1
00:00:15,650 --> 00:00:17,268
Hello, everyone. How are you?

2
00:00:17,728 --> 00:00:19,218
My name is Ataka.

3
00:00:20,289 --> 00:00:26,959
I was asked to come here today
to speak about Japan's current situation,

4
00:00:26,969 --> 00:00:29,797
and our prospects for the future.

5
00:00:29,797 --> 00:00:32,772
My talk is titled "Japan Resurgence."

6
00:00:32,772 --> 00:00:33,587
(Laughter)

7
00:00:33,587 --> 00:00:36,352
Soon you will see why I chose this title.

8
00:00:36,352 --> 00:00:41,272
Let me recap our current situation.

9
00:00:41,883 --> 00:00:47,394
I think the biggest change in the world
over the past 30 years

10
00:00:47,394 --> 00:00:50,506
is that people carry portable computers.

11
00:00:50,506 --> 00:00:54,116
30 years ago, the fastest supercomputer
of the time was built by Cray.

12
00:00:54,116 --> 00:00:57,246
Now, you carry something
over 10 times faster in your pocket.

13
00:00:57,246 --> 00:00:59,336
iPhone 6 is 20 times faster.

14
00:00:59,336 --> 00:01:01,086
iPhone 7 and over is even faster.

15
00:01:01,086 --> 00:01:04,225
Everyone is carrying around 
an unbelievably fast computer.

16
00:01:04,745 --> 00:01:08,766
Bandwidth is also increasing
at a tremendous rate,

17
00:01:08,776 --> 00:01:11,492
over 1,000-fold during the last decade.

18
00:01:11,492 --> 00:01:14,032
In the 20 years since
the start of "Yahoo! Japan,"

19
00:01:14,032 --> 00:01:16,008
it has become over 10,000 times faster.

20
00:01:16,008 --> 00:01:17,955
And with the arrival of 5G,

21
00:01:17,955 --> 00:01:20,822
the speed is set to be
another 1,000 times faster.

22
00:01:20,822 --> 00:01:21,842
OK?

23
00:01:21,842 --> 00:01:25,182
Recently, the biggest packet plan
increased from 15GB to 50GB.

24
00:01:25,182 --> 00:01:29,482
In 5 to 10 years, a 1-month plan
will be something like 5 terabytes.

25
00:01:29,482 --> 00:01:31,486
The age of the 5TB plan is coming!

26
00:01:31,486 --> 00:01:32,326
(Laughter)

27
00:01:32,326 --> 00:01:33,736
What an amazing era!

28
00:01:34,069 --> 00:01:36,119
This is called "big data".

29
00:01:36,119 --> 00:01:39,139
A tremendous amount of data is emerging.

30
00:01:39,139 --> 00:01:44,189
Fortunately, computers are evolving
at an incredible speed.

31
00:01:44,539 --> 00:01:48,009
The number of chips is increasing
with remarkable momentum,

32
00:01:48,009 --> 00:01:50,089
so processing capability increases, too.

33
00:01:50,089 --> 00:01:51,819
Here is another lucky thing,

34
00:01:51,819 --> 00:01:55,467
which is mostly thanks to
the computers and data.

35
00:01:55,897 --> 00:02:00,087
We can now visualize data
which was previously impossible to see.

36
00:02:00,594 --> 00:02:04,374
We are running neural networks,
which couldn't be done before,

37
00:02:04,374 --> 00:02:06,584
and deep learning is up and running.

38
00:02:06,784 --> 00:02:09,776
So, all kinds of abnormal patterns
which were invisible before,

39
00:02:09,776 --> 00:02:13,036
hidden in a huge amount of data,

40
00:02:13,036 --> 00:02:15,516
these patterns are now becoming visibile.

41
00:02:15,516 --> 00:02:21,036
As a result, we live in a unique time,
close to an industrial revolution.

42
00:02:21,766 --> 00:02:26,596
The industrial revolution occurred
over the course of a few hundred years.

