1
00:00:09,480 --> 00:00:10,945
Wow! That was great!

2
00:00:10,945 --> 00:00:13,896
My name is Audrey 
and I'm a neuroscientist at Caltech.

3
00:00:13,896 --> 00:00:17,752
And I want to tell you
about what I work on here at Caltech.

4
00:00:17,752 --> 00:00:19,755
So, what I work on in my lab

5
00:00:19,755 --> 00:00:21,505
is, we actually study sleep.

6
00:00:21,505 --> 00:00:22,759
Let me ask you a question:

7
00:00:22,759 --> 00:00:24,501
If you want to study the brain,

8
00:00:24,501 --> 00:00:25,843
do you want to look at it

9
00:00:25,843 --> 00:00:28,323
where it's kind of enclosed

10
00:00:28,323 --> 00:00:30,033
in a skull you can't see through?

11
00:00:30,033 --> 00:00:33,015
Or do you want to see it 
in a transparent box?

12
00:00:33,015 --> 00:00:34,767
I think it's a transparent box!

13
00:00:34,767 --> 00:00:37,239
So, one thing that's really good

14
00:00:37,239 --> 00:00:40,759
is that we actually have this 
already in nature.

15
00:00:40,759 --> 00:00:42,765
We have a brain in a glass box.

16
00:00:42,765 --> 00:00:44,409
We have something where we can see

17
00:00:44,409 --> 00:00:48,008
straight through an organism
and see their brain.

18
00:00:48,008 --> 00:00:50,286
What we can do 
is zoom in on their brain

19
00:00:50,286 --> 00:00:51,717
and look at these neurons.

20
00:00:51,717 --> 00:00:53,986
What I'm showing you here 
are neurons,

21
00:00:53,986 --> 00:00:56,513
and each one of them
is a different color.

22
00:00:56,513 --> 00:00:58,756
They're neurons in Technicolor.

23
00:00:58,996 --> 00:01:02,499
These neurons are hypocretin neurons.

24
00:01:02,499 --> 00:01:04,840
Does anyone know anybody with narcolepsy?

25
00:01:04,840 --> 00:01:07,497
Have you guys heard of narcolepsy?

26
00:01:07,497 --> 00:01:09,496
Narcolepsy is a sleep disorder

27
00:01:09,496 --> 00:01:12,653
where you're always kind of sleepy,

28
00:01:12,653 --> 00:01:14,996
and when you get stimuli that excite you,

29
00:01:14,996 --> 00:01:17,755
or something that is,
I guess, your brain food,

30
00:01:17,755 --> 00:01:19,460
instead of being excited,

31
00:01:19,460 --> 00:01:21,550
which is kind of the normal thing to do,

32
00:01:21,550 --> 00:01:23,484
you actually fall asleep.

33
00:01:23,484 --> 00:01:25,532
These neurons I'm showing you here

34
00:01:25,532 --> 00:01:28,505
are the neurons
that are involved in that.

35
00:01:28,505 --> 00:01:32,610
So, when we have all these neurons 
in different colors in the brain,

36
00:01:32,610 --> 00:01:33,973
the important thing is:

37
00:01:33,973 --> 00:01:35,805
How do they connect with one another?

38
00:01:35,805 --> 00:01:37,663
You have all these tracks--

39
00:01:37,663 --> 00:01:41,185
how many of you guys here 
have taken public transportation?

40
00:01:42,751 --> 00:01:44,888
When you have public transportation,

41
00:01:44,888 --> 00:01:47,644
what's important is 
where those points of contact are,

42
00:01:47,644 --> 00:01:49,936
and where all those 
different tracks lead.

43
00:01:49,936 --> 00:01:51,481
And so, similar to that,

44
00:01:51,481 --> 00:01:53,745
we are constructing a system map.

45
00:01:53,745 --> 00:01:55,741
One of the ways 
we want to do this,

46
00:01:55,741 --> 00:01:59,504
is to take cues from how
we interact with other humans.

