1
00:00:04,400 --> 00:00:07,800
Speaker 1: Welcome to tech Stuff, a production from I Heart Radio.

2
00:00:12,160 --> 00:00:14,760
Speaker 1: Hey there, and welcome to tech Stuff. I'm your host,

3
00:00:14,880 --> 00:00:18,040
Speaker 1: Johnathan Strickland. I'm an executive producer with I Heart Radio

4
00:00:18,040 --> 00:00:21,440
Speaker 1: and a love of all things tech except for tech news.

5
00:00:21,480 --> 00:00:26,799
Speaker 1: Today's a tech news day for Thursday, October one. And y'all,

6
00:00:27,520 --> 00:00:30,920
Speaker 1: the tech news can just really it's just like reading

7
00:00:30,920 --> 00:00:33,120
Speaker 1: the news in general. He can really beat you down

8
00:00:33,159 --> 00:00:36,480
Speaker 1: after a while. We're really going to focus on Amazon

9
00:00:36,880 --> 00:00:39,800
Speaker 1: and Facebook in this news episode because a lot of

10
00:00:39,800 --> 00:00:43,159
Speaker 1: stuff has come out about both companies and it's not

11
00:00:43,720 --> 00:00:47,000
Speaker 1: great for the most part. So let's start with Amazon.

12
00:00:47,440 --> 00:00:50,680
Speaker 1: Reuter's reports that Amazon has been found to game the

13
00:00:50,760 --> 00:00:53,760
Speaker 1: system in a really big way, leaning heavily on the

14
00:00:53,760 --> 00:00:59,120
Speaker 1: company's ability to gather and analyze data at a huge scale. Now,

15
00:00:59,200 --> 00:01:03,960
Speaker 1: in this particular case, we're talking about Amazon's operations in India.

16
00:01:04,160 --> 00:01:08,559
Speaker 1: Broiters came into possession of various internal documents and emails

17
00:01:09,160 --> 00:01:15,200
Speaker 1: that show Amazon employees were analyzing customer behaviors on the platform. Okay,

18
00:01:15,440 --> 00:01:18,360
Speaker 1: no problem so far. I mean, you would expect Amazon

19
00:01:18,440 --> 00:01:23,880
Speaker 1: to analyze customer behaviors because the company wants to streamline

20
00:01:23,920 --> 00:01:26,600
Speaker 1: the experience, make it as useful as possible so that

21
00:01:26,640 --> 00:01:31,319
Speaker 1: people keep coming back to Amazon. So some analysis makes sense.

22
00:01:32,000 --> 00:01:35,640
Speaker 1: Amazon was also looking at which products were particularly popular.

23
00:01:35,920 --> 00:01:39,800
Speaker 1: Still no problem. I mean, if you identify which products

24
00:01:39,840 --> 00:01:43,880
Speaker 1: are popular, you can feature those and thus sell more units.

25
00:01:43,959 --> 00:01:48,960
Speaker 1: So not necessarily a bad thing here either. However, company

26
00:01:49,000 --> 00:01:55,040
Speaker 1: employees designated these as reference products and then set about

27
00:01:55,080 --> 00:02:00,840
Speaker 1: to make Amazon knockoff versions of those things. Now they

28
00:02:00,840 --> 00:02:04,520
Speaker 1: wouldn't go so far as to actually counterfeit something, but

29
00:02:04,600 --> 00:02:08,720
Speaker 1: they did make products that were virtually identical to existing

30
00:02:08,760 --> 00:02:13,280
Speaker 1: ones on the market. So this wasn't about Amazon saying, hey,

31
00:02:13,320 --> 00:02:17,280
Speaker 1: you know, recharging cables are really popular, Let's make our

32
00:02:17,280 --> 00:02:21,320
Speaker 1: own recharging cable. This was Amazon saying, hey, this very

33
00:02:21,320 --> 00:02:25,400
Speaker 1: specific recharging cable is really popular. Let's make a version

34
00:02:25,440 --> 00:02:29,320
Speaker 1: that is indistinguishable from this, sell it at a lower cost,

35
00:02:29,720 --> 00:02:34,160
Speaker 1: and then promote it and thus make, you know, essentially

36
00:02:34,240 --> 00:02:40,120
Speaker 1: undercut this existing vendor who is dependent upon Amazon as

37
00:02:40,160 --> 00:02:44,280
Speaker 1: an online shop front. So on top of all this,

38
00:02:44,360 --> 00:02:48,240
Speaker 1: Amazon used its algorithm to place the Amazon made versions

39
00:02:48,400 --> 00:02:52,600
Speaker 1: of these products higher in search. So now someone searching

40
00:02:52,639 --> 00:02:55,080
Speaker 1: for a specific product, you know, it could be a

41
00:02:55,080 --> 00:02:57,560
Speaker 1: piece of clothing, it could be an electronics gadget, it

42
00:02:57,600 --> 00:03:01,000
Speaker 1: could be a household item, They're more likely to come

43
00:03:01,040 --> 00:03:05,760
Speaker 1: across the Amazon branded version of that before whichever one

