1
00:00:03,200 --> 00:00:08,480
Speaker 1: Bloomberg Audio Studios, podcasts, radio news.

2
00:00:17,280 --> 00:00:21,279
Speaker 2: It is about having this narrative that democracy only leads

3
00:00:21,280 --> 00:00:24,440
Speaker 2: to chaos and democracy never delivers. That is the overarching

4
00:00:24,600 --> 00:00:28,720
Speaker 2: narrative for the authoritarian attacks. So I think one lesson

5
00:00:29,000 --> 00:00:31,360
Speaker 2: is to let the people know that this has happened.

6
00:00:31,440 --> 00:00:34,120
Speaker 2: Any If you'll let everybody know that this is the narrative,

7
00:00:34,120 --> 00:00:36,879
Speaker 2: they're going to push the people built immune system and

8
00:00:36,960 --> 00:00:40,239
Speaker 2: hype bodies in their mind even before that attack actually happens.

9
00:00:42,800 --> 00:00:47,040
Speaker 3: Welcome to Voteronomics, where politics and markets and everything else collide.

10
00:00:47,200 --> 00:00:50,040
Speaker 3: This year, voters around the world have the ability to

11
00:00:50,120 --> 00:00:54,160
Speaker 3: affect markets, countries, and economies like never before, so it

12
00:00:54,280 --> 00:00:57,160
Speaker 3: created this series Photonomics to help you make sense of

13
00:00:57,160 --> 00:00:59,320
Speaker 3: it all. I'm Stephanie Flanders.

14
00:00:58,880 --> 00:01:01,000
Speaker 4: I'm Adrian Wooldridge, and I'm Allegra Stratton.

15
00:01:03,600 --> 00:01:06,080
Speaker 3: And you heard at the top of the show the

16
00:01:06,600 --> 00:01:10,160
Speaker 3: former recently stepped down Minister of Digital Affairs of Taiwan,

17
00:01:10,280 --> 00:01:13,520
Speaker 3: Audrey Tang. Now you might wonder why on earth, you know,

18
00:01:13,560 --> 00:01:15,800
Speaker 3: we have all these illustrious people on this podcast, not

19
00:01:15,920 --> 00:01:19,120
Speaker 3: least people I'm talking to right now, why would we

20
00:01:19,200 --> 00:01:22,480
Speaker 3: have a recently stepped down Minister of Digital affairs. But

21
00:01:22,600 --> 00:01:26,720
Speaker 3: Audrey Tang for the last several years has been absolutely

22
00:01:26,760 --> 00:01:31,720
Speaker 3: the front line of the misinformation war, which of course

23
00:01:31,720 --> 00:01:35,600
Speaker 3: we're all worried about being above and beneath the surface

24
00:01:35,880 --> 00:01:38,160
Speaker 3: of these big elections that we're facing in the UK

25
00:01:38,240 --> 00:01:41,480
Speaker 3: and the US and other countries, because Taiwan, as she

26
00:01:41,520 --> 00:01:46,000
Speaker 3: says herself, has been the country most subjected to cyber

27
00:01:46,080 --> 00:01:50,600
Speaker 3: onslaught on all fronts, whether it's around the senior US

28
00:01:50,680 --> 00:01:55,320
Speaker 3: politician Nancy Pelosi's visit to Taiwan and most recently the

29
00:01:55,440 --> 00:01:59,200
Speaker 3: election where China clearly didn't want to have someone win

30
00:01:59,600 --> 00:02:02,360
Speaker 3: the president see in Taiwan who was considered to be

31
00:02:02,480 --> 00:02:07,600
Speaker 3: sort of pro independence or some form of independence for Taiwan.

32
00:02:07,720 --> 00:02:09,240
Speaker 3: I mean, allegra you and I spoke to her and

33
00:02:09,280 --> 00:02:11,679
Speaker 3: it was just a fascinating conversation.

34
00:02:11,240 --> 00:02:13,360
Speaker 1: About It was fascinating and it was quite chilling as well,

35
00:02:13,360 --> 00:02:16,839
Speaker 1: because when I was asking her if she the incidences

36
00:02:16,840 --> 00:02:19,799
Speaker 1: that we've seen in the UK of possible disinformation and

37
00:02:21,000 --> 00:02:23,960
Speaker 1: Russian bots stirring up division and so on, and you know,

38
00:02:24,240 --> 00:02:27,280
Speaker 1: and she said yeah, and some like you know, yes,

39
00:02:27,320 --> 00:02:28,840
Speaker 1: and it will have been much more than that. A

40
00:02:29,480 --> 00:02:32,160
Speaker 1: And then the other thing that I that really resonated

41
00:02:32,160 --> 00:02:34,840
Speaker 1: and I liked a lot was this idea of jury

42
00:02:34,880 --> 00:02:39,120
Speaker 1: service on social media, so you could we as citizens

43
00:02:39,120 --> 00:02:42,520
Speaker 1: should all be dipping into various different pieces of social

44
00:02:42,600 --> 00:02:46,520
Speaker 1: media and kind of observing in quite an intense way.

45
00:02:46,639 --> 00:02:49,040
Speaker 4: Yeah, I'm very much looking forward to hit to hearing it.

46
00:02:49,120 --> 00:02:50,520
Speaker 4: Absolutely fascinating subject.

47
00:02:50,919 --> 00:02:53,239
Speaker 3: But of course we're interested in these things, even more

48
00:02:53,280 --> 00:02:56,280
Speaker 3: interested in elections right now because we have one. And

49
00:02:56,360 --> 00:02:58,960
Speaker 3: actually we had that was a snapcast that we did

50
00:02:59,040 --> 00:03:01,320
Speaker 3: earlier this week of voting when we had first got

51
00:03:01,320 --> 00:03:04,799
Speaker 3: that news the announcement for Rissi Sinek of the election

52
00:03:05,080 --> 00:03:07,280
Speaker 3: in early July. And I think that's the one thing

53
00:03:07,280 --> 00:03:09,560
Speaker 3: we didn't talk about, actually, is the risk of misinformation.

54
00:03:10,600 --> 00:03:14,640
Speaker 1: Well, it's interesting. I think that experts are divided. I

55
00:03:14,680 --> 00:03:18,919
Speaker 1: think that they feel that actually, in the elections that

56
00:03:18,960 --> 00:03:21,200
Speaker 1: we've already seen in twenty twenty four around the world,

57
00:03:21,280 --> 00:03:23,640
Speaker 1: that it hasn't actually played a huge role, but it

58
00:03:23,720 --> 00:03:25,160
Speaker 1: is certainly right to be vigilant.

59
00:03:25,280 --> 00:03:27,040
Speaker 4: I think there's absolute truth in that. But I think

60
00:03:27,080 --> 00:03:29,959
Speaker 4: the extent to which this is going on is increasing

61
00:03:30,000 --> 00:03:31,840
Speaker 4: all the time, and the extent to which the sort

62
00:03:31,840 --> 00:03:37,160
Speaker 4: of the Russian model of deliberately spreading its spreading misinformation

63
00:03:37,240 --> 00:03:40,520
Speaker 4: and agit prop is being taken up by China, and

64
00:03:40,640 --> 00:03:44,560
Speaker 4: push forward is really quite important. And also I think

65
00:03:44,600 --> 00:03:46,960
Speaker 4: if you look at the what you might call the

66
00:03:47,600 --> 00:03:51,160
Speaker 4: states that are but Russia, you know, the Baltic States

67
00:03:51,160 --> 00:03:52,800
Speaker 4: and the rest of them, you see it. It's a

68
00:03:52,880 --> 00:03:55,280
Speaker 4: lot more of a problem it has been here. So

69
00:03:55,320 --> 00:03:58,200
Speaker 4: I think the future is it's quite dark when it

70
00:03:58,200 --> 00:03:59,000
Speaker 4: comes to misinformation.

71
00:03:59,080 --> 00:04:01,760
Speaker 3: It's very hard to say definitively what's happening. And that's

72
00:04:01,800 --> 00:04:03,320
Speaker 3: part of the issue is that you have this sort

73
00:04:03,320 --> 00:04:06,360
Speaker 3: of echo chamber of different forms of social media, self

74
00:04:06,400 --> 00:04:11,600
Speaker 3: selected groups who are then having damaging messages amplified. And

75
00:04:11,640 --> 00:04:13,800
Speaker 3: Audrey Tang talks about it. But actually I was struck

76
00:04:13,840 --> 00:04:16,839
Speaker 3: by Anne app Obama's done this fantastic big story for

77
00:04:16,920 --> 00:04:21,800
Speaker 3: The Atlantic and this month's June issue of The Atlantic magazine,

78
00:04:21,800 --> 00:04:25,720
Speaker 3: the New Propaganda War, where she just maps out how China,

79
00:04:25,960 --> 00:04:29,880
Speaker 3: Russia and other sort of semi rogue nations are making

80
00:04:30,000 --> 00:04:33,920
Speaker 3: common cause with sort of make America great again Trumpian

81
00:04:34,000 --> 00:04:38,599
Speaker 3: Republicans to spread a message around sort of undermining liberalism

82
00:04:39,000 --> 00:04:42,279
Speaker 3: and undermining kind of normal democrats.

83
00:04:42,279 --> 00:04:44,360
Speaker 4: I think one of the many fascinating things that she

84
00:04:44,560 --> 00:04:49,320
Speaker 4: said is that these countries are collaborating with each other.

