘ืขื–ืจืช ืžืงืœ ืžืชื›ืช ืืจื•ืš,
06:22
our society will be willing to reward you
156
382260
2000
ื”ื—ื‘ืจื” ืฉืœื ื• ืชืจืฆื” ืœื’ืžื•ืœ ืœื•
06:24
with hundreds of millions of dollars.
157
384260
3000
ื‘ืžืื•ืช ืžื™ืœื™ื•ื ื™ ื“ื•ืœืจื™ื.
06:27
Now what I want to convince you of
158
387260
2000
ื›ืขืช ืžื” ืฉื‘ืจืฆื•ื ื™ ืœืฉื›ื ืข ืืชื›ื ื‘ื•
06:29
is the brain also goes through a lot of effort
159
389260
2000
ื”ื•ื ืฉื”ืžื•ื— ืขื•ืฉื” ื”ืจื‘ื” ืžืืžืฆื™ื
06:31
to reduce the negative consequences
160
391260
2000
ื›ื“ื™ ืœืฆืžืฆื ืืช ื”ื”ืฉืœื›ื•ืช ื”ืฉืœื™ืœื™ื•ืช
06:33
of this sort of noise and variability.
161
393260
2000
ืฉืœ ืจืขืฉื™ื ื•ืฉื™ื ื•ื™ื™ื ื›ืืœื”.
06:35
And to do that, I'm going to tell you about a framework
162
395260
2000
ื›ื“ื™ ืœืขืฉื•ืช ื–ืืช, ืืกืคืจ ืœื›ื ืขืœ ืžืขืจื›ืช
06:37
which is very popular in statistics and machine learning of the last 50 years
163
397260
3000
ืฉื”ื™ื ืžืื•ื“ ื ืคื•ืฆื” ื‘ืกื˜ื˜ื™ืกื˜ื™ืงื” ื•ื‘ืœื™ืžื•ื“ื™ ืžื›ื•ื ื” ื‘-50 ื”ืฉื ื™ื ื”ืื—ืจื•ื ื•ืช
06:40
called Bayesian decision theory.
164
400260
2000
ื•ื”ื ืงืจืืช ืชื™ืื•ืจื™ื” ื‘ื™ื™ืกื™ืื ื™ืช ืœื”ื—ืœื˜ื•ืช.
06:42
And it's more recently a unifying way
165
402260
3000
ื–ื• ื“ืจืš ื›ื•ืœืœื ื™ืช ืœื—ืฉื™ื‘ื”
06:45
to think about how the brain deals with uncertainty.
166
405260
3000
ืขืœ ื›ื™ืฆื“ ื”ืžื•ื— ืžืชืžื•ื“ื“ ืขื ื—ื•ืกืจ ื•ื•ื“ืื•ืช.
06:48
And the fundamental idea is you want to make inferences and then take actions.
167
408260
3000
ื”ืจืขื™ื•ืŸ ื”ืžืจื›ื–ื™ ื”ื•ื ืฉืื“ื ืจื•ืฆื” ืœื”ืกื™ืง ืžืกืงื ื•ืช ื•ืœื ืงื•ื˜ ื‘ืคืขื•ืœื•ืช.
06:51
So let's think about the inference.
168
411260
2000
ืื– ื‘ื•ืื• ื ื—ืฉื•ื‘ ืขืœ ื”ืกืงืช ืžืกืงื ื•ืช.
06:53
You want to generate beliefs about the world.
169
413260
2000
ืื ื• ืจื•ืฆื™ื ืœื™ื™ืฆืจ ืืžื•ื ื•ืช ืœื’ื‘ื™ ื”ืขื•ืœื.
06:55
So what are beliefs?
170
415260
2000
ืื– ืžื”ืŸ ืืžื•ื ื•ืช?
06:57
Beliefs could be: where are my arms in space?
171
417260
2000
ืืžื•ื ื•ืช ื™ื›ื•ืœื•ืช ืœื”ื™ื•ืช: ื”ื™ื›ืŸ ื ืžืฆืื•ืช ื–ืจื•ืขื•ืชื™ื™ ื‘ื—ืœืœ?
06:59
Am I looking at a cat or a fox?
172
419260
2000
ื”ืื ืื ื™ ืžื‘ื™ื˜ ื‘ื—ืชื•ืœ ืื• ื‘ืฉื•ืขืœ?
07:01
But we're going to represent beliefs with probabilities.
173
421260
3000
ืื‘ืœ ืื ื• ื ืฆื™ื’ ืืžื•ื ื•ืช ื‘ืฆื•ืจืช ื”ืกืชื‘ืจื•ื™ื•ืช.
07:04
So we're going to represent a belief
174
424260
2000
ื ืฆื™ื’ ืืžื•ื ื”
07:06
with a number between zero and one --
175
426260
2000
ื‘ืืžืฆืขื•ืช ืžืกืคืจ ื‘ื™ืŸ 0 ืœ-1 --
07:08
zero meaning I don't believe it at all, one means I'm absolutely certain.
176
428260
3000
0 ืžืฉืžืขื•ืชื• ืฉืื™ื ื™ ืžืืžื™ืŸ ื›ืœืœ ื•-1 ืžืฉืžืขื•ืชื• ืฉืื ื™ ื‘ื˜ื•ื— ืœื’ืžืจื™.
07:11
And numbers in between give you the gray levels of uncertainty.
177
431260
3000
ื•ืžืกืคืจื™ื ื‘ื™ื ื™ื”ื ื ื•ืชื ื™ื ื“ืจื’ื•ืช ืฉื•ื ื•ืช ืฉืœ ื—ื•ืกืจ ื•ื•ื“ืื•ืช.
07:14
And the key idea to Bayesian inference
178
434260
2000
ื”ืจืขื™ื•ืŸ ื”ืžืจื›ื–ื™ ืฉืœ ื”ืกืงื” ื‘ื™ื™ืกื™ืื ื™ืช
07:16
is you have two sources of information
179
436260
2000
ื”ื•ื ืฉื™ืฉ ืฉื ื™ ืžืงื•ืจื•ืช ืžื™ื“ืข
07:18
from which to make your inference.
180
438260
2000
ืฉืžื”ื ืžืกื™ืงื™ื ืืช ื”ืžืกืงื ื”.
07:20
You have data,
181
440260
2000
ื™ืฉ ื ืชื•ื ื™ื,
07:22
and data in neuroscience is sensory input.
182
442260
2000
ื•ื ืชื•ื ื™ื ื‘ืžื“ืขื™-ืขืฆื‘ ื–ื” ืงืœื˜ ืฉืœ ื—ื•ืฉื™ื.
07:24
So I have sensory input, which I can take in to make beliefs.
183
444260
3000
ื™ืฉ ืœื ื• ืงืœื˜ ื—ื•ืฉื™ื ืฉืื ื™ ื™ื›ื•ืœ ืœื ืฆืœ ื›ื“ื™ ืœื™ืฆื•ืจ ืืžื•ื ื•ืช.
07:27
But there's another source of information, and that's effectively prior knowledge.
184
447260
3000
ืื‘ืœ ื™ืฉ ืžืงื•ืจ ื ื•ืกืฃ ืฉืœ ืžื™ื“ืข ื•ื”ื•ื ื”ื™ื“ืข ื”ืžื•ืงื“ื ืฉื™ืฉ ืœื ื•.
07:30
You accumulate knowledge throughout your life in memories.
185
450260
3000
ืื ื• ืฆื•ื‘ืจื™ื ื™ื“ืข ืœืื•ืจืš ื—ื™ื™ื ื• ื‘ืฆื•ืจืช ื–ื™ื›ืจื•ื ื•ืช.
07:33
And the point about Bayesian decision theory
186
453260
2000
ื•ื”ืขื™ืงืจ ื‘ืชื™ืื•ืจื™ื” ื‘ื™ื™ืกื™ืื ื™ืช ืœื”ื—ืœื˜ื•ืช ื”ื•ื
07:35
is it gives you the mathematics
187
455260
2000
ืฉื”ื™ื ื ื•ืชื ืช ืœื ื• ืืช ื”ืžืชืžื˜ื™ืงื”
07:37
of the optimal way to combine
188
457260
2000
ืฉืœ ื”ื“ืจืš ื”ืžื™ื˜ื‘ื™ืช ืœืฉื™ืœื•ื‘ ื‘ื™ืŸ
07:39
your prior knowledge with your sensory evidence
189
459260
2000
ื”ื™ื“ืข ื”ืžื•ืงื“ื ื•ื”ื ืชื•ื ื™ื ื”ื—ื•ืฉื™ื™ื
07:41
to generate new beliefs.
190
461260
2000
ื›ื“ื™ ืœื™ื™ืฆืจ ืืžื•ื ื•ืช ื—ื“ืฉื•ืช.
07:43
And I've put the formula up there.
191
463260
2000
ืจืฉืžืชื™ ืืช ื”ื ื•ืกื—ื” ื›ืืŸ.
07:45
I'm not going to explain what that formula is, but it's very beautiful.
192
465260
2000
ืœื ืืกื‘ื™ืจ ืืช ื”ื ื•ืกื—ื”, ืื‘ืœ ื”ื™ื ืžืื•ื“ ื™ืคื”.
07:47
And it has real beauty and real explanatory power.
193
467260
3000
ื™ืฉ ื‘ื” ื™ื•ืคื™ ืืžื™ืชื™ ื•ื™ื›ื•ืœืช ืืžื™ืชื™ืช ืœื”ื‘ื”ื™ืจ ื“ื‘ืจื™ื.
07:50
And what it really says, and what you want to estimate,
194
470260
2000
ื”ื™ื ืื•ืžืจืช ืฉืžื” ืฉืื ื• ืจื•ืฆื™ื ืœืืžื•ื“,
07:52
is the probability of different beliefs
195
472260
2000
ื”ื™ื ื”ื”ืกืชื‘ืจื•ืช ืฉืœ ื”ืืžื•ื ื•ืช ื”ืฉื•ื ื•ืช
07:54
given your sensory input.
196
474260
2000
ื‘ืงืœื˜ ื—ื•ืฉื™ื ื ืชื•ืŸ.
07:56
So let me give you an intuitive example.
197
476260
2000
ืืชืŸ ืœื›ื ื“ื•ื’ืžื ืื™ื ื˜ื•ืื™ื˜ื™ื‘ื™ืช.
07:58
Imagine you're learning to play tennis
198
478260
3000
ืชื—ืฉื‘ื• ืฉืืชื ืœื•ืžื“ื™ื ืœืฉื—ืง ื˜ื ื™ืก
08:01
and you want to decide where the ball is going to bounce
199
481260
2000
ื•ืขืœื™ื›ื ืœื”ื—ืœื™ื˜ ื”ื™ื›ืŸ ื”ื›ื“ื•ืจ ืขื•ืžื“ ืœื™ืคื•ืœ
08:03
as it comes over the net towards you.