43
00:02:26,596 --> 00:02:29,916
Around the turn of the 19th century,

44
00:02:29,916 --> 00:02:33,086
over 90% of Japanese, Americans
and people in other areas

45
00:02:33,086 --> 00:02:36,716
were engaged in primary industry.

46
00:02:36,716 --> 00:02:38,976
96% in the U.S. at that time.

47
00:02:38,976 --> 00:02:41,014
But now, only a small percentage does.

48
00:02:41,014 --> 00:02:44,817
In spite of that, our current productivity
is much higher.

49
00:02:44,817 --> 00:02:50,857
Similarly, we now use most of our time 
for information processing,

50
00:02:50,857 --> 00:02:55,157
but with the next industrial revolution,
we will be released from that burden.

51
00:02:55,157 --> 00:03:00,816
I think we live in an extremely important,
and interesting time.

52
00:03:00,816 --> 00:03:03,906
We need to embrace living in this time.

53
00:03:03,906 --> 00:03:06,753
It is ridiculously lucky,
and super interesting.

54
00:03:06,753 --> 00:03:08,455
It's like living in a bonus round.

55
00:03:08,455 --> 00:03:09,398
(Laughter)

56
00:03:09,398 --> 00:03:10,511
Yes, it is.

57
00:03:10,511 --> 00:03:15,051
A professor at Berkeley claims
to have researched human productivity

58
00:03:15,051 --> 00:03:18,993
for the past 2,000 years, and actually
as far back as a million years.

59
00:03:19,553 --> 00:03:22,966
I don't know if he is correct,
but according to him,

60
00:03:22,966 --> 00:03:26,128
from the Roman Empire
to the industrial revolution,

61
00:03:26,128 --> 00:03:28,698
in those 2,000 years,
human productivity only doubled

62
00:03:28,698 --> 00:03:31,228
However, during the 150 years after that,

63
00:03:31,228 --> 00:03:34,876
the productivity skyrocketed
by nearly 100 times.

64
00:03:34,876 --> 00:03:36,176
But looking carefully,

65
00:03:36,176 --> 00:03:39,276
we are reaching saturation now,
from a macroeconomic perspective.

66
00:03:39,276 --> 00:03:43,646
The way things are right now,
we may soon see another big leap.

67
00:03:44,103 --> 00:03:46,873
Here is a very interesting sign.

68
00:03:46,873 --> 00:03:49,423
Let's look at market
capitalization rankings.

69
00:03:49,423 --> 00:03:51,323
Top ranks go to ICT companies

70
00:03:51,323 --> 00:03:54,353
from the information and communications
technology industry.

71
00:03:54,353 --> 00:03:59,557
Alibaba and Amazon are above Toyota,
one of the world's top automobile makers.

72
00:04:00,982 --> 00:04:03,785
As for GDP, measured
through gross value added,

73
00:04:03,785 --> 00:04:06,275
Japan has had no ICT sector
for the past 20 years,

74
00:04:06,275 --> 00:04:08,415
and we have had a shrinking economy.

75
00:04:08,415 --> 00:04:11,585
Actually, the tendency is similar
across the world:

76
00:04:11,585 --> 00:04:14,845
the ICT sector is the driver for growth.

77
00:04:15,745 --> 00:04:17,238
The interesting thing is,

78
00:04:17,238 --> 00:04:21,164
it is not so much the profit,
but the market cap that is really big.

79
00:04:21,164 --> 00:04:24,544
It seems a sense of "changing the world"
leads to success.

80
00:04:24,544 --> 00:04:29,838
In other words, this is an astonishing era
in which dreaming creates wealth.

81
00:04:29,838 --> 00:04:34,388
It is not hard work, but imagination
that generates national wealth.

82
00:04:34,388 --> 00:04:38,587
Keeping that in mind,
how should we move forward?

83
00:04:39,047 --> 00:04:43,470
I think the following three things are
very important.

84
00:04:43,470 --> 00:04:49,499
First, use a massive amount
of big data in various fields.

