1
00:00:00,120 --> 00:00:05,080
Speaker 1: Welcome to this week's classic episode. Let's play conspiracy Jeopardy.

2
00:00:05,280 --> 00:00:08,600
Speaker 1: The category is Google the Pentagon and AI.

3
00:00:08,560 --> 00:00:11,040
Speaker 2: I don't even know what to say about this one. Guys, like,

4
00:00:11,400 --> 00:00:15,560
Speaker 2: this is again, this is twenty eighteen. We didn't even understand,

5
00:00:15,920 --> 00:00:18,560
Speaker 2: We didn't even understand episode.

6
00:00:19,480 --> 00:00:22,040
Speaker 3: No, it's probably a really teachable moment for all of us.

7
00:00:22,079 --> 00:00:23,720
Speaker 3: We should go back and listen to the whole thing too,

8
00:00:23,800 --> 00:00:25,919
Speaker 3: and learn from our past selves. That's one of the

9
00:00:25,920 --> 00:00:28,000
Speaker 3: things that's weird about doing podcasts, you guess ever think

10
00:00:28,000 --> 00:00:32,200
Speaker 3: about that there's this weird version this history of us

11
00:00:32,600 --> 00:00:35,919
Speaker 3: being kind of varyingly behind and ahead of the eight

12
00:00:36,000 --> 00:00:40,040
Speaker 3: ball over the years, and sometimes it terrifies me to

13
00:00:40,080 --> 00:00:41,680
Speaker 3: go back and revisit it. So y'all out there in

14
00:00:41,760 --> 00:00:43,599
Speaker 3: conspiracy realists think can do it for us.

15
00:00:44,240 --> 00:00:48,720
Speaker 1: From UFOs to psychic powers and government conspiracies, history is

16
00:00:48,800 --> 00:00:53,120
Speaker 1: riddled with unexplained events. You can turn back now or

17
00:00:53,200 --> 00:00:55,160
Speaker 1: learn this stuff they don't want you to know.

18
00:01:08,520 --> 00:01:09,479
Speaker 2: Welcome back to the show.

19
00:01:09,520 --> 00:01:10,880
Speaker 4: My name is My name is Noah.

20
00:01:11,280 --> 00:01:13,840
Speaker 1: They call me Ben. We are joined with our super

21
00:01:13,880 --> 00:01:18,600
Speaker 1: producer Paul Big Apple decand most importantly, you are you.

22
00:01:18,600 --> 00:01:22,679
Speaker 1: You are here, and that makes this stuff they don't

23
00:01:22,720 --> 00:01:25,840
Speaker 1: want you to know. Today, we're going to talk about

24
00:01:25,880 --> 00:01:30,400
Speaker 1: a company called Google, commonly known as Google. That's its

25
00:01:30,400 --> 00:01:33,560
Speaker 1: street name. That's such a popular name, that's become a

26
00:01:33,640 --> 00:01:37,520
Speaker 1: verb in American English. They have a parent company who

27
00:01:37,520 --> 00:01:41,520
Speaker 1: could call it, the real company, which is known as Alphabet.

28
00:01:41,600 --> 00:01:44,160
Speaker 1: And we've talked about this a little bit in the past,

29
00:01:44,760 --> 00:01:49,320
Speaker 1: but we want to whatever we mentioned Google establish that

30
00:01:49,880 --> 00:01:53,600
Speaker 1: from the jump at the top, because it's strange how

31
00:01:54,280 --> 00:02:00,440
Speaker 1: relatively obscure that fact remains. And now, while Google O'tabet

32
00:02:00,560 --> 00:02:02,600
Speaker 1: whatever you want to call it may not be the

33
00:02:02,720 --> 00:02:05,760
Speaker 1: largest company in the world in terms of capital or

34
00:02:05,880 --> 00:02:10,600
Speaker 1: territory or material possessions, it is undoubtedly one of the

35
00:02:10,639 --> 00:02:15,040
Speaker 1: most influential organizations on the planet, at least in terms

36
00:02:15,200 --> 00:02:19,160
Speaker 1: of its connection to human beings or the ways in

37
00:02:19,240 --> 00:02:24,960
Speaker 1: which it influences networks or organizations or webs of human beings.

