1
00:00:00,605 --> 00:00:02,877
Growth is not dead.

2
00:00:02,877 --> 00:00:04,263
(Applause)

3
00:00:04,263 --> 00:00:08,226
Let's start the story 120 years ago,

4
00:00:08,226 --> 00:00:11,858
when American factories began to electrify their operations,

5
00:00:11,858 --> 00:00:15,202
igniting the Second Industrial Revolution.

6
00:00:15,202 --> 00:00:16,313
The amazing thing is

7
00:00:16,313 --> 00:00:19,090
that productivity did not increase in those factories

8
00:00:19,090 --> 00:00:22,346
for 30 years. Thirty years.

9
00:00:22,346 --> 00:00:25,820
That's long enough for a generation of managers to retire.

10
00:00:25,820 --> 00:00:28,042
You see, the first wave of managers

11
00:00:28,042 --> 00:00:31,459
simply replaced their steam engines with electric motors,

12
00:00:31,459 --> 00:00:34,469
but they didn't redesign the factories to take advantage

13
00:00:34,469 --> 00:00:36,810
of electricity's flexibility.

14
00:00:36,810 --> 00:00:40,794
It fell to the next generation to invent new work processes,

15
00:00:40,794 --> 00:00:43,521
and then productivity soared,

16
00:00:43,521 --> 00:00:47,186
often doubling or even tripling in those factories.

17
00:00:47,186 --> 00:00:51,909
Electricity is an example of a general purpose technology,

18
00:00:51,909 --> 00:00:54,139
like the steam engine before it.

19
00:00:54,139 --> 00:00:57,555
General purpose technologies drive most economic growth,

20
00:00:57,555 --> 00:01:01,009
because they unleash cascades of complementary innovations,

21
00:01:01,009 --> 00:01:04,641
like lightbulbs and, yes, factory redesign.

22
00:01:04,641 --> 00:01:08,251
Is there a general purpose technology of our era?

23
00:01:08,251 --> 00:01:10,759
Sure. It's the computer.

24
00:01:10,759 --> 00:01:13,418
But technology alone is not enough.

25
00:01:13,418 --> 00:01:16,184
Technology is not destiny.

26
00:01:16,184 --> 00:01:17,764
We shape our destiny,

27
00:01:17,764 --> 00:01:20,280
and just as the earlier generations of managers

28
00:01:20,280 --> 00:01:22,578
needed to redesign their factories,

29
00:01:22,578 --> 00:01:24,807
we're going to need to reinvent our organizations

30
00:01:24,807 --> 00:01:27,362
and even our whole economic system.

31
00:01:27,362 --> 00:01:30,964
We're not doing as well at that job as we should be.

32
00:01:30,964 --> 00:01:32,194
As we'll see in a moment,

33
00:01:32,194 --> 00:01:34,916
productivity is actually doing all right,

34
00:01:34,916 --> 00:01:38,778
but it has become decoupled from jobs,

35
00:01:38,778 --> 00:01:43,197
and the income of the typical worker is stagnating.

36
00:01:43,197 --> 00:01:45,716
These troubles are sometimes misdiagnosed

37
00:01:45,716 --> 00:01:49,428
as the end of innovation,

38
00:01:49,428 --> 00:01:51,557
but they are actually the growing pains

39
00:01:51,557 --> 00:01:57,147
of what Andrew McAfee and I call the new machine age.

40
00:01:57,147 --> 00:01:59,029
Let's look at some data.

41
00:01:59,029 --> 00:02:01,931
So here's GDP per person in America.

42
00:02:01,931 --> 00:02:04,697
There's some bumps along the way, but the big story

43
00:02:04,697 --> 00:02:07,412
is you could practically fit a ruler to it.

44
00:02:07,412 --> 00:02:10,688
This is a log scale, so what looks like steady growth

45
00:02:10,688 --> 00:02:13,731
is actually an acceleration in real terms.

46
00:02:13,731 --> 00:02:15,891
And here's productivity.

