1
00:00:11,495 --> 00:00:13,095
Last December,

2
00:00:13,095 --> 00:00:16,245
me and my fellow Nobel Laureates
were asked by a journalist

3
00:00:16,245 --> 00:00:19,825
if there was one thing
that we could teach the world,

4
00:00:19,825 --> 00:00:21,185
what would it be?

5
00:00:21,185 --> 00:00:22,135
And to my surprise,

6
00:00:22,135 --> 00:00:27,645
two economists, two biologists,
a chemist, and three physicists

7
00:00:27,645 --> 00:00:29,535
gave the same answer.

8
00:00:29,535 --> 00:00:33,085
And that answer was about uncertainty.

9
00:00:33,085 --> 00:00:36,395
So I'm going to talk to you today
about uncertainty.

10
00:00:37,382 --> 00:00:42,762
To understand anything,
you must understand its uncertainty.

11
00:00:42,762 --> 00:00:47,812
Uncertainty is at the heart
of the fabric of the Universe.

12
00:00:47,812 --> 00:00:51,222
I'm going to illustrate this with a laser.

13
00:00:51,894 --> 00:00:53,894
A laser puts out a small,

14
00:00:53,894 --> 00:00:57,894
but not infinitesimally small
point of light.

15
00:00:58,312 --> 00:01:00,992
You might think that if I go through

16
00:01:00,992 --> 00:01:04,762
and I try to make
that point of light smaller

17
00:01:04,762 --> 00:01:08,762
by, for example,
bringing two jars of a slit together,

18
00:01:08,762 --> 00:01:11,122
that I could make that point
as small as I want.

19
00:01:11,122 --> 00:01:15,332
I just want to make
those slits closer and closer.

20
00:01:15,712 --> 00:01:18,672
So let's see what happens
when I do this for real.

21
00:01:20,122 --> 00:01:22,952
My friends at Mount Stromlo gave a call

22
00:01:22,952 --> 00:01:26,952
and made up a nice little invent,
a little here.

23
00:01:28,084 --> 00:01:33,234
By essentially adjusting
the laser, the slit -

24
00:01:33,234 --> 00:01:36,174
we're going to go through
and we are going to see what happens

25
00:01:36,174 --> 00:01:38,274
when I close the jaws of the slit.

26
00:01:38,765 --> 00:01:40,345
The more I close it,

27
00:01:41,885 --> 00:01:45,885
instead of getting smaller,
the laser gets spread out.

28
00:01:47,487 --> 00:01:51,487
So it works exactly the opposite
of what I was expecting.

29
00:01:52,247 --> 00:01:56,247
And that's due to something known
as Heisenberg's Uncertainty Principle.

30
00:01:57,275 --> 00:01:59,285
Heisenberg's Uncertainty Principle states

31
00:01:59,285 --> 00:02:03,285
that you can't know exactly
where something is

32
00:02:03,285 --> 00:02:06,645
and know its momentum at the same time.

33
00:02:06,645 --> 00:02:10,285
Light's momentum is really its direction.

34
00:02:10,741 --> 00:02:15,551
So, as I bring those slits
closer and closer together,

35
00:02:16,263 --> 00:02:19,123
I actually constrain where the light is.

36
00:02:19,752 --> 00:02:24,642
But the quantum world says
you can't do that.

37
00:02:24,642 --> 00:02:26,802
The light then has an uncertain direction.

38
00:02:26,802 --> 00:02:32,262
So instead of being a smaller point,
the light has a randomness put out to it,

39
00:02:32,262 --> 00:02:34,762
which is that pattern that we saw.

40
00:02:37,230 --> 00:02:42,340
Many things in life you can think of
as a series of little decisions.

41
00:02:42,340 --> 00:02:46,662
For example, if I start at a point,
and I can go left or right,

42
00:02:46,662 --> 00:02:50,782
well, I do it, let's say, 50% of the time
I can go left or right.