200
483260
2000
ื›ืืฉืจ ื”ื•ื ืขื•ื‘ืจ ืืช ื”ืจืฉืช ื‘ื“ืจื›ื• ืืœื™ื›ื.
08:05
There are two sources of information
201
485260
2000
ื™ืฉ ืฉื ื™ ืžืงื•ืจื•ืช ืžื™ื“ืข,
08:07
Bayes' rule tells you.
202
487260
2000
ื›ืš ืื•ืžืจ ื”ื›ืœืœ ื”ื‘ื™ื™ืกื™ืื ื™.
08:09
There's sensory evidence -- you can use visual information auditory information,
203
489260
3000
ื™ืฉ ื ืชื•ื ื™ื ืžื”ื—ื•ืฉื™ื -- ื ื™ืชืŸ ืœื”ืฉืชืžืฉ ื‘ืžื™ื“ืข ื—ื–ื•ืชื™ ืื• ืฉืžื™ืขืชื™,
08:12
and that might tell you it's going to land in that red spot.
204
492260
3000
ื•ื–ื” ื™ื›ื•ืœ ืœื•ืžืจ ืœืš ืœื”ืชืจื›ื– ื‘ื ืงื•ื“ื” ื”ืื“ื•ืžื” ื”ื”ื™ื.
08:15
But you know that your senses are not perfect,
205
495260
3000
ืื‘ืœ ืืชื” ื™ื•ื“ืข ืฉื—ื•ืฉื™ืš ืื™ื ื ืžื•ืฉืœืžื™ื, ื•ืœื›ืŸ ืงื™ื™ืžืช ืื™-ื•ื•ื“ืื•ืช
08:18
and therefore there's some variability of where it's going to land
206
498260
2000
ืžืกื•ื™ื™ืžืช ืœื’ื‘ื™ ื”ื™ื›ืŸ ื”ื›ื“ื•ืจ ื™ืคื•ืœ
08:20
shown by that cloud of red,
207
500260
2000
ื•ื”ื™ื ืžืชื•ืืจืช ื‘ืืžืฆืขื•ืช ื”ื›ืชื ื”ืื“ื•ื,
08:22
representing numbers between 0.5 and maybe 0.1.
208
502260
3000
ื”ืžื™ื™ืฆื’ ืžืกืคืจื™ื ื‘ื™ืŸ 0.5 ืœ-0.1 ื‘ืขืจืš.
08:26
That information is available in the current shot,
209
506260
2000
ืžื™ื“ืข ื–ื” ื–ืžื™ืŸ ื‘ื—ื‘ื˜ื” ื”ื ื•ื›ื—ื™ืช,
08:28
but there's another source of information
210
508260
2000
ืื‘ืœ ื™ืฉื ื• ืžืงื•ืจ ืžื™ื“ืข ื ื•ืกืฃ
08:30
not available on the current shot,
211
510260
2000
ืฉืื™ื ื• ื–ืžื™ืŸ ื‘ื—ื‘ื˜ื” ื”ื ื•ื›ื—ื™ืช,
08:32
but only available by repeated experience in the game of tennis,
212
512260
3000
ืืœื ื”ื•ื ื–ืžื™ืŸ ืจืง ื‘ืืžืฆืขื•ืช ื ื™ืกื™ื•ื ื•ืช ื—ื•ื–ืจื™ื ื‘ืžืฉื—ืง ื”ื˜ื ื™ืก,
08:35
and that's that the ball doesn't bounce
213
515260
2000
ื•ื”ื•ื ืฉื”ื›ื“ื•ืจ ืื™ื ื• ื ื•ืคืœ ื‘ื”ืกืชื‘ืจื•ืช ืฉื•ื•ื”
08:37
with equal probability over the court during the match.
214
517260
2000
ื‘ื›ืœ ื”ืžื’ืจืฉ ื‘ื–ืžืŸ ื”ืžืฉื—ืง.
08:39
If you're playing against a very good opponent,
215
519260
2000
ืื ืืชื” ืžืฉื—ืง ื ื’ื“ ืฉื—ืงืŸ ื‘ืจืžื” ื’ื‘ื•ื”ื”,
08:41
they may distribute it in that green area,
216
521260
2000
ื”ื ื™ื›ื•ืœื™ื ืœื™ืคื•ืœ ื‘ืื–ื•ืจ ื”ื™ืจื•ืง ื”ื”ื•ื,
08:43
which is the prior distribution,
217
523260
2000
ืฉื–ื” ืคื™ื–ื•ืจ ืชืื•ืจื˜ื™ ื”ืงื™ื™ื ืžืจืืฉ,
08:45
making it hard for you to return.
218
525260
2000
ื”ืžืงืฉื” ืขืœื™ืš ืœื”ื—ื–ื™ืจ.
08:47
Now both these sources of information carry important information.
219
527260
2000
ืฉื ื™ ืžืงื•ืจื•ืช ื”ืžื™ื“ืข ื ื•ืฉืื™ื ืขื™ืžื ืžื™ื“ืข ื—ืฉื•ื‘.
08:49
And what Bayes' rule says
220
529260
2000
ืžื” ืฉื—ื•ืง ื‘ื™ื™ืกื™ืื ื™ ืื•ืžืจ ื”ื•ื ืฉืขืœื™ื™
08:51
is that I should multiply the numbers on the red by the numbers on the green
221
531260
3000
ืœื”ื›ืคื™ืœ ืืช ื”ืžืกืคืจื™ื ืฉื‘ืื“ื•ื ื‘ืžืกืคืจื™ื ืฉื‘ื™ืจื•ืง
08:54
to get the numbers of the yellow, which have the ellipses,
222
534260
3000
ื›ื“ื™ ืœืงื‘ืœ ืืช ื”ืžืกืคืจื™ื ื‘ืฆื”ื•ื‘, ืฉื™ืฉ ืฉื ืืช ื”ืืœื™ืคืกื•ืช,
08:57
and that's my belief.
223
537260
2000
ื•ื–ื• ื”ืืžื•ื ื” ืฉืœื™.
08:59
So it's the optimal way of combining information.
224
539260
3000
ื–ื•ื”ื™ ื”ื“ืจืš ื”ืžื™ื˜ื‘ื™ืช ืœืฉื™ืœื•ื‘ ื‘ื™ืŸ ืžืงื•ืจื•ืช ืžื™ื“ืข.
09:02
Now I wouldn't tell you all this if it wasn't that a few years ago,
225
542260
2000
ืœื ื”ื™ื™ืชื™ ืžืกืคืจ ืืช ื›ืœ ื–ื” ืื ืœืคื ื™ ื›ืžื” ืฉื ื™ื
09:04
we showed this is exactly what people do
226
544260
2000
ืœื ื”ื™ื™ื ื• ืžื•ื›ื™ื—ื™ื ืฉื–ื” ื‘ื“ื™ื•ืง ืžื” ืฉืื ืฉื™ื ืขื•ืฉื™ื
09:06
when they learn new movement skills.
227
546260
2000
ื›ืืฉืจ ื”ื ืœื•ืžื“ื™ื ืžื™ื•ืžื ื•ื™ื•ืช ืชื ื•ืขื” ื—ื“ืฉื•ืช.
09:08
And what it means
228
548260
2000
ื•ืžื” ืฉื–ื” ืื•ืžืจ ื”ื•ื ืฉืื ื—ื ื• ื‘ืืžืช
09:10
is we really are Bayesian inference machines.
229
550260
2000
ืžื›ื•ื ื•ืช ื‘ื™ื™ืกื™ืื ื™ื•ืช ืœื”ืกืงืช ืžืกืงื ื•ืช.
09:12
As we go around, we learn about statistics of the world and lay that down,
230
552260
4000
ื‘ืขื•ื“ื ื• ืžืชื ื”ืœื™ื ื‘ืขื•ืœื, ืื ื• ืœื•ืžื“ื™ื ืขืœื™ื• ืกื˜ื˜ื™ืกื˜ื™ืงื•ืช ื•ื–ื•ื›ืจื™ื ืื•ืชืŸ,
09:16
but we also learn
231
556260
2000
ืื‘ืœ ืื ื• ื’ื ืœื•ืžื“ื™ื ืขื“ ื›ืžื”
09:18
about how noisy our own sensory apparatus is,
232
558260
2000
ืกื•ืื ืช ื”ื™ื ื”ืžืขืจื›ืช ื”ื—ื•ืฉื™ืช ืฉืœื ื•,
09:20
and then combine those
233
560260
2000
ื•ืื– ืžืฉืœื‘ื™ื ื‘ื™ื ื™ื”ืŸ
09:22
in a real Bayesian way.
234
562260
2000
ืžืžืฉ ื‘ื“ืจืš ื‘ื™ื™ืกื™ืื ื™ืช.
09:24
Now a key part to the Bayesian is this part of the formula.
235
564260
3000
ื”ืžืคืชื— ืขื‘ื•ืจ ื”ื‘ื™ื™ืกื™ืื ื™ืช ื–ื”ื• ื”ื—ืœืง ื”ื–ื” ื‘ื ื•ืกื—ื”.
09:27
And what this part really says
236
567260
2000
ื•ืžื” ืฉื—ืœืง ื–ื” ืื•ืžืจ ื”ื•ื
09:29
is I have to predict the probability
237
569260
2000
ืฉืขืœื™ื™ ืœื—ื–ื•ืช ืืช ื”ื”ืกืชื‘ืจื•ืช
09:31
of different sensory feedbacks
238
571260
2000
ืฉืœ ืžืฉื•ื‘ื™ื ื—ื•ืฉื™ื™ื ืฉื•ื ื™ื
09:33
given my beliefs.
239
573260
2000
ื‘ื”ืชืื ืœืืžื•ื ื•ืช ืฉืœื™.
09:35
So that really means I have to make predictions of the future.
240
575260
3000
ื–ื” ืื•ืžืจ ืฉืขืœื™ื™ ืžืžืฉ ืœื—ื–ื•ืช ืืช ื”ืขืชื™ื“.
09:38
And I want to convince you the brain does make predictions
241
578260
2000
ื•ื‘ืจืฆื•ื ื™ ืœืฉื›ื ืข ืืชื›ื ืฉื”ืžื•ื— ืฉืœื ื• ืื›ืŸ ืขื•ืฉื” ืชื—ื–ื™ื•ืช
09:40
of the sensory feedback it's going to get.
242
580260
2000
ืขืœ ื”ืžืฉื•ื‘ ื”ื—ื•ืฉื™ ืฉื”ื•ื ืขื•ืžื“ ืœืงื‘ืœ.
09:42
And moreover, it profoundly changes your perceptions
243
582260
2000
ื‘ื ื•ืกืฃ, ื”ื•ื ืžืฉื ื” ืžื”ื•ืชื™ืช ืืช ื”ืชืคื™ืกื•ืช ืฉืœื ื•
09:44
by what you do.
