WEBVTT

1
00:00:00.000 --> 00:00:05.540
<v Chair>Good afternoon everyone. My name is Susanna Elliott from the Australian Science Media Centre

2
00:00:05.540 --> 00:00:14.780
<v Chair>and the session today is The Choice with Voice with Professor Peter Frey. Peter Frey is one of

3
00:00:14.780 --> 00:00:21.940
<v Chair>those journalists who has seen it all. As a print journalist in the 1980s with Fairfax,

4
00:00:22.660 --> 00:00:28.180
<v Chair>he's seen just about everything that you can imagine happening in the media right through

5
00:00:28.180 --> 00:00:36.440
<v Chair>to today's incredible convergent multimedia extravaganza that we see today in the news media.

6
00:00:36.860 --> 00:00:42.240
<v Chair>He's been editor of the Sunday Age, the Canberra Times and the Sydney Morning Herald and he even

7
00:00:42.240 --> 00:00:49.900
<v Chair>did a stint at the Australian as deputy editor there. He founded a website in Australia called

8
00:00:49.900 --> 00:00:56.320
<v Chair>PolitiFact which was a very important political fact-checking website and he paints and sculpts

9
00:00:56.320 --> 00:00:57.440
<v Chair>in his spare time.

10
00:00:58.340 --> 00:01:07.480
<v Chair>He is currently professor of the practice of journalism and is co-director of the Centre for Media Transition

11
00:01:07.480 --> 00:01:14.040
<v Chair>at the University of Technology in Sydney. So he's come all the way over from Sydney just to talk to you guys today.

12
00:01:14.260 --> 00:01:15.920
<v Chair>So please welcome Peter Frey.

13
00:01:16.860 --> 00:01:23.460
<v Professor Peter Fray>Thank you so much. Thank you Susanna. Great to see you. I know you're all here for Benjamin Law who's just

14
00:01:23.460 --> 00:01:27.800
<v Professor Peter Fray>sitting there and stay around because he's got a great talk as well. I'm going to do that.

15
00:01:28.180 --> 00:01:28.280
<v Professor Peter Fray>So Peter Frey, you're the director of the Science Media Centre and you're the director of the Science Media Centre.

16
00:01:28.280 --> 00:01:35.600
<v Professor Peter Fray>So Peter Frey is the science media centre who has kindly been a sponsor of this session and there's me in case you're wondering what I look like.

17
00:01:37.060 --> 00:01:43.660
<v Professor Peter Fray>So it's great to be back in Adelaide. Last time I was here was about four years ago and I was talking about the joys of fact-checking,

18
00:01:44.880 --> 00:01:51.340
<v Professor Peter Fray>fact-checking politicians to make our political debate more honest and our politicians more accountable and civil.

19
00:01:52.020 --> 00:01:56.680
<v Professor Peter Fray>You might remember PolitiFact Australia. Clearly that worked.

20
00:01:58.180 --> 00:02:05.020
<v Professor Peter Fray>I'm actually quite happy to talk about at the end about why it didn't work and interested in thoughts about how it might work better.

21
00:02:05.580 --> 00:02:20.820
<v Professor Peter Fray>But today I want to take us on a journey, a journey that is both once technological, it's very commercial and it really at its heart is a challenge to how we interact with artificial intelligence and smart machines.

22
00:02:21.580 --> 00:02:26.320
<v Professor Peter Fray>And in doing and going through this journey and hopefully this is, we can talk about this afterwards.

23
00:02:26.800 --> 00:02:27.860
<v Professor Peter Fray>I want us to,

24
00:02:27.860 --> 00:02:32.680
<v Professor Peter Fray>I'm gonna argue that we need to do some deep thinking of our own about how we interact with smart machines.

25
00:02:33.200 --> 00:02:39.780
<v Professor Peter Fray>So what I'm really talking about is this new age of interaction with voice interaction, interaction machines.

26
00:02:40.180 --> 00:02:49.380
<v Professor Peter Fray>So the stars of the show are, are many of the tech giants who brought us the mobile phone, probably the most important device we have so far, I think, in history.

27
00:02:50.540 --> 00:02:56.080
<v Professor Peter Fray>It's certainly the most important device in terms of how journalism is iterated in, in our current day.

28
00:02:56.200 --> 00:02:57.660
<v Professor Peter Fray>So the machines I speak of,

29
00:02:57.660 --> 00:03:00.200
<v Professor Peter Fray>oh, it's not Scarlett Johansson.

30
00:03:00.320 --> 00:03:01.880
<v Professor Peter Fray>The machines on which I speak of today

31
00:03:01.880 --> 00:03:03.340
<v Professor Peter Fray>are not Scarlett Johansson.

32
00:03:04.860 --> 00:03:07.980
<v Professor Peter Fray>They are increasingly going to be

33
00:03:07.980 --> 00:03:09.840
<v Professor Peter Fray>how we discover all sorts of content,

34
00:03:10.020 --> 00:03:11.260
<v Professor Peter Fray>including the news.

35
00:03:11.900 --> 00:03:14.640
<v Professor Peter Fray>They will help the likes of Facebook and Google

36
00:03:14.640 --> 00:03:18.560
<v Professor Peter Fray>further enmesh themselves into our lives.

37
00:03:19.020 --> 00:03:21.140
<v Professor Peter Fray>They will increasingly be the place

38
00:03:21.140 --> 00:03:23.100
<v Professor Peter Fray>where the attention and money of consumers

39
00:03:23.100 --> 00:03:25.520
<v Professor Peter Fray>and advertisers will go.

40
00:03:26.440 --> 00:03:28.100
<v Professor Peter Fray>And they will bring into play,

41
00:03:28.600 --> 00:03:29.880
<v Professor Peter Fray>and into many homes,

42
00:03:30.040 --> 00:03:31.720
<v Professor Peter Fray>the retail giant of Amazon,

43
00:03:32.240 --> 00:03:34.360
<v Professor Peter Fray>which we haven't really seen the full impact of Amazon

44
00:03:34.360 --> 00:03:35.240
<v Professor Peter Fray>in this country yet,

45
00:03:35.340 --> 00:03:37.660
<v Professor Peter Fray>but I really think that this is at the heart

46
00:03:37.660 --> 00:03:38.680
<v Professor Peter Fray>of something we should talk about.