85
00:04:49,819 --> 00:04:56,269
Second, process the data
with tremendous power at affordable costs.

86
00:04:56,429 --> 00:05:01,949
Third, of course we need
world-class scientists and engineers

87
00:05:01,949 --> 00:05:04,365
to work very hard.

88
00:05:04,365 --> 00:05:10,145
Here we will need to face an inconvenient
truth that is rarely mentioned in Japan.

89
00:05:10,672 --> 00:05:12,322
The first data is...

90
00:05:12,322 --> 00:05:14,838
To tell the truth,

91
00:05:14,838 --> 00:05:17,838
it's really no competition.

92
00:05:17,838 --> 00:05:24,348
We aren't even close, in any field,
when it comes to using data.

93
00:05:24,348 --> 00:05:28,715
Of course there are autonomous cars
and other advanced technologies,

94
00:05:28,715 --> 00:05:34,015
but in terms of room sharing, car sharing,
and other industries that use data,

95
00:05:34,015 --> 00:05:37,325
personal information protection laws
make them impossible.

96
00:05:37,325 --> 00:05:38,805
In Japan, anyways.

97
00:05:38,805 --> 00:05:42,375
With autonomous cars, for example,
there are other problems.

98
00:05:42,375 --> 00:05:44,965
There is a high possibility
that Japan could make

99
00:05:44,965 --> 00:05:47,095
the world's top autonomous car.

100
00:05:47,095 --> 00:05:49,185
But look at Japanese roads -

101
00:05:49,752 --> 00:05:51,122
impossible!

102
00:05:51,266 --> 00:05:52,186
(Laughter)

103
00:05:52,257 --> 00:05:53,338
It is impossible!

104
00:05:53,338 --> 00:05:54,271
(Laughter)

105
00:05:55,891 --> 00:05:58,311
An old lady comes driving the other way

106
00:05:58,311 --> 00:06:00,831
yelling at you to get out of the way,

107
00:06:00,831 --> 00:06:05,691
so you have to pull into a driveway,
or back up 10 meters to pass each other.

108
00:06:05,691 --> 00:06:08,511
Automatic cars can't do that,
they completely stop.

109
00:06:08,511 --> 00:06:11,311
Some say to solve it with technology.

110
00:06:11,313 --> 00:06:15,723
Sure, we could make flying cars
or cars with legs.

111
00:06:15,723 --> 00:06:19,123
But the car would cost 30 million yen,
so it's impossible.

112
00:06:19,123 --> 00:06:21,713
Even if we try flying drones,
look at this.

113
00:06:21,713 --> 00:06:22,501
See?

114
00:06:22,501 --> 00:06:23,896
This is Minato-ku, Tokyo.

115
00:06:23,896 --> 00:06:27,266
It is a mixture of one-story
and 50-story buildings.

116
00:06:27,792 --> 00:06:30,992
Most of buildings in Paris are flat,
except for the Eiffel Tower.

117
00:06:30,992 --> 00:06:34,602
So they can use drone zones
to fly, but we can't.

118
00:06:35,349 --> 00:06:38,449
As for data processing,

119
00:06:39,493 --> 00:06:44,013
Basically, the only expense
for data processing should be electricity.

120
00:06:44,013 --> 00:06:45,653
But we pay a lot.

121
00:06:45,653 --> 00:06:48,403
Data processing costs are huge
in Japan for some reason.

122
00:06:48,403 --> 00:06:51,223
And how about data processing solutions?

123
00:06:51,223 --> 00:06:58,093
These are only a few of the options,
but there is no Japanese presence.

124
00:06:59,313 --> 00:07:02,853
So, Japan is far behind
on data processing.

125
00:07:02,853 --> 00:07:04,973
In fact, the situation is terrible.

126
00:07:04,973 --> 00:07:11,842
One of the core technologies,
deep learning, has a few key centers,

127
00:07:12,387 --> 00:07:16,447
and experts of ITC industry
gather in these areas,

128
00:07:16,447 --> 00:07:18,025
they are famous.