244
584260
2000
ืœืคื™ ืžื” ืฉืื ื• ืขื•ืฉื™ื ื‘ืคื•ืขืœ.
09:46
And to do that, I'll tell you
245
586260
2000
ื›ื“ื™ ืœืขืฉื•ืช ื–ืืช, ืืกืคืจ ืœื›ื
09:48
about how the brain deals with sensory input.
246
588260
2000
ื›ื™ืฆื“ ื”ืžื•ื— ืžืชืžื•ื“ื“ ืขื ืงืœื˜ ื—ื•ืฉื™.
09:50
So you send a command out,
247
590260
3000
ืื ื• ืฉื•ืœื—ื™ื ืคืงื•ื“ื” ื”ื—ื•ืฆื”,
09:53
you get sensory feedback back,
248
593260
2000
ืžืงื‘ืœื™ื ื‘ื—ื–ืจื” ืงืœื˜ ื—ื•ืฉื™,
09:55
and that transformation is governed
249
595260
2000
ื•ืชื”ืœื™ืš ื–ื” ื ืฉืœื˜
09:57
by the physics of your body and your sensory apparatus.
250
597260
3000
ืขืœ-ื™ื“ื™ ื”ืคื™ื–ื™ื•ืœื•ื’ื™ื” ืฉืœ ื’ื•ืคื ื• ื•ื”ืžืขืจื›ืช ื”ื—ื•ืฉื™ืช.
10:00
But you can imagine looking inside the brain.
251
600260
2000
ืื‘ืœ ื ื“ืžื™ื™ืŸ ืฉื ื™ืชืŸ ืœื”ืชื‘ื•ื ืŸ ืœืชื•ืš ื”ืžื•ื—.
10:02
And here's inside the brain.
252
602260
2000
ื–ื” ืคื ื™ื ื”ืžื•ื—.
10:04
You might have a little predictor, a neural simulator,
253
604260
2000
ื™ื›ื•ืœ ืœื”ื™ื•ืช ืฉื™ืฉ ืืฆืœื ื• ื—ื–ืื™ ืงื˜ืŸ, ืกื™ืžื•ืœื˜ื•ืจ ืขืฆื‘ื™,
10:06
of the physics of your body and your senses.
254
606260
2000
ืœืคื™ื–ื™ื•ืœื•ื’ื™ื” ืฉืœ ื’ื•ืคื ื• ื•ื—ื•ืฉื™ื ื•.
10:08
So as you send a movement command down,
255
608260
2000
ื›ืืฉืจ ืฉื•ืœื—ื™ื ืคืงื•ื“ืช ืชื ื•ืขื” ืœื’ื•ืฃ,
10:10
you tap a copy of that off
256
610260
2000
ืœื•ืงื—ื™ื ื”ืขืชืง ืฉืœื”
10:12
and run it into your neural simulator
257
612260
2000
ื•ืžืจื™ืฆื™ื ืื•ืชื” ื‘ืกื™ืžื•ืœื˜ื•ืจ ื”ืขืฆื‘ื™ ืฉืœื ื•
10:14
to anticipate the sensory consequences of your actions.
258
614260
4000
ื›ื“ื™ ืœื”ืงื“ื™ื ื•ืœื—ื–ื•ืช ืืช ื”ืชื•ืฆืื•ืช ื”ื—ื•ืฉื™ื•ืช ืฉืœ ืคืขื•ืœืชื ื•.
10:18
So as I shake this ketchup bottle,
259
618260
2000
ื›ืš ืฉื›ืืฉืจ ืื ื™ ืžื ืขืจ ื‘ืงื‘ื•ืง ืงื˜ืฉื•ืค,
10:20
I get some true sensory feedback as the function of time in the bottom row.
260
620260
3000
ืื ื™ ืžืงื‘ืœ ืžืฉื•ื‘ ื—ื•ืฉื™ ืืžื™ืชื™ ื›ืคื•ื ืงืฆื™ื” ืฉืœ ื–ืžืŸ ื‘ืฆื™ืจ ื”ืื•ืคืงื™.
10:23
And if I've got a good predictor, it predicts the same thing.
261
623260
3000
ืื ื™ืฉ ืืฆืœื™ ื—ื–ืื™ ื˜ื•ื‘, ื”ื•ื ื—ื•ื–ื” ืืช ืื•ืชื• ื”ื“ื‘ืจ.
10:26
Well why would I bother doing that?
262
626260
2000
ืžื“ื•ืข ืฉืื˜ืจื— ืœืขืฉื•ืช ืืช ื›ืœ ื–ื”?
10:28
I'm going to get the same feedback anyway.
263
628260
2000
ื‘ื›ืœ ืžืงืจื” ืื ื™ ื”ื•ืœืš ืœืงื‘ืœ ืื•ืชื• ืžืฉื•ื‘.
10:30
Well there's good reasons.
264
630260
2000
ื™ืฉ ืœื–ื” ืกื™ื‘ื•ืช ื˜ื•ื‘ื•ืช.
10:32
Imagine, as I shake the ketchup bottle,
265
632260
2000
ื ื“ืžื™ื™ืŸ ืฉื‘ืขื•ื“ื™ ืžื ืขืจ ืืช ื‘ืงื‘ื•ืง ื”ืงื˜ืฉื•ืค,
10:34
someone very kindly comes up to me and taps it on the back for me.
266
634260
3000
ืžื™ืฉื”ื• ื‘ื ื•ื˜ื•ืคื— ืขืœื™ื• ืžืื—ื•ืจ ื›ื“ื™ ืœืขื–ื•ืจ ืœื™.
10:37
Now I get an extra source of sensory information
267
637260
2000
ื›ืขืช ื™ืฉ ืœื™ ืžืงื•ืจ ื ื•ืกืฃ ืฉืœ ืžื™ื“ืข ื—ื•ืฉื™
10:39
due to that external act.
268
639260
2000
ื‘ื’ืœืœ ื”ืคืขื•ืœื” ื”ื—ื™ืฆื•ื ื™ืช ื”ื–ื•.
10:41
So I get two sources.
269
641260
2000
ืœื›ืŸ ื™ืฉ ืœื™ ืฉื ื™ ืžืงื•ืจื•ืช.
10:43
I get you tapping on it, and I get me shaking it,
270
643260
3000
ืื ื™ ืžืงื‘ืœ ืžื”ื˜ืคื™ื—ื” ื•ืื ื™ ืžืงื‘ืœ ืžื”ื ื™ืขื•ืจ,
10:46
but from my senses' point of view,
271
646260
2000
ืื‘ืœ ืžื‘ื—ื™ื ืช ื”ื—ื•ืฉื™ื ืฉืœื™,
10:48
that is combined together into one source of information.
272
648260
3000
ื–ื” ืžืฉืชืœื‘ ื‘ื™ื—ื“ ืœืžืงื•ืจ ืžื™ื“ืข ืื—ื“.
10:51
Now there's good reason to believe
273
651260
2000
ื™ืฉ ืกื™ื‘ื” ื˜ื•ื‘ื” ืœื”ืืžื™ืŸ ืฉื”ื™ื™ื ื• ืจื•ืฆื™ื
10:53
that you would want to be able to distinguish external events from internal events.
274
653260
3000
ืœื”ื™ื•ืช ืžืกื•ื’ืœื™ื ืœื”ื‘ื—ื™ืŸ ื‘ื™ืŸ ืื™ืจื•ืขื™ื ื—ื™ืฆื•ื ื™ื™ื ืœืื™ืจื•ืขื™ื ืคื ื™ืžื™ื™ื.
10:56
Because external events are actually much more behaviorally relevant
275
656260
3000
ื›ื™ ืื™ืจื•ืขื™ื ื—ื™ืฆื•ื ื™ื™ื ื ื•ื’ืขื™ื ื”ืจื‘ื” ื™ื•ืชืจ ืœื”ืชื ื”ื’ื•ืช
10:59
than feeling everything that's going on inside my body.
276
659260
3000
ืœืขื•ืžืช ื”ืชื—ื•ืฉื” ืฉืœ ื›ืœ ืžื” ืฉืงื•ืจื” ื‘ื’ื•ืฃ ืฉืœื ื•.
11:02
So one way to reconstruct that
277
662260
2000
ืœื›ืŸ ื“ืจืš ื—ื“ืฉื” ืœืกื“ืจ ืืช ื–ื”
11:04
is to compare the prediction --
278
664260
2000
ื”ื™ื ืœื”ืฉื•ื•ืช ืืช ื”ืชื—ื–ื™ืช --
11:06
which is only based on your movement commands --
279
666260
2000
ื”ืžืชื‘ืกืกืช ืจืง ืขืœ ืคืงื•ื“ื•ืช ื”ืชื ื•ืขื” ืฉืœื ื• --
11:08
with the reality.
280
668260
2000
ืขื ื”ืžืฆื™ืื•ืช.
11:10
Any discrepancy should hopefully be external.
281
670260
3000
ืจืฆื•ื™ ืฉื›ืœ ืื™-ื”ืชืืžื” ืชื”ื™ื” ื—ื™ืฆื•ื ื™ืช.
11:13
So as I go around the world,
282
673260
2000
ื›ืš ืฉื‘ืขื•ื“ื™ ืžืชื”ืœืš ื‘ืขื•ืœื,
11:15
I'm making predictions of what I should get, subtracting them off.
283
675260
3000
ืื ื™ ืขื•ืฉื” ืชื—ื–ื™ื•ืช ืฉืœ ืžื” ืฉืื ื™ ืืžื•ืจ ืœื”ื™ืชืงืœ ื‘ื• ื•ืžื—ืกื™ืจ ืื•ืชืŸ.
11:18
Everything left over is external to me.
284
678260
2000
ื›ืœ ืžื” ืฉื ื•ืชืจ ื”ื•ื ื—ื™ืฆื•ื ื™ ืขื‘ื•ืจื™.
11:20
What evidence is there for this?
285
680260
2000
ืื™ื–ื” ืจืื™ื•ืช ื™ืฉ ืœื ื• ืขืœ ื–ื”?
11:22
Well there's one very clear example
286
682260
2000
ื™ืฉ ื“ื•ื’ืžื ืื—ืช ืžืื•ื“ ื‘ืจื•ืจื”
11:24
where a sensation generated by myself feels very different
287
684260
2000
ื‘ื” ืชื—ื•ืฉื” ืฉืื ื™ ื‘ืขืฆืžื™ ื™ื•ืฆืจ, ื’ื•ืจืžืช ื”ืจื’ืฉื” ืžืื•ื“ ืฉื•ื ื”
11:26
then if generated by another person.