43
00:02:51,369 --> 00:02:55,519
Let's say I have another decision tree
down below that.

44
00:02:55,519 --> 00:03:00,619
I can go left, I can go right,
or I can go to the middle.

45
00:03:00,619 --> 00:03:04,229
Because I've had two chances
to go to the middle from above,

46
00:03:04,229 --> 00:03:07,029
I would do that 50% of the time.

47
00:03:07,029 --> 00:03:10,779
I only go one quarter to the left
and one quarter all the way to the right.

48
00:03:10,779 --> 00:03:14,429
And you can build up such a decision tree,
and Pascal did this.

49
00:03:14,429 --> 00:03:16,549
It's called Pascal's triangle.

50
00:03:16,549 --> 00:03:20,549
You get a probability
of where you are going to end up.

51
00:03:20,549 --> 00:03:23,369
I brought something like this
with me today.

52
00:03:24,698 --> 00:03:27,668
It's this machine right here.

53
00:03:28,759 --> 00:03:33,079
This is a machine you can put balls into
and you can randomly see what happens.

54
00:03:33,079 --> 00:03:36,469
So, for example, if I put a ball in here,

55
00:03:36,469 --> 00:03:40,039
it'll bounce down
and it'll end up somewhere.

56
00:03:40,039 --> 00:03:43,469
It's essentially an enactment
of Pascal's triangle.

57
00:03:43,469 --> 00:03:45,719
I need two people
from the audience to help me,

58
00:03:45,719 --> 00:03:49,169
and I think I am going to have
Sly and Jon right there

59
00:03:49,169 --> 00:03:51,079
come up and help me if that's okay.

60
00:03:51,079 --> 00:03:52,299
You know who you are.

61
00:03:52,299 --> 00:03:53,949
(Laughter)

62
00:03:53,949 --> 00:03:55,359
What they are going to do

63
00:03:55,359 --> 00:03:58,089
is they are going to, as fast as they can,

64
00:03:58,089 --> 00:04:01,689
- faster than they are going right now,
because I only have 18 minutes -

65
00:04:01,689 --> 00:04:02,559
(Laughter)

66
00:04:02,559 --> 00:04:06,689
put balls through this machine,
and we're going to see what happens.

67
00:04:07,507 --> 00:04:09,747
This machine counts things where they end.

68
00:04:09,747 --> 00:04:13,237
So you guys have to go through
as fast as you can.

69
00:04:13,237 --> 00:04:14,517
Work together,

70
00:04:14,517 --> 00:04:17,567
and during the rest of my talk,
you are going to build up this.

71
00:04:17,567 --> 00:04:20,727
And the more you do,
the better it is, okay?

72
00:04:20,727 --> 00:04:24,357
So go for it, and I'll keep on going.

73
00:04:24,357 --> 00:04:25,127
(Laughter)

74
00:04:25,127 --> 00:04:25,887
Alright.

75
00:04:26,307 --> 00:04:30,307
It turns out that if you have
a series of random events in life,

76
00:04:30,307 --> 00:04:33,667
you end up with something
called a Bell Shaped Curve,

77
00:04:33,667 --> 00:04:38,157
which we also call a Normal Distribution
or Gaussian Distribution.

78
00:04:39,027 --> 00:04:42,197
So, for example, if you have
just a few random events,

79
00:04:42,197 --> 00:04:44,677
you don't get something
that really looks like that.

80
00:04:44,677 --> 00:04:46,027
But if you do more and more,

81
00:04:46,027 --> 00:04:49,707
they add up to give you
this very characteristic pattern

82
00:04:49,707 --> 00:04:53,677
which Gauss famously
wrote down mathematically.

83
00:04:53,677 --> 00:04:55,447
It turns out that in most cases

84
00:04:55,447 --> 00:05:00,607
a series of random events
gives you this bell-shaped curve.

85
00:05:02,259 --> 00:05:04,059
It doesn't really matter what it is.