129
00:07:18,025 --> 00:07:22,485
They all live abroad,
mostly in the U.S. and U.K.

130
00:07:22,905 --> 00:07:24,455
This is the situation.

131
00:07:24,455 --> 00:07:28,272
You may think
there are engineers in Japan.

132
00:07:28,272 --> 00:07:30,410
Yes, there are many.

133
00:07:30,410 --> 00:07:32,400
But only 1/3 as many as the U.S.

134
00:07:32,400 --> 00:07:35,090
Notice the hidden area in the middle?

135
00:07:35,090 --> 00:07:38,327
We may think subconsciously
that Japan ranks second, after the U.S.

136
00:07:38,327 --> 00:07:39,457
That is not true.

137
00:07:39,457 --> 00:07:41,497
We are losing to China and India.

138
00:07:42,244 --> 00:07:43,314
This is our reality.

139
00:07:43,314 --> 00:07:45,174
Even fewer people are using big data,

140
00:07:45,174 --> 00:07:47,697
so we might actually
rank around 7th or 8th.

141
00:07:47,697 --> 00:07:52,987
There is a shortage
of people who work with numbers.

142
00:07:54,355 --> 00:08:00,075
Japan has 100,000 less yearly graduates
in science and technology than Korea,

143
00:08:00,075 --> 00:08:03,575
which has a population
of only 50 million.

144
00:08:03,575 --> 00:08:09,445
Korea and Germany consider themselves
to be tech-oriented nations;

145
00:08:10,165 --> 00:08:12,635
over 60% of their
college students major in science

146
00:08:12,635 --> 00:08:14,025
but only 20% in Japan.

147
00:08:14,025 --> 00:08:17,101
People like me who study
to get PhDs are considered strange,

148
00:08:17,101 --> 00:08:19,501
like some freak of nature.

149
00:08:19,501 --> 00:08:22,690
So what caused this situation?

150
00:08:22,690 --> 00:08:26,980
The number of the people
with analytical training is very small,

151
00:08:26,990 --> 00:08:29,707
1/3 of the number in the U.S.
per 100 people.

152
00:08:29,707 --> 00:08:31,908
Compared to Poland,
a math-oriented nation,

153
00:08:31,908 --> 00:08:33,748
we are one digit column lower.

154
00:08:33,748 --> 00:08:38,878
We lack not only people,
but programs as well.

155
00:08:39,488 --> 00:08:42,578
The U.S. didn't have any
data science programs either,

156
00:08:42,578 --> 00:08:45,228
but in just one year
they started over 500.

157
00:08:45,228 --> 00:08:49,798
In Japan, Shiga University will finally
open a program in spring, 2017.

158
00:08:49,798 --> 00:08:51,488
We have none at all.

159
00:08:51,488 --> 00:08:55,918
It is a complete shortage of personnel.

160
00:08:55,918 --> 00:08:59,488
Not only are people difficult to find,

161
00:08:59,488 --> 00:09:02,328
they don't have much enterprising spirit.

162
00:09:02,328 --> 00:09:04,978
We live in an era where anyone
could be the next Edison,

163
00:09:04,978 --> 00:09:06,478
and yet people lack motivation.

164
00:09:06,478 --> 00:09:10,574
We don't need otaku, we need people
who want to change the world.

165
00:09:10,574 --> 00:09:14,527
Of course we need hackers and geeks;

166
00:09:14,527 --> 00:09:18,843
we don't have enough of them,
and that is why we can't compete globally.

167
00:09:18,843 --> 00:09:20,203
(Laughter)

168
00:09:20,613 --> 00:09:24,043
We have had such an era before.

169
00:09:24,043 --> 00:09:27,033
We have had very difficult times.

170
00:09:27,353 --> 00:09:31,353
Once again, we have
almost nothing to offer.

171
00:09:31,877 --> 00:09:34,487
Do we have no hope in Japan?