47
00:02:15,891 --> 00:02:18,562
You can see a little bit of a slowdown there in the mid-'70s,

48
00:02:18,562 --> 00:02:22,300
but it matches up pretty well with the Second Industrial Revolution,

49
00:02:22,300 --> 00:02:24,991
when factories were learning how to electrify their operations.

50
00:02:24,991 --> 00:02:29,120
After a lag, productivity accelerated again.

51
00:02:29,120 --> 00:02:31,691
So maybe "history doesn't repeat itself,

52
00:02:31,691 --> 00:02:34,259
but sometimes it rhymes."

53
00:02:34,259 --> 00:02:37,395
Today, productivity is at an all-time high,

54
00:02:37,395 --> 00:02:39,372
and despite the Great Recession,

55
00:02:39,372 --> 00:02:43,624
it grew faster in the 2000s than it did in the 1990s,

56
00:02:43,624 --> 00:02:47,760
the roaring 1990s, and that was faster than the '70s or '80s.

57
00:02:47,760 --> 00:02:51,434
It's growing faster than it did during the Second Industrial Revolution.

58
00:02:51,434 --> 00:02:53,177
And that's just the United States.

59
00:02:53,177 --> 00:02:56,425
The global news is even better.

60
00:02:56,425 --> 00:02:58,785
Worldwide incomes have grown at a faster rate

61
00:02:58,785 --> 00:03:01,281
in the past decade than ever in history.

62
00:03:01,281 --> 00:03:06,332
If anything, all these numbers actually understate our progress,

63
00:03:06,332 --> 00:03:08,244
because the new machine age

64
00:03:08,244 --> 00:03:09,908
is more about knowledge creation

65
00:03:09,908 --> 00:03:12,239
than just physical production.

66
00:03:12,239 --> 00:03:15,177
It's mind not matter, brain not brawn,

67
00:03:15,177 --> 00:03:17,239
ideas not things.

68
00:03:17,239 --> 00:03:19,809
That creates a problem for standard metrics,

69
00:03:19,809 --> 00:03:23,311
because we're getting more and more stuff for free,

70
00:03:23,311 --> 00:03:25,952
like Wikipedia, Google, Skype,

71
00:03:25,952 --> 00:03:29,015
and if they post it on the web, even this TED Talk.

72
00:03:29,015 --> 00:03:32,318
Now getting stuff for free is a good thing, right?

73
00:03:32,318 --> 00:03:34,083
Sure, of course it is.

74
00:03:34,083 --> 00:03:37,951
But that's not how economists measure GDP.

75
00:03:37,951 --> 00:03:43,543
Zero price means zero weight in the GDP statistics.

76
00:03:43,543 --> 00:03:45,655
According to the numbers, the music industry

77
00:03:45,655 --> 00:03:48,655
is half the size that it was 10 years ago,

78
00:03:48,655 --> 00:03:52,311
but I'm listening to more and better music than ever.

79
00:03:52,311 --> 00:03:54,503
You know, I bet you are too.

80
00:03:54,503 --> 00:03:57,226
In total, my research estimates

81
00:03:57,226 --> 00:04:01,980
that the GDP numbers miss over 300 billion dollars per year

82
00:04:01,980 --> 00:04:05,326
in free goods and services on the Internet.

83
00:04:05,326 --> 00:04:07,115
Now let's look to the future.

84
00:04:07,115 --> 00:04:09,378
There are some super smart people

85
00:04:09,378 --> 00:04:14,397
who are arguing that we've reached the end of growth,

86
00:04:14,397 --> 00:04:17,955
but to understand the future of growth,

87
00:04:17,955 --> 00:04:20,638
we need to make predictions

88
00:04:20,638 --> 00:04:23,928
about the underlying drivers of growth.

89
00:04:23,928 --> 00:04:27,734
I'm optimistic, because the new machine age

90
00:04:27,734 --> 00:04:32,764
is digital, exponential and combinatorial.