85
00:04:49,440 --> 00:04:53,040
Speaker 4: I mean both and said this, and Audrey Tang said this,

86
00:04:53,240 --> 00:04:55,400
Speaker 4: And I think we should be looking certainly to what

87
00:04:55,480 --> 00:04:57,520
Speaker 4: Taiwan has been doing. But again, I also think we

88
00:04:57,520 --> 00:04:59,839
Speaker 4: should be looking to what some of the Baltic states

89
00:05:00,120 --> 00:05:02,160
Speaker 4: or some of the states on that are closest to

90
00:05:02,200 --> 00:05:04,400
Speaker 4: Russia have been doing. I think a really interesting example

91
00:05:04,440 --> 00:05:07,640
Speaker 4: of that is Finland, which has in its schools a

92
00:05:07,680 --> 00:05:13,919
Speaker 4: civic education program which teaches people how to resist foreign

93
00:05:13,960 --> 00:05:18,279
Speaker 4: propaganda because they're being constantly bullied and acted by the Russians,

94
00:05:18,320 --> 00:05:21,240
Speaker 4: and that's actually woven into their education system. So I

95
00:05:21,240 --> 00:05:23,880
Speaker 4: think when we're thinking about what our defenses should be,

96
00:05:24,440 --> 00:05:26,520
Speaker 4: we shouldn't just be thinking about the military. We should

97
00:05:26,520 --> 00:05:29,920
Speaker 4: also be thinking about a much broader set of defenses

98
00:05:29,960 --> 00:05:33,799
Speaker 4: against foreign interference, which include public opinion, public education.

99
00:05:34,080 --> 00:05:35,159
Speaker 1: Do you think it would work here?

100
00:05:35,360 --> 00:05:37,400
Speaker 3: Yes, well, I think it's just and I'm also just struck.

101
00:05:37,440 --> 00:05:39,880
Speaker 3: We had our guest last week who was saying, when

102
00:05:39,920 --> 00:05:42,200
Speaker 3: he goes to talk to students, and they all are

103
00:05:42,240 --> 00:05:44,600
Speaker 3: sort of many of them are parroting the same sort

104
00:05:44,640 --> 00:05:49,159
Speaker 3: of somewhat sympathetic to Russia line about Ukraine, and she

105
00:05:49,240 --> 00:05:55,039
Speaker 3: actually Anne Applebam sites having had talking to other Western

106
00:05:55,080 --> 00:05:58,440
Speaker 3: diplomats who have sort of been been in Africa and

107
00:05:58,440 --> 00:06:01,680
Speaker 3: then are sort of shocked by everywhere they go people

108
00:06:01,720 --> 00:06:06,200
Speaker 3: with a completely different narrative around the war in Ukraine,

109
00:06:06,440 --> 00:06:09,880
Speaker 3: and many of them believing this particular story around the

110
00:06:10,000 --> 00:06:14,160
Speaker 3: US using biological warfare early on in helping Ukraine to

111
00:06:14,600 --> 00:06:17,600
Speaker 3: commit biological warfare at the beginning of the war. So

112
00:06:17,760 --> 00:06:20,080
Speaker 3: I think it is there is a fear that even

113
00:06:20,120 --> 00:06:22,920
Speaker 3: if you're doing all these things within your core core

114
00:06:23,279 --> 00:06:29,680
Speaker 3: education system and on your mainstream media, the segmentation of

115
00:06:29,720 --> 00:06:30,920
Speaker 3: the voices that people.

116
00:06:30,720 --> 00:06:33,880
Speaker 4: Are getting you just have there's an enemy agate problem.

117
00:06:33,880 --> 00:06:37,720
Speaker 4: But there's also just rumors and just nonsense that that

118
00:06:37,880 --> 00:06:44,880
Speaker 4: that is there on that we have Arnold's and I think,

119
00:06:44,920 --> 00:06:47,800
Speaker 4: you know, as this British election unfolds, I think it's

120
00:06:47,880 --> 00:06:50,680
Speaker 4: very important that we take a pretty broad view of

121
00:06:51,360 --> 00:06:53,880
Speaker 4: public opinion, not just reading the Times, but also looking

122
00:06:53,920 --> 00:06:56,160
Speaker 4: at some of the weirdest stuff coming out on the

123
00:06:56,160 --> 00:06:58,279
Speaker 4: Internet or gb News and the rest of it.

124
00:07:06,880 --> 00:07:10,320
Speaker 3: While we're talking about you know, there's the risks of

125
00:07:10,360 --> 00:07:12,760
Speaker 3: what might happen in the UK election I think you know,

126
00:07:12,800 --> 00:07:15,640
Speaker 3: we also have that US election. We've been running a

127
00:07:15,680 --> 00:07:18,840
Speaker 3: swing state pole because as you as you both will know,

128
00:07:19,800 --> 00:07:22,320
Speaker 3: the seven swing states, they are the ones that will

129
00:07:22,320 --> 00:07:25,400
Speaker 3: decide the outcome of this election. And actually the latest

130
00:07:25,520 --> 00:07:29,240
Speaker 3: poll has just come out had some quite interesting results,

131
00:07:29,520 --> 00:07:33,160
Speaker 3: particularly on this subject of foreign interference and also the

132
00:07:33,280 --> 00:07:36,320
Speaker 3: risk of violence in and around the elections. We can

133
00:07:36,360 --> 00:07:39,800
Speaker 3: go back to a friend of voter nomics, Nancy Cook,

134
00:07:40,240 --> 00:07:44,920
Speaker 3: our top reporter in Washington covering all things campaign and

135
00:07:45,040 --> 00:07:49,080
Speaker 3: economic qualicy and much else. Nancy, I know you've we've

136
00:07:49,120 --> 00:07:52,520
Speaker 3: had quite a few of these swing state poles, but

137
00:07:52,600 --> 00:07:54,280
Speaker 3: what are the top lines from this one?

138
00:07:54,360 --> 00:07:58,240
Speaker 5: Well, the taplines are that former President Trump still leads

139
00:07:58,240 --> 00:08:01,640
Speaker 5: Biden about forty eight percent to forty four percent across

140
00:08:01,680 --> 00:08:05,320
Speaker 5: the seven swing states that we have been surveying for

141
00:08:05,360 --> 00:08:08,160
Speaker 5: the last several months. And you know, this sort of

142
00:08:08,200 --> 00:08:11,360
Speaker 5: builds on the success that we have seen in other

143
00:08:11,400 --> 00:08:15,160
Speaker 5: polls with him where he has been leading Biden. I

144
00:08:15,200 --> 00:08:17,800
Speaker 5: think what's interesting is that about forty six percent of

145
00:08:17,840 --> 00:08:21,880
Speaker 5: the people we surveyed expressed some concern about foreign interference.

146
00:08:22,440 --> 00:08:24,800
Speaker 5: And then the other sort of big takeaway is that

147
00:08:24,840 --> 00:08:29,880
Speaker 5: Trump is leading in states like North Carolina, Arizona, Georgia,

148
00:08:29,920 --> 00:08:32,520
Speaker 5: which is a state with a lot of African Americans,

149
00:08:32,520 --> 00:08:34,880
Speaker 5: which will be key for Biden to win, and then

150
00:08:35,160 --> 00:08:37,000
Speaker 5: they're tied in Nevada, Nancy.

151
00:08:37,120 --> 00:08:39,280
Speaker 1: One of the things we're picking up on during this

152
00:08:39,720 --> 00:08:42,360
Speaker 1: voter nomics is this idea of echo chambers and that

153
00:08:42,520 --> 00:08:45,199
Speaker 1: different voter groups are sort of talking to with themselves

154
00:08:45,200 --> 00:08:49,160
Speaker 1: and not really straying outside. And the poll underscoores that

155
00:08:49,240 --> 00:08:51,480
Speaker 1: it's deeply partisan. So you have nine out of ten

156
00:08:51,520 --> 00:08:54,280
Speaker 1: Republicans saying the economy was better off under Trump and

157
00:08:54,360 --> 00:08:56,640
Speaker 1: only one out of ten Democrats thinking that. So you've

158
00:08:56,640 --> 00:09:02,600
Speaker 1: got this very very strong expression of how partisan this

159
00:09:02,679 --> 00:09:03,240
Speaker 1: election is.

160
00:09:03,640 --> 00:09:06,480
Speaker 5: Absolutely and you just see that across the board with

161
00:09:06,559 --> 00:09:09,679
Speaker 5: almost every policy area. I mean, you see that, you know,

162
00:09:09,760 --> 00:09:13,000
Speaker 5: Trump really earns high marks on the economy from you know,

163
00:09:13,080 --> 00:09:18,199
Speaker 5: all Republican voters. You see that also with foreign policy, housing, immigration.

164
00:09:19,400 --> 00:09:22,120
Speaker 5: It's just it is so polarized that it's almost like

165
00:09:22,160 --> 00:09:25,400
Speaker 5: two different countries at this point, the Democratic States and

166
00:09:25,440 --> 00:09:28,520
Speaker 5: the Republican States. And that's why we have focused on

167
00:09:28,559 --> 00:09:31,440
Speaker 5: these swing states, because you know, there's so much of

168
00:09:31,480 --> 00:09:34,040
Speaker 5: the US that's basically you already know how people are

169
00:09:34,040 --> 00:09:36,680
Speaker 5: going to vote, and these states are the ones that

170
00:09:36,679 --> 00:09:37,840
Speaker 5: are kind of up for grabs.