44
00:03:06,000 --> 00:03:10,240
Speaker 1: Amazon employees were using as a reference product. So you

45
00:03:10,240 --> 00:03:14,760
Speaker 1: can find something that is almost identical to a very

46
00:03:14,880 --> 00:03:20,160
Speaker 1: popular product that's branded by Amazon. And because the algorithm,

47
00:03:20,240 --> 00:03:24,079
Speaker 1: you're seeing the Amazon one first, and their Amazon is

48
00:03:24,160 --> 00:03:28,120
Speaker 1: essentially undercutting these other merchants. Now, I'm pretty sure you

49
00:03:28,120 --> 00:03:29,800
Speaker 1: would be hard pressed to come up with a way

50
00:03:30,120 --> 00:03:34,440
Speaker 1: to defend Amazon against a charge of being anti competitive

51
00:03:35,040 --> 00:03:38,920
Speaker 1: in this case. Uh the internal documents showed that company

52
00:03:38,920 --> 00:03:43,800
Speaker 1: employees wanted to displace existing products on the market. According

53
00:03:43,800 --> 00:03:46,680
Speaker 1: to Reuters, the documents also showed that employees even sought

54
00:03:46,680 --> 00:03:50,200
Speaker 1: out partnerships with the companies that were making the reference

55
00:03:50,280 --> 00:03:54,680
Speaker 1: products because some of those products had final touches or

56
00:03:54,800 --> 00:03:58,560
Speaker 1: unique processes that meant that those products were set apart

57
00:03:58,960 --> 00:04:02,440
Speaker 1: and then it was harder to to make a you know,

58
00:04:02,520 --> 00:04:07,440
Speaker 1: a reasonable copy of that thing. So employees needed to

59
00:04:07,440 --> 00:04:10,560
Speaker 1: be able to find out what these unique processes were

60
00:04:11,040 --> 00:04:16,479
Speaker 1: in order to replicate those products. For Amazon. That's pretty

61
00:04:16,560 --> 00:04:19,760
Speaker 1: much industrial espionage when you get down to it again.

62
00:04:19,800 --> 00:04:22,760
Speaker 1: This is all according to Reuters, and this also contradicts

63
00:04:22,800 --> 00:04:25,640
Speaker 1: claims that Jeff Bezos made when he appeared before the

64
00:04:25,680 --> 00:04:29,880
Speaker 1: United States Congress in twenty twenty. During that testimony, Bezos

65
00:04:29,960 --> 00:04:33,039
Speaker 1: argued that Amazon doesn't use the data it collects to

66
00:04:33,080 --> 00:04:36,080
Speaker 1: give its own branded products a leg up on other

67
00:04:36,200 --> 00:04:39,360
Speaker 1: sellers that would be anti competitive, nor that the company

68
00:04:39,480 --> 00:04:43,640
Speaker 1: uses its algorithm to boost Amazon brands in search results

69
00:04:43,680 --> 00:04:47,520
Speaker 1: at the expense of other brands. But hey, maybe maybe

70
00:04:47,560 --> 00:04:50,320
Speaker 1: he just meant, oh except in India. Maybe that's what

71
00:04:50,440 --> 00:04:52,800
Speaker 1: he just forgot to say that part. Even if all

72
00:04:52,839 --> 00:04:55,880
Speaker 1: of that is the case, the Reuters report is already

73
00:04:55,960 --> 00:05:00,000
Speaker 1: stirring up activity in US politics again. Senator Elizabeth Warren

74
00:05:00,520 --> 00:05:04,520
Speaker 1: sent a tweet saying, quote, these documents show what we

75
00:05:04,680 --> 00:05:08,840
Speaker 1: feared about Amazon's monopoly power. That the company is willing

76
00:05:08,960 --> 00:05:11,920
Speaker 1: and able to rig its platform to benefit its bottom

77
00:05:11,960 --> 00:05:16,240
Speaker 1: line while stifling small businesses and entrepreneurs. This is one

78
00:05:16,279 --> 00:05:18,480
Speaker 1: of the many reasons we need to break it up.

79
00:05:18,839 --> 00:05:24,360
Speaker 1: End quote. Similarly, the American Economic Liberties Project criticized Amazon,

80
00:05:24,800 --> 00:05:28,080
Speaker 1: pointing out that there have been reports from various sellers

81
00:05:28,120 --> 00:05:32,880
Speaker 1: that alleged the company has copied product designs, indicating that

82
00:05:32,920 --> 00:05:36,560
Speaker 1: Amazon had likely used its powerful data collection and analysis

83
00:05:36,600 --> 00:05:40,200
Speaker 1: tools to identify products that could be big sellers, and

84
00:05:40,240 --> 00:05:43,719
Speaker 1: then followed a similar path as we saw in India

85
00:05:44,240 --> 00:05:49,720
Speaker 1: and the that that organization UH cited a an article

86
00:05:49,800 --> 00:05:51,920
Speaker 1: from the Wall Street Journal that was published in twenty

87
00:05:52,000 --> 00:05:56,160
Speaker 1: twenty that alleged Amazon employees had done pretty much the

88
00:05:56,200 --> 00:06:00,520
Speaker 1: same thing that was going on in India over here, Oudweight,

89
00:06:01,120 --> 00:06:05,159
Speaker 1: we are not done reporting on Amazon just yet. See.