288
686260
2000
ืžืืฉืจ ืื•ืชื” ืชื—ื•ืฉื” ื”ื ื•ืฆืจืช ืขืœ-ื™ื“ื™ ืื“ื ืื—ืจ.
11:28
And so we decided the most obvious place to start
289
688260
2000
ืœื›ืŸ ื”ื—ืœื˜ื ื• ืฉื”ื“ื‘ืจ ื”ืžื•ื‘ืŸ ืžืืœื™ื• ืœื”ืชื—ื™ืœ ื‘ื•
11:30
was with tickling.
290
690260
2000
ื”ื•ื ื“ื™ื’ื“ื•ื’.
11:32
It's been known for a long time, you can't tickle yourself
291
692260
2000
ื™ื“ื•ืข ื›ื‘ืจ ื–ืžืŸ ืจื‘ ืฉืœื ื ื™ืชืŸ ืœื“ื’ื“ื’ ืืช ืขืฆืžืš
11:34
as well as other people can.
292
694260
2000
ื‘ืื•ืชื” ืžื™ื“ื” ืฉื‘ื” ืื—ืจื™ื ื™ื›ื•ืœื™ื.
11:36
But it hasn't really been shown, it's because you have a neural simulator,
293
696260
3000
ืื‘ืœ ื–ื” ืืฃ ืคืขื ืœื ืžืžืฉ ื”ื•ื“ื’ื ื›ื™ ื™ืฉ ืืฆืœื ื• ืกื™ืžื•ืœื˜ื•ืจ ื˜ื‘ืขื™,
11:39
simulating your own body
294
699260
2000
ื”ืžื“ืžื” ืืช ื”ื’ื•ืฃ ืฉืœื ื• ืขืฆืžื ื•
11:41
and subtracting off that sense.
295
701260
2000
ื•ื”ืžื—ืกื™ืจ ืืช ื”ื—ื•ืฉ ื”ื–ื”.
11:43
So we can bring the experiments of the 21st century
296
703260
3000
ื›ืขืช ืื ื• ืžื‘ื™ืื™ื ืืช ื”ื ื™ืกื•ื™ื™ื ืœืžืื” ื”-21
11:46
by applying robotic technologies to this problem.
297
706260
3000
ื‘ืืžืฆืขื•ืช ื™ื™ืฉื•ื ื˜ื›ื ื•ืœื•ื’ื™ื•ืช ืจื•ื‘ื•ื˜ื™ื•ืช ืœื‘ืขื™ื” ื–ื•.
11:49
And in effect, what we have is some sort of stick in one hand attached to a robot,
298
709260
3000
ืœืžืขืฉื”, ื™ืฉ ืœื ื• ืžื™ืŸ ืžืงืœ ื‘ื™ื“ ืื—ืช ื”ืžื—ื•ื‘ืจืช ืœืจื•ื‘ื•ื˜,
11:52
and they're going to move that back and forward.
299
712260
2000
ื•ื”ื ื”ื•ืœื›ื™ื ืœื”ื–ื™ื– ืื•ืชื” ืงื“ื™ืžื” ื•ืื—ื•ืจื”.
11:54
And then we're going to track that with a computer
300
714260
2000
ืื ื• ื”ื•ืœื›ื™ื ืœืขืงื•ื‘ ืื—ืจื™ ื–ื” ื‘ืขื–ืจืช ืžื—ืฉื‘
11:56
and use it to control another robot,
301
716260
2000
ื•ืœื”ืฉืชืžืฉ ื‘ื• ื›ื“ื™ ืœืฉืœื•ื˜ ืขืœ ืจื•ื‘ื•ื˜ ืื—ืจ,
11:58
which is going to tickle their palm with another stick.
302
718260
2000
ืืฉืจ ืขื•ืžื“ ืœื“ื’ื“ื’ ืืช ื›ืฃ-ื™ื“ื ืขื ืžืงืœ ืื—ืจ.
12:00
And then we're going to ask them to rate a bunch of things
303
720260
2000
ืื ื• ื ื‘ืงืฉื ืœื“ืจื’ ืžืกืคืจ ื“ื‘ืจื™ื
12:02
including ticklishness.
304
722260
2000
ื›ื•ืœืœ ื”ืจื’ื™ืฉื•ืช ืœื“ื™ื’ื“ื•ื’.
12:04
I'll show you just one part of our study.
305
724260
2000
ืืฆื™ื’ ืœื›ื ืจืง ื—ืœืง ืื—ื“ ืฉืœ ื”ืžื—ืงืจ.
12:06
And here I've taken away the robots,
306
726260
2000
ื›ืืŸ ื”ืจื—ืงื ื• ืืช ื”ืจื•ื‘ื•ื˜ื™ื,
12:08
but basically people move with their right arm sinusoidally back and forward.
307
728260
3000
ืื‘ืœ ืขื“ื™ื™ืŸ ืื ืฉื™ื ืžื–ื™ื–ื™ื ืืช ื–ืจื•ืขื ืงื“ื™ืžื” ื•ืื—ื•ืจื”
12:11
And we replay that to the other hand with a time delay.
308
731260
3000
ื•ืื ื• ืžืฉื—ื–ืจื™ื ืืช ื”ืชื ื•ืขื” ืขืœ ื’ื‘ื™ ื”ื™ื“ ื”ืฉื ื™ื” ื‘ืื™ื—ื•ืจ ืฉืœ ื–ืžืŸ.
12:14
Either no time delay,
309
734260
2000
ืื• ืœืœื ืื™ื—ื•ืจ ื–ืžืŸ,
12:16
in which case light would just tickle your palm,
310
736260
2000
ืฉื‘ืžืงืจื” ื›ื–ื” ื›ืื™ืœื• ื”ืื•ืจ ืžื“ื’ื“ื’ ืืช ื›ืฃ-ื”ื™ื“,
12:18
or with a time delay of two-tenths of three-tenths of a second.
311
738260
4000
ืื• ื‘ืื™ื—ื•ืจ ื–ืžืŸ ืฉืœ ืฉืชื™ื™ื ืื• ืฉืœื•ืฉ ืขืฉื™ืจื™ื•ืช ืฉื ื™ื”.
12:22
So the important point here
312
742260
2000
ื”ื ืงื•ื“ื” ื”ื—ืฉื•ื‘ื” ื›ืืŸ ื”ื™ื
12:24
is the right hand always does the same things -- sinusoidal movement.
313
744260
3000
ืฉื™ื“ ื™ืžื™ืŸ ืขื•ืฉื” ืชืžื™ื“ ืื•ืชื• ื”ื“ื‘ืจ -- ืชื ื•ืขื” ื’ืœื™ืช.
12:27
The left hand always is the same and puts sinusoidal tickle.
314
747260
3000
ื™ื“ ืฉืžืืœ ื”ื™ื ื‘ืื•ืชื• ืžืฆื‘ ื•ืจื•ืฉืžืช ืืช ื”ื“ื™ื’ื“ื•ื’ ื”ื’ืœื™.
12:30
All we're playing with is a tempo causality.
315
750260
2000
ื›ืœ ืžื” ืฉืื ื• ืžืฉื ื™ื ื–ื” ื”ืชื™ื–ืžื•ืŸ.
12:32
And as we go from naught to 0.1 second,
316
752260
2000
ื›ืืฉืจ ืขื•ื‘ืจื™ื ืžืืคืก ืœ-0.1 ืฉื ื™ื”,
12:34
it becomes more ticklish.
317
754260
2000
ื”ืจื’ื™ืฉื•ืช ืœื“ื™ื’ื“ื•ื’ ืขื•ืœื”.
12:36
As you go from 0.1 to 0.2,
318
756260
2000
ื›ืืฉืจ ืขื•ื‘ืจื™ื ืž-0.1 ืœ-0.2,
12:38
it becomes more ticklish at the end.
319
758260
2000
ื”ืจื’ื™ืฉื•ืช ืขื•ืœื” ืขื•ื“.
12:40
And by 0.2 of a second,
320
760260
2000
ื‘-0.2 ืฉื ื™ื•ืช,
12:42
it's equivalently ticklish
321
762260
2000
ื–ื” ืžื“ื’ื“ื’ ื‘ืื•ืชื” ืžื™ื“ื”
12:44
to the robot that just tickled you without you doing anything.
322
764260
2000
ื›ืžื• ื”ืจื•ื‘ื•ื˜ ืฉื“ื™ื’ื“ื’ ืžื‘ืœื™ ืฉื”ืžืฉืชืชืฃ ืขืฉื” ืžืฉื”ื•.
12:46
So whatever is responsible for this cancellation
323
766260
2000
ืœื ืžืฉื ื” ืžื” ืื—ืจืื™ ืœื‘ื™ื˜ื•ืœ ื–ื”,
12:48
is extremely tightly coupled with tempo causality.
324
768260
3000
ื–ื” ืงืฉื•ืจ ื‘ืื•ืคืŸ ื”ื“ื•ืง ืœืชื™ื–ืžื•ืŸ.
12:51
And based on this illustration, we really convinced ourselves in the field
325
771260
3000
ื‘ื”ืชื‘ืกืก ืขืœ ื”ืžื—ืฉื” ื–ื•, ื”ื’ืขื ื• ืœืžืกืงื ื”
12:54
that the brain's making precise predictions
326
774260
2000
ืฉื”ืžื•ื— ืขื•ืฉื” ื—ื™ื–ื•ื™ื™ื ืžื“ื•ื™ื™ืงื™ื
12:56
and subtracting them off from the sensations.
327
776260
3000
ื•ืžื—ืกื™ืจ ืื•ืชื ืžื”ืชื—ื•ืฉื•ืช.
12:59
Now I have to admit, these are the worst studies my lab has ever run.
328
779260
3000
ืื‘ืœ ืขืœื™ื™ ืœื”ื•ื“ื•ืช ืฉืืœื” ื”ืžื—ืงืจื™ื ื”ื›ื™ ื’ืจื•ืขื™ื ืฉื‘ื•ืฆืขื• ื‘ืžืขื‘ื“ื” ืฉืœื™.
13:02
Because the tickle sensation on the palm comes and goes,
329
782260
2000
ื›ื™ ื‘ื’ืœืœ ืฉืชื—ื•ืฉืช ื”ื“ื™ื’ื“ื•ื’ ืขืœ ื›ืฃ-ื”ื™ื“ ื‘ืื” ื•ื”ื•ืœื›ืช,
13:04
you need large numbers of subjects
330
784260
2000
ื™ืฉ ืฆื•ืจืš ื‘ืžืกืคืจ ื’ื“ื•ืœ ืฉืœ ืžืฉืชืชืคื™ื
13:06
with these stars making them significant.
331
786260
2000
ื›ื“ื™ ืฉื–ื” ื™ื”ื™ื” ืžืฉืžืขื•ืชื™.