171
00:09:38,320 --> 00:09:42,120
Speaker 4: Nancy, are you surprised that the current Trump trial hasn't

172
00:09:42,160 --> 00:09:45,240
Speaker 4: had more of an impact on voting intentions so far?

173
00:09:45,480 --> 00:09:50,640
Speaker 5: I I'm actually not because I think that one, we're

174
00:09:50,679 --> 00:09:53,240
Speaker 5: still many months away from the election. Two, I think

175
00:09:53,280 --> 00:09:57,040
Speaker 5: Americans are very disengaged right now from this particular election.

176
00:09:57,679 --> 00:10:01,120
Speaker 5: I think most Americans don't want either President Biden or

177
00:10:01,280 --> 00:10:03,720
Speaker 5: former President Trump to be their candidates, and so I

178
00:10:03,720 --> 00:10:06,360
Speaker 5: think people aren't paying a ton of attention to the trial.

179
00:10:06,600 --> 00:10:08,199
Speaker 5: The fact of the matter is is that the first

180
00:10:08,240 --> 00:10:10,960
Speaker 5: indictment that Trump had actually was great for him. It

181
00:10:11,000 --> 00:10:13,720
Speaker 5: helped him with fundraising, It helped him with Republican base,

182
00:10:13,800 --> 00:10:16,960
Speaker 5: it was something that boosted him actually, So you know,

183
00:10:17,160 --> 00:10:18,760
Speaker 5: I just think it hasn't hurt him at all. And

184
00:10:18,800 --> 00:10:20,960
Speaker 5: I think if there is a conviction, which we have

185
00:10:21,000 --> 00:10:23,240
Speaker 5: to wait and see, it will only help him.

186
00:10:23,480 --> 00:10:26,640
Speaker 3: But that is interesting, Nancy, because as you know, polls

187
00:10:26,679 --> 00:10:30,160
Speaker 3: of Republicans had suggested that around a third of his

188
00:10:30,360 --> 00:10:36,600
Speaker 3: support would consider it material if he got a fatherly conviction.

189
00:10:36,760 --> 00:10:38,200
Speaker 3: You think that's just gone away now.

190
00:10:38,480 --> 00:10:41,120
Speaker 5: I'm just not sure that that will hold and that

191
00:10:41,160 --> 00:10:44,240
Speaker 5: will be the thing that will sway people on election

192
00:10:44,400 --> 00:10:48,040
Speaker 5: day versus things like immigration and the economy. I think

193
00:10:48,080 --> 00:10:50,880
Speaker 5: that people's sense of the economy, in their dismay with

194
00:10:50,920 --> 00:10:53,280
Speaker 5: it is so strong that I'm just not sure that

195
00:10:53,280 --> 00:10:56,360
Speaker 5: that will overshadow a conviction.

196
00:10:57,679 --> 00:10:59,320
Speaker 3: The very top line I saw in one of the

197
00:10:59,320 --> 00:11:02,360
Speaker 3: headlines around the poll is that you know, a half

198
00:11:02,559 --> 00:11:08,439
Speaker 3: and that's Democrats and Republicans fear violence around the election,

199
00:11:09,200 --> 00:11:11,960
Speaker 3: which I'm not sure whether we were asking people that

200
00:11:12,120 --> 00:11:16,280
Speaker 3: in the last election or certainly you know, last twenty years,

201
00:11:16,360 --> 00:11:17,959
Speaker 3: but that seems very striking.

202
00:11:18,000 --> 00:11:19,880
Speaker 5: It is striking, and I think that's all because of

203
00:11:19,960 --> 00:11:23,760
Speaker 5: January sixth and the violence and the attempted insurrection that

204
00:11:23,840 --> 00:11:27,360
Speaker 5: happened after the last election. It has been so interesting

205
00:11:27,440 --> 00:11:30,439
Speaker 5: covering politics in the week of that because a ton

206
00:11:30,559 --> 00:11:33,560
Speaker 5: of the Republican Party has basically rewritten history on that

207
00:11:33,720 --> 00:11:36,400
Speaker 5: and come to see them as a protest. You know,

208
00:11:36,600 --> 00:11:39,440
Speaker 5: not that big of a deal. But I think that

209
00:11:39,840 --> 00:11:43,360
Speaker 5: regardless of who wins in November, Democrats or Republicans, there's

210
00:11:43,360 --> 00:11:46,040
Speaker 5: a fear that whoever is the loser, there will be

211
00:11:46,080 --> 00:11:48,160
Speaker 5: a lot of protests and there will be violence. And

212
00:11:48,440 --> 00:11:52,880
Speaker 5: I think it's really a stark data point in this poll.

213
00:11:53,559 --> 00:11:57,760
Speaker 3: I did notice we had a fascinating piece by Josh

214
00:11:57,840 --> 00:12:01,960
Speaker 3: Green earlier the last couple of days that the posters

215
00:12:02,040 --> 00:12:06,440
Speaker 3: and focus people taking focus groups on both sides weren't

216
00:12:06,440 --> 00:12:09,280
Speaker 3: looking for it, but kept on tripping up over people

217
00:12:09,520 --> 00:12:13,920
Speaker 3: concerned that President Trump, if re elected, would then try

218
00:12:14,000 --> 00:12:16,800
Speaker 3: to stay in office beyond the four years and violate

219
00:12:17,880 --> 00:12:19,840
Speaker 3: the constitutional limit on two terms.

220
00:12:20,040 --> 00:12:22,120
Speaker 5: Yeah. I love that piece as well, and I think

221
00:12:22,160 --> 00:12:25,600
Speaker 5: that it is a concern among particularly Independents, and it's

222
00:12:25,600 --> 00:12:28,360
Speaker 5: something that Trump keeps joking about. I mean, it's not

223
00:12:28,520 --> 00:12:31,400
Speaker 5: really something that's just in voter's minds. It's something that

224
00:12:31,440 --> 00:12:36,400
Speaker 5: he keeps talking about. He spoke before the National Rifle

225
00:12:36,440 --> 00:12:40,560
Speaker 5: Association in Dallas, Texas, on Saturday, and he joked about,

226
00:12:41,000 --> 00:12:43,240
Speaker 5: you know, winning a third term. You know, in the

227
00:12:43,320 --> 00:12:45,880
Speaker 5: US you can only win two terms. That's what the

228
00:12:45,960 --> 00:12:48,760
Speaker 5: law says now, and so he joked about serving even

229
00:12:48,840 --> 00:12:51,199
Speaker 5: beyond that. I mean, he has made these comments before,

230
00:12:51,240 --> 00:12:53,439
Speaker 5: so it's not really out of left field that this

231
00:12:53,480 --> 00:12:55,920
Speaker 5: is something that is on the minds of American voters.

232
00:12:56,160 --> 00:12:58,000
Speaker 5: But I thought it was interesting that it's coming up

233
00:12:58,040 --> 00:12:59,080
Speaker 5: in these focus groups.

234
00:12:59,120 --> 00:13:02,920
Speaker 1: Now as Joshua, it's NANSI. Do you enjoy the TV

235
00:13:03,000 --> 00:13:03,880
Speaker 1: debate spin rooms?

236
00:13:04,840 --> 00:13:07,800
Speaker 5: I think they're fine. I mean, you know, it's interesting though,

237
00:13:07,880 --> 00:13:11,200
Speaker 5: is that the debates that they recently agree to, you know,

238
00:13:11,679 --> 00:13:13,400
Speaker 5: they will have the spin rooms, I'm sure, but they're

239
00:13:13,440 --> 00:13:15,480
Speaker 5: not going to be before an audience, and so it

240
00:13:15,559 --> 00:13:18,240
Speaker 5: will be interesting to see, you know, how those ended

241
00:13:18,320 --> 00:13:21,439
Speaker 5: up looking on TV, but then also what they look

242
00:13:21,559 --> 00:13:22,479
Speaker 5: like afterwards.

243
00:13:22,600 --> 00:13:24,079
Speaker 1: Why have they not going for an audience?

244
00:13:24,320 --> 00:13:26,920
Speaker 5: I think that they It was one of the conditions

245
00:13:26,920 --> 00:13:29,840
Speaker 5: that Biden and his team put forward to debate Trump.

246
00:13:29,960 --> 00:13:32,959
Speaker 5: I think Trump is so masterful at sort of owning

247
00:13:33,000 --> 00:13:36,920
Speaker 5: the media, owning the room, sort of overtaking a situation,

248
00:13:37,040 --> 00:13:39,640
Speaker 5: and we saw that in a CNN town hall that

249
00:13:39,720 --> 00:13:42,000
Speaker 5: he did just with him and a CNN moderator. He

250
00:13:42,120 --> 00:13:45,359
Speaker 5: kind of just flipped the whole situation to his advantage.

251
00:13:45,800 --> 00:13:48,360
Speaker 5: And it was a very sympathetic audience and they kept

252
00:13:48,400 --> 00:13:51,040
Speaker 5: sort of interrupting the moderator, interrupting the follow up questions,

253
00:13:51,080 --> 00:13:53,040
Speaker 5: and I think the Biden people wanted to feel like

254
00:13:53,080 --> 00:13:55,520
Speaker 5: it was you know, President Biden really had a fair

255
00:13:55,520 --> 00:13:58,120
Speaker 5: shot to make his points, and that people stuck to

256
00:13:58,160 --> 00:14:00,319
Speaker 5: the time limits, and they felt like if there was

257
00:14:00,320 --> 00:14:03,200
Speaker 5: a studio audience that person, you know, it would have

258
00:14:03,320 --> 00:14:04,640
Speaker 5: just been even more disrupted.