90
00:06:05,960 --> 00:06:11,080
Speaker 1: Amazon also recently scrapped plans to build a distribution center

91
00:06:11,200 --> 00:06:15,159
Speaker 1: in San Diego County in San Diego, California. Now, the

92
00:06:15,240 --> 00:06:19,239
Speaker 1: company has not cited a specific reason for backing out

93
00:06:19,279 --> 00:06:22,800
Speaker 1: of this plan, like this was something that was not

94
00:06:22,920 --> 00:06:28,560
Speaker 1: a done deal, but it was heading towards that direction. Critics, however,

95
00:06:28,720 --> 00:06:32,080
Speaker 1: point to how San Diego recently passed legislation called the

96
00:06:32,160 --> 00:06:36,920
Speaker 1: Working Families Ordinance. Now, that law would require companies that

97
00:06:37,000 --> 00:06:41,720
Speaker 1: operate within San Diego County to pay the prevailing wage,

98
00:06:41,960 --> 00:06:46,040
Speaker 1: and the prevailing wage takes union wages as the benchmark,

99
00:06:46,920 --> 00:06:52,039
Speaker 1: and also any company that employs people within San Diego County.

100
00:06:52,040 --> 00:06:54,839
Speaker 1: Any company that operates within San Diego County has to

101
00:06:54,880 --> 00:06:59,400
Speaker 1: provide a baseline level of benefits and worker safety guarantees.

102
00:06:59,760 --> 00:07:03,600
Speaker 1: So in other words, if you don't meet these basic criteria,

103
00:07:03,839 --> 00:07:09,720
Speaker 1: you cannot operate within San Diego County. So these critics conclude,

104
00:07:10,320 --> 00:07:14,280
Speaker 1: it looks like Amazon said, huh, if we have to

105
00:07:14,600 --> 00:07:17,000
Speaker 1: if we build something here, if we build our distribution

106
00:07:17,040 --> 00:07:20,200
Speaker 1: center in San Diego County, we would have to pay

107
00:07:20,240 --> 00:07:22,960
Speaker 1: people more money, and we would have to spend more

108
00:07:23,000 --> 00:07:25,760
Speaker 1: money to make sure that employees are safe, and we'd

109
00:07:25,760 --> 00:07:27,960
Speaker 1: have to give them a minimum of fifty six hours

110
00:07:28,000 --> 00:07:35,040
Speaker 1: of annual sick leaf. No thanks now again, Amazon, I

111
00:07:35,040 --> 00:07:39,160
Speaker 1: should stress, has not referenced the law when addressing the

112
00:07:39,200 --> 00:07:42,040
Speaker 1: fact that the company is no longer building a distribution

113
00:07:42,120 --> 00:07:45,800
Speaker 1: center in San Diego County. But based on other recent

114
00:07:45,840 --> 00:07:49,360
Speaker 1: activities that we've seen in which Amazon has taken a

115
00:07:49,440 --> 00:07:55,400
Speaker 1: pretty tough stance against efforts of worker organization and unionization,

116
00:07:55,920 --> 00:07:59,960
Speaker 1: I feel like maybe the critics who are pointing at

117
00:08:00,120 --> 00:08:03,760
Speaker 1: this they might be onto something here. The company behind

118
00:08:03,760 --> 00:08:07,880
Speaker 1: the development of the warehouse, essentially the real estate company,

119
00:08:07,960 --> 00:08:11,240
Speaker 1: Chestnut Properties, sent a message out that also gets me

120
00:08:11,280 --> 00:08:14,640
Speaker 1: a bit angry. That message says, quote, just the threat

121
00:08:15,320 --> 00:08:19,400
Speaker 1: mention of this ordinance has already cost over four hundred

122
00:08:19,440 --> 00:08:22,560
Speaker 1: great jobs for the weld property that I have been

123
00:08:22,560 --> 00:08:26,720
Speaker 1: working on for over five years end quote. Now that

124
00:08:26,840 --> 00:08:31,000
Speaker 1: seems to argue that the ordinance is actively harming the

125
00:08:31,040 --> 00:08:35,559
Speaker 1: community because it's discouraging businesses from moving in and creating jobs,

126
00:08:35,840 --> 00:08:39,280
Speaker 1: and that this therefore is a bad thing. Now, keep

127
00:08:39,280 --> 00:08:43,520
Speaker 1: in mind, the ordinance demands a baseline level of compensation

128
00:08:43,720 --> 00:08:48,200
Speaker 1: and benefits to employees, and without that legislation, companies like

129
00:08:48,440 --> 00:08:52,360
Speaker 1: Amazon could be paying out much lower wages with fewer

130
00:08:52,360 --> 00:08:56,560
Speaker 1: benefits and fewer guarantees of workers safety. I would argue

131
00:08:56,880 --> 00:08:59,719
Speaker 1: that that kind of job does not qualify as a

132
00:09:00,160 --> 00:09:04,080
Speaker 1: unquote great job, you know, as Chestnut Properties has argued.