13:08
So we were looking for a much more objective way
332
788260
2000
ืœื›ืŸ ืื ื• ืžื—ืคืฉื™ื ื“ืจืš ื”ืจื‘ื” ื™ื•ืชืจ ืื•ื‘ื™ื™ืงื˜ื™ื‘ื™ืช
13:10
to assess this phenomena.
333
790260
2000
ื›ื“ื™ ืœืืžื•ื“ ืืช ื”ืชื•ืคืขื”.
13:12
And in the intervening years I had two daughters.
334
792260
2000
ื‘ื™ื ืชื™ื™ื ื ื•ืœื“ื• ืœื™ ืฉืชื™ ื‘ื ื•ืช.
13:14
And one thing you notice about children in backseats of cars on long journeys,
335
794260
3000
ื•ืื—ื“ ื”ื“ื‘ืจื™ื ืฉืžื‘ื—ื™ื ื™ื ื‘ื”ื ื‘ื ื•ื’ืข ืœื™ืœื“ื™ื ื‘ืžื•ืฉื‘ื™ื ืื—ื•ืจื™ื™ื ืฉืœ ืžื›ื•ื ื™ืช
13:17
they get into fights --
336
797260
2000
ื”ื•ื ืฉื”ื ืžืชื—ื™ืœื™ื ืœืจื™ื‘ --
13:19
which started with one of them doing something to the other, the other retaliating.
337
799260
3000
ื–ื” ืžืชื—ื™ืœ ืขื ืื—ื“ ืฉืขื•ืฉื” ืžืฉื”ื• ืœืื—ืจ ื•ื–ื” ืžื’ื™ื‘.
13:22
It quickly escalates.
338
802260
2000
ื•ื–ื” ืžืกืœื™ื ื‘ืžื”ื™ืจื•ืช.
13:24
And children tend to get into fights which escalate in terms of force.
339
804260
3000
ื™ืœื“ื™ื ื ื•ื˜ื™ื ืœื”ืชื—ื™ืœ ืžืจื™ื‘ื•ืช ื”ืžืกืœื™ืžื•ืช ื‘ืžื•ื ื—ื™ื ืฉืœ ื›ื•ื—.
13:27
Now when I screamed at my children to stop,
340
807260
2000
ื›ืืฉืจ ืฆืจื—ืชื™ ืขืœ ื™ืœื“ื•ืชื™ื™ ืœื”ืคืกื™ืง,
13:29
sometimes they would both say to me
341
809260
2000
ืœืคืขืžื™ื ืฉืชื™ื”ืŸ ื”ื™ื• ืื•ืžืจื•ืช ืœื™
13:31
the other person hit them harder.
342
811260
3000
ืฉื”ืฉื ื™ื” ื”ืจื‘ื™ืฆื” ื™ื•ืชืจ ื—ื–ืง.
13:34
Now I happen to know my children don't lie,
343
814260
2000
ืื ื™ ื™ื•ื“ืข ืฉื™ืœื“ื•ืชื™ื™ ืœื ืžืฉืงืจื•ืช,
13:36
so I thought, as a neuroscientist,
344
816260
2000
ืœื›ืŸ ื—ืฉื‘ืชื™, ื‘ืชื•ืจ ื—ื•ืงืจ ืžืขืจื›ื•ืช ืขืฆื‘,
13:38
it was important how I could explain
345
818260
2000
ืฉื–ื” ื—ืฉื•ื‘ ื™ื”ื™ื” ืœื”ืกื‘ื™ืจ
13:40
how they were telling inconsistent truths.
346
820260
2000
ื›ื™ืฆื“ ื”ืŸ ืกื™ืคืจื• ืืžื™ืชื•ืช ืœื ืงื•ื”ืจื ื˜ื™ื•ืช.
13:42
And we hypothesize based on the tickling study
347
822260
2000
ื”ื ื—ื ื• ื‘ื”ืชื‘ืกืก ืขืœ ืžื—ืงืจ ื”ื“ื™ื’ื“ื•ื’
13:44
that when one child hits another,
348
824260
2000
ืฉื›ืืฉืจ ื™ืœื“ ืื—ื“ ืžื›ื” ืืช ื”ืฉื ื™,
13:46
they generate the movement command.
349
826260
2000
ื”ื•ื ืžื™ื™ืฆืจ ืคืงื•ื“ืช ืชื ื•ืขื”.
13:48
They predict the sensory consequences and subtract it off.
350
828260
3000
ื”ื•ื ื—ื•ื–ื” ืืช ื”ืชื•ืฆืื•ืช ื”ื—ื•ืฉื™ื•ืช ื•ืžื—ืกื™ืจ ืื•ืชืŸ.
13:51
So they actually think they've hit the person less hard than they have --
351
831260
2000
ืœื›ืŸ ื”ื•ื ืœืžืขืฉื” ื—ื•ืฉื‘ ืฉื”ื•ื ื”ื™ื›ื” ืคื—ื•ืช ื—ื–ืง ืžืืฉืจ ื‘ืคื•ืขืœ --
13:53
rather like the tickling.
352
833260
2000
ื‘ืขืจืš ื›ืžื• ื”ื“ื™ื’ื“ื•ื’.
13:55
Whereas the passive recipient
353
835260
2000
ื‘ืขื•ื“ ืฉื”ืกื•ืคื’ ื”ืคืกื™ื‘ื™
13:57
doesn't make the prediction, feels the full blow.
354
837260
2000
ืื™ื ื• ืขื•ืฉื” ื—ื™ื–ื•ื™ ื•ืœื›ืŸ ืžืจื’ื™ืฉ ืืช ืžืœื•ื ื”ืขื•ืฆืžื”.
13:59
So if they retaliate with the same force,
355
839260
2000
ืื ื”ื•ื ืžื—ื–ื™ืจ ื‘ืื•ืชื” ืขื•ืฆืžื”,
14:01
the first person will think it's been escalated.
356
841260
2000
ื”ืจืืฉื•ืŸ ื™ื—ืฉื•ื‘ ืฉื”ืžืฆื‘ ื”ืกืœื™ื.
14:03
So we decided to test this in the lab.
357
843260
2000
ืœื›ืŸ ื”ื—ืœื˜ื ื• ืœื‘ื“ื•ืง ื–ืืช ื‘ืžืขื‘ื“ื”.
14:05
(Laughter)
358
845260
3000
(ืฆื—ื•ืง)
14:08
Now we don't work with children, we don't work with hitting,
359
848260
2000
ืื ื• ืœื ืขื•ื‘ื“ื™ื ืขื ื™ืœื“ื™ื ื•ืœื ืขื ืžื›ื•ืช,
14:10
but the concept is identical.
360
850260
2000
ืื‘ืœ ื”ืจืขื™ื•ืŸ ื ืฉืืจ.
14:12
We bring in two adults. We tell them they're going to play a game.
361
852260
3000
ืžื‘ื™ืื™ื ืฉื ื™ ื‘ื•ื’ืจื™ื. ืžืกืคืจื™ื ืœื”ื ืฉื”ื ื”ื•ืœื›ื™ื ืœืฉื—ืง ืžืฉื—ืง.
14:15
And so here's player one and player two sitting opposite to each other.
362
855260
2000
ื”ื ื” ืฉื ื™ ื”ืฉื—ืงื ื™ื ื”ื™ื•ืฉื‘ื™ื ื–ื” ืžื•ืœ ื–ื”.
14:17
And the game is very simple.
363
857260
2000
ื”ืžืฉื—ืง ืžืื•ื“ ืคืฉื•ื˜.
14:19
We started with a motor
364
859260
2000
ื”ืชื—ืœื ื• ืขื ืžื ื•ืข
14:21
with a little lever, a little force transfuser.
365
861260
2000
ืฉื™ืฉ ืœื• ื™ื“ื™ืช ื”ืžืขื‘ื™ืจื” ื›ื•ื—.
14:23
And we use this motor to apply force down to player one's fingers
366
863260
2000
ืžืฉืชืžืฉื™ื ื‘ืžื ื•ืข ืœื”ืคืขื™ืœ ื›ื•ื— ื›ืœืคื™ ืžื˜ื” ืขืœ ืืฆื‘ืข
14:25
for three seconds and then it stops.
367
865260
3000
ืฉืœ ืฉื—ืงืŸ 1 ื‘ืžืฉืš 3 ืฉื ื™ื•ืช ื•ืื– ืœื”ืคืกื™ืง.
14:28
And that player's been told, remember the experience of that force
368
868260
3000
ืžื‘ืงืฉื™ื ืžืื•ืชื• ืฉื—ืงืŸ ืฉื™ื–ื›ื•ืจ ืืช ืชื—ื•ืฉืช ื”ื›ื•ื—
14:31
and use your other finger
369
871260
2000
ื•ืฉื™ืฉืชืžืฉ ื‘ืืฆื‘ืขื• ื”ืื—ืจืช
14:33
to apply the same force
370
873260
2000
ืœื”ืคืขื™ืœ ืืช ืื•ืชื• ื”ื›ื•ื—
14:35
down to the other subject's finger through a force transfuser -- and they do that.
371
875260
3000
ืขืœ ืืฆื‘ืข ื”ืฉื—ืงืŸ ื”ืื—ืจ ื‘ืืžืฆืขื•ืช ื”ื™ื“ื™ืช, ื•ื›ืš ื”ื ืขืฉื•.
14:38
And player two's been told, remember the experience of that force.
372
878260
3000
ืœืฉื—ืงืŸ 2 ืื•ืžืจื™ื ืฉื™ื–ื›ื•ืจ ืืช ืชื—ื•ืฉืช ื”ื›ื•ื—.
14:41
Use your other hand to apply the force back down.
373
881260
3000
ืฉื™ืฉืชืžืฉ ื‘ื™ื“ื• ื”ืฉื ื™ื” ืœื”ืคืขื™ืœ ื›ื•ื— ื‘ื—ื–ืจื”.
14:44
And so they take it in turns
374
884260
2000
ื•ื›ืš ื–ื” ืงื•ืจื” ื‘ืžื—ื–ื•ืจื™ื•ืช
14:46
to apply the force they've just experienced back and forward.
375
886260
2000
ื‘ื” ื”ื ืžืคืขื™ืœื™ื ื›ื•ื— ืฉื”ื ื—ื•ื• ื–ื” ืขืชื”.
14:48
But critically,
376
888260
2000
ืื‘ืœ ื™ืฉ ื“ื‘ืจ ืื—ื“ ื—ืฉื•ื‘,
14:50
they're briefed about the rules of the game in separate rooms.
377
890260
3000
ื”ื ืžืชื•ื“ืจื›ื™ื ื‘ื ืคืจื“ ื–ื” ืžื–ื” ืขืœ ื—ื•ืงื™ ื”ืžืฉื—ืง.
14:53
So they don't know the rules the other person's playing by.