259
00:14:04,840 --> 00:14:08,360
Speaker 3: But it's a huge gamble which we will be looking

260
00:14:08,360 --> 00:14:10,680
Speaker 3: at to see how it turns out. Nancy Cook, thank

261
00:14:10,679 --> 00:14:21,160
Speaker 3: you so much, thanks for having me. So now onto

262
00:14:21,160 --> 00:14:27,000
Speaker 3: that interview with Audrey Tang, Taiwan's former Minister of Digital Affairs.

263
00:14:27,000 --> 00:14:31,680
Speaker 3: She's literally just stepped down and she feels Taiwan has

264
00:14:31,720 --> 00:14:33,720
Speaker 3: been many years ahead of the rest of the world

265
00:14:33,760 --> 00:14:38,560
Speaker 3: in having to respond to the threat of misinformation and

266
00:14:38,760 --> 00:14:42,800
Speaker 3: cyber wars influencing elections. So I started by asking her

267
00:14:43,040 --> 00:14:45,280
Speaker 3: how does she think we're doing so far and what

268
00:14:45,320 --> 00:14:47,920
Speaker 3: do we need to learn before going to the ballot

269
00:14:47,920 --> 00:14:48,960
Speaker 3: boxes later in the year.

270
00:14:50,200 --> 00:14:52,720
Speaker 2: Taiwan has been on the forefront because we're the most

271
00:14:52,760 --> 00:14:56,840
Speaker 2: heavily targeted according to v them for the past decade.

272
00:14:57,000 --> 00:14:58,840
Speaker 2: Every year, we're at the top of the world in

273
00:14:58,920 --> 00:15:02,360
Speaker 2: terms of the our information manipulation received, and we know

274
00:15:02,520 --> 00:15:08,560
Speaker 2: something about information manipulation and it's not just elections. When

275
00:15:08,640 --> 00:15:11,720
Speaker 2: the Ministry of Digital Affairs first started in August twenty

276
00:15:11,760 --> 00:15:14,960
Speaker 2: twenty two, a few weeks before that, Sensi Pelosi, then

277
00:15:15,080 --> 00:15:18,280
Speaker 2: US Speaker visits the Taiwan in historic visits, and we

278
00:15:18,360 --> 00:15:22,480
Speaker 2: saw the whole spectrum of destructive attacked, tack and leak,

279
00:15:22,720 --> 00:15:27,400
Speaker 2: encouraging physical protests and all that in a very coordinated way.

280
00:15:27,560 --> 00:15:32,160
Speaker 2: It's about creating a information vacuum. It is about having

281
00:15:32,200 --> 00:15:35,840
Speaker 2: this narrative that democracy only leads to chaos and democracy

282
00:15:35,880 --> 00:15:40,320
Speaker 2: never delivers. That is the overarching narrative for the authoritarian attacks.

283
00:15:40,440 --> 00:15:43,480
Speaker 2: So I think one lesson is just to pre bunk,

284
00:15:43,800 --> 00:15:46,280
Speaker 2: to let the people know that this is happening. This

285
00:15:46,320 --> 00:15:49,640
Speaker 2: is not about pro a candidate or against a candidate.

286
00:15:49,880 --> 00:15:53,920
Speaker 2: This is about an overarching narrative that says democracy never delivers,

287
00:15:54,040 --> 00:15:56,520
Speaker 2: democracy only leads to chaos. And if you pre bunk it,

288
00:15:56,560 --> 00:15:59,040
Speaker 2: if you'll let everybody know that this is the narrative,

289
00:15:59,040 --> 00:16:02,440
Speaker 2: they're going to push the people built immune system antibodies

290
00:16:02,520 --> 00:16:05,160
Speaker 2: in their mind even before that attack actually happens.

291
00:16:05,800 --> 00:16:08,520
Speaker 3: So just to explain that Nancy Pelosi visit, you know,

292
00:16:08,640 --> 00:16:12,080
Speaker 3: obviously the Chinese government had not wanted any senior US

293
00:16:12,200 --> 00:16:15,560
Speaker 3: politician to visit Taiwan, so that's why you had that

294
00:16:15,800 --> 00:16:19,000
Speaker 3: kind of concerted attack. And the more recent time when

295
00:16:19,040 --> 00:16:22,120
Speaker 3: people were concerned about Taiwan was around the time of

296
00:16:22,160 --> 00:16:24,840
Speaker 3: the election at the start of this year, were you

297
00:16:25,080 --> 00:16:27,760
Speaker 3: applying some of the same lessons for that.

298
00:16:28,000 --> 00:16:32,280
Speaker 2: Yeah, definitely. For example, in every election, we already know

299
00:16:32,800 --> 00:16:35,960
Speaker 2: that the counting process is going to be the main

300
00:16:36,000 --> 00:16:39,840
Speaker 2: attack for not just every attack, but mostly information operation.

301
00:16:40,480 --> 00:16:44,000
Speaker 2: People would accuse that the counting was rigged. People would

302
00:16:44,040 --> 00:16:47,960
Speaker 2: accuse the election officials of siding with one particular candidate,

303
00:16:48,000 --> 00:16:50,800
Speaker 2: People would dispute the result of the election, and so on.

304
00:16:50,840 --> 00:16:55,760
Speaker 2: These are all traditional information operation targets. In Taiwan, we

305
00:16:55,920 --> 00:17:00,160
Speaker 2: countered that by participatory counting. So for the past few

306
00:17:00,240 --> 00:17:03,720
Speaker 2: years we restructured our accounting process. We have a paper

307
00:17:03,840 --> 00:17:08,400
Speaker 2: only ballot, and so when people count those paper only ballots,

308
00:17:08,440 --> 00:17:11,560
Speaker 2: they take each and everyone out, they show it to

309
00:17:11,880 --> 00:17:15,240
Speaker 2: three different angles, and the three major parties, their YouTubers,

310
00:17:15,240 --> 00:17:19,680
Speaker 2: their quarders, their participatory counters, they all film the process,

311
00:17:19,840 --> 00:17:22,920
Speaker 2: and so each and every paper ballot was counted this way.

312
00:17:23,160 --> 00:17:26,240
Speaker 2: So we didn't use digital technology to speed up the

313
00:17:26,320 --> 00:17:30,600
Speaker 2: tallying process. Even though maybe electronic talenting can finish what

314
00:17:31,040 --> 00:17:34,080
Speaker 2: used to take four hours in forty minutes, we conclude

315
00:17:34,080 --> 00:17:36,640
Speaker 2: it's not worth it. The participation is important because when

316
00:17:36,720 --> 00:17:41,800
Speaker 2: the inevitable information operation happens right after the counting process

317
00:17:41,840 --> 00:17:44,880
Speaker 2: concluded that there was rigged counting and so and so forth.

318
00:17:45,080 --> 00:17:47,679
Speaker 2: Actually it didn't spread very far, even though we have

319
00:17:47,800 --> 00:17:51,080
Speaker 2: evidence that foreign information manipulators put a lot of resource

320
00:17:51,359 --> 00:17:55,080
Speaker 2: in getting all those conspiracy theories to spread. All the

321
00:17:55,119 --> 00:17:59,119
Speaker 2: three party leaders, they have their YouTubers and their campaign

322
00:17:59,160 --> 00:18:03,679
Speaker 2: people in that particular counting station. It's all documented. So

323
00:18:03,800 --> 00:18:06,600
Speaker 2: all they can say after the fact is that there's

324
00:18:06,680 --> 00:18:10,080
Speaker 2: no riaking of the election. So it limits the basic

325
00:18:10,119 --> 00:18:14,840
Speaker 2: reproduction number that are zero value of those information manipulations.

326
00:18:14,840 --> 00:18:17,520
Speaker 2: So even though they threw a lot of resources into

327
00:18:17,640 --> 00:18:21,640
Speaker 2: spreading this particular message, we fixed this problem by inviting

328
00:18:21,680 --> 00:18:25,320
Speaker 2: everybody into the counting process. This is something that extronic

329
00:18:25,760 --> 00:18:29,240
Speaker 2: tally cannot do as easily, oh do.

330
00:18:29,359 --> 00:18:33,760
Speaker 1: The way you're describing recent events is very sophisticated digital literacy.

331
00:18:34,400 --> 00:18:38,439
Speaker 1: Tell us your observations looking at these huge elections in

332
00:18:38,480 --> 00:18:41,360
Speaker 1: the UK in America and how resilient you think those

333
00:18:41,400 --> 00:18:42,440
Speaker 1: electorates really are.