38
00:02:25,240 --> 00:02:30,080
Speaker 1: Most people on this planet, literally most people think about

39
00:02:30,120 --> 00:02:33,760
Speaker 1: that have heard of Google, even if they have somehow

40
00:02:33,800 --> 00:02:38,560
Speaker 1: not directly interacted with it, And arguably, on a slightly

41
00:02:38,600 --> 00:02:42,640
Speaker 1: more controversial note, most people have in some way to

42
00:02:42,800 --> 00:02:46,840
Speaker 1: some degree been affected by the actions of the US military,

43
00:02:46,960 --> 00:02:50,320
Speaker 1: whether that's something that happened generations ago in their own

44
00:02:50,400 --> 00:02:53,280
Speaker 1: country or an adjacent country, or to an ancestor, or

45
00:02:53,320 --> 00:02:56,400
Speaker 1: whether something is happening to them now as we record

46
00:02:56,480 --> 00:03:01,639
Speaker 1: this episode, and so today begins the question spoiler alert.

47
00:03:01,680 --> 00:03:04,560
Speaker 1: It also ends with a question, and our beginning question

48
00:03:04,840 --> 00:03:10,200
Speaker 1: is this what happens when these groups Google Alphabet, whatever

49
00:03:10,240 --> 00:03:12,560
Speaker 1: you want to call it, and the US military or

50
00:03:12,760 --> 00:03:18,480
Speaker 1: Uncle Sam, what happens when they join forces? Because they

51
00:03:18,480 --> 00:03:24,280
Speaker 1: have so, they have vultrons up, and they have vultroned

52
00:03:24,360 --> 00:03:32,560
Speaker 1: up in a again controversial way, it involves unmanned aerial systems,

53
00:03:33,080 --> 00:03:34,920
Speaker 1: also known as drones.

54
00:03:35,040 --> 00:03:37,720
Speaker 4: You are the ones that like deliver your Amazon packages, right.

55
00:03:37,640 --> 00:03:41,240
Speaker 1: Yeah, and some variation of them. Yeah, they're commonly called drones,

56
00:03:41,280 --> 00:03:45,840
Speaker 1: of course, but unmanned aerial vehicles or UAVs have been

57
00:03:45,960 --> 00:03:50,960
Speaker 1: a long running trope in science fiction, and surprisingly for

58
00:03:51,080 --> 00:03:55,160
Speaker 1: many people, some version of this technology existed long before

59
00:03:55,320 --> 00:03:59,760
Speaker 1: the modern day. There wasn't a history rattling moment when

60
00:03:59,800 --> 00:04:03,760
Speaker 1: the first drone ever went airborne. And that's in large

61
00:04:03,800 --> 00:04:08,320
Speaker 1: part because it's very difficult for us to isolate exactly

62
00:04:08,440 --> 00:04:11,880
Speaker 1: when the early version of an invention ends and the

63
00:04:11,920 --> 00:04:16,240
Speaker 1: first modern versions begin. Matt when you and I worked

64
00:04:16,240 --> 00:04:19,600
Speaker 1: on the invention show Stuff of Genius. We ran into

65
00:04:19,640 --> 00:04:23,560
Speaker 1: this pretty often. Somebody comes up with a rough approximation

66
00:04:23,680 --> 00:04:27,240
Speaker 1: of an idea and then over a course of years, decades,

67
00:04:27,320 --> 00:04:30,520
Speaker 1: or even centuries in some cases, people continually make small

68
00:04:30,600 --> 00:04:36,120
Speaker 1: improvements until it becomes the thing we know. Most inventions

69
00:04:36,480 --> 00:04:41,920
Speaker 1: are not a Eureka moment in reality. I mean, they

70
00:04:41,960 --> 00:04:44,960
Speaker 1: may be a Eureka moment in terms of the idea

71
00:04:45,160 --> 00:04:49,560
Speaker 1: or the epiphany, but most inventions in reality, and when

72
00:04:49,600 --> 00:04:52,359
Speaker 1: we get to the hardware phases are going to be

73
00:04:54,400 --> 00:04:57,159
Speaker 1: created by people standing on the shoulders of giants.

74
00:04:57,480 --> 00:04:59,479
Speaker 2: Yeah, it ends up being the person who makes it

75
00:04:59,520 --> 00:05:01,800
Speaker 2: marketable rather than just creates it.