172
00:09:34,487 --> 00:09:36,637
Looking back to the Industrial Revolution,

173
00:09:36,637 --> 00:09:39,161
it actually had three phases.

174
00:09:39,161 --> 00:09:43,431
New technologies were invented
in the late 1750's,

175
00:09:43,431 --> 00:09:45,761
and then they were put to practical use.

176
00:09:45,761 --> 00:09:48,571
Finally they came together even further,

177
00:09:48,571 --> 00:09:52,225
forming a kind of intricate ecosystem.

178
00:09:52,225 --> 00:09:54,205
What was Japan doing at that time?

179
00:09:54,205 --> 00:09:58,115
During Phase 1,
samurai were tying up their hair.

180
00:09:58,115 --> 00:10:00,975
Over 90% of people worked in rice fields.

181
00:10:00,975 --> 00:10:04,903
That continued for over 100 years,
so Japan didn't participate in Phase 1.

182
00:10:04,903 --> 00:10:08,243
Then Japan suddenly adopted new technology

183
00:10:08,243 --> 00:10:10,633
and made tremendous progress.

184
00:10:10,934 --> 00:10:15,724
We made bullet trains and super computers,

185
00:10:15,724 --> 00:10:18,111
and Japan became a high-tech nation.

186
00:10:18,111 --> 00:10:22,011
Once again,
we have obviously missed Phase 1,

187
00:10:22,011 --> 00:10:23,971
but let's not worry about that.

188
00:10:23,971 --> 00:10:25,251
(Laughter)

189
00:10:25,613 --> 00:10:31,763
We should focus on preparing
for the upcoming Phase 2 and 3.

190
00:10:31,903 --> 00:10:34,263
This is my proposal.

191
00:10:34,263 --> 00:10:35,433
(Laughter)

192
00:10:36,351 --> 00:10:38,617
We still have potential.

193
00:10:38,617 --> 00:10:41,485
Japan didn't experience
Phase 1 in the past, either.

194
00:10:41,485 --> 00:10:43,545
So it's not a problem.

195
00:10:43,545 --> 00:10:44,385
(Laughter)

196
00:10:44,385 --> 00:10:47,301
The other big waves,
Phase 2 and 3, are coming!

197
00:10:47,301 --> 00:10:50,921
We should prepare to ride those waves.

198
00:10:52,061 --> 00:10:56,721
We are full of great concepts in Japan.

199
00:10:56,721 --> 00:10:59,772
We have Ghost in the Shell, Doraemon...

200
00:10:59,772 --> 00:11:01,972
there are so many excellent ideas.

201
00:11:01,972 --> 00:11:04,688
Wouldn't these would be perfect
for Phase 2 and Phase 3?

202
00:11:04,688 --> 00:11:05,405
(Laughter)

203
00:11:05,405 --> 00:11:06,662
We have material.

204
00:11:06,662 --> 00:11:08,342
We have ideas.

205
00:11:08,342 --> 00:11:09,802
We can win again!

206
00:11:09,802 --> 00:11:11,442
(Laughter)

207
00:11:12,162 --> 00:11:17,122
Here is an emotional line
from a recent film.

208
00:11:19,195 --> 00:11:22,685
"Japan rose up through scrap and build,

209
00:11:22,685 --> 00:11:24,927
and we will rise again."

210
00:11:24,927 --> 00:11:27,957
These are lines
from "Godzilla Resurgence."

211
00:11:27,957 --> 00:11:30,807
We are faced with extreme circumstances,

212
00:11:30,807 --> 00:11:33,227
and we are almost scrapped.

213
00:11:33,227 --> 00:11:35,977
But let's try again!

214
00:11:35,977 --> 00:11:40,997
Let's start fresh,
and enjoy the competition.

215
00:11:40,997 --> 00:11:42,609
That's all I have to say.

216
00:11:42,869 --> 00:11:44,049
Thank you.

217
00:11:44,049 --> 00:11:45,989
(Applause)