86
00:05:04,059 --> 00:05:05,879
For example, if I were going to go out

87
00:05:05,879 --> 00:05:10,469
and have a million scales across Australia

88
00:05:10,469 --> 00:05:12,649
measure my weight.

89
00:05:12,649 --> 00:05:14,469
Well, there's some randomness to that,

90
00:05:14,469 --> 00:05:19,119
and you'll get a bell-shaped curve
of what my weight actually is.

91
00:05:19,747 --> 00:05:22,267
If I were instead to go through

92
00:05:22,267 --> 00:05:25,287
and ask a million Australian males
what their weight is,

93
00:05:25,287 --> 00:05:26,467
and actually measure it,

94
00:05:26,467 --> 00:05:29,477
I would also get a bell-shaped curve,

95
00:05:29,477 --> 00:05:32,187
because that is also made up
of a series of random events

96
00:05:32,187 --> 00:05:34,207
which determine people's weight.

97
00:05:34,914 --> 00:05:38,314
So the way a bell-shaped curve
is characterized

98
00:05:38,314 --> 00:05:42,004
is by its mean
- that's the most likely value -

99
00:05:42,004 --> 00:05:46,474
and its width, which we call
a standard deviation.

100
00:05:47,121 --> 00:05:49,311
This is a very important concept

101
00:05:49,311 --> 00:05:53,071
because the width
and how close you are to the mean

102
00:05:53,071 --> 00:05:54,361
you can characterize,

103
00:05:54,361 --> 00:05:57,201
so the likelihood of things is occurring.

104
00:05:57,795 --> 00:06:01,905
So it turns out if you are within
one standard deviation,

105
00:06:01,905 --> 00:06:06,045
that happens 68.3% of the time.

106
00:06:06,045 --> 00:06:10,002
I'm going to illustrate how this works
for work example in just a second.

107
00:06:10,729 --> 00:06:14,519
If you have two standard deviations,
that happens 95.4% of the time;

108
00:06:14,519 --> 00:06:15,779
you're within two.

109
00:06:15,779 --> 00:06:19,759
99.73% within three standard deviations.

110
00:06:19,759 --> 00:06:24,829
This is a very powerful way for us
to describe things in the world.

111
00:06:24,829 --> 00:06:28,769
So, it turns out this means
that I can go out

112
00:06:28,769 --> 00:06:31,569
and make a measurement of, for example,

113
00:06:31,569 --> 00:06:33,249
how much I weigh,

114
00:06:33,249 --> 00:06:36,319
and if I use more and more
scales in Australia,

115
00:06:36,319 --> 00:06:39,199
I will get a better and better answer,

116
00:06:39,199 --> 00:06:41,179
provided they are good scales.

117
00:06:41,731 --> 00:06:46,531
It turns out the more trials I do,
or the more measurements I make,

118
00:06:46,531 --> 00:06:49,421
the better I will make that measurement.

119
00:06:49,421 --> 00:06:51,661
And the accuracy increases

120
00:06:51,661 --> 00:06:56,601
as the square root of the number
of times I make the measurement.

121
00:06:56,601 --> 00:06:59,281
That's why I am having these guys 
do what they are doing

122
00:06:59,281 --> 00:07:00,321
as fast as they can.

123
00:07:00,321 --> 00:07:01,141
(Laughter)

124
00:07:01,141 --> 00:07:04,851
So let's apply this
to a real world problem we all see:

125
00:07:04,851 --> 00:07:08,061
the approval rating
of the Prime Minister of Australia.

126
00:07:08,739 --> 00:07:10,629
Over the past 15 months,

127
00:07:10,629 --> 00:07:13,859
every couple of weeks, we hear news poll

128
00:07:13,859 --> 00:07:19,269
go out and ask the people of Australia:
"Do you approve of the Prime Minister?"