76
00:05:02,040 --> 00:05:05,760
Speaker 1: Oh good call, and a very sad and capitalist call. Right,

77
00:05:05,800 --> 00:05:10,320
Speaker 1: it's the person who popularizes it. So if we look

78
00:05:10,360 --> 00:05:14,880
Speaker 1: at the first modern version of a drone, an unmanned vehicle,

79
00:05:14,960 --> 00:05:17,039
Speaker 1: would it be one of those bombs in the eighteen

80
00:05:17,120 --> 00:05:19,880
Speaker 1: hundreds that was strapped to a balloon and then just

81
00:05:19,920 --> 00:05:22,479
Speaker 1: sort of put off like something bad will happen, We

82
00:05:22,560 --> 00:05:25,440
Speaker 1: can't steer it. There wasn't a person on that and

83
00:05:25,480 --> 00:05:27,479
Speaker 1: it was being used as a weapon. Or was it

84
00:05:27,520 --> 00:05:31,440
Speaker 1: the initial V one rockets that Germany deployed during World

85
00:05:31,480 --> 00:05:35,320
Speaker 1: War Two. In the early nineteen hundreds, military groups did

86
00:05:35,400 --> 00:05:38,520
Speaker 1: use drones. They had radio controlled versions that were just

87
00:05:38,560 --> 00:05:43,840
Speaker 1: for target practice. Engineers developed unmanned aircraft that again had

88
00:05:43,960 --> 00:05:48,159
Speaker 1: munitions of some sort. These weren't drones the way we

89
00:05:48,240 --> 00:05:52,880
Speaker 1: think of them in the current age. But there's you know,

90
00:05:52,920 --> 00:05:56,080
Speaker 1: there are other examples, like cruise missiles. These things were

91
00:05:56,120 --> 00:05:59,520
Speaker 1: the first cruise missiles, often called flying torpedoes. They were

92
00:05:59,520 --> 00:06:02,799
Speaker 1: not meant to return to base, and drones are meant

93
00:06:02,920 --> 00:06:08,120
Speaker 1: to function like very very smart, sometimes very very deadly boomerangs.

94
00:06:08,200 --> 00:06:11,279
Speaker 1: They come out, they ain't come back, and somebody.

95
00:06:10,839 --> 00:06:15,200
Speaker 2: Claps yeah, generally not the people being targeted.

96
00:06:15,160 --> 00:06:19,239
Speaker 1: Generally, yes, yes, generally not. You are correct, my friend.

97
00:06:19,320 --> 00:06:22,680
Speaker 1: During the Cold War and the Vietnam War or the

98
00:06:22,760 --> 00:06:25,720
Speaker 1: series of conflicts collectively known as the Vietnam War here

99
00:06:25,720 --> 00:06:29,799
Speaker 1: in the US, Uncle Sam ramped up research on drones.

100
00:06:30,080 --> 00:06:33,240
Speaker 1: It was no longer just an interesting idea posed by

101
00:06:33,240 --> 00:06:36,359
Speaker 1: a few eccentric scientists and engineers. In fact, by the

102
00:06:36,400 --> 00:06:40,240
Speaker 1: time of the Vietnam War, unmanned drones flew thousands of

103
00:06:40,320 --> 00:06:44,360
Speaker 1: high risk reconnaissance missions, and they got shot down all

104
00:06:44,440 --> 00:06:48,239
Speaker 1: the time. But In that process, they saved the lives

105
00:06:48,279 --> 00:06:51,680
Speaker 1: of human pilots who otherwise would have been a board.

106
00:06:51,960 --> 00:06:54,440
Speaker 1: And I thought about I thought about you, Matt when

107
00:06:54,480 --> 00:06:58,160
Speaker 1: I was looking into this because another precedent for drones,

108
00:06:58,400 --> 00:07:00,680
Speaker 1: which we won't go into in this episode, another president

109
00:07:00,720 --> 00:07:04,880
Speaker 1: for drones where weaponized bats. Did you ever hear about this?

110
00:07:06,200 --> 00:07:07,679
Speaker 4: The oneted to drop the little bombs?

111
00:07:07,920 --> 00:07:10,640
Speaker 1: Yeah, well, yes, that's a version of it. Yeah, they

112
00:07:10,640 --> 00:07:13,600
Speaker 1: said bats can fly, we have access to them.