334
00:18:43,600 --> 00:18:47,160
Speaker 2: In Taiwan, we don't even use the word digital literacy

335
00:18:47,359 --> 00:18:50,920
Speaker 2: or media literacy anymore in our curriculum. We always say

336
00:18:51,000 --> 00:18:54,960
Speaker 2: digital media competence because literacy is when you are viewing

337
00:18:55,040 --> 00:18:57,800
Speaker 2: the news, maybe with a critical eye, but still mostly

338
00:18:57,800 --> 00:19:01,080
Speaker 2: as a consumer, whereas competent is when you're a co

339
00:19:01,160 --> 00:19:04,360
Speaker 2: producer of information. And the great thing about getting high

340
00:19:04,400 --> 00:19:07,720
Speaker 2: schoolers or even primary schoolers into the collaborative fact checking

341
00:19:08,040 --> 00:19:11,760
Speaker 2: is that it changes their family dynamic, usually the children

342
00:19:11,840 --> 00:19:14,800
Speaker 2: when they introduce a new platform, a new ecosystem, and

343
00:19:14,840 --> 00:19:18,440
Speaker 2: so the parents and grandparents are sympathetic to it. And

344
00:19:18,520 --> 00:19:20,600
Speaker 2: what we have found is that it is not the

345
00:19:20,720 --> 00:19:24,000
Speaker 2: check facts that inoculates the mind. It is the act

346
00:19:24,119 --> 00:19:27,359
Speaker 2: of going through fact checking, a thinking like a journalist

347
00:19:27,520 --> 00:19:30,720
Speaker 2: that inoculates the mind, which is why it's so important

348
00:19:30,800 --> 00:19:34,280
Speaker 2: to combine that with the curriculum and with this or

349
00:19:34,480 --> 00:19:35,720
Speaker 2: family exercise.

350
00:19:36,080 --> 00:19:39,080
Speaker 1: When you look at America and the UK and the

351
00:19:39,160 --> 00:19:42,000
Speaker 1: many many other polls taking place this year around the world,

352
00:19:42,480 --> 00:19:43,920
Speaker 1: do you think people know what's coming?

353
00:19:44,920 --> 00:19:48,200
Speaker 2: I think what's really needed is that it comes from

354
00:19:48,480 --> 00:19:52,000
Speaker 2: not just credibly neutral sources, which I understand in the

355
00:19:52,119 --> 00:19:56,320
Speaker 2: US it's difficult to combine now, but rather from each

356
00:19:56,320 --> 00:19:59,679
Speaker 2: and every party, from each and every denomination, from each

357
00:19:59,680 --> 00:20:03,119
Speaker 2: and ever even religion and so on, community leaders and

358
00:20:03,160 --> 00:20:05,679
Speaker 2: so on, such as published a book on this is

359
00:20:05,840 --> 00:20:12,160
Speaker 2: about collaborating across people who were like ideologically or politically

360
00:20:12,200 --> 00:20:16,760
Speaker 2: opposite of each other, but each side agreed that democracy

361
00:20:17,000 --> 00:20:20,639
Speaker 2: is a process that requires some commonly agreed facts, and

362
00:20:20,680 --> 00:20:23,760
Speaker 2: for that we need cross checking instead of just taking

363
00:20:23,800 --> 00:20:27,240
Speaker 2: one institute and upholding it as the most rigorous. So

364
00:20:27,320 --> 00:20:31,080
Speaker 2: a cross checking ecosystem is always superior to just a

365
00:20:31,160 --> 00:20:34,880
Speaker 2: single like this Information Overside institute or things like that

366
00:20:35,040 --> 00:20:37,879
Speaker 2: when it comes to pre bunking and debunking issues. And

367
00:20:37,920 --> 00:20:42,280
Speaker 2: the other thing, especially for career public servants, is that

368
00:20:42,680 --> 00:20:46,159
Speaker 2: even if you didn't anticipate something and you cannot pre

369
00:20:46,240 --> 00:20:49,520
Speaker 2: bunk it, as long as you respond within an hour,

370
00:20:50,080 --> 00:20:53,280
Speaker 2: it's as good as if you've pre bunked it. So

371
00:20:53,400 --> 00:20:56,479
Speaker 2: we ensure in Taiwan that if we missed some pre

372
00:20:56,560 --> 00:21:01,040
Speaker 2: bunking targets, we always respond with a hopeful, slightly humorous

373
00:21:01,080 --> 00:21:05,880
Speaker 2: message mes media sixteen minutes or less after we detected

374
00:21:05,880 --> 00:21:08,800
Speaker 2: that it is going to trend. And the reason why

375
00:21:08,960 --> 00:21:11,720
Speaker 2: is that the career of public service through the press

376
00:21:11,720 --> 00:21:15,840
Speaker 2: release channels, actually reaches more people than conspiracy theories, and

377
00:21:15,880 --> 00:21:18,639
Speaker 2: if it's a little bit humorous, it also spreads faster

378
00:21:19,160 --> 00:21:22,680
Speaker 2: than the conspiracy theories. So for everag person they actually

379
00:21:23,000 --> 00:21:27,240
Speaker 2: received the clarification first before they received this information.

380
00:21:28,000 --> 00:21:33,000
Speaker 3: You were a pioneer, Audrey, developing open source technology before

381
00:21:33,359 --> 00:21:36,399
Speaker 3: going into government. But also you've been a force for

382
00:21:36,440 --> 00:21:41,440
Speaker 3: this very collaborative, transparent process which you were just describing.

383
00:21:41,720 --> 00:21:43,520
Speaker 3: I mean, I think that's built up a level of

384
00:21:43,640 --> 00:21:48,360
Speaker 3: trust which you're not always going to get with other

385
00:21:48,480 --> 00:21:51,680
Speaker 3: kinds of government fact checking in other countries. In fact,

386
00:21:51,880 --> 00:21:54,800
Speaker 3: I was struck by a recent example in Senegal where

387
00:21:54,840 --> 00:21:59,040
Speaker 3: there'd been a fake news law that actually was associated

388
00:21:59,119 --> 00:22:02,439
Speaker 3: with the cumbent government in effect trying to prevent a

389
00:22:02,440 --> 00:22:06,080
Speaker 3: lot of opposition discussion on TikTok and in other places.

390
00:22:06,280 --> 00:22:10,080
Speaker 3: And funny enough, that law backfired because this opposition leader

391
00:22:10,560 --> 00:22:12,600
Speaker 3: had been somewhat silenced by those laws, but there was

392
00:22:12,640 --> 00:22:15,159
Speaker 3: so much curiosity about him that people went out and

393
00:22:15,160 --> 00:22:17,040
Speaker 3: found out about him and they ended up voting him

394
00:22:17,080 --> 00:22:21,520
Speaker 3: into office. But in a lot of cases, you know,

395
00:22:21,720 --> 00:22:24,399
Speaker 3: don't we see that it is in fact quite authoritarian

396
00:22:24,440 --> 00:22:28,080
Speaker 3: governments that are then using the language of fake news

397
00:22:28,440 --> 00:22:31,879
Speaker 3: and talking about citizens needing to be protected from it.

398
00:22:31,920 --> 00:22:36,960
Speaker 3: So how do you develop trust and avoid that problem?

399
00:22:37,200 --> 00:22:40,360
Speaker 2: Yes, I think this is a great observation. How one

400
00:22:40,760 --> 00:22:43,760
Speaker 2: is the tub of Asia when it comes to not

401
00:22:43,840 --> 00:22:47,800
Speaker 2: just freedom on the net, but also journalistic freedom. And

402
00:22:48,080 --> 00:22:51,120
Speaker 2: because my parents were both journalists, I would never say

403
00:22:51,160 --> 00:22:54,280
Speaker 2: the words fake news to them or to anyone really,

404
00:22:54,480 --> 00:23:00,399
Speaker 2: because to me, information manipulation is fundamentally about getting people

405
00:23:00,440 --> 00:23:04,439
Speaker 2: to distrust the democratic process. This is not about pro

406
00:23:04,600 --> 00:23:08,320
Speaker 2: or anti establishment pro or any parties and so on,

407
00:23:08,600 --> 00:23:11,560
Speaker 2: or the polarization attacks that we have seen for most

408
00:23:11,560 --> 00:23:15,960
Speaker 2: of this year. After the generative AI capabilities getting enjoyed

409
00:23:16,080 --> 00:23:20,000
Speaker 2: by all the information manipulators, the easiest way is not

410
00:23:20,240 --> 00:23:23,560
Speaker 2: to push for fake narratives, but just to amplify the

411
00:23:23,600 --> 00:23:28,880
Speaker 2: parts of the opinions that looks like hate messages across

412
00:23:28,880 --> 00:23:32,760
Speaker 2: party lines, across ideological lines, and just put a lot

413
00:23:32,760 --> 00:23:36,000
Speaker 2: of resource in amplifying those polarized messages and then the

414
00:23:36,000 --> 00:23:40,520
Speaker 2: polarization person the democracy by yourself and fact checks can

415
00:23:40,640 --> 00:23:46,880
Speaker 2: do little about such polarization attacks because it's not factually wrong, right,

416
00:23:47,119 --> 00:23:49,280
Speaker 2: So I think a lot of the work that we're

417
00:23:49,320 --> 00:23:53,160
Speaker 2: now doing is again pre bunking. We introduce new laws

418
00:23:53,280 --> 00:23:56,880
Speaker 2: like the Anti Fraud Act currently in the Parliament that says,

419
00:23:56,920 --> 00:23:59,919
Speaker 2: if you want to have freedom of reach through everard,

420
00:24:00,840 --> 00:24:03,760
Speaker 2: you have to digitally sign your message, and not just

421
00:24:03,760 --> 00:24:07,240
Speaker 2: the people paying it, but also the people appearing in it.

422
00:24:07,240 --> 00:24:09,920
Speaker 2: It's from a celebrity, the celebrity need to sign it digitally,

423
00:24:10,200 --> 00:24:14,880
Speaker 2: and that applies to foreign advertisements too, so there's no exceptions.