378
893260
2000
ืœื›ืŸ ื”ื ืื™ื ื ืžื›ื™ืจื™ื ืืช ื”ื—ื•ืงื™ื ืœืคื™ื”ื ื”ืื—ืจ ืžืฉื—ืง.
14:55
And what we've measured
379
895260
2000
ื•ืžื” ืฉืžื“ื“ื ื•
14:57
is the force as a function of terms.
380
897260
2000
ื–ื” ืืช ื”ื›ื•ื— ื›ืคื•ื ืงืฆื™ื” ืฉืœ ืžื—ื–ื•ืจื™ื.
14:59
And if we look at what we start with,
381
899260
2000
ืื ืžืกืชื›ืœื™ื ืขื ืžื” ื”ืชื—ืœื ื•,
15:01
a quarter of a Newton there, a number of turns,
382
901260
2000
ืจื‘ืข ื ื™ื•ื˜ื•ืŸ ืฉื, ืžืก' ืžื—ื–ื•ืจื™ื,
15:03
perfect would be that red line.
383
903260
2000
ื”ื“ื‘ืจ ื”ืžื•ืฉืœื ื–ื” ื”ืงื• ื”ืื“ื•ื ืฉื.
15:05
And what we see in all pairs of subjects is this --
384
905260
3000
ื•ืžื” ืฉืจื•ืื™ื ื‘ื›ืœ ื”ื–ื•ื’ื•ืช ืฉื”ืฉืชืชืคื• ื–ื” --
15:08
a 70 percent escalation in force
385
908260
2000
ืขืœื™ื” ืฉืœ 70 ืื—ื•ื– ื‘ื›ื•ื—
15:10
on each go.
386
910260
2000
ื‘ื›ืœ ืžื—ื–ื•ืจ.
15:12
So it really suggests, when you're doing this --
387
912260
2000
ื–ื” ืžืจืžื– ืฉื›ืืฉืจ ืขื•ืฉื™ื ืืช ื–ื” --
15:14
based on this study and others we've done --
388
914260
2000
ืœืคื™ ืžื—ืงืจ ื–ื” ื•ืžื—ืงืจื™ื ืื—ืจื™ื ืฉืขืฉื™ื ื• --
15:16
that the brain is canceling the sensory consequences
389
916260
2000
ื”ืžื•ื— ืžื‘ื˜ืœ ืืช ื”ืชื—ื•ืฉื•ืช ื”ื—ื•ืฉื™ื•ืช
15:18
and underestimating the force it's producing.
390
918260
2000
ื•ืžืžืขื™ื˜ ื‘ืขืจืš ืฉืœ ื”ื›ื•ื— ืฉื”ื•ื ืžื™ื™ืฆืจ.
15:20
So it re-shows the brain makes predictions
391
920260
2000
ื–ื” ืžื•ื›ื™ื— ืžื—ื“ืฉ ืฉื”ืžื•ื— ืขื•ืฉื” ืชื—ื–ื™ื•ืช
15:22
and fundamentally changes the precepts.
392
922260
3000
ื•ืžื™ืกื•ื“ื• ืžืฉื ื” ืืช ื”ืคืงื•ื“ื•ืช.
15:25
So we've made inferences, we've done predictions,
393
925260
3000
ืื ื›ืš, ื”ืกืงื ื• ืžืกืงื ื•ืช, ืขืฉื™ื ื• ืชื—ื–ื™ื•ืช,
15:28
now we have to generate actions.
394
928260
2000
ื›ืขืช ืขืœื™ื ื• ืœื™ื™ืฆืจ ืคืขื•ืœื•ืช.
15:30
And what Bayes' rule says is, given my beliefs,
395
930260
2000
ื•ืžื” ืฉื—ื•ืง ื‘ื™ื™ืกื™ืื ื™ ืื•ืžืจ ื”ื•ื ืฉื‘ืืžื•ื ื•ืชื™ื™ ื”ื ืชื•ื ื•ืช,
15:32
the action should in some sense be optimal.
396
932260
2000
ื”ืคืขื•ืœื” ื‘ืžื•ื‘ืŸ ืžืกื•ื™ื™ื ืฆืจื™ื›ื” ืœื”ื™ื•ืช ื”ืžื™ื˜ื‘ื™ืช.
15:34
But we've got a problem.
397
934260
2000
ืื‘ืœ ื™ืฉ ื›ืืŸ ื‘ืขื™ื”.
15:36
Tasks are symbolic -- I want to drink, I want to dance --
398
936260
3000
ืžืฉื™ืžื•ืช ื”ืŸ ืกื™ืžื‘ื•ืœื™ื•ืช -- ืื ื™ ืจื•ืฆื” ืœืฉืชื•ืช, ืจื•ืฆื” ืœืจืงื•ื“ --
15:39
but the movement system has to contract 600 muscles
399
939260
2000
ืื‘ืœ ืžืขืจื›ืช ื”ืชื ื•ืขื” ืฆืจื™ื›ื” ืœื›ื•ื•ืฅ 600 ืฉืจื™ืจื™ื
15:41
in a particular sequence.
400
941260
2000
ื‘ืกื“ืจ ืžืื•ื“ ืžืกื•ื™ื™ื.
15:43
And there's a big gap
401
943260
2000
ื•ื–ื” ื”ืคืขืจ ื”ื’ื“ื•ืœ
15:45
between the task and the movement system.
402
945260
2000
ื‘ื™ืŸ ื”ืžืฉื™ืžื” ื•ืžืขืจื›ืช ื”ืชื ื•ืขื”.
15:47
So it could be bridged in infinitely many different ways.
403
947260
2000
ื ื™ืชืŸ ืœื’ืฉืจ ืขืœ ื–ื” ื‘ื”ืžื•ืŸ ื“ืจื›ื™ื ืฉื•ื ื•ืช.
15:49
So think about just a point to point movement.
404
949260
2000
ืชื—ืฉื‘ื• ืจืง ืขืœ ืชื ื•ืขื” ืžื ืงื•ื“ื” ืื—ืช ืœืฉื ื™ื”.
15:51
I could choose these two paths
405
951260
2000
ื ื™ืชืŸ ืœื‘ื—ื•ืจ ื‘ืฉื ื™ ืžืกืœื•ืœื™ื ื”ืœืœื•
15:53
out of an infinite number of paths.
406
953260
2000
ืžืชื•ืš ืžืกืคืจ ืื™ืŸ-ืกื•ืคื™ ืฉืœ ืžืกืœื•ืœื™ื.
15:55
Having chosen a particular path,
407
955260
2000
ื‘ื”ื™ื‘ื—ืจ ืžืกืœื•ืœ ืžืกื•ื™ื™ื,
15:57
I can hold my hand on that path
408
957260
2000
ืืคืฉืจ ืœื”ื—ื–ื™ืง ืืช ื”ื™ื“ ืžืขืœ ืื•ืชื• ืžืกืœื•ืœ
15:59
as infinitely many different joint configurations.
409
959260
2000
ื‘ืื•ืคื ื™ื ืฉื•ื ื™ื ื›ืžืกืคืจ ืžืฆื‘ื™ ื”ืžืคืจืง ื”ืฉื•ื ื™ื.
16:01
And I can hold my arm in a particular joint configuration
410
961260
2000
ื•ื ื™ืชืŸ ืœื”ื—ื–ื™ืง ืืช ื”ื™ื“ ื‘ืชื ื•ื—ืช ืžืคืจืง ืžืกื•ื™ื™ืžืช,
16:03
either very stiff or very relaxed.
411
963260
2000
ืื• ื‘ืžืฆื‘ ืงืฉื™ื— ืื• ื‘ืžืฆื‘ ื ืจืคื”.
16:05
So I have a huge amount of choice to make.
412
965260
3000
ื›ืš ืฉื™ืฉ ืžืกืคืจ ืขืฆื•ื ืฉืœ ื‘ื—ื™ืจื•ืช ืœืขืฉื•ืช.
16:08
Now it turns out, we are extremely stereotypical.
413
968260
3000
ื›ืขืช ืžืชื‘ืจืจ ืฉืื ื• ืžืื•ื“ ืฆืคื•ื™ื™ื.
16:11
We all move the same way pretty much.
414
971260
3000
ื›ื•ืœื ื• ื ืขื™ื ื‘ืงื™ืจื•ื‘ ื‘ืื•ืชื• ืื•ืคืŸ.
16:14
And so it turns out we're so stereotypical,
415
974260
2000
ืžืกืชื‘ืจ ืฉืื ื• ื›ื” ืฆืคื•ื™ื™ื
16:16
our brains have got dedicated neural circuitry
416
976260
2000
ืฉื‘ืžื•ื—ื ื• ืงื™ื™ื ืžืขื’ืœ ืขืฆื‘ื™ ื™ื™ืขื•ื“ื™
16:18
to decode this stereotyping.
417
978260
2000
ื›ื“ื™ ืœืคืขื ื— ืืช ืฆื•ืคืŸ ื”ื”ืชื ื”ื’ื•ืช ื”ืฆืคื•ื™ื” ื”ื–ื•.
16:20
So if I take some dots
418
980260
2000
ื›ืš ืฉืื ืœื•ืงื—ื™ื ื›ืžื” ื ืงื•ื“ื•ืช
16:22
and set them in motion with biological motion,
419
982260
3000
ื•ืžื ื™ืขื™ื ืื•ืชืŸ ื‘ืชื ื•ืขื” ื‘ื™ื•ืœื•ื’ื™ืช,
16:25
your brain's circuitry would understand instantly what's going on.
420
985260
3000
ื”ืžืขื’ืœ ื‘ืžื•ื—ื ื• ื™ืชืคื•ืก ืžื™ื™ื“ ืžื” ืงื•ืจื” ื›ืืŸ.
16:28
Now this is a bunch of dots moving.
421
988260
2000
ื–ื• ืงื‘ื•ืฆืช ื ืงื•ื“ื•ืช ื ืขื”.
16:30
You will know what this person is doing,
422
990260
3000
ื™ืฉืจ ื™ื•ื“ืขื™ื ืžื” ื”ื“ืžื•ืช ื”ื–ื• ืขื•ืฉื”,
16:33
whether happy, sad, old, young -- a huge amount of information.
423
993260
3000
ืื ืฉืžื—ื”, ืขืฆื•ื‘ื”, ืงืฉื™ืฉื”, ืฆืขื™ืจื” -- ื›ืžื•ืช ืžื™ื“ืข ืื“ื™ืจื”.
16:36
If these dots were cars going on a racing circuit,
424
996260
2000
ืื ื ืงื•ื“ื•ืช ื”ืœืœื• ื”ื™ื• ืžื›ื•ื ื™ื•ืช ื‘ืžืกืœื•ืœ ืžืจื•ืฆื™ื,
16:38
you would have absolutely no idea what's going on.