47
00:03:39.180 --> 00:03:41.600
<v Professor Peter Fray>And I'm kind of tempted to say that Amazon's Echo

48
00:03:42.300 --> 00:03:44.100
<v Professor Peter Fray>is a kind of Trojan horse,

49
00:03:44.140 --> 00:03:45.720
<v Professor Peter Fray>but it isn't a Trojan horse at all,

50
00:03:45.800 --> 00:03:49.240
<v Professor Peter Fray>because there's nothing secretive or subtle about Echo

51
00:03:49.240 --> 00:03:52.220
<v Professor Peter Fray>and the interface, which is known as Alexa,

52
00:03:53.500 --> 00:03:55.460
<v Professor Peter Fray>and who often gets mistaken

53
00:03:55.460 --> 00:03:56.920
<v Professor Peter Fray>for Scarlett Johansson,

54
00:03:57.700 --> 00:03:59.260
<v Professor Peter Fray>in part, I think, because,

55
00:03:59.260 --> 00:04:01.680
<v Professor Peter Fray>as my friend Deirdre Mack in The Writer pointed out,

56
00:04:01.820 --> 00:04:04.520
<v Professor Peter Fray>that Johansson did the voice in this movie,

57
00:04:05.240 --> 00:04:05.720
<v Professor Peter Fray>her,

58
00:04:06.040 --> 00:04:08.720
<v Professor Peter Fray>and has forever been seemingly embodiment

59
00:04:08.720 --> 00:04:10.780
<v Professor Peter Fray>of sort of how computers talk to us.

60
00:04:11.340 --> 00:04:12.740
<v Professor Peter Fray>That's her words, not mine.

61
00:04:13.180 --> 00:04:15.560
<v Professor Peter Fray>But you can ask Alexa,

62
00:04:15.840 --> 00:04:17.300
<v Professor Peter Fray>this on my left here,

63
00:04:17.440 --> 00:04:20.420
<v Professor Peter Fray>you can ask Alexa about Scarlett Johansson

64
00:04:20.420 --> 00:04:21.260
<v Professor Peter Fray>if you really want to.

65
00:04:21.360 --> 00:04:23.140
<v Professor Peter Fray>So there's a kind of squaring of the circle.

66
00:04:23.920 --> 00:04:25.180
<v Professor Peter Fray>So what I'm really talking about,

67
00:04:25.220 --> 00:04:25.420
<v Professor Peter Fray>is,

68
00:04:25.420 --> 00:04:26.680
<v Professor Peter Fray>there's the rise and rise of these things,

69
00:04:26.840 --> 00:04:27.820
<v Professor Peter Fray>smart devices.

70
00:04:28.420 --> 00:04:30.980
<v Professor Peter Fray>And they're increasingly going to be sitting in our homes

71
00:04:30.980 --> 00:04:33.340
<v Professor Peter Fray>and interacting with us via voice commands.

72
00:04:34.020 --> 00:04:35.180
<v Professor Peter Fray>And this is what they look like.

73
00:04:35.980 --> 00:04:37.180
<v Professor Peter Fray>They are Google Home,

74
00:04:37.380 --> 00:04:38.240
<v Professor Peter Fray>Amazon Echo,

75
00:04:39.740 --> 00:04:40.440
<v Professor Peter Fray>aka Alexa,

76
00:04:40.700 --> 00:04:41.720
<v Professor Peter Fray>Apple HomePod,

77
00:04:41.920 --> 00:04:43.160
<v Professor Peter Fray>and coming down the pipe,

78
00:04:44.460 --> 00:04:45.340
<v Professor Peter Fray>Facebook's Aloha,

79
00:04:45.500 --> 00:04:47.160
<v Professor Peter Fray>which is a video chat product,

80
00:04:47.360 --> 00:04:48.140
<v Professor Peter Fray>but like the others,

81
00:04:48.220 --> 00:04:51.680
<v Professor Peter Fray>will be enabled to help us buy things at home.

82
00:04:52.960 --> 00:04:55.080
<v Professor Peter Fray>Here is a slide of Mark Zuckerberg,

83
00:04:55.420 --> 00:04:56.340
<v Professor Peter Fray>talking about Aloha,

84
00:04:57.100 --> 00:04:58.080
<v Professor Peter Fray>and how he will,

85
00:04:58.640 --> 00:05:00.260
<v Professor Peter Fray>Aloha will help his company

86
00:05:00.260 --> 00:05:02.060
<v Professor Peter Fray>onto another pathway to world domination.

87
00:05:03.160 --> 00:05:05.520
<v Professor Peter Fray>So let me first deal with the commercial side of this.

88
00:05:05.560 --> 00:05:06.480
<v Professor Peter Fray>I'm going to go really quick,

89
00:05:06.560 --> 00:05:08.940
<v Professor Peter Fray>because I'm going to bombard you with some facts and figures,

90
00:05:09.060 --> 00:05:11.540
<v Professor Peter Fray>but it's important that we kind of get a sense of the size of this.

91
00:05:12.160 --> 00:05:12.560
<v Professor Peter Fray>So,

92
00:05:13.500 --> 00:05:14.280
<v Professor Peter Fray>in the US,

93
00:05:14.760 --> 00:05:16.540
<v Professor Peter Fray>smart speakers are all the rage.

94
00:05:18.700 --> 00:05:20.960
<v Professor Peter Fray>People seem to love interacting with them.

95
00:05:21.260 --> 00:05:23.620
<v Professor Peter Fray>And if you've ever seen a teenager interacting with Siri,

96
00:05:23.840 --> 00:05:24.620
<v Professor Peter Fray>you'll know what I mean.

97
00:05:25.440 --> 00:05:28.160
<v Professor Peter Fray>They are coming to our country real quick.

98
00:05:29.080 --> 00:05:30.380
<v Professor Peter Fray>Obviously, we're behind the US.

99
00:05:31.720 --> 00:05:34.400
<v Professor Peter Fray>Some news media is latching on,

100
00:05:34.420 --> 00:05:35.100
<v Professor Peter Fray>are latching on.