129
00:07:19,269 --> 00:07:20,427
Over the last 15 months,

130
00:07:20,427 --> 00:07:24,097
they have done this 28 times,
and they asked 1100 people.

131
00:07:24,097 --> 00:07:26,887
They don't ask
about 22 million Australians

132
00:07:26,887 --> 00:07:28,707
because it's too expensive to do that,

133
00:07:28,707 --> 00:07:30,557
so they ask 1100 people.

134
00:07:30,557 --> 00:07:33,087
The square root of 1100 is 33,

135
00:07:33,087 --> 00:07:36,317
and so it turns out
their answers are uncertain

136
00:07:36,317 --> 00:07:40,537
by plus or minus 33 people
when they asked these 1100 people.

137
00:07:40,537 --> 00:07:44,937
That's a 3% error.
That's 33 divided by 1100.

138
00:07:44,937 --> 00:07:46,930
So let's see what they get.

139
00:07:46,930 --> 00:07:49,290
Here is last fifteen months.

140
00:07:49,290 --> 00:07:52,870
You can see it seems that some time
in the middle of the last year

141
00:07:52,870 --> 00:07:55,820
the Prime Minister had a very bad week,

142
00:07:55,820 --> 00:08:00,250
followed a few weeks later
by what appears to be a very good week.

143
00:08:00,968 --> 00:08:04,568
Of course, you could look at it
in another way.

144
00:08:04,568 --> 00:08:08,028
You could say, "What would happened
if the Prime Minister's popularity

145
00:08:08,028 --> 00:08:12,908
hasn't changed at all
in the last fifteen months?"

146
00:08:12,908 --> 00:08:14,608
Well, then there's an average,

147
00:08:14,608 --> 00:08:19,548
and that mean turns out
to be 29.6% for this set of polls.

148
00:08:19,548 --> 00:08:23,338
So she hasn't been very popular
over the last 15 months.

149
00:08:23,338 --> 00:08:27,599
And we know that, if a basis bell curve,

150
00:08:27,599 --> 00:08:31,039
that's 68.3% of the time,

151
00:08:31,039 --> 00:08:34,489
it should lie within plus or minus 3%,

152
00:08:34,489 --> 00:08:38,489
because of the number
of people we're asking.

153
00:08:38,489 --> 00:08:43,639
So that means we expect it turns out
between 15 and 23 of the time.

154
00:08:43,639 --> 00:08:46,809
So it should lie within plus or minus 3%.

155
00:08:46,809 --> 00:08:49,899
And the actual number of times is 24.

156
00:08:51,064 --> 00:08:52,874
What about those really extreme cases

157
00:08:52,874 --> 00:08:56,764
when she seems to have
a really bad or really good week?

158
00:08:56,764 --> 00:09:01,604
Well, you actually expect
zero to two times,

159
00:09:01,604 --> 00:09:03,274
so 5% of the time,

160
00:09:03,274 --> 00:09:06,814
to be more than 6% discrepant
from the mean.

161
00:09:06,814 --> 00:09:08,114
And what do we see?

162
00:09:08,754 --> 00:09:09,543
Two.

163
00:09:09,876 --> 00:09:12,576
In other words, over the last 15 months

164
00:09:12,576 --> 00:09:15,986
the polls are completely consistent

165
00:09:15,986 --> 00:09:20,236
with the Prime Minister's popularity
not changing a bit.

166
00:09:21,763 --> 00:09:25,273
Alright. And let's see what the news is.

167
00:09:25,273 --> 00:09:27,393
For example, just last week.

168
00:09:27,393 --> 00:09:31,203
Well, approval rating,
big headline in the Australians,

169
00:09:31,203 --> 00:09:33,763
dropped from 29 to 27%,

170
00:09:33,763 --> 00:09:38,823
even though the error on that
is at least 3% even for that single poll.

171
00:09:38,823 --> 00:09:40,733
It's not just Australia that does this;

172
00:09:40,733 --> 00:09:43,403
it's all the news agencies.