424
00:24:14,960 --> 00:24:18,800
Speaker 2: It wants to ensure the authenticity of the actor without

425
00:24:18,960 --> 00:24:22,080
Speaker 2: saying anything about the content. We've seen more and more

426
00:24:22,160 --> 00:24:25,680
Speaker 2: that the content level laws do a lot of collateral harm,

427
00:24:25,720 --> 00:24:28,520
Speaker 2: even with the best intentions, exactly as you pointed out,

428
00:24:28,760 --> 00:24:32,560
Speaker 2: because it also interferes with the journalistic work. And I

429
00:24:32,640 --> 00:24:35,000
Speaker 2: understand that there are similar laws in the works in

430
00:24:35,040 --> 00:24:37,840
Speaker 2: both US and the UK, and my sincere hope is

431
00:24:37,880 --> 00:24:41,280
Speaker 2: that these can be passed before the election, because otherwise

432
00:24:41,359 --> 00:24:44,479
Speaker 2: it's not just fake versus true anymore. It is a

433
00:24:44,480 --> 00:24:48,600
Speaker 2: lot about just inauthentic portrayal of celebrities and so on.

434
00:24:49,600 --> 00:24:54,080
Speaker 3: Audrey, I would say, when we've spoken in Taipei the

435
00:24:54,160 --> 00:24:56,240
Speaker 3: end of last year, I would say you were quite

436
00:24:56,280 --> 00:24:59,720
Speaker 3: calm about some of the sort of straightforward deep fakes

437
00:25:00,080 --> 00:25:01,879
Speaker 3: that people have talked a lot about here in the

438
00:25:01,880 --> 00:25:04,600
Speaker 3: context of the election, Things coming out just before the

439
00:25:04,640 --> 00:25:07,879
Speaker 3: election that show one of the candidates saying something or

440
00:25:07,920 --> 00:25:12,080
Speaker 3: doing something but I noted that you were much more

441
00:25:12,200 --> 00:25:16,800
Speaker 3: concerned about the potential for Ai to produce very kind

442
00:25:16,840 --> 00:25:22,480
Speaker 3: of individualized fakes. So you know, not just a fake video,

443
00:25:23,359 --> 00:25:28,760
Speaker 3: but technology allowing manipulator to build up a relationship with

444
00:25:28,800 --> 00:25:32,800
Speaker 3: an individual person. So maybe just talk us through talk

445
00:25:32,840 --> 00:25:34,640
Speaker 3: us through that so we too can be frightened.

446
00:25:35,320 --> 00:25:39,080
Speaker 2: So after the election in January, I think I was

447
00:25:39,200 --> 00:25:43,280
Speaker 2: right in not worried about the polarization attacks because the

448
00:25:43,480 --> 00:25:47,080
Speaker 2: effective polarization, how people hit each other cross party line,

449
00:25:47,240 --> 00:25:50,480
Speaker 2: went into a historic love All three parties felt that

450
00:25:50,560 --> 00:25:53,240
Speaker 2: they have won a little bit, so there's no winner

451
00:25:53,280 --> 00:25:56,160
Speaker 2: takes all, loser loses all dynamic and it's very good

452
00:25:56,240 --> 00:25:59,280
Speaker 2: for the health of our democracy and mental health for

453
00:25:59,440 --> 00:26:03,960
Speaker 2: everyone involved. But I do see that this direct communication,

454
00:26:04,200 --> 00:26:08,840
Speaker 2: this precision persuation as an ongoing threat. The idea is

455
00:26:08,880 --> 00:26:12,080
Speaker 2: that if you have an addictive app that people use

456
00:26:12,400 --> 00:26:15,879
Speaker 2: in Taiwan, for example, the short form videos, people like

457
00:26:15,960 --> 00:26:18,960
Speaker 2: it just like any other country. We're a little bit

458
00:26:19,040 --> 00:26:21,640
Speaker 2: fortunate in that we've been seeing for quite a few

459
00:26:21,680 --> 00:26:25,760
Speaker 2: years that TikTok is a harmful product for cybersecurity. So

460
00:26:25,960 --> 00:26:29,920
Speaker 2: it only has about a quarter of installation based in Taiwan.

461
00:26:30,280 --> 00:26:32,840
Speaker 2: And only about one tenth of that use it as

462
00:26:32,880 --> 00:26:36,280
Speaker 2: their primary platform. So we're not as addicted as some

463
00:26:36,400 --> 00:26:39,800
Speaker 2: other jurisdictions are to TikTok. But TikTok is just one

464
00:26:39,800 --> 00:26:43,880
Speaker 2: example because what it does is already has a reward model, right,

465
00:26:43,960 --> 00:26:48,040
Speaker 2: it knows what takes you and what addicts you. And

466
00:26:48,119 --> 00:26:51,760
Speaker 2: so instead of just spreading one single message to try

467
00:26:51,800 --> 00:26:54,240
Speaker 2: to get people to retweet or whatever, that was like

468
00:26:54,280 --> 00:26:59,160
Speaker 2: the previous wave of information manipulation, nowadays decision targeting AI

469
00:26:59,480 --> 00:27:03,919
Speaker 2: can just push a series of messages, let you feel

470
00:27:04,000 --> 00:27:07,480
Speaker 2: a little bit more disenfranchised, a little bit more disempowered,

471
00:27:07,520 --> 00:27:10,480
Speaker 2: a little bit more empathy around the democratic process. But

472
00:27:10,640 --> 00:27:12,920
Speaker 2: if you are addicted to it, and if you watch

473
00:27:12,920 --> 00:27:15,959
Speaker 2: it hours and hours every day, then the end result

474
00:27:16,160 --> 00:27:19,560
Speaker 2: is that you lose energy to participate in the democratic processes.

475
00:27:19,800 --> 00:27:23,520
Speaker 2: So it is even more insidious and chronic than polarization.

476
00:27:24,040 --> 00:27:26,440
Speaker 2: And I think this is something that we need really

477
00:27:26,520 --> 00:27:27,320
Speaker 2: to watch out for.

478
00:27:28,520 --> 00:27:30,520
Speaker 3: Talk us through that. But how can one pre bunk

479
00:27:30,640 --> 00:27:33,920
Speaker 3: or pre inoculate against that. It just feels so sophisticated.

480
00:27:35,600 --> 00:27:39,400
Speaker 2: Yeah, one way is simply to let people know that

481
00:27:39,480 --> 00:27:43,199
Speaker 2: this form of addiction, everything else being equal, is not

482
00:27:43,280 --> 00:27:45,439
Speaker 2: good for him. In the US, of course, as we know,

483
00:27:45,800 --> 00:27:49,760
Speaker 2: or TikTok in particular, they are now setting a timeline

484
00:27:49,800 --> 00:27:52,560
Speaker 2: so that it needs to divest right and need to

485
00:27:52,760 --> 00:27:56,840
Speaker 2: sell to a US operator who hopefully will instill a

486
00:27:57,440 --> 00:28:02,400
Speaker 2: more digital democracy capabilities into the app, because if they're

487
00:28:02,440 --> 00:28:06,280
Speaker 2: already addicted to it. Nowadays, there's research that says if

488
00:28:06,320 --> 00:28:11,040
Speaker 2: you adjust the sorting algorithm a little bit to promote,

489
00:28:11,119 --> 00:28:14,920
Speaker 2: for example, authentic self disclosure, or promote a little bit

490
00:28:14,960 --> 00:28:19,159
Speaker 2: more vulnerability, a little bit more ideas of communicating across

491
00:28:19,200 --> 00:28:22,879
Speaker 2: society differences, a little bit more collective sense making and

492
00:28:22,960 --> 00:28:26,680
Speaker 2: all that instead of the addictive entertainment stuff, it actually

493
00:28:26,720 --> 00:28:30,920
Speaker 2: results in the same amount of engagement, whereas it makes

494
00:28:30,920 --> 00:28:34,159
Speaker 2: people care more about democracies or maybe OI techs is

495
00:28:34,240 --> 00:28:37,800
Speaker 2: a few hyper parameter changes and then it gets people

496
00:28:37,920 --> 00:28:40,960
Speaker 2: on social media to be more pro social rather than antisocial.

497
00:28:41,120 --> 00:28:45,280
Speaker 2: So maybe that's another experiment that whatever cause TikTok can try.

498
00:28:46,000 --> 00:28:48,479
Speaker 3: Got it. I love the idea of after an election,

499
00:28:48,720 --> 00:28:52,360
Speaker 3: all parties feeling like they want a bit and feeling

500
00:28:52,360 --> 00:28:55,080
Speaker 3: like they hate each other less. We can only hope

501
00:28:56,040 --> 00:28:58,760
Speaker 3: you mentioned the US approach. There's obviously been a specific

502
00:28:58,840 --> 00:29:02,840
Speaker 3: bill there addressing to which may end up applying to

503
00:29:02,880 --> 00:29:05,719
Speaker 3: other platforms. The EU has had its own package of

504
00:29:06,480 --> 00:29:10,880
Speaker 3: digital services legislation which it brought out at the end

505
00:29:10,880 --> 00:29:14,720
Speaker 3: of the year. I mean, how would you compare the

506
00:29:14,800 --> 00:29:18,120
Speaker 3: EUSE approach to that of say the US or the UK.