133
00:09:04,360 --> 00:09:07,319
Speaker 1: I mean, it's a job, so it is a job.

134
00:09:07,920 --> 00:09:11,319
Speaker 1: But if it's a job that keeps workers living below

135
00:09:11,360 --> 00:09:15,400
Speaker 1: the poverty line and workers are on the hook even

136
00:09:15,440 --> 00:09:19,600
Speaker 1: if they get sick or injured, that's not a great job.

137
00:09:20,400 --> 00:09:25,400
Speaker 1: That's that's like indentured servitude. It might have been great

138
00:09:25,440 --> 00:09:27,920
Speaker 1: for the developer. It might have been great for Chestnut

139
00:09:28,000 --> 00:09:30,400
Speaker 1: Properties because they saw to make a lot of money

140
00:09:30,520 --> 00:09:34,120
Speaker 1: for you know, essentially selling off this this space for

141
00:09:34,200 --> 00:09:37,880
Speaker 1: a distribution center. But that filter, so no, it's not

142
00:09:37,960 --> 00:09:40,640
Speaker 1: great for them. I mean I get that, but I've

143
00:09:40,640 --> 00:09:42,960
Speaker 1: said it a few times in this podcast, I since

144
00:09:43,600 --> 00:09:46,920
Speaker 1: that there is a general labor movement that's growing in

145
00:09:46,960 --> 00:09:49,559
Speaker 1: the United States, and it's something that has been overdue

146
00:09:49,640 --> 00:09:52,240
Speaker 1: because there have been a lot of big companies that

147
00:09:52,320 --> 00:09:55,559
Speaker 1: have grown exponentially at the expense of the folks who

148
00:09:55,559 --> 00:09:58,360
Speaker 1: work for those companies. One of the frequent things you'll

149
00:09:58,400 --> 00:10:02,440
Speaker 1: hear is, you know, it's hard to look at a

150
00:10:02,480 --> 00:10:06,720
Speaker 1: company that pays its employees at such a low level

151
00:10:07,240 --> 00:10:10,319
Speaker 1: that the employees can't afford whatever it is the company makes.

152
00:10:10,600 --> 00:10:13,120
Speaker 1: That's not necessarily the case with Amazon. Amazon makes a

153
00:10:13,160 --> 00:10:15,000
Speaker 1: lot of stuff, and it makes a lot of stuff

154
00:10:15,040 --> 00:10:18,280
Speaker 1: that's super cheap. But you get the idea, right, Like,

155
00:10:19,000 --> 00:10:22,240
Speaker 1: you know, the the argument about paying out, say fifteen

156
00:10:22,280 --> 00:10:26,079
Speaker 1: dollars an hour, Well that's pretty much the minimum wage

157
00:10:26,080 --> 00:10:28,760
Speaker 1: in California, and California is a state that has a

158
00:10:28,920 --> 00:10:34,160
Speaker 1: very high cost of living. So while that might sound

159
00:10:34,200 --> 00:10:36,160
Speaker 1: great if you're in a place that has a very

160
00:10:36,200 --> 00:10:40,240
Speaker 1: low cost of living in California by comparison, it does

161
00:10:40,280 --> 00:10:42,800
Speaker 1: not mean you're making a ton of money. And if

162
00:10:42,880 --> 00:10:45,600
Speaker 1: you're in an environment that could potentially lead to you

163
00:10:45,640 --> 00:10:49,600
Speaker 1: getting sick or hurt and you have no real benefits

164
00:10:49,640 --> 00:10:54,080
Speaker 1: to recover from that. That's not great. So a trio

165
00:10:54,280 --> 00:11:00,280
Speaker 1: of Amazon stories that are pretty rough. When we come act,

166
00:11:00,600 --> 00:11:03,160
Speaker 1: we're gonna do the same thing all over again, except

167
00:11:03,160 --> 00:11:14,439
Speaker 1: this time we'll talk about Facebook. Okay. So yeah, Facebook,

168
00:11:14,480 --> 00:11:17,000
Speaker 1: like I said on Tuesday's episode, had a bad week

169
00:11:17,120 --> 00:11:19,880
Speaker 1: last week. Well, that bad week has stretched into a

170
00:11:19,880 --> 00:11:23,040
Speaker 1: couple of bad weeks. You probably know that the company

171
00:11:23,040 --> 00:11:26,520
Speaker 1: has been dealing with consequences after some employees have leaked

172
00:11:26,559 --> 00:11:31,960
Speaker 1: internal documents to reporters, law enforcement agencies, and political bodies

173
00:11:32,000 --> 00:11:35,040
Speaker 1: like the United States Congress, and next week it will

174
00:11:35,120 --> 00:11:40,200
Speaker 1: be regulatory body bodies in the European Union. Well, when

175
00:11:40,240 --> 00:11:44,079
Speaker 1: that happens to a company, there are a few things

176
00:11:44,120 --> 00:11:48,160
Speaker 1: that a company can do. There are a few different um,

177
00:11:48,200 --> 00:11:52,319
Speaker 1: you know, the courses of action that you can pursue.