113
00:07:13,840 --> 00:07:15,960
Speaker 2: Well yeah, I mean you could go back to carrier

114
00:07:16,000 --> 00:07:18,280
Speaker 2: pigeons as being a type of drone as well.

115
00:07:18,720 --> 00:07:20,880
Speaker 1: Yeah, I mean there's not a person aboard.

116
00:07:21,440 --> 00:07:23,400
Speaker 5: There's a really cool episode of the podcast of the

117
00:07:23,440 --> 00:07:26,920
Speaker 5: Memory Palace about the bats that's called idy Bitty Bombs.

118
00:07:27,240 --> 00:07:28,000
Speaker 4: It's quite good.

119
00:07:28,160 --> 00:07:32,120
Speaker 1: Yeah, yeah, so check that one out for more information.

120
00:07:32,720 --> 00:07:35,920
Speaker 1: Around the same time Vietnam War, it goes this far back.

121
00:07:35,960 --> 00:07:39,160
Speaker 1: Around the same time of Vietnam War, engineers started adding

122
00:07:39,240 --> 00:07:45,880
Speaker 1: real time surveillance ability to drones, so cameras. Right today

123
00:07:46,160 --> 00:07:49,080
Speaker 1: we look at I guess the closest thing we have

124
00:07:49,160 --> 00:07:51,880
Speaker 1: to a military version of the first drone would be

125
00:07:52,200 --> 00:07:57,520
Speaker 1: the MQ one Predator, was first unveiled in nineteen ninety five,

126
00:07:57,880 --> 00:07:59,800
Speaker 1: which is weird when you think about it because that

127
00:07:59,880 --> 00:08:02,880
Speaker 1: was such a long time ago. And in two thousand

128
00:08:02,880 --> 00:08:05,840
Speaker 1: and two, the CIA used the MQ one Predator to

129
00:08:05,840 --> 00:08:09,240
Speaker 1: make its first successful kill of a quote unquote enemy

130
00:08:09,360 --> 00:08:13,680
Speaker 1: combatant in Afghanistan. Since then, since just two thousand and two,

131
00:08:13,720 --> 00:08:18,200
Speaker 1: this technology has grown by cartoonishly extreme leaps and bounds,

132
00:08:18,640 --> 00:08:22,120
Speaker 1: and it's currently on the bleeding edge of scientific advancement. Yeah.

133
00:08:22,160 --> 00:08:25,040
Speaker 2: And if that MQ one Predator drone, that's the drone

134
00:08:25,080 --> 00:08:26,960
Speaker 2: that you have seen before, that's the one that is

135
00:08:27,000 --> 00:08:30,080
Speaker 2: in pictures all over the place, that is the I mean,

136
00:08:30,160 --> 00:08:32,000
Speaker 2: I guess being one of the first, it just became

137
00:08:32,200 --> 00:08:34,120
Speaker 2: iconic in that way. Just gonna want to throw that

138
00:08:34,160 --> 00:08:34,440
Speaker 2: in there.

139
00:08:34,559 --> 00:08:38,960
Speaker 1: Yeah, it looks like a windowless commercial plane with an

140
00:08:39,160 --> 00:08:43,480
Speaker 1: upside down tail fin, right, two upside down tail fins.

141
00:08:43,559 --> 00:08:47,319
Speaker 2: Yeah, it's kind of The shape is slightly slightly odd,

142
00:08:47,320 --> 00:08:48,960
Speaker 2: but yeah, it's it's pretty cool.

143
00:08:49,440 --> 00:08:53,120
Speaker 1: And as these drones become more and more common in

144
00:08:53,240 --> 00:08:56,880
Speaker 1: theaters of war or in commercial industries like Amazon using them,

145
00:08:56,960 --> 00:09:00,920
Speaker 1: or recreational use like you buy a when one is

146
00:09:00,960 --> 00:09:06,480
Speaker 1: a gift, they're also doing the inorganic version of Specie

147
00:09:06,480 --> 00:09:09,760
Speaker 1: eighting you know, speci eating is when a relatively common

148
00:09:09,800 --> 00:09:18,440
Speaker 1: ancestor through generations and generations, produces different species of the

149
00:09:18,480 --> 00:09:22,920
Speaker 1: same template. Right, So these drones are becoming increasingly specialized.