173
00:09:44,756 --> 00:09:46,546
Now, the other thing is that

174
00:09:46,546 --> 00:09:48,996
news polls are not the only people
who do this.

175
00:09:48,996 --> 00:09:51,856
For example, Nielsen
does this for Fairfax,

176
00:09:51,856 --> 00:09:53,876
and here are their polls.

177
00:09:53,876 --> 00:09:55,616
Same question,

178
00:09:55,616 --> 00:09:59,016
and you'll see that it seems
that they are also consistent

179
00:09:59,016 --> 00:10:03,286
with the Prime Minister's popularity 
not changing over time.

180
00:10:03,286 --> 00:10:05,216
But they seem to get a different answer.

181
00:10:05,216 --> 00:10:09,216
They get 36.5% approval over that period.

182
00:10:09,778 --> 00:10:12,788
We are not talking about 1,000 people here

183
00:10:12,788 --> 00:10:14,948
when we compare these two things.

184
00:10:14,948 --> 00:10:16,778
We're talking about 30,000,

185
00:10:16,778 --> 00:10:19,248
because we get to add up all those people.

186
00:10:19,248 --> 00:10:23,248
So, the uncertainty in these measurement
is well less than 1%,

187
00:10:23,248 --> 00:10:26,328
and yet they disagree by 6%.

188
00:10:26,328 --> 00:10:29,438
That's because
not all uncertainty is random.

189
00:10:29,438 --> 00:10:34,208
It can be done to just make
mistakes or errors.

190
00:10:34,208 --> 00:10:39,078
It turns out it really hard to ask
1,100 people across Australia

191
00:10:39,078 --> 00:10:42,818
who are representative
of the average Australian.

192
00:10:42,818 --> 00:10:46,728
So, there is an additional uncertainty
caused by just error

193
00:10:46,728 --> 00:10:51,788
that is making a scientific
or a polling error which we see here.

194
00:10:51,788 --> 00:10:53,778
You might ask yourself,

195
00:10:53,778 --> 00:10:57,278
"Why don't they just ask more people,
like 10,000 people,

196
00:10:57,278 --> 00:10:59,958
less frequently, once a month?"

197
00:10:59,958 --> 00:11:00,968
And a cynic might say

198
00:11:00,968 --> 00:11:06,328
because there’s no news in telling people
that the popularity is the same

199
00:11:06,328 --> 00:11:07,958
month after month after month.

200
00:11:07,958 --> 00:11:10,038
(Laughter)

201
00:11:10,038 --> 00:11:10,988
Alright.

202
00:11:10,988 --> 00:11:13,978
Not all things, though,
become more accurate

203
00:11:13,978 --> 00:11:15,548
the more you measure them,

204
00:11:15,548 --> 00:11:20,268
and such systems we call
as exhibiting chaotic behavior.

205
00:11:20,268 --> 00:11:24,518
I happen to have something
that exhibits chaotic behavior here,

206
00:11:24,518 --> 00:11:27,528
which is a double pendulum.

207
00:11:27,528 --> 00:11:28,708
A double pendulum -

208
00:11:28,708 --> 00:11:32,928
this was made up for the people
by me at Questacon,

209
00:11:32,928 --> 00:11:34,808
and I thank them for that.

210
00:11:35,492 --> 00:11:38,732
A double pendulum is just two pendulums
connected to each other.

211
00:11:38,732 --> 00:11:42,522
And the beautiful thing is
this doesn't always exhibit chaos.

212
00:11:42,522 --> 00:11:44,162
Let me show you what happens here.

213
00:11:44,162 --> 00:11:45,562
If I just start this thing,

214
00:11:45,562 --> 00:11:49,482
these things swing
back and forth in unison

215
00:11:49,482 --> 00:11:51,602
because there is no chaos here.