91
00:04:32,764 --> 00:04:35,028
When goods are digital, they can be replicated

92
00:04:35,028 --> 00:04:39,537
with perfect quality at nearly zero cost,

93
00:04:39,537 --> 00:04:43,555
and they can be delivered almost instantaneously.

94
00:04:43,555 --> 00:04:46,355
Welcome to the economics of abundance.

95
00:04:46,355 --> 00:04:50,045
But there's a subtler benefit to the digitization of the world.

96
00:04:50,045 --> 00:04:54,645
Measurement is the lifeblood of science and progress.

97
00:04:54,645 --> 00:04:56,793
In the age of big data,

98
00:04:56,793 --> 00:05:01,079
we can measure the world in ways we never could before.

99
00:05:01,079 --> 00:05:05,174
Secondly, the new machine age is exponential.

100
00:05:05,174 --> 00:05:11,109
Computers get better faster than anything else ever.

101
00:05:11,109 --> 00:05:14,677
A child's Playstation today is more powerful

102
00:05:14,677 --> 00:05:18,733
than a military supercomputer from 1996.

103
00:05:18,733 --> 00:05:21,940
But our brains are wired for a linear world.

104
00:05:21,940 --> 00:05:25,828
As a result, exponential trends take us by surprise.

105
00:05:25,828 --> 00:05:28,430
I used to teach my students that there are some things,

106
00:05:28,430 --> 00:05:30,364
you know, computers just aren't good at,

107
00:05:30,364 --> 00:05:32,749
like driving a car through traffic.

108
00:05:32,749 --> 00:05:34,762
(Laughter)

109
00:05:34,762 --> 00:05:38,253
That's right, here's Andy and me grinning like madmen

110
00:05:38,253 --> 00:05:40,637
because we just rode down Route 101

111
00:05:40,637 --> 00:05:44,306
in, yes, a driverless car.

112
00:05:44,306 --> 00:05:46,889
Thirdly, the new machine age is combinatorial.

113
00:05:46,889 --> 00:05:50,937
The stagnationist view is that ideas get used up,

114
00:05:50,937 --> 00:05:52,793
like low-hanging fruit,

115
00:05:52,793 --> 00:05:55,956
but the reality is that each innovation

116
00:05:55,956 --> 00:05:59,212
creates building blocks for even more innovations.

117
00:05:59,212 --> 00:06:02,557
Here's an example. In just a matter of a few weeks,

118
00:06:02,557 --> 00:06:04,629
an undergraduate student of mine

119
00:06:04,629 --> 00:06:08,740
built an app that ultimately reached 1.3 million users.

120
00:06:08,740 --> 00:06:10,439
He was able to do that so easily

121
00:06:10,439 --> 00:06:12,266
because he built it on top of Facebook,

122
00:06:12,266 --> 00:06:14,199
and Facebook was built on top of the web,

123
00:06:14,199 --> 00:06:15,897
and that was built on top of the Internet,

124
00:06:15,897 --> 00:06:18,315
and so on and so forth.

125
00:06:18,315 --> 00:06:23,080
Now individually, digital, exponential and combinatorial

126
00:06:23,080 --> 00:06:25,430
would each be game-changers.

127
00:06:25,430 --> 00:06:27,620
Put them together, and we're seeing a wave

128
00:06:27,620 --> 00:06:29,013
of astonishing breakthroughs,

129
00:06:29,013 --> 00:06:32,073
like robots that do factory work or run as fast as a cheetah

130
00:06:32,073 --> 00:06:34,869
or leap tall buildings in a single bound.

131
00:06:34,869 --> 00:06:37,101
You know, robots are even revolutionizing

132
00:06:37,101 --> 00:06:38,930
cat transportation.

133
00:06:38,930 --> 00:06:41,200
(Laughter)

134
00:06:41,200 --> 00:06:43,932
But perhaps the most important invention,

135
00:06:43,932 --> 00:06:48,997
the most important invention is machine learning.

136
00:06:48,997 --> 00:06:52,373
Consider one project: IBM's Watson.