47
00:01:59,504 --> 00:02:01,012
When we have another human,

48
00:02:01,012 --> 00:02:03,517
one thing that we do
is give a handshake.

49
00:02:03,517 --> 00:02:06,021
So what professor Cori Bargmann did

50
00:02:06,021 --> 00:02:09,682
is that she designed a way
to do a molecular grasp.

51
00:02:09,682 --> 00:02:11,609
So for example, I'm neuron Audrey,

52
00:02:11,609 --> 00:02:13,361
this is neuron Ella,

53
00:02:13,361 --> 00:02:17,509
and we each have a glove
that's not lighting up.

54
00:02:17,509 --> 00:02:19,502
But when we make contact,

55
00:02:19,502 --> 00:02:23,255
we actually have our gloves lighting up.

56
00:02:23,255 --> 00:02:25,044
And when we are further away,

57
00:02:25,044 --> 00:02:27,510
the lights go off.

58
00:02:27,510 --> 00:02:29,523
So, this is a way to figure out

59
00:02:29,523 --> 00:02:31,759
how those neurons
are connecting to one other,

60
00:02:31,759 --> 00:02:33,000
and we can actually see...

61
00:02:33,000 --> 00:02:35,279
(Audience laughs at off-camera event)

62
00:02:35,279 --> 00:02:37,248
[Cori and her friends
Molecular Grasp]

63
00:02:37,248 --> 00:02:39,988
...we see when the neurons
are talking to each other,

64
00:02:39,988 --> 00:02:41,875
when they're close enough
to one other,

65
00:02:41,875 --> 00:02:44,223
and when they're further apart 
from each other.

66
00:02:44,231 --> 00:02:47,906
I want to talk about
my motivation to study science.

67
00:02:47,906 --> 00:02:49,369
I want to study science

68
00:02:49,369 --> 00:02:52,867
because I like to test ideas
and observe what happens.

69
00:02:53,487 --> 00:02:57,301
What we want to do is make sure 
we're making a fair comparison.

70
00:02:57,301 --> 00:03:01,517
When we have an apple,
we want to compare it to an apple.

71
00:03:01,527 --> 00:03:03,917
When you make a comparison,
you want to make sure

72
00:03:03,917 --> 00:03:06,004
you're comparing
an orange to an orange.

73
00:03:06,004 --> 00:03:10,434
So in controlled conditions,
we have lots of different controls.

74
00:03:10,434 --> 00:03:12,776
We have positive controls 
and negative controls;

75
00:03:12,776 --> 00:03:14,768
you have positive controls
to make sure

76
00:03:14,768 --> 00:03:18,022
when you're actually testing something,
you can see an effect.

77
00:03:18,026 --> 00:03:21,499
You have negative controls
to make sure you don't have auto-activation,

78
00:03:21,499 --> 00:03:23,833
so that when you don't have introduction

79
00:03:23,833 --> 00:03:27,176
of the experimental condition
or stimuli you're testing,

80
00:03:27,176 --> 00:03:29,247
it's not going to just go off.

81
00:03:29,247 --> 00:03:32,256
And then you have
your experimental condition.

82
00:03:32,256 --> 00:03:35,505
And when you design
an experiment,

83
00:03:35,505 --> 00:03:37,517
you have lots of positive controls,

84
00:03:37,517 --> 00:03:39,644
you have lots of negative controls,

85
00:03:39,644 --> 00:03:42,234
and you have your experimental condition.

86
00:03:42,234 --> 00:03:44,760
One trend that we're seeing 
in neuroscience

87
00:03:44,760 --> 00:03:49,497
is that we can observe
and test our idea.

88
00:03:49,497 --> 00:03:50,983
And what we scientists are,

89
00:03:50,983 --> 00:03:53,997
is we're kind of like little ninjas,

90
00:03:53,997 --> 00:03:57,505
just quietly looking at the brain.