216
00:11:51,602 --> 00:11:54,612
If I make measurements,
better and better measurements,

217
00:11:54,612 --> 00:11:58,742
I can predict exactly
what is going on here.

218
00:11:58,742 --> 00:12:00,952
The better I do, the better I will know

219
00:12:00,952 --> 00:12:03,682
what pendulum is going to be
in the future.

220
00:12:03,682 --> 00:12:08,270
But if I take a double pendulum
and I swing it a lot,

221
00:12:08,270 --> 00:12:10,510
then something else happens.

222
00:12:10,511 --> 00:12:12,811
They don't do the same thing,

223
00:12:12,811 --> 00:12:15,071
and there is nothing I can do,

224
00:12:15,071 --> 00:12:17,381
no matter how many measurements I make,

225
00:12:17,381 --> 00:12:22,391
that I can predict what is going to happen
with these pendulums,

226
00:12:22,391 --> 00:12:27,651
because infinite testable differences
lead to different outcomes.

227
00:12:27,906 --> 00:12:29,556
Not is all lost here.

228
00:12:29,556 --> 00:12:32,126
It turns out there are things
we can learn.

229
00:12:32,126 --> 00:12:36,526
For example, I can know
through my measurements,

230
00:12:36,526 --> 00:12:40,526
what the likelihood of the things
swinging all the way around is,

231
00:12:40,526 --> 00:12:42,546
how often that's going to happen.

232
00:12:42,546 --> 00:12:45,375
So, you can know things
about chaotic systems,

233
00:12:45,375 --> 00:12:49,045
but you cannot predict exactly
what they're going to do.

234
00:12:49,950 --> 00:12:53,110
Alright, so what is a chaotic system
that we are used to?

235
00:12:53,110 --> 00:12:59,120
Well, it turns out the Earth's climate
is a good example of a chaotic system.

236
00:12:59,120 --> 00:13:03,460
I show you here the temperature record
from Antarctic ice cores

237
00:13:03,460 --> 00:13:06,300
over the last 650,000 years.

238
00:13:06,300 --> 00:13:10,300
You can see in grey regions times
when the Earth is quite warm,

239
00:13:10,300 --> 00:13:13,390
and then it seemingly cools down.

240
00:13:13,390 --> 00:13:14,540
And why does it do that?

241
00:13:14,540 --> 00:13:20,630
It's a chaotic process that is related
to how the Earth goes around the Sun

242
00:13:20,630 --> 00:13:23,510
in a quite complex way.

243
00:13:23,510 --> 00:13:27,680
So it's very difficult to predict exactly
what the Earth is going to do

244
00:13:27,680 --> 00:13:29,460
at any given time.

245
00:13:30,533 --> 00:13:34,213
Also, it's just hard to measure
what's going on with the Earth.

246
00:13:34,213 --> 00:13:35,563
For the last thousand years,

247
00:13:35,563 --> 00:13:38,953
here are temperature reconstructions
from different groups.

248
00:13:38,953 --> 00:13:41,303
You can see over the last thousand years,

249
00:13:41,303 --> 00:13:45,383
we get quite different answers
back in time.

250
00:13:45,383 --> 00:13:48,003
We more or less agree
where we have better information,

251
00:13:48,003 --> 00:13:51,013
which is in the last hundred years or so,

252
00:13:51,013 --> 00:13:54,743
that the Earth is warmed up
about 8/10 of a degree.

253
00:13:55,772 --> 00:13:58,302
So, modeling and measuring
the climate is hard.

254
00:13:58,926 --> 00:14:01,646
The consensus view of just using the data

255
00:14:01,646 --> 00:14:06,516
is that we are 90% sure
that the warming trend is not an accident,

256
00:14:06,516 --> 00:14:08,199
that it is actually caused

257
00:14:08,199 --> 00:14:12,909
by anthropogenic
or man-made carbon dioxide.

258
00:14:12,909 --> 00:14:15,329
As a scientist trying
to make an experiment,

259
00:14:15,329 --> 00:14:19,599
90% isn't a very good result.