101
00:05:35.180 --> 00:05:36.480
<v Professor Peter Fray>This is NPR.

102
00:05:38.300 --> 00:05:41.740
<v Professor Peter Fray>Sorry, NPR is streaming a lot of stuff through Alexa at the moment.

103
00:05:42.880 --> 00:05:45.520
<v Professor Peter Fray>And the news part of this offering,

104
00:05:45.720 --> 00:05:47.500
<v Professor Peter Fray>I guess, which is the thing I'm most concerned about,

105
00:05:47.600 --> 00:05:49.800
<v Professor Peter Fray>the news part of this offering is important,

106
00:05:50.000 --> 00:05:51.040
<v Professor Peter Fray>and it will be growing.

107
00:05:52.120 --> 00:05:55.000
<v Professor Peter Fray>And there is a substitution effect going on here.

108
00:05:55.420 --> 00:05:58.300
<v Professor Peter Fray>More home device equals less radio, less mobile,

109
00:05:58.520 --> 00:05:59.880
<v Professor Peter Fray>less tablet, and less newspaper.

110
00:06:01.620 --> 00:06:04.700
<v Professor Peter Fray>So, by and large, the news media industry knows

111
00:06:04.700 --> 00:06:07.040
<v Professor Peter Fray>that it has to work with these devices.

112
00:06:07.500 --> 00:06:10.740
<v Professor Peter Fray>And one of its futures is going to be voice.

113
00:06:11.460 --> 00:06:14.400
<v Professor Peter Fray>But already, it's showing signs of concern

114
00:06:14.920 --> 00:06:16.560
<v Professor Peter Fray>about the impact that voice is going to have.

115
00:06:17.180 --> 00:06:19.720
<v Professor Peter Fray>Because, not simply because of the reach of these devices,

116
00:06:19.880 --> 00:06:22.500
<v Professor Peter Fray>and any substitution effect, as of that slide,

117
00:06:23.140 --> 00:06:25.400
<v Professor Peter Fray>but really, and this is the guts of it,

118
00:06:25.420 --> 00:06:28.180
<v Professor Peter Fray>about the narrow, the potential for these devices

119
00:06:28.180 --> 00:06:30.900
<v Professor Peter Fray>to narrow the choice of news,

120
00:06:31.160 --> 00:06:34.360
<v Professor Peter Fray>and place more power in the hands of the tech platforms.

121
00:06:34.760 --> 00:06:35.860
<v Professor Peter Fray>Think about it this way.

122
00:06:36.220 --> 00:06:39.340
<v Professor Peter Fray>When you type in a search word in Google,

123
00:06:39.540 --> 00:06:41.920
<v Professor Peter Fray>you get thousands of answers, right?

124
00:06:42.000 --> 00:06:44.380
<v Professor Peter Fray>You may only read two or three or four or five of them,

125
00:06:44.420 --> 00:06:45.700
<v Professor Peter Fray>but you get a lot of answers.

126
00:06:46.200 --> 00:06:49.360
<v Professor Peter Fray>When you ask a machine, you'll get one answer back.

127
00:06:50.000 --> 00:06:52.260
<v Professor Peter Fray>Just the one answer back, right, initially.

128
00:06:52.440 --> 00:06:54.520
<v Professor Peter Fray>If you say, what's the news today, Google, Alexa?

129
00:06:54.520 --> 00:06:57.880
<v Professor Peter Fray>You'll get a Google, Alexa will tell you what the news today is.

130
00:06:59.340 --> 00:07:03.280
<v Professor Peter Fray>Now, the tech giants tend to see smart speakers at the moment

131
00:07:03.280 --> 00:07:06.500
<v Professor Peter Fray>as an add-on to the range of its platforms and channels,

132
00:07:06.580 --> 00:07:07.200
<v Professor Peter Fray>which is fair enough.

133
00:07:07.680 --> 00:07:10.480
<v Professor Peter Fray>But when I recently asked an executive from one of them

134
00:07:10.480 --> 00:07:13.660
<v Professor Peter Fray>about the potential to narrow the choice for news consumers,

135
00:07:13.840 --> 00:07:16.540
<v Professor Peter Fray>she replied, yes, that's a real challenge.

136
00:07:16.580 --> 00:07:17.900
<v Professor Peter Fray>This is a developing area.

137
00:07:18.240 --> 00:07:20.680
<v Professor Peter Fray>And it really is a very fast developing area.

138
00:07:21.160 --> 00:07:23.640
<v Professor Peter Fray>So, News Corporation recently made a submission

139
00:07:23.640 --> 00:07:27.520
<v Professor Peter Fray>to the ACCC's inquiry into the power of social media platforms.

140
00:07:27.760 --> 00:07:29.460
<v Professor Peter Fray>It's a very thorough piece of work,

141
00:07:29.500 --> 00:07:32.260
<v Professor Peter Fray>and if you want to jump into it, there's a lot of good stuff in there.

142
00:07:33.220 --> 00:07:35.500
<v Professor Peter Fray>But it includes the following data about Amazon.

143
00:07:35.700 --> 00:07:39.040
<v Professor Peter Fray>By September 2017, Amazon's Alexa in the US

144
00:07:39.040 --> 00:07:42.580
<v Professor Peter Fray>was receiving 50 billion searches a month.

145
00:07:42.980 --> 00:07:46.520
<v Professor Peter Fray>By 2020, this will grow to 200 billion,

146
00:07:47.120 --> 00:07:49.900
<v Professor Peter Fray>and there are over 25,000 skills,

147
00:07:50.040 --> 00:07:52.940
<v Professor Peter Fray>what we would call functions, attached to the platform.

148
00:07:53.640 --> 00:07:56.800
<v Professor Peter Fray>The News Corp submission says, and I quote,

149
00:07:57.080 --> 00:07:59.620
<v Professor Peter Fray>the Alexa voice service is always in the cloud

150
00:07:59.620 --> 00:08:03.740
<v Professor Peter Fray>and can be built into any hardware with a microphone and a speaker,

151
00:08:03.840 --> 00:08:04.920
<v Professor Peter Fray>such as, say, cars.