178
00:11:53,000 --> 00:11:55,040
Speaker 1: One thing a company could try to do is it

179
00:11:55,120 --> 00:11:59,480
Speaker 1: can try to address systemic problems within the organization. They

180
00:11:59,480 --> 00:12:03,280
Speaker 1: can try to operate in a more ethical, transparent, and

181
00:12:03,440 --> 00:12:07,160
Speaker 1: responsible way. It can try to improve in other words,

182
00:12:08,320 --> 00:12:11,320
Speaker 1: or we can do what Facebook has done and try

183
00:12:11,400 --> 00:12:16,240
Speaker 1: to just clamp down on leaks by instituting policies that

184
00:12:16,480 --> 00:12:19,720
Speaker 1: limit access to certain internal group assets on the company's

185
00:12:19,760 --> 00:12:26,160
Speaker 1: internal communications platform. Facebook uses Workplace. I've also used Workplace.

186
00:12:26,360 --> 00:12:31,600
Speaker 1: It's a platform for all sorts of internal company uh processes.

187
00:12:32,200 --> 00:12:34,840
Speaker 1: I'm actually amazed that folks have time to access the

188
00:12:34,920 --> 00:12:40,040
Speaker 1: darned thing to, you know, have conversations. UM. I find

189
00:12:40,120 --> 00:12:44,760
Speaker 1: navigating it somewhat tedious, but that's my own experience. And

190
00:12:45,000 --> 00:12:47,800
Speaker 1: in Facebook, it is a way for employees to be

191
00:12:47,840 --> 00:12:51,280
Speaker 1: able to access and share information internally within the company.

192
00:12:51,400 --> 00:12:54,800
Speaker 1: None the past, employees were pretty much able to communicate

193
00:12:54,840 --> 00:12:59,439
Speaker 1: across different topics within workplace without issue. They could jump

194
00:12:59,520 --> 00:13:03,800
Speaker 1: into a front channels essentially that are dedicated to specific

195
00:13:04,880 --> 00:13:10,439
Speaker 1: issues or concerns or corporate areas of interest. But now

196
00:13:10,640 --> 00:13:14,720
Speaker 1: certain groups that focus on specific topics, ones that Facebook

197
00:13:14,760 --> 00:13:18,720
Speaker 1: has labeled as being sensitive, have institute a kind of

198
00:13:18,720 --> 00:13:21,600
Speaker 1: members only kind of approach. In other words, it's a

199
00:13:21,679 --> 00:13:24,120
Speaker 1: need to know basis sort of thing. If you are

200
00:13:24,240 --> 00:13:28,080
Speaker 1: part of that department, you have access, Otherwise you do not.

201
00:13:28,920 --> 00:13:32,880
Speaker 1: Um and multiple outlets, including The New York Times, have

202
00:13:32,960 --> 00:13:36,359
Speaker 1: reported on this, showing that the effort to stop leaks

203
00:13:36,559 --> 00:13:40,400
Speaker 1: was then immediately leaked to the media. So what does

204
00:13:40,440 --> 00:13:43,120
Speaker 1: that mean for Facebook? Well, I would say it probably

205
00:13:43,240 --> 00:13:45,880
Speaker 1: means that for one thing, some employees at the company

206
00:13:46,040 --> 00:13:48,600
Speaker 1: might not be so thrilled about what Facebook has been

207
00:13:48,720 --> 00:13:53,600
Speaker 1: up to, and their discontent is enough to be an

208
00:13:53,679 --> 00:13:58,760
Speaker 1: ongoing issue, and that perhaps if Facebook wants to fix this,

209
00:13:59,280 --> 00:14:01,960
Speaker 1: they need to look at the root causes as opposed

210
00:14:01,960 --> 00:14:05,840
Speaker 1: to the leaks. The Verge has another Facebook story that

211
00:14:06,000 --> 00:14:10,440
Speaker 1: is interesting. The company has been working on various AI

212
00:14:10,559 --> 00:14:15,920
Speaker 1: projects that ultimately would relate to Facebook's augmented reality initiatives.

213
00:14:16,120 --> 00:14:18,080
Speaker 1: So the idea is to create a system that has

214
00:14:18,080 --> 00:14:22,000
Speaker 1: a type of almost like a memory connected to it.

215
00:14:22,120 --> 00:14:27,080
Speaker 1: Like the system itself would be able to go back

216
00:14:27,320 --> 00:14:31,800
Speaker 1: over data over time and glean important things from that.