91
00:03:57,505 --> 00:04:00,260
And that's what
Alex and his friends have done.

92
00:04:00,260 --> 00:04:03,745
What they do is take these brains
that are in glass boxes,

93
00:04:03,745 --> 00:04:05,777
and the brains are
swimming, or sleeping,

94
00:04:05,777 --> 00:04:07,417
or doing whatever they want,

95
00:04:07,417 --> 00:04:10,093
and the team, meanwhile,
captures the neural activity.

96
00:04:10,093 --> 00:04:11,443
The way they do this

97
00:04:11,443 --> 00:04:13,490
is that they've
genetically engineered

98
00:04:13,490 --> 00:04:17,523
each of these neurons
to give off a photon of light,

99
00:04:17,523 --> 00:04:19,476
and the more active
these brains are,

100
00:04:19,476 --> 00:04:22,008
the more photons come out,

101
00:04:22,008 --> 00:04:24,798
so you see a greater luminescence.

102
00:04:24,808 --> 00:04:28,249
And so, while these fish
are doing their regular business,

103
00:04:28,249 --> 00:04:31,789
while these brains in glass boxes
are doing their regular business,

104
00:04:31,789 --> 00:04:35,787
and we can observe the neural activity.

105
00:04:35,787 --> 00:04:38,518
Let's say we want to take this
a little bit further,

106
00:04:38,518 --> 00:04:41,284
to the entire human population.

107
00:04:41,284 --> 00:04:45,759
What if we look at
the entire human population,

108
00:04:45,759 --> 00:04:47,516
in your native environment,

109
00:04:47,516 --> 00:04:51,756
and let's say we take away
those positive and negative controls

110
00:04:51,756 --> 00:04:54,536
and just have experimental conditions?

111
00:04:54,536 --> 00:04:57,331
What if we have something 
like a wiki log,

112
00:04:57,331 --> 00:04:59,131
a wiki lab notebook,

113
00:04:59,131 --> 00:05:02,149
where everyone contributes their ideas,

114
00:05:02,149 --> 00:05:06,307
and as people are contributing,
everyone writes over one another,

115
00:05:06,307 --> 00:05:08,510
kind of like Wikipedia.
Who would have thought

116
00:05:08,510 --> 00:05:10,957
that WIkipedia would have
taken off the way it did,

117
00:05:10,957 --> 00:05:13,904
with all its success?
But it did.

118
00:05:13,904 --> 00:05:16,008
And what if we could
apply it to science?

119
00:05:16,008 --> 00:05:18,852
It's slightly different than
how we think about things now.

120
00:05:18,852 --> 00:05:21,030
We have these apples and oranges,

121
00:05:21,030 --> 00:05:23,765
and usually we can't compare
them to one another.

122
00:05:23,765 --> 00:05:26,064
But now we have 
all these different conditions,

123
00:05:26,064 --> 00:05:28,007
and things like maybe weird fruits,

124
00:05:28,007 --> 00:05:29,990
like dragon fruit,

125
00:05:29,990 --> 00:05:31,999
and we want to compare those.

126
00:05:31,999 --> 00:05:34,250
So, I want to give you a challenge:

127
00:05:34,250 --> 00:05:36,886
to be the ones,
to be the mathematicians

128
00:05:36,886 --> 00:05:39,085
and the statisticians to design


129
00:05:39,085 --> 00:05:40,733
these types of algorithms;

130
00:05:40,733 --> 00:05:42,517
and to decide whether or not

131
00:05:42,517 --> 00:05:43,785
we can extrapolate trends

132
00:05:43,785 --> 00:05:47,623
and figure out the observations
without all those controls.

133
00:05:47,623 --> 00:05:50,152
So, change the way we do science.
Go change the world.

134
00:05:50,152 --> 00:05:51,436
Thanks!

135
00:05:51,436 --> 00:05:54,104
(Applause)