425
998260
3000
ืœื ื”ื™ื” ืœื›ื ืžื•ืฉื’ ืžื” ืงื•ืจื” ื›ืืŸ.
16:41
So why is it
426
1001260
2000
ืื– ืžื“ื•ืข
16:43
that we move the particular ways we do?
427
1003260
2000
ืื ื• ื ืขื™ื ื‘ืžืกืœื•ืœื™ื ืžืกื•ื™ื™ืžื™ื?
16:45
Well let's think about what really happens.
428
1005260
2000
ื”ื‘ื” ื ื—ืฉื•ื‘ ืžื” ื‘ืืžืช ืžืชืจื—ืฉ ื›ืืŸ.
16:47
Maybe we don't all quite move the same way.
429
1007260
3000
ืื•ืœื™ ืื ื• ืœื ืœื’ืžืจื™ ื ืขื™ื ื‘ืื•ืชื• ืžืกืœื•ืœ.
16:50
Maybe there's variation in the population.
430
1010260
2000
ืื•ืœื™ ื™ืฉ ืฉื•ื ื™ ื‘ืื•ื›ืœื•ืกื™ื”.
16:52
And maybe those who move better than others
431
1012260
2000
ื•ืื•ืœื™ ื›ืืœื” ื”ื ืขื™ื ื™ื•ืชืจ ื˜ื•ื‘ ืžืื—ืจื™ื
16:54
have got more chance of getting their children into the next generation.
432
1014260
2000
ื”ื ื‘ืขืœื™ ืกื™ื›ื•ื™ื™ื ื™ื•ืชืจ ื˜ื•ื‘ื™ื ืœื”ื•ืœื™ื“ ืฆืืฆืื™ื.
16:56
So in evolutionary scales, movements get better.
433
1016260
3000
ืœื›ืŸ ื‘ืงื ื”-ืžื™ื“ื” ืื‘ื•ืœื•ืฆื™ื•ื ื™, ื”ืชื ื•ืขื•ืช ืžืฉืชืคืจื•ืช.
16:59
And perhaps in life, movements get better through learning.
434
1019260
3000
ื•ืื•ืœื™ ื‘ื—ื™ื™ื, ื”ืชื ื•ืขื•ืช ืžืฉืชืคืจื•ืช ื“ืจืš ืœืžื™ื“ื”.
17:02
So what is it about a movement which is good or bad?
435
1022260
2000
ืื– ืžื” ื˜ื•ื‘ ื•ืžื” ืจืข ื‘ืชื ื•ืขื”?
17:04
Imagine I want to intercept this ball.
436
1024260
2000
ืชื—ืฉื‘ื• ืฉื‘ืจืฆื•ื ื™ ืœืชืคื•ืก ื›ื“ื•ืจ ื–ื”.
17:06
Here are two possible paths to that ball.
437
1026260
3000
ื”ื ื” ืฉื ื™ ืžืกืœื•ืœื™ื ืืคืฉืจื™ื™ื ืืœ ืื•ืชื• ื›ื“ื•ืจ.
17:09
Well if I choose the left-hand path,
438
1029260
2000
ืื ืื ื™ ื‘ื•ื—ืจ ืืช ื”ืžืกืœื•ืœ ื”ืฉืžืืœื™,
17:11
I can work out the forces required
439
1031260
2000
ืื•ื›ืœ ืœื—ืฉื‘ ืืช ื”ื›ื•ื—ื•ืช ื”ื“ืจื•ืฉื™ื
17:13
in one of my muscles as a function of time.
440
1033260
2000
ื‘ืื—ื“ ืžืฉืจื™ืจื™ื™ ื›ืคื•ื ืงืฆื™ื” ืฉืœ ื–ืžืŸ.
17:15
But there's noise added to this.
441
1035260
2000
ืื‘ืœ ื ื•ืกืฃ ืœื–ื” ืจืขืฉ.
17:17
So what I actually get, based on this lovely, smooth, desired force,
442
1037260
3000
ืœื›ืŸ ืžื” ืฉืืงื‘ืœ, ื‘ื”ืชื‘ืกืก ืขืœ ื”ื›ื•ื— ื”ื—ืžื•ื“, ื”ื—ืœืง ื•ื”ืจืฆื•ื™,
17:20
is a very noisy version.
443
1040260
2000
ื–ื• ื’ื™ืจืกื” ืžืื•ื“ ืจื•ืขืฉืช.
17:22
So if I pick the same command through many times,
444
1042260
3000
ืœื›ืŸ ืื ืื‘ื—ืจ ื‘ืื•ืชื” ืคืงื•ื“ื” ื”ืจื‘ื” ืคืขืžื™ื,
17:25
I will get a different noisy version each time, because noise changes each time.
445
1045260
3000
ืืงื‘ืœ ื›ืœ ืคืขื ื’ื™ืจืกื” ืจื•ืขืฉืช ืื—ืจืช, ื›ื™ ื”ืจืขืฉ ืžืฉืชื ื” ื‘ื›ืœ ืคืขื.
17:28
So what I can show you here
446
1048260
2000
ืœื›ืŸ ืžื” ืฉืื•ื›ืœ ืœื”ืจืื•ืช ืœื›ื ื›ืืŸ
17:30
is how the variability of the movement will evolve
447
1050260
2000
ื–ื” ื›ื™ืฆื“ ืชืชืคืชื— ื”ื”ืฉืชื ื•ืช ืฉืœ ื”ืชื ื•ืขื”
17:32
if I choose that way.
448
1052260
2000
ืื ืื ื™ ื‘ื•ื—ืจ ื‘ื“ืจืš ื”ื”ื™ื.
17:34
If I choose a different way of moving -- on the right for example --
449
1054260
3000
ืื ืื‘ื—ืจ ื‘ื“ืจืš ืื—ืจืช ืฉืœ ืชื ื•ืขื” -- ื”ื™ืžื ื™ืช ืœื“ื•ื’ืžื --
17:37
then I'll have a different command, different noise,
450
1057260
2000
ืชื”ื™ื” ืœื™ ืคืงื•ื“ื” ืื—ืจืช, ืจืขืฉ ืื—ืจ.
17:39
playing through a noisy system, very complicated.
451
1059260
3000
ืžืื•ื“ ืžืกื•ื‘ืš ืœืชืคืงื“ ื‘ืžืขืจื›ืช ืจื•ืขืฉืช.
17:42
All we can be sure of is the variability will be different.
452
1062260
3000
ื ื™ืชืŸ ืจืง ืœื”ื™ื•ืช ื‘ื˜ื•ื— ืฉื”ื”ืฉืชื ื•ืช ืชื”ื™ื” ืฉื•ื ื”.
17:45
If I move in this particular way,
453
1065260
2000
ืื ืื ื•ืข ื‘ืžืกืœื•ืœ ืžืกื•ื™ื™ื ื–ื”,
17:47
I end up with a smaller variability across many movements.
454
1067260
3000
ืืกื™ื™ื ื‘ื”ืฉืชื ื•ืช ื™ื•ืชืจ ื ืžื•ื›ื” ืœืื—ืจ ืชื ื•ืขื•ืช ืจื‘ื•ืช.
17:50
So if I have to choose between those two,
455
1070260
2000
ืœื›ืŸ ืื ืขืœื™ื™ ืœื‘ื—ื•ืจ ื‘ื™ืŸ ืฉืชื™ ืืœื•,
17:52
I would choose the right one because it's less variable.
456
1072260
2000
ืื‘ื—ืจ ืืช ื”ื™ืžื ื™ืช ื›ื™ ื”ื™ื ืคื—ื•ืช ืžืฉืชื ื”.
17:54
And the fundamental idea
457
1074260
2000
ื”ืจืขื™ื•ืŸ ื”ื‘ืกื™ืกื™ ื”ื•ื
17:56
is you want to plan your movements
458
1076260
2000
ืฉืจื•ืฆื™ื ืœืชื›ื ืŸ ืืช ื”ืชื ื•ืขื•ืช ื›ืš
17:58
so as to minimize the negative consequence of the noise.
459
1078260
3000
ืฉื”ื”ืฉืคืขื” ื”ืฉืœื™ืœื™ืช ืฉืœ ื”ืจืขืฉ ืชื”ื™ื” ืžื™ื ื™ืžืœื™ืช.
18:01
And one intuition to get
460
1081260
2000
ืชื•ื‘ื ื” ืื—ืช ืฉืžืชืงื‘ืœืช ื”ื™ื
18:03
is actually the amount of noise or variability I show here
461
1083260
2000
ืฉื›ืžื•ืช ื”ืจืขืฉ ืื• ื”ื”ืฉืชื ื•ืช ืฉืื ื™ ืžืจืื” ื›ืืŸ
18:05
gets bigger as the force gets bigger.
462
1085260
2000
ืขื•ืœื” ื›ื›ืœ ืฉื”ื›ื•ื— ืžืชื’ื‘ืจ.
18:07
So you want to avoid big forces as one principle.
463
1087260
3000
ืœื›ืŸ ื›ื“ืื™ ืœื”ื™ืžื ืข ืžื›ื•ื—ื•ืช ื’ื“ื•ืœื™ื ื‘ืชื•ืจ ืขื™ืงืจื•ืŸ ืจืืฉื•ืŸ.
18:10
So we've shown that using this,
464
1090260
2000
ื”ืจืื ื• ืฉื‘ืืžืฆืขื•ืช ื–ื”,
18:12
we can explain a huge amount of data --
465
1092260
2000
ื ื™ืชืŸ ืœื”ืกื‘ื™ืจ ื›ืžื•ืช ืื“ื™ืจื” ืฉืœ ื ืชื•ื ื™ื --
18:14
that exactly people are going about their lives planning movements
466
1094260
3000
ืฉืื ืฉื™ื ืžืชื ื”ืœื™ื ื‘ื—ื™ื™ื”ื ื•ืžืชื›ื ื ื™ื ืชื ื•ืขื•ืช
18:17
so as to minimize negative consequences of noise.
467
1097260
3000
ื›ืš ืฉื”ื”ืฉืคืขื•ืช ื”ืฉืœื™ืœื™ื•ืช ืฉืœ ืจืขืฉ ื™ื”ื™ื• ืžื™ื ื™ืžืœื™ื•ืช.
18:20
So I hope I've convinced you the brain is there
468
1100260
2000
ืื ื™ ืžืงื•ื” ืฉืฉื™ื›ื ืขืชื™ ืืชื›ื ืฉื”ืžื•ื— ืงื™ื™ื
18:22
and evolved to control movement.
469
1102260
2000
ื•ื”ืชืคืชื— ื›ื“ื™ ืœืฉืœื•ื˜ ื‘ืชื ื•ืขื”.
18:24
And it's an intellectual challenge to understand how we do that.