507
00:29:18,320 --> 00:29:22,560
Speaker 2: Yeah, I think that you approach of assigning more liabilities

508
00:29:22,680 --> 00:29:26,920
Speaker 2: and more duties to the gatekeepers is definitely a robust

509
00:29:27,080 --> 00:29:30,960
Speaker 2: approach because as soon as some platform reaches a certain

510
00:29:31,000 --> 00:29:33,920
Speaker 2: protest percentage of the people, it doesn't matter where it

511
00:29:33,960 --> 00:29:36,600
Speaker 2: comes from, it now has to comply with these duties.

512
00:29:36,840 --> 00:29:39,840
Speaker 2: In fact, in our INTI Fraud Act, that's precisely the

513
00:29:39,920 --> 00:29:43,520
Speaker 2: approach that we take. So even if you're a foreign operator,

514
00:29:43,680 --> 00:29:46,560
Speaker 2: as long as you hit x percent of active Tawanese

515
00:29:46,640 --> 00:29:50,240
Speaker 2: user space and new advertisements, you're required.

516
00:29:51,040 --> 00:29:53,680
Speaker 1: Audre, you've been really eloquent describing all the different policy

517
00:29:53,760 --> 00:29:56,320
Speaker 1: leavers and mechanisms that democracy can put in place to

518
00:29:56,320 --> 00:29:59,560
Speaker 1: try and protect itself or in your phrase, inoculate itself

519
00:29:59,600 --> 00:30:02,440
Speaker 1: against the the problems. But can you just reflect for

520
00:30:02,560 --> 00:30:06,440
Speaker 1: us just taking a step back on the threat, the enemy,

521
00:30:06,480 --> 00:30:09,760
Speaker 1: the stimulus, the people that are seeking to spread these untruths,

522
00:30:10,240 --> 00:30:12,120
Speaker 1: and I want to give you an example in the

523
00:30:12,200 --> 00:30:15,920
Speaker 1: UK a few years back. There's obviously been incidents since,

524
00:30:16,040 --> 00:30:18,520
Speaker 1: but there was a very high profile example where our

525
00:30:18,560 --> 00:30:22,080
Speaker 1: footballers missed a bunch of penalties in the Euros a

526
00:30:22,120 --> 00:30:25,520
Speaker 1: few information I prov on football tournament. There was abuse

527
00:30:26,080 --> 00:30:29,280
Speaker 1: and they huge abuse. Some of that abuse may have

528
00:30:29,320 --> 00:30:33,480
Speaker 1: come their way anyway from domestic social media accounts, but

529
00:30:33,680 --> 00:30:36,080
Speaker 1: equally there was evidence that some of it was also

530
00:30:36,280 --> 00:30:40,680
Speaker 1: stirred up from abroad. It seemed to be compounded, whipped up,

531
00:30:40,720 --> 00:30:44,880
Speaker 1: amplified by foreign social media accounts. Can you just describe

532
00:30:44,960 --> 00:30:48,120
Speaker 1: for us? Is that kind of phenomenon here to stay?

533
00:30:48,320 --> 00:30:49,880
Speaker 1: Is it going to get worse? Is it already very

534
00:30:49,920 --> 00:30:52,520
Speaker 1: bad and we just don't realize it. Just give us

535
00:30:52,560 --> 00:30:53,200
Speaker 1: a stock take.

536
00:30:53,760 --> 00:30:56,680
Speaker 2: So when it comes to elections, we can always start

537
00:30:56,720 --> 00:31:01,080
Speaker 2: from the ordinary, right, the classic sever trusions so we're

538
00:31:01,200 --> 00:31:05,080
Speaker 2: enabled to vote, Tempering and authorized access data theft and

539
00:31:05,160 --> 00:31:08,719
Speaker 2: things like that, and ransomware and so on, all all

540
00:31:08,760 --> 00:31:11,480
Speaker 2: these are still here to stay. They're not going away.

541
00:31:11,880 --> 00:31:15,960
Speaker 2: What worries us is that it's now evolving into what's

542
00:31:16,000 --> 00:31:21,320
Speaker 2: called hybrid capabilities. So instead of just addos attack denying

543
00:31:21,360 --> 00:31:26,920
Speaker 2: the service of governmental agencies, websites, coupling that with amplifying

544
00:31:27,120 --> 00:31:30,600
Speaker 2: information operation material so that when fact checkers want to

545
00:31:30,680 --> 00:31:33,160
Speaker 2: access the source of truth they found the source of

546
00:31:33,200 --> 00:31:36,560
Speaker 2: truth is denied service, whether it's a media outlet or

547
00:31:36,560 --> 00:31:40,360
Speaker 2: whether it's a ministrial website or as we mentioned, a

548
00:31:40,560 --> 00:31:44,040
Speaker 2: heck and leak about a campaign and so on, but

549
00:31:44,360 --> 00:31:47,960
Speaker 2: then using the leak material to do AI abuse and

550
00:31:48,040 --> 00:31:51,360
Speaker 2: defaikes and so on, as literally putting words into the

551
00:31:51,440 --> 00:31:53,920
Speaker 2: heck victim that they didn't say. But because the heck

552
00:31:54,040 --> 00:31:58,040
Speaker 2: material contains maybe videos or sufficient amount of sound files,

553
00:31:58,200 --> 00:32:00,600
Speaker 2: it sounds and looks very real and so on. So

554
00:32:00,640 --> 00:32:04,920
Speaker 2: these are the new capabilities new as of last year

555
00:32:05,040 --> 00:32:08,160
Speaker 2: or so that were extra word because then the severate

556
00:32:08,240 --> 00:32:11,960
Speaker 2: intrusions are bad enough. But previously it would take a

557
00:32:12,160 --> 00:32:16,600
Speaker 2: very resourceful adversary and many different trines before it reaches

558
00:32:16,680 --> 00:32:21,080
Speaker 2: the common people. But now amplified by this precision persuasion

559
00:32:21,120 --> 00:32:24,560
Speaker 2: and abuse and defails, they can afford to do each

560
00:32:24,640 --> 00:32:27,880
Speaker 2: try with minimal or negligible cost, and so they can

561
00:32:28,080 --> 00:32:30,920
Speaker 2: just keep trying until a narrative hits and then we

562
00:32:31,000 --> 00:32:34,080
Speaker 2: have a viral information manipulation. So which is why it

563
00:32:34,160 --> 00:32:37,480
Speaker 2: is so important to have this digital signature scheme because

564
00:32:37,480 --> 00:32:40,720
Speaker 2: it means that one person can only have one account.

565
00:32:41,000 --> 00:32:44,160
Speaker 2: If you want to open pseudonyms, that's fine, but you

566
00:32:44,320 --> 00:32:46,720
Speaker 2: sign it in a way it's called zero knowledge that

567
00:32:46,840 --> 00:32:49,680
Speaker 2: still lets people know that it is just a single person,

568
00:32:49,720 --> 00:32:53,080
Speaker 2: maybe anonymous, a single person behind it. No longer can

569
00:32:53,200 --> 00:32:56,800
Speaker 2: one simply register ten thousand accounts and wait for most

570
00:32:56,800 --> 00:32:59,200
Speaker 2: of them getting fund or whetherver it, but still activate

571
00:32:59,240 --> 00:33:02,400
Speaker 2: the rest of the account to look like it's astroturfing

572
00:33:02,440 --> 00:33:05,000
Speaker 2: and so on, look like there's a ground swell support

573
00:33:05,040 --> 00:33:07,000
Speaker 2: and so on, and all these are going to be

574
00:33:07,160 --> 00:33:10,840
Speaker 2: very difficult to tell from actual groundswell support because if

575
00:33:10,880 --> 00:33:14,920
Speaker 2: you click into each account's profile, live history, interaction, it

576
00:33:14,960 --> 00:33:19,520
Speaker 2: looks completely real. It's now generative AI made. So I

577
00:33:19,560 --> 00:33:22,920
Speaker 2: think digital signature really is one of the new bedrocks

578
00:33:23,000 --> 00:33:26,440
Speaker 2: that we need to instill into those larger platforms this year.

579
00:33:27,000 --> 00:33:29,040
Speaker 1: At the time, there was a debate, you know, on

580
00:33:29,080 --> 00:33:32,080
Speaker 1: the one hand, with suggesting that this has been uniquely

581
00:33:32,080 --> 00:33:34,760
Speaker 1: whipped up abroad and that this would never have happened

582
00:33:34,760 --> 00:33:37,600
Speaker 1: if it weren't for foreign media accounts making things worse.

583
00:33:38,080 --> 00:33:41,200
Speaker 1: But equally we have to focus and analyze ourselves, don't we.

584
00:33:41,520 --> 00:33:45,000
Speaker 2: That is certainly true. And also if you're on x

585
00:33:45,040 --> 00:33:48,320
Speaker 2: dot com or Twitter, there's this jury duty that you

586
00:33:48,400 --> 00:33:51,800
Speaker 2: can enter called community notes, and I participate in the

587
00:33:51,800 --> 00:33:55,600
Speaker 2: community notes just to improve my media competence. And if

588
00:33:55,640 --> 00:33:59,160
Speaker 2: you participate in the community note community, very quickly you

589
00:33:59,200 --> 00:34:02,760
Speaker 2: can see all sort of information manipulators going on, trying

590
00:34:02,800 --> 00:34:06,240
Speaker 2: to weigh the public opinion one way or another. And

591
00:34:07,120 --> 00:34:11,200
Speaker 2: I attended the Google Jigsaw conference a couple months ago,

592
00:34:11,480 --> 00:34:14,160
Speaker 2: and there were a lot of conversations around these and

593
00:34:14,280 --> 00:34:18,200
Speaker 2: Google Jigsaw did release a new set of capabilities, a

594
00:34:18,239 --> 00:34:21,600
Speaker 2: new set of API. Instead of just detecting the toxicity

595
00:34:21,719 --> 00:34:26,520
Speaker 2: and so on, it now also measures democratic bridge making abilities.