217
00:14:32,600 --> 00:14:34,800
Speaker 1: You may have seen videos of Google kind of showing

218
00:14:34,800 --> 00:14:39,600
Speaker 1: off a similar approach to how like the Assistant program,

219
00:14:39,640 --> 00:14:43,960
Speaker 1: the virtual assistant in Google can continue to answer questions

220
00:14:44,000 --> 00:14:47,560
Speaker 1: about a subject. So you ask one question, then after

221
00:14:47,640 --> 00:14:49,840
Speaker 1: you get the answer, you follow up with another question,

222
00:14:50,400 --> 00:14:54,000
Speaker 1: and without you know, specifically referencing that you're still talking

223
00:14:54,040 --> 00:14:57,080
Speaker 1: about the same topic, assistant can then draw the conclusion

224
00:14:57,160 --> 00:14:59,520
Speaker 1: that the second question is about the same subject as

225
00:14:59,560 --> 00:15:02,800
Speaker 1: the first. Let me give you an example. Uh, using

226
00:15:02,800 --> 00:15:05,520
Speaker 1: this version of assistant, I could say something like, what's

227
00:15:05,560 --> 00:15:09,240
Speaker 1: the weather like at Disneyland tomorrow, and I get the answer,

228
00:15:09,520 --> 00:15:11,600
Speaker 1: and then I might follow that up with what time

229
00:15:11,640 --> 00:15:14,880
Speaker 1: does it open? Well, I didn't say Disneyland, I said

230
00:15:14,880 --> 00:15:18,240
Speaker 1: what time does it open? Using this kind of approach

231
00:15:18,280 --> 00:15:22,240
Speaker 1: to AI, the assistant might be able to glean the

232
00:15:22,280 --> 00:15:25,520
Speaker 1: fact that when I say it, I'm referring to the

233
00:15:26,160 --> 00:15:30,040
Speaker 1: subject from my previous question, that is Disneyland, and thus

234
00:15:30,040 --> 00:15:33,360
Speaker 1: give me the correct answer. Now this is trivial for

235
00:15:33,480 --> 00:15:35,880
Speaker 1: human beings, right, but it's actually a big deal for

236
00:15:35,960 --> 00:15:40,080
Speaker 1: artificial intelligence. AI does not natively have the ability to

237
00:15:40,240 --> 00:15:46,680
Speaker 1: understand context. So Facebook is working on several similar prop projects,

238
00:15:46,720 --> 00:15:51,320
Speaker 1: some of which are really ambitious. Now, the downside of

239
00:15:51,360 --> 00:15:53,720
Speaker 1: this is that it could lead to some pretty nasty

240
00:15:53,840 --> 00:15:58,800
Speaker 1: privacy and security issues. So, for example, imagine that someone

241
00:15:59,000 --> 00:16:02,640
Speaker 1: is wearing a pair of those a R ray band glasses,

242
00:16:02,640 --> 00:16:05,120
Speaker 1: except these are a more advanced version. They're one that

243
00:16:05,160 --> 00:16:10,200
Speaker 1: are have you know, more augmented reality type of features

244
00:16:10,200 --> 00:16:13,000
Speaker 1: built into them. They're not just like a camera built

245
00:16:13,000 --> 00:16:17,040
Speaker 1: into glasses. And let's say this person wearing the glasses

246
00:16:17,640 --> 00:16:20,960
Speaker 1: goes by a secure facility of some sort, maybe it's

247
00:16:20,960 --> 00:16:24,760
Speaker 1: a bank, right, and they look into the bank and

248
00:16:24,800 --> 00:16:27,440
Speaker 1: the camera and the glasses captures images of a person

249
00:16:27,480 --> 00:16:30,640
Speaker 1: who's accessing a secure part of that facility. So someone

250
00:16:30,680 --> 00:16:35,480
Speaker 1: who's actually authorized to go into some area that otherwise

251
00:16:35,520 --> 00:16:39,880
Speaker 1: you would not have access to. Okay, if you have

252
00:16:40,120 --> 00:16:43,280
Speaker 1: this capability and you pair it with something like facial recognition,

253
00:16:43,320 --> 00:16:47,040
Speaker 1: you could compromise that person's security. It could say, all right,

254
00:16:48,000 --> 00:16:52,160
Speaker 1: this person so and so has access to this secure

255
00:16:52,280 --> 00:16:55,760
Speaker 1: part of the bank. Here's some more information about so

256
00:16:55,800 --> 00:16:58,240
Speaker 1: and so that we're able to pull from the Internet.

257
00:16:58,400 --> 00:17:04,000
Speaker 1: Because of online mind social networking platforms and Twitter profiles

258
00:17:04,000 --> 00:17:06,600
Speaker 1: and all that kind of stuff, you could easily see

259
00:17:06,880 --> 00:17:11,959
Speaker 1: where this can quickly become a threat to privacy and security. Now,

260
00:17:12,000 --> 00:17:14,840
Speaker 1: that is a pretty dramatic example, right, That's something that's

261
00:17:14,840 --> 00:17:18,400
Speaker 1: obviously not going to apply to everyone. But the reality

262
00:17:18,480 --> 00:17:23,160
Speaker 1: is that this technology, unless it's implemented very carefully, could

263
00:17:23,160 --> 00:17:26,120
Speaker 1: become a huge threat to privacy for people in general.