152
00:08:05.200 --> 00:08:10.040
<v Professor Peter Fray>This end-to-end integration of voice, search, and e-commerce

153
00:08:10.720 --> 00:08:14.900
<v Professor Peter Fray>via Amazon Shopping Service makes Alexa a very compelling consumer offer.

154
00:08:15.240 --> 00:08:17.320
<v Professor Peter Fray>Now, of course, these are the end-to-end capabilities

155
00:08:18.060 --> 00:08:20.300
<v Professor Peter Fray>that news executives dream of.

156
00:08:20.700 --> 00:08:23.620
<v Professor Peter Fray>Certainly, I used to dream of them when I was publisher of the Sydney Mail.

157
00:08:23.640 --> 00:08:24.100
<v Professor Peter Fray>I was a good morning Herald.

158
00:08:24.240 --> 00:08:26.260
<v Professor Peter Fray>And of course, in a way, it's what they used to have

159
00:08:26.260 --> 00:08:29.580
<v Professor Peter Fray>when they controlled print end-to-end.

160
00:08:30.180 --> 00:08:32.440
<v Professor Peter Fray>But now this game is all over, as we all know,

161
00:08:32.580 --> 00:08:35.880
<v Professor Peter Fray>and now the power was going to lie more and more in these devices.

162
00:08:36.160 --> 00:08:36.920
<v Professor Peter Fray>So imagine this.

163
00:08:37.580 --> 00:08:39.760
<v Professor Peter Fray>These devices will provide us the information.

164
00:08:40.600 --> 00:08:42.740
<v Professor Peter Fray>They will provide that information to us

165
00:08:42.740 --> 00:08:45.920
<v Professor Peter Fray>when we are in a selling environment,

166
00:08:46.400 --> 00:08:49.680
<v Professor Peter Fray>and they will enable us to have the means to buy stuff

167
00:08:49.680 --> 00:08:52.540
<v Professor Peter Fray>all in the same device in the same environment.

168
00:08:53.640 --> 00:08:56.280
<v Professor Peter Fray>Thus, they will have the intimate knowledge about who is buying,

169
00:08:56.440 --> 00:08:57.200
<v Professor Peter Fray>i.e. us,

170
00:08:57.340 --> 00:08:59.780
<v Professor Peter Fray>and possibly why they are doing so.

171
00:09:00.680 --> 00:09:03.600
<v Professor Peter Fray>So back to my friend Deirdre Mackin,

172
00:09:03.620 --> 00:09:05.760
<v Professor Peter Fray>who recently wrote a very funny column about this.

173
00:09:05.920 --> 00:09:07.540
<v Professor Peter Fray>So they recently bought an Amazon.

174
00:09:08.760 --> 00:09:10.020
<v Professor Peter Fray>And she writes,

175
00:09:10.260 --> 00:09:12.860
<v Professor Peter Fray>Sometimes I forget Alexa is around,

176
00:09:13.100 --> 00:09:15.140
<v Professor Peter Fray>but then I see the green light on top of the speaker,

177
00:09:15.400 --> 00:09:16.940
<v Professor Peter Fray>and then I realize she is listening.

178
00:09:17.480 --> 00:09:18.980
<v Professor Peter Fray>And if she is always listening,

179
00:09:19.100 --> 00:09:20.920
<v Professor Peter Fray>she might be keeping track of what we say.

180
00:09:21.300 --> 00:09:22.500
<v Professor Peter Fray>And if she's keeping track,

181
00:09:22.500 --> 00:09:25.080
<v Professor Peter Fray>then someone in the cloud is keeping track.

182
00:09:25.560 --> 00:09:27.620
<v Professor Peter Fray>And they know what our household is like.

183
00:09:28.820 --> 00:09:31.020
<v Professor Peter Fray>Now I doubt Jeff Bezos is desperate.

184
00:09:31.140 --> 00:09:32.740
<v Professor Peter Fray>Jeff Bezos, the CEO of Amazon,

185
00:09:32.920 --> 00:09:35.240
<v Professor Peter Fray>is desperate to know how my husband and I live,

186
00:09:35.440 --> 00:09:38.500
<v Professor Peter Fray>but his Amazon algorithm might want to know.

187
00:09:39.320 --> 00:09:40.620
<v Professor Peter Fray>Let's look at it another way.

188
00:09:41.500 --> 00:09:45.400
<v Professor Peter Fray>An advertiser wishes to reach a 25-year-old woman with a new product.

189
00:09:45.640 --> 00:09:47.700
<v Professor Peter Fray>In the past, a newspaper would say,

190
00:09:47.800 --> 00:09:50.080
<v Professor Peter Fray>Sure, we have 25-year-old women.

191
00:09:50.240 --> 00:09:51.520
<v Professor Peter Fray>Everyone buys a newspaper.

192
00:09:51.520 --> 00:09:52.580
<v Professor Peter Fray>Everyone reads all the ads.

193
00:09:52.740 --> 00:09:53.580
<v Professor Peter Fray>Give us your money.

194
00:09:53.820 --> 00:09:55.260
<v Professor Peter Fray>Now we all know that game is up.

195
00:09:55.460 --> 00:09:56.620
<v Professor Peter Fray>A website would say,

196
00:09:56.680 --> 00:10:00.320
<v Professor Peter Fray>Hell yeah, our analytics indicate that we have 25-year-old women,

197
00:10:00.460 --> 00:10:04.860
<v Professor Peter Fray>and they read lifestyle and general news and the bit of world advertised here.

198
00:10:06.060 --> 00:10:07.060
<v Professor Peter Fray>Facebook would say,

199
00:10:07.320 --> 00:10:09.760
<v Professor Peter Fray>Sure, we have 25-year-old women,

200
00:10:09.820 --> 00:10:12.040
<v Professor Peter Fray>but how many of them do you want to target,

201
00:10:12.200 --> 00:10:14.020
<v Professor Peter Fray>and where do you want them to live?