470
1104260
3000
ื•ื–ื” ืืชื’ืจ ืžื—ืฉื‘ืชื™ ืœื”ื‘ื™ืŸ ื›ื™ืฆื“ ืื ื• ืžืฉื™ื’ื™ื ื–ืืช.
18:27
But it's also relevant
471
1107260
2000
ืื‘ืœ ื–ื” ื’ื ื ื•ื’ืข
18:29
for disease and rehabilitation.
472
1109260
2000
ืœืžื—ืœื•ืช ื•ืฉื™ืงื•ื.
18:31
There are many diseases which effect movement.
473
1111260
3000
ื™ืฉื ืŸ ื”ืจื‘ื” ืžื—ืœื•ืช ื”ืžืฉืคื™ืขื•ืช ืขืœ ืชื ื•ืขื”.
18:34
And hopefully if we understand how we control movement,
474
1114260
2000
ื›ื•ืœื™ ืชืงื•ื” ืฉืื ื ื‘ื™ืŸ ื›ื™ืฆื“ ืฉื•ืœื˜ื™ื ื‘ืชื ื•ืขื”,
18:36
we can apply that to robotic technology.
475
1116260
2000
ื ื•ื›ืœ ืœื™ื™ืฉื ื–ืืช ืขืœ ื˜ื›ื ื•ืœื•ื’ื™ื” ืจื•ื‘ื•ื˜ื™ืช.
18:38
And finally, I want to remind you,
476
1118260
2000
ืœื‘ืกื•ืฃ, ื‘ืจืฆื•ื ื™ ืœื”ื–ื›ื™ืจื›ื, ืฉื›ืืฉืจ ืืชื
18:40
when you see animals do what look like very simple tasks,
477
1120260
2000
ืจื•ืื™ื ื—ื™ื•ืช ื”ืขื•ืฉื•ืช ืžื” ืฉื ืจืื” ื›ืžืฉื™ืžื” ืคืฉื•ื˜ื”,
18:42
the actual complexity of what is going on inside their brain
478
1122260
2000
ื”ืžื•ืจื›ื‘ื•ืช ื‘ืคื•ืขืœ ืฉืœ ืžื” ืฉืžืชืจื—ืฉ ื‘ืžื•ื—ืŸ
18:44
is really quite dramatic.
479
1124260
2000
ื”ื™ื ื‘ืืžืช ืžื“ื”ื™ืžื”.
18:46
Thank you very much.
480
1126260
2000
ืชื•ื“ื” ืจื‘ื” ืœื›ื.
18:48
(Applause)
481
1128260
8000
(ืžื—ื™ืื•ืช ื›ืคื™ื™ื)
18:56
Chris Anderson: Quick question for you, Dan.
482
1136260
2000
ื›ืจื™ืก ืื ื“ืจืกื•ืŸ: ืฉืืœื” ืงืฆืจื” ื“ืŸ.
18:58
So you're a movement -- (DW: Chauvinist.) -- chauvinist.
483
1138260
4000
ื•ื‘ื›ืŸ ืืชื” ืงื ืื™ ืœืชื ื•ืขื” -- (ื“.ื•.: ืงื ืื™) .
19:02
Does that mean that you think that the other things we think our brains are about --
484
1142260
3000
ื”ืื ื–ื” ืื•ืžืจ ืฉืœืคื™ ื“ืขืชืš ื”ื“ื‘ืจื™ื ื”ืื—ืจื™ื ืฉืื ื• ืกื‘ื•ืจื™ื ืฉื”ืžื•ื— ืงื™ื™ื ืขื‘ื•ืจื --
19:05
the dreaming, the yearning, the falling in love and all these things --
485
1145260
3000
ื”ื—ืœื•ืžื•ืช, ื”ื’ืขื’ื•ืข, ื”ื”ืชืื”ื‘ื•ืช ื•ื“ื‘ืจื™ื ื›ืืœื” --
19:08
are a kind of side show, an accident?
486
1148260
3000
ื”ื ืžื™ืŸ ืžื•ืคืข ืฆื“ื“ื™, ืžืงืจื™ื•ืช?
19:11
DW: No, no, actually I think they're all important
487
1151260
2000
ื“.ื•.: ืœื, ืœื, ืœืžืขืฉื” ืื ื™ ืกื‘ื•ืจ ืฉื”ื ื›ื•ืœื ื—ืฉื•ื‘ื™ื
19:13
to drive the right movement behavior to get reproduction in the end.
488
1153260
3000
ื›ื“ื™ ืœื›ื•ื•ืŸ ืœื”ืชื ื”ื’ื•ืช ืชื ื•ืขืชื™ืช ื ื›ื•ื ื” ืœืžืขืŸ ื”ืจื‘ื™ื™ื” ื‘ืกื•ืฃ.
19:16
So I think people who study sensation or memory
489
1156260
3000
ืœื›ืŸ ืื ื™ ื—ื•ืฉื‘ ืฉืื ืฉื™ื ื”ื—ื•ืงืจื™ื ืชื—ื•ืฉื•ืช ืื• ื–ื™ื›ืจื•ืŸ
19:19
without realizing why you're laying down memories of childhood.
490
1159260
2000
ืžื‘ืœื™ ืœื’ืœื•ืช ืžื“ื•ืข ืื ื• ืฉื•ื›ื—ื™ื ื–ื™ื›ืจื•ื ื•ืช ื™ืœื“ื•ืช.
19:21
The fact that we forget most of our childhood, for example,
491
1161260
3000
ื”ืขื•ื‘ื“ื” ืฉืื ื• ืฉื•ื›ื—ื™ื ืืช ืจื•ื‘ ื™ืœื“ื•ืชื™ื ื•, ืœื“ื•ื’ืžื,
19:24
is probably fine, because it doesn't effect our movements later in life.
492
1164260
3000
ื–ื” ื›ื ืจืื” ื‘ืกื“ืจ ื’ืžื•ืจ, ื›ื™ ื–ื” ืœื ืžืฉืคื™ืข ืขืœ ืชื ื•ืขื•ืชื™ื ื• ื‘ืฉืœื‘ ื™ื•ืชืจ ืžืื•ื—ืจ ื‘ื—ื™ื™ื.
19:27
You only need to store things which are really going to effect movement.
493
1167260
3000
ืื ื• ืจืง ื–ืงื•ืงื™ื ืœืื™ื—ืกื•ืŸ ืฉืœ ืžื” ืฉื‘ืืžืช ื”ื•ืœืš ืœื”ืฉืคื™ืข ืขืœ ืชื ื•ืขื”.
19:30
CA: So you think that people thinking about the brain, and consciousness generally,
494
1170260
3000
ื›.ื.: ืื– ืืชื” ืกื‘ื•ืจ ืฉืื ืฉื™ื ื”ื—ื•ืฉื‘ื™ื ืขืœ ื”ืžื•ื— ื•ืชื•ื“ืขื”,
19:33
could get real insight
495
1173260
2000
ื™ื›ื•ืœื™ื ืœื”ื’ื™ืข ืœืชื•ื‘ื ื•ืช ืžืžืฉื™ื•ืช
19:35
by saying, where does movement play in this game?
496
1175260
2000
ืื ื™ื—ืฉื‘ื• ืื™ืš ื”ืชื ื•ืขื” ืžืžืœืืช ื›ืืŸ ืชืคืงื™ื“?
19:37
DW: So people have found out for example
497
1177260
2000
ื“.ื•.: ื›ืš ืื ืฉื™ื ื’ื™ืœื• ืœื“ื•ื’ืžื ืฉื—ืงืจ ื”ืจืื™ื”
19:39
that studying vision in the absence of realizing why you have vision
498
1179260
2000
ืœืœื ื”ื‘ื ืช ื”ืกื™ื‘ื” ืœืงื™ื•ื ื”ืจืื™ื”
19:41
is a mistake.
499
1181260
2000
ื”ื•ื ืฉื’ื™ืื”.
19:43
You have to study vision with the realization
500
1183260
2000
ืฆืจื™ืš ืœื—ืงื•ืจ ืจืื™ื” ื‘ื”ืงืฉืจ ืฉืœ
19:45
of how the movement system is going to use vision.
501
1185260
2000
ื›ื™ืฆื“ ืžืขืจื›ืช ื”ืชื ื•ืขื” ืžืฉืชืžืฉืช ื‘ืจืื™ื”.
19:47
And it uses it very differently once you think about it that way.
502
1187260
2000
ื•ื”ื™ื ืžืฉืชืžืฉืช ื‘ื” ื‘ืื•ืคืŸ ืžืื•ื“ ืฉื•ื ื” ื›ืืฉืจ ื—ื•ืฉื‘ื™ื ืขืœื™ื” ื‘ื”ืงืฉืจ ื–ื”.
19:49
CA: Well that was quite fascinating. Thank you very much indeed.
503
1189260
3000
ื›.ื.: ื–ื” ืžืจืชืง. ืชื•ื“ื” ืจื‘ื” ืœืš.
19:52
(Applause)
504
1192260
2000
(ืžื—ื™ืื•ืช ื›ืคื™ื™ื)
ืขืœ ืืชืจ ื–ื”

ืืชืจ ื–ื” ื™ืฆื™ื’ ื‘ืคื ื™ื›ื ืกืจื˜ื•ื ื™ YouTube ื”ืžื•ืขื™ืœื™ื ืœืœื™ืžื•ื“ ืื ื’ืœื™ืช. ืชื•ื›ืœื• ืœืจืื•ืช ืฉื™ืขื•ืจื™ ืื ื’ืœื™ืช ื”ืžื•ืขื‘ืจื™ื ืขืœ ื™ื“ื™ ืžื•ืจื™ื ืžื”ืฉื•ืจื” ื”ืจืืฉื•ื ื” ืžืจื—ื‘ื™ ื”ืขื•ืœื. ืœื—ืฅ ืคืขืžื™ื™ื ืขืœ ื”ื›ืชื•ื‘ื™ื•ืช ื‘ืื ื’ืœื™ืช ื”ืžื•ืฆื’ื•ืช ื‘ื›ืœ ื“ืฃ ื•ื™ื“ืื• ื›ื“ื™ ืœื”ืคืขื™ืœ ืืช ื”ืกืจื˜ื•ืŸ ืžืฉื. ื”ื›ืชื•ื‘ื™ื•ืช ื’ื•ืœืœื•ืช ื‘ืกื ื›ืจื•ืŸ ืขื ื”ืคืขืœืช ื”ื•ื•ื™ื“ืื•. ืื ื™ืฉ ืœืš ื”ืขืจื•ืช ืื• ื‘ืงืฉื•ืช, ืื ื ืฆื•ืจ ืื™ืชื ื• ืงืฉืจ ื‘ืืžืฆืขื•ืช ื˜ื•ืคืก ื™ืฆื™ืจืช ืงืฉืจ ื–ื”.

https://forms.gle/WvT1wiN1qDtmnspy7