137
00:06:52,373 --> 00:06:53,962
These little dots here,

138
00:06:53,962 --> 00:06:58,822
those are all the champions on the quiz show "Jeopardy."

139
00:06:58,822 --> 00:07:01,366
At first, Watson wasn't very good,

140
00:07:01,366 --> 00:07:06,988
but it improved at a rate faster than any human could,

141
00:07:06,988 --> 00:07:09,675
and shortly after Dave Ferrucci showed this chart

142
00:07:09,675 --> 00:07:11,327
to my class at MIT,

143
00:07:11,327 --> 00:07:14,869
Watson beat the world "Jeopardy" champion.

144
00:07:14,869 --> 00:07:18,858
At age seven, Watson is still kind of in its childhood.

145
00:07:18,858 --> 00:07:24,176
Recently, its teachers let it surf the Internet unsupervised.

146
00:07:24,176 --> 00:07:30,122
The next day, it started answering questions with profanities.

147
00:07:30,122 --> 00:07:32,396
Damn. (Laughter)

148
00:07:32,396 --> 00:07:34,676
But you know, Watson is growing up fast.

149
00:07:34,676 --> 00:07:38,888
It's being tested for jobs in call centers, and it's getting them.

150
00:07:38,888 --> 00:07:42,612
It's applying for legal, banking and medical jobs,

151
00:07:42,612 --> 00:07:44,562
and getting some of them.

152
00:07:44,562 --> 00:07:46,451
Isn't it ironic that at the very moment

153
00:07:46,451 --> 00:07:48,685
we are building intelligent machines,

154
00:07:48,685 --> 00:07:52,134
perhaps the most important invention in human history,

155
00:07:52,134 --> 00:07:56,109
some people are arguing that innovation is stagnating?

156
00:07:56,109 --> 00:07:58,528
Like the first two industrial revolutions,

157
00:07:58,528 --> 00:08:01,662
the full implications of the new machine age

158
00:08:01,662 --> 00:08:04,344
are going to take at least a century to fully play out,

159
00:08:04,344 --> 00:08:07,376
but they are staggering.

160
00:08:07,376 --> 00:08:10,712
So does that mean we have nothing to worry about?

161
00:08:10,712 --> 00:08:14,392
No. Technology is not destiny.

162
00:08:14,392 --> 00:08:16,961
Productivity is at an all time high,

163
00:08:16,961 --> 00:08:19,944
but fewer people now have jobs.

164
00:08:19,944 --> 00:08:23,064
We have created more wealth in the past decade than ever,

165
00:08:23,064 --> 00:08:26,968
but for a majority of Americans, their income has fallen.

166
00:08:26,968 --> 00:08:29,280
This is the great decoupling

167
00:08:29,280 --> 00:08:32,256
of productivity from employment,

168
00:08:32,256 --> 00:08:35,360
of wealth from work.

169
00:08:35,360 --> 00:08:37,706
You know, it's not surprising that millions of people

170
00:08:37,706 --> 00:08:40,552
have become disillusioned by the great decoupling,

171
00:08:40,552 --> 00:08:42,299
but like too many others,

172
00:08:42,299 --> 00:08:45,396
they misunderstand its basic causes.

173
00:08:45,396 --> 00:08:48,006
Technology is racing ahead,

174
00:08:48,006 --> 00:08:51,556
but it's leaving more and more people behind.

175
00:08:51,556 --> 00:08:55,075
Today, we can take a routine job,

176
00:08:55,075 --> 00:08:58,166
codify it in a set of machine-readable instructions,

177
00:08:58,166 --> 00:09:00,993
and then replicate it a million times.

178
00:09:00,993 --> 00:09:03,272
You know, I recently overheard a conversation

179
00:09:03,272 --> 00:09:05,224
that epitomizes these new economics.

180
00:09:05,224 --> 00:09:09,421
This guy says, "Nah, I don't use H&R Block anymore.

181
00:09:09,421 --> 00:09:11,869
TurboTax does everything that my tax preparer did,

182
00:09:11,869 --> 00:09:16,427
but it's faster, cheaper and more accurate."