202
00:10:14.900 --> 00:10:16.300
<v Professor Peter Fray>And Amazon would say,

203
00:10:16.900 --> 00:10:18.580
<v Professor Peter Fray>We have 25-year-old women,

204
00:10:18.700 --> 00:10:21.500
<v Professor Peter Fray>but our intelligence gathered by our AI management

205
00:10:21.520 --> 00:10:22.720
<v Professor Peter Fray>and our computer machines, including Alexa,

206
00:10:23.020 --> 00:10:26.920
<v Professor Peter Fray>tells us that your product is better sold to 18-year-old men.

207
00:10:27.500 --> 00:10:28.620
<v Professor Peter Fray>We have them too.

208
00:10:29.000 --> 00:10:32.320
<v Professor Peter Fray>Here they are, and by the way, they are in buying mode right now.

209
00:10:32.800 --> 00:10:34.560
<v Professor Peter Fray>And here's the news to go with it.

210
00:10:34.760 --> 00:10:36.480
<v Professor Peter Fray>So I'm not really trying to be alarmist.

211
00:10:36.560 --> 00:10:37.520
<v Professor Peter Fray>I know it sounds a little bit alarmist.

212
00:10:37.560 --> 00:10:39.760
<v Professor Peter Fray>I'm not trying to be alarmist, and I'm really not trying to be negative.

213
00:10:40.180 --> 00:10:41.680
<v Professor Peter Fray>This could all end really well.

214
00:10:42.300 --> 00:10:45.300
<v Professor Peter Fray>There is, for instance, a tendency for smart speaker users

215
00:10:45.300 --> 00:10:47.280
<v Professor Peter Fray>to think of their device as a friend.

216
00:10:48.320 --> 00:10:50.620
<v Professor Peter Fray>And again, that's a news executive's dream,

217
00:10:50.620 --> 00:10:52.240
<v Professor Peter Fray>and we all need friends, right?

218
00:10:52.640 --> 00:10:56.580
<v Professor Peter Fray>And interestingly, unlike some products,

219
00:10:56.800 --> 00:10:58.980
<v Professor Peter Fray>smart devices seem to grow on people.

220
00:10:59.200 --> 00:11:02.020
<v Professor Peter Fray>The more we have them, the more we use them.

221
00:11:02.220 --> 00:11:04.760
<v Professor Peter Fray>And that's why this is all really about choice,

222
00:11:04.880 --> 00:11:08.700
<v Professor Peter Fray>the choices we make, and how smart we, as humans, can be

223
00:11:08.700 --> 00:11:12.020
<v Professor Peter Fray>in articulating them and interacting with the machines,

224
00:11:12.140 --> 00:11:15.900
<v Professor Peter Fray>because what we do has a massive influence on what we receive.

225
00:11:16.260 --> 00:11:19.940
<v Professor Peter Fray>And here is a quick history of choice in the music industry.

226
00:11:19.940 --> 00:11:22.300
<v Professor Peter Fray>So, you know, first up we have albums,

227
00:11:22.480 --> 00:11:25.720
<v Professor Peter Fray>and we had to buy a whole album just because we wanted one nice good track.

228
00:11:25.920 --> 00:11:27.640
<v Professor Peter Fray>This is a particular favorite of mine,

229
00:11:27.700 --> 00:11:31.360
<v Professor Peter Fray>the Funkadelic One Nation on a Groove is 95% crap,

230
00:11:31.540 --> 00:11:34.040
<v Professor Peter Fray>but with one really amazing track on it.

231
00:11:34.580 --> 00:11:38.300
<v Professor Peter Fray>Then, of course, we have CDs, where we could choose to move around in CDs

232
00:11:38.300 --> 00:11:41.460
<v Professor Peter Fray>and pick the tracks we wanted by one of these or other means.

233
00:11:41.680 --> 00:11:45.500
<v Professor Peter Fray>And now, of course, we get our music because the algorithm tells us, right?

234
00:11:45.560 --> 00:11:48.940
<v Professor Peter Fray>The algorithm by our actions tells us what we like.

235
00:11:49.340 --> 00:11:50.720
<v Professor Peter Fray>And the same is true of news.

236
00:11:51.080 --> 00:11:55.260
<v Professor Peter Fray>You know, in the past, we used to get this thing, this brick, you know,

237
00:11:55.320 --> 00:11:59.320
<v Professor Peter Fray>and all the choices were made by someone who looked a bit like me.

238
00:11:59.740 --> 00:12:01.040
<v Professor Peter Fray>So you didn't really have a lot of choice.

239
00:12:01.400 --> 00:12:04.240
<v Professor Peter Fray>And then we have websites where, of course, we have a bit more choice,

240
00:12:04.240 --> 00:12:06.180
<v Professor Peter Fray>and the news is streamed for us a little bit more.

241
00:12:06.240 --> 00:12:08.460
<v Professor Peter Fray>So if you only want to look at sport or entertainment, you can do that.

242
00:12:08.680 --> 00:12:12.200
<v Professor Peter Fray>And now, of course, we get it all from our news feed, if you want to.

243
00:12:13.280 --> 00:12:17.440
<v Professor Peter Fray>So, obviously, increasingly, the news we get is predetermined by an algorithm

244
00:12:18.500 --> 00:12:18.920
<v Professor Peter Fray>that knows what we're doing.

245
00:12:18.940 --> 00:12:19.540
<v Professor Peter Fray>It knows our actions.

246
00:12:20.200 --> 00:12:22.780
<v Professor Peter Fray>So that's our future in front of us with voice.

247
00:12:23.240 --> 00:12:27.480
<v Professor Peter Fray>But it's on steroids in a way because the intimate relationship we're going to have

248
00:12:27.480 --> 00:12:31.780
<v Professor Peter Fray>with voice machines will constantly know what we want and what we're thinking.

249
00:12:33.080 --> 00:12:36.260
<v Professor Peter Fray>And as I mentioned before, voice will give us one answer at a time.

250
00:12:36.660 --> 00:12:42.140
<v Professor Peter Fray>And then it will be up to us to change it and to think about what we're actually doing.

251
00:12:42.620 --> 00:12:45.660
<v Professor Peter Fray>And this is where I think where the wood is on us.

252
00:12:45.880 --> 00:12:47.700
<v Professor Peter Fray>So here's an example, another example.