596
00:34:26,719 --> 00:34:29,400
Speaker 2: So for each and every note, you can actually gauge

597
00:34:29,560 --> 00:34:33,000
Speaker 2: how much pro social it is compared to your ordinary

598
00:34:33,080 --> 00:34:36,040
Speaker 2: hate messages and so on. So it's not just weeding

599
00:34:36,080 --> 00:34:39,560
Speaker 2: out the toxic or the hate, but actually nudging people

600
00:34:39,760 --> 00:34:43,200
Speaker 2: to provide pro social services online. And I think this

601
00:34:43,280 --> 00:34:45,880
Speaker 2: is also very helpful if it can be applied to

602
00:34:45,960 --> 00:34:47,600
Speaker 2: more social media platforms.

603
00:34:48,000 --> 00:34:50,800
Speaker 3: So I guess I mean, before we finished that, just

604
00:34:50,880 --> 00:34:53,040
Speaker 3: there's one fundamental thing which comes out of what you

605
00:34:53,080 --> 00:34:55,240
Speaker 3: said right at the beginning, and you said the common

606
00:34:55,280 --> 00:34:57,800
Speaker 3: feature of a lot of these attacks. What you're trying

607
00:34:57,840 --> 00:35:02,200
Speaker 3: to pre bunk is the idea that democracy doesn't deliver,

608
00:35:02,280 --> 00:35:05,239
Speaker 3: that it produces chaos. I mean, I guess you know

609
00:35:05,320 --> 00:35:07,120
Speaker 3: you have to then also, as well as doing all

610
00:35:07,120 --> 00:35:08,840
Speaker 3: the things you're talking about as a government, you have

611
00:35:08,920 --> 00:35:12,200
Speaker 3: to work on that not being true, actually on actually

612
00:35:12,239 --> 00:35:16,759
Speaker 3: delivering for people, because if their lived experience is chaos

613
00:35:17,000 --> 00:35:22,520
Speaker 3: and political dysfunction we've seen in many places, you know,

614
00:35:22,520 --> 00:35:26,120
Speaker 3: there's no point trying to debunk that if actually people

615
00:35:26,160 --> 00:35:29,960
Speaker 3: are feeling in their daily lives like democracy is not delivering.

616
00:35:30,760 --> 00:35:36,319
Speaker 2: Yeah, I think exactly as you observed. If people think

617
00:35:36,360 --> 00:35:41,160
Speaker 2: of democracy only as voting every two years or four years,

618
00:35:41,400 --> 00:35:45,440
Speaker 2: very small bandwidth, just a few bits, very long latency

619
00:35:45,600 --> 00:35:48,560
Speaker 2: for change to happen or to see it to actually

620
00:35:48,600 --> 00:35:52,919
Speaker 2: not happen, then obviously there's going to be a pathy. Right.

621
00:35:53,280 --> 00:35:56,600
Speaker 2: But I think a lot of the online ecosystems, the

622
00:35:56,680 --> 00:36:02,400
Speaker 2: fact checking ecosystem, the covacs, the participatorying, the initiatives of

623
00:36:02,640 --> 00:36:06,080
Speaker 2: not just referenda but community initiatives and so on, they're

624
00:36:06,200 --> 00:36:09,960
Speaker 2: there because it's a continuous democracy. It is something that

625
00:36:10,000 --> 00:36:14,000
Speaker 2: people can feel every day. As long as people employ

626
00:36:14,280 --> 00:36:18,400
Speaker 2: those democratic tools in their lives more, maybe in their community,

627
00:36:18,400 --> 00:36:21,200
Speaker 2: in their schools, in their churches. I think the word

628
00:36:21,239 --> 00:36:25,280
Speaker 2: democracy will have a much more nuanced meaning. The more

629
00:36:25,440 --> 00:36:30,720
Speaker 2: people have the stage to show affinity, compassion, curiosity, newance,

630
00:36:30,960 --> 00:36:34,880
Speaker 2: sharing personal stories, reasoning, and respect of each other, the

631
00:36:34,880 --> 00:36:37,759
Speaker 2: more people are going to associate the word democracy with

632
00:36:37,880 --> 00:36:41,359
Speaker 2: some lived experience that is actually pro social. So I'm

633
00:36:41,400 --> 00:36:44,920
Speaker 2: not saying that let's just protect the election. I'm saying that,

634
00:36:45,160 --> 00:36:48,160
Speaker 2: leading up to the election, let's protect our lift experience

635
00:36:48,239 --> 00:36:52,239
Speaker 2: around democracy on a smaller scale, lowercase the democracy.

636
00:36:53,040 --> 00:36:55,560
Speaker 3: But yeah, but the elected politicians actually have to deliver

637
00:36:55,640 --> 00:36:58,120
Speaker 3: on things people care about, like education and health.

638
00:36:58,200 --> 00:37:00,279
Speaker 2: Yes, which is why we need to do both. We

639
00:37:00,320 --> 00:37:04,399
Speaker 2: need to build a grassroots legitimacy that's hopefully higher than

640
00:37:04,440 --> 00:37:06,759
Speaker 2: any of the particul parties, and then we need to

641
00:37:06,840 --> 00:37:10,279
Speaker 2: hold those parties into accounts actually delivering what the people want.

642
00:37:11,000 --> 00:37:12,920
Speaker 3: So, Audrey, you've now stepped down. I think you were

643
00:37:12,920 --> 00:37:15,880
Speaker 3: one of the youngest ministers when you joined the government,

644
00:37:16,120 --> 00:37:18,480
Speaker 3: and you're now we're hearing from you going to be

645
00:37:18,520 --> 00:37:21,000
Speaker 3: traveling a lot more outside Taiwan, so I hope you're

646
00:37:21,000 --> 00:37:23,040
Speaker 3: going to be able to spend lots of time with

647
00:37:24,000 --> 00:37:28,640
Speaker 3: our governments and help them protect everyone against protect ourselves

648
00:37:29,680 --> 00:37:32,000
Speaker 3: against some of the things that you're talking about.

649
00:37:32,600 --> 00:37:36,600
Speaker 2: Yeah, how I can help and stand ready to help.

650
00:37:37,400 --> 00:37:42,480
Speaker 2: In fact, I'm going through a bunch of European countries

651
00:37:42,880 --> 00:37:47,840
Speaker 2: as we are airing this episode, going across seven different

652
00:37:47,880 --> 00:37:51,200
Speaker 2: countries in just twenty two days. The idea is that

653
00:37:51,680 --> 00:37:55,120
Speaker 2: trying to come to each jurisdiction like three or four

654
00:37:55,200 --> 00:37:58,400
Speaker 2: months before they are having a election, and just connect

655
00:37:58,520 --> 00:38:03,520
Speaker 2: with the officials, societyal leaders to share what worked in Taiwan,

656
00:38:03,840 --> 00:38:07,640
Speaker 2: and also to share the kind of community grassroots toolkits

657
00:38:07,840 --> 00:38:11,719
Speaker 2: that everybody can just adopt and if people coordinate, I'm

658
00:38:11,800 --> 00:38:16,799
Speaker 2: sure that this year will be not a democratic setback,

659
00:38:16,960 --> 00:38:19,080
Speaker 2: but rather a democratic rebatalization.

660
00:38:19,719 --> 00:38:21,879
Speaker 3: I hope so too. Maybe in a few months we'll

661
00:38:21,920 --> 00:38:24,399
Speaker 3: have to we'll have to find out from you how

662
00:38:24,440 --> 00:38:26,879
Speaker 3: you mark each of these countries that you've met.

663
00:38:26,920 --> 00:38:29,640
Speaker 2: Great, happy to be back on the podcast.

664
00:38:29,400 --> 00:38:31,200
Speaker 3: Audrey Tang, Thank you so.

665
00:38:31,200 --> 00:38:32,760
Speaker 2: Much, Thank you Liveline Prosper.

666
00:38:36,560 --> 00:38:39,960
Speaker 3: Okay, that is more than enough for this week's second

667
00:38:40,040 --> 00:38:43,600
Speaker 3: episode of Voter Nomics. Thanks for listening. This episode was

668
00:38:43,640 --> 00:38:47,880
Speaker 3: hosted by me Stephanie Flanders, Alegra Stratton and Adrian Woodridge.

669
00:38:48,280 --> 00:38:51,680
Speaker 3: It was produced by Sammasadi with help from Chris Marklu

670
00:38:52,080 --> 00:38:57,000
Speaker 3: and Julia Manns, editorial direction from Victoria Wakely. Brendan Francis

671
00:38:57,080 --> 00:39:01,200
Speaker 3: Newnan is our executive producer, and Sage Bowman is Head

672
00:39:01,280 --> 00:39:04,960
Speaker 3: of Bloomberg Podcasts. With special thanks to Nancy Cook and

673
00:39:05,040 --> 00:39:08,000
Speaker 3: Audrey Tank. Please, if you like it, subscribe, rate, and

674
00:39:08,080 --> 00:39:18,560
Speaker 3: review this show wherever you listen to your podcasts.