260
00:14:19,599 --> 00:14:21,319
You're not very sure about it.

261
00:14:21,319 --> 00:14:24,909
However, if someone's trying to figure out
my future of my life,

262
00:14:24,909 --> 00:14:27,479
90% is a pretty big risk factor.

263
00:14:27,479 --> 00:14:31,429
So, that's a very different thing
between those two things.

264
00:14:31,429 --> 00:14:33,401
But from my point as a scientist,

265
00:14:33,401 --> 00:14:37,491
I am 99.99999% sure

266
00:14:37,491 --> 00:14:41,611
that physics tells us
that adding CO2 to the atmosphere

267
00:14:41,611 --> 00:14:45,811
causes sunlight to be more effectively
trapped in our atmosphere,

268
00:14:45,811 --> 00:14:48,021
raising the temperature a bit.

269
00:14:48,021 --> 00:14:49,941
The hard part is

270
00:14:49,941 --> 00:14:51,761
- and what we are much less sure of -

271
00:14:51,761 --> 00:14:53,771
is how many clouds there are going to be,

272
00:14:53,771 --> 00:14:55,611
how much water vapor will be released,

273
00:14:55,611 --> 00:14:57,971
which warms the Earth up even more,

274
00:14:57,971 --> 00:15:00,291
how many methane releases will follow,

275
00:15:00,291 --> 00:15:03,841
and precisely how the oceans
will interact with all this

276
00:15:03,841 --> 00:15:07,281
to trap the CO2 and hold the warmth.

277
00:15:07,281 --> 00:15:12,811
Of course, we have no idea really
how much CO2 we will emit into the future.

278
00:15:14,064 --> 00:15:16,064
So here is our best estimate.

279
00:15:16,514 --> 00:15:19,144
The red curve shows
what we think will happen

280
00:15:19,144 --> 00:15:23,684
if we don't do anything
about our CO2 emission into the future.

281
00:15:23,684 --> 00:15:25,334
We're going to burn more and more

282
00:15:25,334 --> 00:15:27,844
as we become
more and more developed as a world.

283
00:15:28,401 --> 00:15:33,471
The blue line shows a very aggressive
carbon reduction strategy

284
00:15:33,471 --> 00:15:35,961
proposed by the IPCC.

285
00:15:36,665 --> 00:15:39,225
And then we can estimate
using our best physics

286
00:15:39,225 --> 00:15:41,315
of what we think is going to happen.

287
00:15:41,315 --> 00:15:43,785
Here is the outcome of the two ideas.

288
00:15:43,785 --> 00:15:48,745
The blue curve shows what happens
if we do that very aggressive drop.

289
00:15:48,745 --> 00:15:51,985
It keeps the rise of temperature
over the next century

290
00:15:51,985 --> 00:15:56,915
to less than 2 degrees C
with about 90% confidence.

291
00:15:57,799 --> 00:16:00,089
On the other hand,
if we let things keep going,

292
00:16:00,089 --> 00:16:04,139
the best prediction is, of course,
that it's going to get warmer and warmer,

293
00:16:04,139 --> 00:16:08,759
with a great deal of uncertainty
of about exactly how warm we'll go.

294
00:16:09,352 --> 00:16:12,412
According to the Australian
Academy of Science

295
00:16:12,412 --> 00:16:15,748
they say, "Expect climate surprises,"

296
00:16:15,748 --> 00:16:16,628
and we should,

297
00:16:16,628 --> 00:16:20,358
because the Earth's climate
is a chaotic system.

298
00:16:20,358 --> 00:16:22,728
We don't exactly know
what it's going to do,

299
00:16:22,728 --> 00:16:26,148
and that is what scares
the hell out of me.

300
00:16:27,253 --> 00:16:30,103
So, life is not black and white.

301
00:16:30,679 --> 00:16:33,689
Life is really shades of grey.