253
00:12:48.120 --> 00:12:48.920
<v Professor Peter Fray>Does anyone remember this?

254
00:12:48.940 --> 00:12:49.460
<v Professor Peter Fray>Does anyone remember Tay?

255
00:12:50.140 --> 00:12:54.320
<v Professor Peter Fray>Tay was a Twitter bot released in March 2016 by Microsoft.

256
00:12:55.160 --> 00:13:00.460
<v Professor Peter Fray>She was intended to be the next evolutionary step in developing a truly human-like chatbot.

257
00:13:00.940 --> 00:13:04.040
<v Professor Peter Fray>She was equipped with a complex learning algorithm.

258
00:13:04.300 --> 00:13:05.760
<v Professor Peter Fray>She had the capacity for language.

259
00:13:06.160 --> 00:13:08.400
<v Professor Peter Fray>She seemed in some ways unmachine-like.

260
00:13:08.600 --> 00:13:09.880
<v Professor Peter Fray>She had randomness.

261
00:13:10.000 --> 00:13:10.680
<v Professor Peter Fray>She had humor.

262
00:13:10.960 --> 00:13:17.160
<v Professor Peter Fray>And even what was termed as she had positions on things, positions on things.

263
00:13:17.160 --> 00:13:18.780
<v Professor Peter Fray>She was a bit political.

264
00:13:19.080 --> 00:13:24.700
<v Professor Peter Fray>And as you may recall, she was destined for a very short and very inglorious life.

265
00:13:25.380 --> 00:13:26.580
<v Professor Peter Fray>This is how she started.

266
00:13:27.600 --> 00:13:29.220
<v Professor Peter Fray>Innocent enough, she said, hello world.

267
00:13:30.240 --> 00:13:32.340
<v Professor Peter Fray>And she went to sleep on the first day.

268
00:13:32.580 --> 00:13:33.860
<v Professor Peter Fray>And this is how she ended.

269
00:13:34.320 --> 00:13:38.660
<v Professor Peter Fray>So Tay was given the personality of a sassy 18 to 24-year-old American woman.

270
00:13:38.980 --> 00:13:40.900
<v Professor Peter Fray>I'm not sure why robots are always women.

271
00:13:40.940 --> 00:13:42.920
<v Professor Peter Fray>But anyway, that's another talk, not by me.

272
00:13:43.520 --> 00:13:47.140
<v Professor Peter Fray>So yeah, 16 hours after she started interacting with, and I'm not sure why, robots are always women.

273
00:13:47.140 --> 00:13:55.100
<v Professor Peter Fray>She had learned from Twitter users, Tay had attracted over 93,000 tweets, and she had become an offensive Jew-hating racist.

274
00:13:55.480 --> 00:14:02.340
<v Professor Peter Fray>She was taken offline, and after a second attempt, it was all aborted, and she's been retired ever since.

275
00:14:03.000 --> 00:14:07.120
<v Professor Peter Fray>So what does Tay tell us about human interaction with robots?

276
00:14:07.500 --> 00:14:14.920
<v Professor Peter Fray>Does it tell us that, given the chance, a bunch of Twitter trolls, online hate-mongers will ruin your bots day?

277
00:14:15.100 --> 00:14:16.020
<v Professor Peter Fray>Well, it might tell you that.

278
00:14:16.020 --> 00:14:18.860
<v Professor Peter Fray>but I suspect it's a little bit more complex than that.

279
00:14:19.080 --> 00:14:22.780
<v Professor Peter Fray>A couple of scholars called Jim and F and Peter Nagy asked this question.

280
00:14:22.980 --> 00:14:25.540
<v Professor Peter Fray>Who was really responsible for Tay's behavior?

281
00:14:26.080 --> 00:14:28.560
<v Professor Peter Fray>Should the blame be located with Tay?

282
00:14:29.040 --> 00:14:31.320
<v Professor Peter Fray>Should it be with her coders?

283
00:14:31.700 --> 00:14:36.420
<v Professor Peter Fray>Should it be with all Twitter users or just the internet pranksters who turned her into a racist?

284
00:14:36.760 --> 00:14:41.540
<v Professor Peter Fray>Or the Microsoft executives who commissioned her or some other agent

285
00:14:41.540 --> 00:14:43.460
<v Professor Peter Fray>or a combination of all the things I've mentioned?

286
00:14:43.460 --> 00:14:47.540
<v Professor Peter Fray>They argue, and I think this is what we're really going to see with voice,

287
00:14:47.740 --> 00:14:51.660
<v Professor Peter Fray>is what they argue is that Tay formed a kind of symbiotic agency

288
00:14:51.660 --> 00:14:54.460
<v Professor Peter Fray>between the machine and the human.

289
00:14:54.760 --> 00:15:00.380
<v Professor Peter Fray>That the machines are learning and changing, but at the same time, so are we.

290
00:15:00.720 --> 00:15:03.480
<v Professor Peter Fray>Just like the guy who told Alexa to suck it, right?

291
00:15:03.640 --> 00:15:06.600
<v Professor Peter Fray>We are having different types of relationships with machines.

292
00:15:07.080 --> 00:15:09.260
<v Professor Peter Fray>Some of us are telling machines to suck it,

293
00:15:09.320 --> 00:15:12.540
<v Professor Peter Fray>and some of us think it's fun to make a Twitter bot a racist.

294
00:15:12.540 --> 00:15:15.680
<v Professor Peter Fray>And our conversations not only differ between how we talk to machines

295
00:15:15.680 --> 00:15:19.180
<v Professor Peter Fray>and say how we talk to fellow human beings,

296
00:15:19.340 --> 00:15:22.600
<v Professor Peter Fray>or in the case of our friend on YouTube there, his dog,

297
00:15:23.300 --> 00:15:27.040
<v Professor Peter Fray>but they change as the machine itself changes and learns with us.

298
00:15:27.300 --> 00:15:29.780
<v Professor Peter Fray>So it ain't passive. This is active.

299
00:15:30.160 --> 00:15:34.520
<v Professor Peter Fray>Going back to the scholars Neff and Nagy, what they say, and I quote,

300
00:15:34.760 --> 00:15:40.580
<v Professor Peter Fray>we must watch for how agency flows among and between an algorithm's designer.