183
00:09:16,427 --> 00:09:18,226
How can a skilled worker

184
00:09:18,226 --> 00:09:21,235
compete with a $39 piece of software?

185
00:09:21,235 --> 00:09:23,202
She can't.

186
00:09:23,202 --> 00:09:25,982
Today, millions of Americans do have faster,

187
00:09:25,982 --> 00:09:28,369
cheaper, more accurate tax preparation,

188
00:09:28,369 --> 00:09:29,855
and the founders of Intuit

189
00:09:29,855 --> 00:09:32,348
have done very well for themselves.

190
00:09:32,348 --> 00:09:36,562
But 17 percent of tax preparers no longer have jobs.

191
00:09:36,562 --> 00:09:38,640
That is a microcosm of what's happening,

192
00:09:38,640 --> 00:09:43,317
not just in software and services, but in media and music,

193
00:09:43,317 --> 00:09:47,003
in finance and manufacturing, in retailing and trade --

194
00:09:47,003 --> 00:09:50,898
in short, in every industry.

195
00:09:50,898 --> 00:09:53,993
People are racing against the machine,

196
00:09:53,993 --> 00:09:57,083
and many of them are losing that race.

197
00:09:57,083 --> 00:10:00,969
What can we do to create shared prosperity?

198
00:10:00,969 --> 00:10:03,986
The answer is not to try to slow down technology.

199
00:10:03,986 --> 00:10:06,543
Instead of racing against the machine,

200
00:10:06,543 --> 00:10:10,220
we need to learn to race with the machine.

201
00:10:10,220 --> 00:10:13,349
That is our grand challenge.

202
00:10:13,349 --> 00:10:15,673
The new machine age

203
00:10:15,673 --> 00:10:18,786
can be dated to a day 15 years ago

204
00:10:18,786 --> 00:10:21,664
when Garry Kasparov, the world chess champion,

205
00:10:21,664 --> 00:10:25,370
played Deep Blue, a supercomputer.

206
00:10:25,370 --> 00:10:27,382
The machine won that day,

207
00:10:27,382 --> 00:10:30,350
and today, a chess program running on a cell phone

208
00:10:30,350 --> 00:10:32,646
can beat a human grandmaster.

209
00:10:32,646 --> 00:10:36,011
It got so bad that, when he was asked

210
00:10:36,011 --> 00:10:38,574
what strategy he would use against a computer,

211
00:10:38,574 --> 00:10:42,590
Jan Donner, the Dutch grandmaster, replied,

212
00:10:42,590 --> 00:10:44,361
"I'd bring a hammer."

213
00:10:44,361 --> 00:10:48,041
(Laughter)

214
00:10:48,041 --> 00:10:52,585
But today a computer is no longer the world chess champion.

215
00:10:52,585 --> 00:10:55,239
Neither is a human,

216
00:10:55,239 --> 00:10:58,818
because Kasparov organized a freestyle tournament

217
00:10:58,818 --> 00:11:00,734
where teams of humans and computers

218
00:11:00,734 --> 00:11:02,833
could work together,

219
00:11:02,833 --> 00:11:05,990
and the winning team had no grandmaster,

220
00:11:05,990 --> 00:11:08,455
and it had no supercomputer.

221
00:11:08,455 --> 00:11:12,630
What they had was better teamwork,

222
00:11:12,630 --> 00:11:17,646
and they showed that a team of humans and computers,

223
00:11:17,646 --> 00:11:20,694
working together, could beat any computer

224
00:11:20,694 --> 00:11:24,214
or any human working alone.

225
00:11:24,214 --> 00:11:25,878
Racing with the machine

226
00:11:25,878 --> 00:11:28,221
beats racing against the machine.

227
00:11:28,221 --> 00:11:30,785
Technology is not destiny.

228
00:11:30,785 --> 00:11:32,527
We shape our destiny.

229
00:11:32,527 --> 00:11:33,974
Thank you.

230
00:11:33,974 --> 00:11:38,990
(Applause)