302
00:16:34,909 --> 00:16:37,059
But it's not all bad.

303
00:16:37,059 --> 00:16:39,359
You guys have done an excellent job,

304
00:16:39,359 --> 00:16:41,379
so what I want you to do now is to stop,

305
00:16:41,379 --> 00:16:43,669
and we are going to read out
your numbers here,

306
00:16:43,669 --> 00:16:45,189
and we're going to compare them

307
00:16:45,189 --> 00:16:48,759
to what I thought
which we were going to predict, okay?

308
00:16:48,759 --> 00:16:53,519
So I have here hopefully
a functioning computer.

309
00:16:53,519 --> 00:16:57,519
So what I need you to do
is to just go through from the left

310
00:16:57,519 --> 00:17:00,169
and read out the numbers
that you have achieved.

311
00:17:00,834 --> 00:17:02,354
Assistant: 5. Brian Schmidt: 5.

312
00:17:02,751 --> 00:17:04,471
A: 10. BS: 10.

313
00:17:04,471 --> 00:17:07,101
A: 21. BS:21.

314
00:17:07,101 --> 00:17:12,011
A: 21. BS:21 again? A: That's right. 24.

315
00:17:12,011 --> 00:17:15,481
BS: 24? A: Yes. Then 30. BS: 30.

316
00:17:15,481 --> 00:17:17,491
A: 37. BS: 37.

317
00:17:17,491 --> 00:17:19,431
A: 47. BS: 47.

318
00:17:19,431 --> 00:17:21,161
A: 41. BS: 41.

319
00:17:21,161 --> 00:17:22,921
A: 43. BS: 43.

320
00:17:22,921 --> 00:17:25,291
A: 29. BS: 29.

321
00:17:25,291 --> 00:17:27,281
A: 21. BS: 21.

322
00:17:27,281 --> 00:17:29,141
A: 8. BS: 8.

323
00:17:29,141 --> 00:17:31,011
A:10. BS: 10.

324
00:17:31,011 --> 00:17:33,221
A: 3. BS: 3.

325
00:17:33,221 --> 00:17:37,751
Well, I am proud to say
you guys were completely random.

326
00:17:37,751 --> 00:17:38,661
It was perfect.

327
00:17:38,661 --> 00:17:40,351
(Laughter)

328
00:17:40,351 --> 00:17:45,231
I show here the prediction
of what should happen and what happened.

329
00:17:45,231 --> 00:17:46,111
Bang on.

330
00:17:46,111 --> 00:17:47,981
(Applause)

331
00:17:47,981 --> 00:17:51,011
There is certainty in uncertainty,

332
00:17:51,011 --> 00:17:52,081
(Laughter)

333
00:17:52,081 --> 00:17:53,731
and that is the beauty of it.

334
00:17:53,731 --> 00:17:59,881
But to make policy decisions
based on what we know about science,

335
00:17:59,881 --> 00:18:01,861
about what we know about economics,

336
00:18:01,861 --> 00:18:06,761
requires our politicians,
our policy makers, and our citizens

337
00:18:06,761 --> 00:18:08,891
to understand uncertainty.

338
00:18:08,891 --> 00:18:12,211
I'm going to finish with the words
of Richard Feynman,

339
00:18:12,211 --> 00:18:14,651
with words that really could be my own,

340
00:18:14,651 --> 00:18:17,911
which is, "I can live
with doubt and uncertainty.

341
00:18:17,911 --> 00:18:21,181
I think it's much more interesting
to live not knowing

342
00:18:21,181 --> 00:18:23,671
than to have answers
which might be wrong."

343
00:18:23,671 --> 00:18:25,061
Thank you very much.

344
00:18:25,061 --> 00:18:28,061
(Applause)

345
00:18:28,699 --> 00:18:31,269
Thank you. Excellent.

346
00:18:31,269 --> 00:18:33,299
(Applause)