301
00:15:40.760 --> 00:15:42.200
<v Professor Peter Fray>The designed algorithm.

302
00:15:42.540 --> 00:15:45.600
<v Professor Peter Fray>Human users and the resulting content interaction and conversations.

303
00:15:46.140 --> 00:15:48.660
<v Professor Peter Fray>So there's a much more complex thing going on here.

304
00:15:48.740 --> 00:15:52.160
<v Professor Peter Fray>As I say, we are active participants in this relationship.

305
00:15:52.440 --> 00:15:55.800
<v Chair>Which kind of brings us back to the age of journalism and voice activation.

306
00:15:56.320 --> 00:16:00.100
<v Professor Peter Fray>So as a journalist, I am very excited about the potential of working with machines,

307
00:16:00.420 --> 00:16:02.340
<v Professor Peter Fray>especially AI and machine learning.

308
00:16:02.840 --> 00:16:05.260
<v Professor Peter Fray>I'm not a technophode, although I'm not a technologist.

309
00:16:06.280 --> 00:16:09.000
<v Professor Peter Fray>And I certainly don't wish that I could return to a time

310
00:16:09.000 --> 00:16:11.560
<v Professor Peter Fray>when people like me decided what you read, heard, and listened to.

311
00:16:11.680 --> 00:16:12.520
<v Professor Peter Fray>I have my own little book, called, The

312
00:16:12.520 --> 00:16:12.660
<v Professor Peter Fray>Little Project.

313
00:16:13.100 --> 00:16:16.060
<v Professor Peter Fray>My own little project is with Rocco Fazzari.

314
00:16:16.260 --> 00:16:18.800
<v Professor Peter Fray>He's a fine son of Adelaide, I should mention.

315
00:16:19.040 --> 00:16:23.480
<v Professor Peter Fray>And a computer engineering PhD student at UTS at my university called Wanming Huang.

316
00:16:23.900 --> 00:16:27.180
<v Professor Peter Fray>And we're developing a machine, learning powered,

317
00:16:27.860 --> 00:16:33.240
<v Professor Peter Fray>but which will automatically attach the relevant emotion to the tweets of Donald Trump.

318
00:16:33.480 --> 00:16:36.960
<v Professor Peter Fray>So we haven't launched it yet, but here's what it's going to look like.

319
00:16:38.640 --> 00:16:42.100
<v Professor Peter Fray>So essentially, Trump will tweet, as you know,

320
00:16:42.520 --> 00:16:43.660
<v Professor Peter Fray>Trump tweets a lot basically.

321
00:16:43.920 --> 00:16:48.560
<v Professor Peter Fray>We're all, you know, global policy is being decided through Twitter really by Donald Trump.

322
00:16:48.780 --> 00:16:54.020
<v Professor Peter Fray>And the tweet bot will attach to it a rating of what we think,

323
00:16:54.060 --> 00:16:56.900
<v Professor Peter Fray>what the machine thinks of this Trump's tweet.

324
00:16:57.240 --> 00:17:01.700
<v Professor Peter Fray>So either it'll be scary Trump, happy Trump, sad Trump.

325
00:17:01.840 --> 00:17:03.000
<v Professor Peter Fray>Those are the three we trialed.

326
00:17:03.240 --> 00:17:06.920
<v Professor Peter Fray>And essentially, it just involves creating a database of Trump's tweets

327
00:17:06.920 --> 00:17:09.340
<v Professor Peter Fray>and programming the machine to interact to them.

328
00:17:10.280 --> 00:17:12.040
<v Professor Peter Fray>Now that's just a bit of fun, right?

329
00:17:12.980 --> 00:17:17.480
<v Professor Peter Fray>But there's clearly a heap of positives which may flow from the rise of these machines.

330
00:17:18.960 --> 00:17:22.340
<v Professor Peter Fray>But the moment is fast approaching that it's really,

331
00:17:22.420 --> 00:17:24.240
<v Professor Peter Fray>we have to understand that this is about choice.

332
00:17:24.420 --> 00:17:27.640
<v Professor Peter Fray>And it's a choice for us, and it's a choice for what the platforms do,

333
00:17:27.820 --> 00:17:29.720
<v Professor Peter Fray>and in a way, it's a choice for the machines.

334
00:17:30.380 --> 00:17:36.420
<v Professor Peter Fray>So for decades, the big platforms have told us they are not much more than conduits to content.

335
00:17:36.620 --> 00:17:38.180
<v Professor Peter Fray>That's essentially what they tell us.

336
00:17:39.220 --> 00:17:41.840
<v Professor Peter Fray>That they are enablers of fruitful interactions.

337
00:17:42.840 --> 00:17:46.260
<v Professor Peter Fray>But with the voice box in the room, they are in our homes.

338
00:17:47.220 --> 00:17:51.280
<v Professor Peter Fray>And we see them as both intermediaries and the packages of journalism,

339
00:17:51.900 --> 00:17:53.500
<v Professor Peter Fray>as the home of news.

340
00:17:54.340 --> 00:17:59.720
<v Professor Peter Fray>They host the content and then arrange it to make it searchable and available to us

341
00:17:59.720 --> 00:18:01.520
<v Professor Peter Fray>for our personal use.

342
00:18:02.140 --> 00:18:08.600
<v Professor Peter Fray>They are, if you like, going to be our very own Channel 7, 9, 10, ABC, all rolled up into one.