264
00:17:26,640 --> 00:17:29,000
Speaker 1: You know, it could even be like who did I

265
00:17:29,040 --> 00:17:33,000
Speaker 1: see yesterday? Well, a question like that might be answered with, oh,

266
00:17:33,040 --> 00:17:35,400
Speaker 1: you saw your buddy Jim, and you also saw these

267
00:17:35,440 --> 00:17:37,960
Speaker 1: other five people you don't know, but I know them

268
00:17:38,000 --> 00:17:40,760
Speaker 1: because I have facial recognition technology and I was able

269
00:17:40,760 --> 00:17:43,720
Speaker 1: to cross reference it with the profiles that are available

270
00:17:43,760 --> 00:17:47,240
Speaker 1: on Facebook. So now you know all the people you encountered,

271
00:17:47,440 --> 00:17:49,680
Speaker 1: whether they wanted you to know them or not. They

272
00:17:49,680 --> 00:17:53,280
Speaker 1: were just going about their daily lives, So that is

273
00:17:53,320 --> 00:17:56,119
Speaker 1: a potential threat. Now Facebook is looking into some pretty

274
00:17:56,160 --> 00:18:00,240
Speaker 1: incredible AI applications such as hand and object manipulation shtion,

275
00:18:00,760 --> 00:18:02,680
Speaker 1: And in that case, you might be trying to learn

276
00:18:02,720 --> 00:18:05,560
Speaker 1: a specific skill and you could use an augmented reality

277
00:18:05,640 --> 00:18:08,199
Speaker 1: system to kind of teach you the steps you need

278
00:18:08,280 --> 00:18:10,439
Speaker 1: to do in order to build your skills. But you

279
00:18:10,440 --> 00:18:13,840
Speaker 1: could also use it potentially to analyze your own performance

280
00:18:13,960 --> 00:18:16,240
Speaker 1: and then and you know, give you tips on how

281
00:18:16,280 --> 00:18:18,879
Speaker 1: you can do that skill better and get you know,

282
00:18:19,000 --> 00:18:22,840
Speaker 1: improve upon it, which is really cool. So this sort

283
00:18:22,840 --> 00:18:27,280
Speaker 1: of hyper focused AI is really you know, compelling, but

284
00:18:27,400 --> 00:18:30,720
Speaker 1: you also start to see where the need for things

285
00:18:30,760 --> 00:18:34,600
Speaker 1: like privacy protection has to step in. This, by the way,

286
00:18:34,720 --> 00:18:37,320
Speaker 1: is all part of an initiative that Facebook calls Ego

287
00:18:37,520 --> 00:18:41,320
Speaker 1: four D. The Verge has a great piece that goes

288
00:18:41,320 --> 00:18:43,280
Speaker 1: into much further detail if you want to learn more

289
00:18:43,320 --> 00:18:47,600
Speaker 1: about this. It is titled Facebook is researching AI systems

290
00:18:47,640 --> 00:18:51,040
Speaker 1: that see here and remember everything you do. It's by

291
00:18:51,160 --> 00:18:54,760
Speaker 1: James Vincent. Highly recommend you check it out more than

292
00:18:54,840 --> 00:18:57,960
Speaker 1: for The organizations have banded together to launch a campaign

293
00:18:58,000 --> 00:19:01,560
Speaker 1: called how to Stop Facebook. So yeah, that hard week

294
00:19:01,640 --> 00:19:05,640
Speaker 1: does just keep extending. The group advocates for regulations that

295
00:19:05,680 --> 00:19:09,399
Speaker 1: would restrict how Facebook collects and uses data, citing the

296
00:19:09,400 --> 00:19:12,200
Speaker 1: concerns that we've seen pop up due to Facebook's reliance

297
00:19:12,200 --> 00:19:17,680
Speaker 1: on engagement driving algorithms, and those algorithms frequently favor content

298
00:19:17,760 --> 00:19:21,000
Speaker 1: that can be harmful in various ways because it drives engagement,

299
00:19:21,480 --> 00:19:25,240
Speaker 1: and as the campaign strategies director of Media Justice Masha

300
00:19:25,440 --> 00:19:30,000
Speaker 1: Hayes has said, quote Facebook's surveillance capitalist business model is

301
00:19:30,080 --> 00:19:35,800
Speaker 1: fundamentally incompatible with basic human rights end quote. That's a

302
00:19:35,880 --> 00:19:38,600
Speaker 1: heck of a statement. Hard for me to find fault

303
00:19:38,640 --> 00:19:43,480
Speaker 1: in it. Um, especially the way Facebook pursues its you know,

304
00:19:43,720 --> 00:19:48,200
Speaker 1: business model. Uh, it appears that it is at least

305
00:19:48,240 --> 00:19:52,760
Speaker 1: indifferent to basic human rights. The group is calling upon leaders,

306
00:19:52,920 --> 00:19:56,000
Speaker 1: like political leaders to intervene and essentially force Facebook to

307
00:19:56,040 --> 00:19:59,480
Speaker 1: fundamentally change how it operates. Um. And there are a

308
00:19:59,480 --> 00:20:01,560
Speaker 1: lot of different groups that are all part of this,

309
00:20:01,760 --> 00:20:04,240
Speaker 1: and a lot of them are are human rights groups

310
00:20:04,320 --> 00:20:07,720
Speaker 1: and things along those lines that are very much concerned

311
00:20:08,080 --> 00:20:12,640
Speaker 1: that Facebook's performance in the last you know a few