343
00:18:09.820 --> 00:18:12.500
<v Professor Peter Fray>I really like how the writer, Talton Gillespie, says,

344
00:18:12.500 --> 00:18:20.740
<v Professor Peter Fray>recently expressed this. He said, in a way, the choices are the product. They're meant to draw in

345
00:18:20.740 --> 00:18:28.520
<v Professor Peter Fray>users and keep them on the platform. And we pay with our attention to the advertisers, and we pay

346
00:18:28.520 --> 00:18:35.140
<v Professor Peter Fray>by handing over ever more personal data. So the platforms are not producing the content, but they

347
00:18:35.140 --> 00:18:41.400
<v Professor Peter Fray>are far more than a simple conduit to it. With smart machines, this relationship is going to get

348
00:18:41.400 --> 00:18:46.520
<v Professor Peter Fray>more pronounced and more powerful and probably more confusing all at the same time. And I don't

349
00:18:46.520 --> 00:18:52.040
<v Professor Peter Fray>think for a second the platforms are going to fix any of it. It's not in their interest to do so,

350
00:18:52.180 --> 00:18:58.840
<v Professor Peter Fray>and that is why it is going to be up to us. With the rise of voice activation, humans have a clear

351
00:18:58.840 --> 00:19:05.180
<v Professor Peter Fray>choice. We can work with the machines to create meaning, or we can roll over and let the machines

352
00:19:05.180 --> 00:19:10.960
<v Professor Peter Fray>feed us what they think we like. Newspapers were, in a large part, a passive medium. We accepted

353
00:19:10.960 --> 00:19:11.380
<v Professor Peter Fray>something that was not a passive medium. We accepted something that was not a passive medium.

354
00:19:11.380 --> 00:19:11.500
<v Professor Peter Fray>We accepted something that was not a passive medium. We accepted something that was not a passive medium.

355
00:19:11.500 --> 00:19:17.620
<v Professor Peter Fray>But someone else's choices in those days are very much gone. There is no more important place to

356
00:19:17.620 --> 00:19:22.520
<v Professor Peter Fray>exercise this choice and create these meanings than in the area of journalism and its attendant

357
00:19:22.520 --> 00:19:27.520
<v Professor Peter Fray>areas such as civic debate and democracy. And I think we are approaching a pivotal moment.

358
00:19:27.820 --> 00:19:33.440
<v Professor Peter Fray>Up to now, the digital world has largely been about what machines can do for us to create meaning

359
00:19:33.440 --> 00:19:40.220
<v Professor Peter Fray>and purpose. Go back to the examples of music in journalism. But we need now, as we bring the likes

360
00:19:40.220 --> 00:19:41.360
<v Professor Peter Fray>of Alexa into our lives, to be able to create meaning and purpose. And I think that's what we need

361
00:19:41.360 --> 00:19:48.200
<v Professor Peter Fray>to do. We need to be mindful that how we interact with machines says much about what sort of world

362
00:19:48.200 --> 00:19:54.040
<v Professor Peter Fray>we want to live in and how we create meaning and purpose ourselves. So you may, of course,

363
00:19:54.060 --> 00:19:59.820
<v Professor Peter Fray>exercise choice and keep the machines out. You don't have to have them. But many of us will let

364
00:19:59.820 --> 00:20:05.640
<v Professor Peter Fray>them in. And it's not all of us. Not all of us are going to tell Alexa to suck it. So when we think

365
00:20:05.640 --> 00:20:11.040
<v Professor Peter Fray>about journalism, much of journalism's authority stems from it being witness and chronicler of news

366
00:20:11.040 --> 00:20:11.340
<v Professor Peter Fray>affairs. And I think that's a very important thing. And I think that's a very important thing.

367
00:20:13.560 --> 00:20:15.900
<v Professor Peter Fray>And I think that's a very important thing. And I think that's a very important thing.

368
00:20:16.220 --> 00:20:21.200
<v Professor Peter Fray>And I think that's a very important thing. And I think that's a very important thing.

369
00:20:21.680 --> 00:20:23.540
<v Professor Peter Fray>And I think that's a very important thing.

370
00:20:24.080 --> 00:20:25.660
<v Professor Peter Fray>And I think that's a very important thing.

371
00:20:25.680 --> 00:20:26.240
<v Professor Peter Fray>And I think that's a very important thing.

372
00:20:26.240 --> 00:20:26.800
<v Professor Peter Fray>And I think that's a very important thing.

373
00:20:26.980 --> 00:20:28.360
<v Professor Peter Fray>And I think that's a very important thing.

374
00:20:28.360 --> 00:20:28.840
<v Professor Peter Fray>And I think that's a very important thing.

375
00:20:28.840 --> 00:20:29.000
<v Professor Peter Fray>And I think that's a very important thing.

376
00:20:29.460 --> 00:20:29.700
<v Professor Peter Fray>And I think that's a very important thing.

377
00:20:30.160 --> 00:20:30.540
<v Professor Peter Fray>And I think that's a very important thing.

378
00:20:31.220 --> 00:20:33.060
<v Professor Peter Fray>And I think that's a very important thing.

379
00:20:33.460 --> 00:20:35.140
<v Professor Peter Fray>And I think that's a very important thing.

380
00:20:35.720 --> 00:20:37.820
<v Professor Peter Fray>And I think that's a very important thing.

381
00:20:38.320 --> 00:20:39.120
<v Professor Peter Fray>And I think that's a very important thing.

382
00:20:39.120 --> 00:20:44.920
<v Professor Peter Fray>choice now than you ever have in the past. So I think when we get these machines into our homes,

383
00:20:45.100 --> 00:20:51.320
<v Professor Peter Fray>we have to think about and choose how we interact with them. And we have to think about what those

384
00:20:51.320 --> 00:21:00.480
<v Professor Peter Fray>choices say about us. And I think we can choose to be active consumers of news, but it really is

385
00:21:00.480 --> 00:21:06.560
<v Professor Peter Fray>time for us to give choice a new voice. And that really is over to us. Thank you.

386
00:21:06.560 --> 00:21:12.340
<v Chair>And if I can just have one last word here, one practical tip that I have for you is never to let

387
00:21:12.340 --> 00:21:19.720
<v Chair>your kids use your phone because it starts to think that you're actually interested in gossip

388
00:21:19.720 --> 00:21:26.960
<v Chair>girls and, or your computer for that matter. So if you've got your kids doing searches for you,

389
00:21:27.040 --> 00:21:30.600
<v Chair>you know, you'll end up being dished up all sorts of stuff that you really don't want.

390
00:21:31.680 --> 00:21:36.020
<v Chair>So can please join me in thanking Professor Peter Frey today.

391
00:21:36.560 --> 00:21:36.720
<v Chair>Thank you.