312
00:20:12,760 --> 00:20:16,560
Speaker 1: years has shown it to be an organization that has

313
00:20:16,640 --> 00:20:20,159
Speaker 1: the capacity to do a disproportionate amount of harm to

314
00:20:20,240 --> 00:20:24,440
Speaker 1: vulnerable populations. Finally, the cherry on top for bad Facebook

315
00:20:24,480 --> 00:20:27,919
Speaker 1: news comes out of Washington State. The Attorney General of

316
00:20:28,000 --> 00:20:31,399
Speaker 1: Washington has filed a lawsuit against Facebook, saying that the

317
00:20:31,440 --> 00:20:35,960
Speaker 1: company representatives have provided false testimony during a previous case

318
00:20:36,440 --> 00:20:41,440
Speaker 1: about whether or not the platform violated Washington's campaign finance laws.

319
00:20:41,480 --> 00:20:43,640
Speaker 1: So at the heart of the matter is how Facebook

320
00:20:43,680 --> 00:20:46,920
Speaker 1: sold political ads in the state of Washington. Washington state

321
00:20:47,000 --> 00:20:51,359
Speaker 1: law requires that any platform that sells ads space to

322
00:20:51,520 --> 00:20:56,040
Speaker 1: political campaigns has to provide information about the ads who

323
00:20:56,080 --> 00:21:00,879
Speaker 1: bought them, uh that their address. It has to also

324
00:21:00,920 --> 00:21:04,240
Speaker 1: have who were the ads meant to target, how many

325
00:21:04,320 --> 00:21:07,240
Speaker 1: views did those ads get. All of that information is

326
00:21:07,280 --> 00:21:11,000
Speaker 1: required by state law to be publicly available upon request.

327
00:21:11,480 --> 00:21:13,960
Speaker 1: Facebook is not complied with the full extent of that law.

328
00:21:14,560 --> 00:21:18,200
Speaker 1: You can request information about Facebook ads, but it doesn't

329
00:21:18,240 --> 00:21:22,600
Speaker 1: include all the points of data that Washington State law

330
00:21:22,720 --> 00:21:27,560
Speaker 1: calls for. The Company's lawyers facebooks lawyers argue that Washington's

331
00:21:27,720 --> 00:21:31,680
Speaker 1: law is unconstitutional, but the heart of this matter is

332
00:21:31,720 --> 00:21:36,840
Speaker 1: that the Attorney General says in that case, Facebook representatives

333
00:21:37,440 --> 00:21:42,480
Speaker 1: specifically gave false statements to questions asked of them, and

334
00:21:42,560 --> 00:21:45,520
Speaker 1: that therefore they are guilty of perjury. So this is

335
00:21:45,560 --> 00:21:49,240
Speaker 1: another part layer on top of everything else. So this

336
00:21:49,359 --> 00:21:51,880
Speaker 1: legal battle is ongoing. We will have to check back

337
00:21:51,920 --> 00:21:57,320
Speaker 1: in the future. Okay, there were other news stories today,

338
00:21:57,359 --> 00:22:00,679
Speaker 1: but honestly, all the Amazon and face Book stuff kind

339
00:22:00,720 --> 00:22:03,040
Speaker 1: of took all the wind out of my sales just

340
00:22:03,040 --> 00:22:08,159
Speaker 1: because it was just so much rough reporting on different things.

341
00:22:08,560 --> 00:22:11,960
Speaker 1: So we're gonna wrap it up here. Plus I can

342
00:22:12,640 --> 00:22:15,840
Speaker 1: here that the lawn maintenance folks are outside of my

343
00:22:15,880 --> 00:22:19,000
Speaker 1: house now, so it's gonna get progressively louder. My dog's

344
00:22:19,000 --> 00:22:21,400
Speaker 1: gonna join into it's just gonna be a free for all.

345
00:22:21,480 --> 00:22:25,159
Speaker 1: So um, while the noise level continually goes up in

346
00:22:25,160 --> 00:22:28,159
Speaker 1: this episode, if you have suggestions for topics I should

347
00:22:28,240 --> 00:22:31,120
Speaker 1: cover in the future, maybe like how weed whackers work,

348
00:22:31,280 --> 00:22:33,840
Speaker 1: for example. Don't know why that jumped to mind, let

349
00:22:33,920 --> 00:22:37,160
Speaker 1: me know. Send me a message on Twitter. The handle

350
00:22:37,200 --> 00:22:40,720
Speaker 1: for the show is text stuff h s W and

351
00:22:40,920 --> 00:22:44,960
Speaker 1: I'll garden again. I'm sorry, I'll talk to you again

352
00:22:45,800 --> 00:22:54,040
Speaker 1: really soon. Text Stuff is an i heart Radio production.

353
00:22:54,280 --> 00:22:57,119
Speaker 1: For more podcasts from my heart Radio, visit the i

354
00:22:57,200 --> 00:23:00,480
Speaker 1: heart Radio app, Apple Podcasts, or wherever you listen to

355
00:23:00,480 --> 00:23:01,399
Speaker 1: your favorite shows.