150
00:09:22,960 --> 00:09:26,320
Speaker 1: They have varying ranges of abilities and odds. Are you've

151
00:09:26,320 --> 00:09:28,840
Speaker 1: seen one, right, We've all seen toy drones.

152
00:09:28,520 --> 00:09:29,920
Speaker 4: The little quad copter guys.

153
00:09:30,280 --> 00:09:32,520
Speaker 5: Yeah, and you gotta wonder too, like how does that

154
00:09:32,600 --> 00:09:34,920
Speaker 5: work patent wise? Like did someone come up with that

155
00:09:35,000 --> 00:09:38,760
Speaker 5: design and it was just generic enough to be copied?

156
00:09:39,200 --> 00:09:41,160
Speaker 5: Like I always wonder about things like that, like the

157
00:09:41,160 --> 00:09:43,120
Speaker 5: fidget spinner, Like did that guy just do a bad

158
00:09:43,240 --> 00:09:45,840
Speaker 5: job of patenting the original fidget spinner? Or is it

159
00:09:45,880 --> 00:09:48,360
Speaker 5: so generic that that doesn't even apply. I'm wondering that

160
00:09:48,360 --> 00:09:50,559
Speaker 5: that's the same thing about the design of those quad

161
00:09:51,120 --> 00:09:51,800
Speaker 5: toy drones.

162
00:09:52,280 --> 00:09:55,040
Speaker 1: Entirely speculative, but it's probably at the very least a

163
00:09:55,200 --> 00:09:59,160
Speaker 1: series of patents. Yeah, they have to multiple things in play,

164
00:09:59,200 --> 00:09:59,960
Speaker 1: I would imagine.

165
00:10:00,120 --> 00:10:02,200
Speaker 5: Pro tip speaking of a play, if you're gonna get

166
00:10:02,200 --> 00:10:03,800
Speaker 5: one of those for your kids, get a good one

167
00:10:03,840 --> 00:10:05,800
Speaker 5: because the cheap ones you might think you got a deal,

168
00:10:06,200 --> 00:10:08,080
Speaker 5: they just don't fly, They suck.

169
00:10:08,520 --> 00:10:12,480
Speaker 1: And they're battery vampires too.

170
00:10:12,600 --> 00:10:14,840
Speaker 5: Oh, my god, that was the thing, Like we got one, like, oh,

171
00:10:15,080 --> 00:10:18,160
Speaker 5: it flies for about four minutes, four minutes, Like that's

172
00:10:18,200 --> 00:10:21,840
Speaker 5: not fun tops tops. Yeah, but what kind of batteries

173
00:10:21,880 --> 00:10:26,080
Speaker 5: do the big boys use? Are they fueled or are

174
00:10:26,080 --> 00:10:27,959
Speaker 5: they do you know that's classified?

175
00:10:28,080 --> 00:10:29,640
Speaker 2: Yeah, sorry, we can't talk about it.

176
00:10:29,679 --> 00:10:30,200
Speaker 4: You can't tell me.

177
00:10:30,320 --> 00:10:33,079
Speaker 2: No, I thought we were pals. Google it all right.

178
00:10:33,240 --> 00:10:39,920
Speaker 1: Oh yeah, so some of the on a serious note,

179
00:10:39,920 --> 00:10:44,559
Speaker 1: some of those power systems would vary, and by the

180
00:10:44,679 --> 00:10:48,040
Speaker 1: time we dive into one and this episode comes out,

181
00:10:48,240 --> 00:10:51,160
Speaker 1: it may have changed. Interesting, that's how quickly they're moving

182
00:10:51,200 --> 00:10:55,120
Speaker 1: on these. So we've seen these, right, and we've probably

183
00:10:55,160 --> 00:11:01,360
Speaker 1: seen photographs or videos of the larger of these tiny drones,

184
00:11:01,440 --> 00:11:04,520
Speaker 1: things like the Predator, like Matt just described. You can

185
00:11:05,120 --> 00:11:08,680
Speaker 1: find a photograph of it pretty easily. The differences are

186
00:11:08,720 --> 00:11:13,480
Speaker 1: stark and astonishing, and missiles are not the biggest difference.

187
00:11:13,559 --> 00:11:17,120
Speaker 1: The fact that they're weaponized. While that is incredibly dangerous

188
00:11:17,679 --> 00:11:20,600
Speaker 1: and possibly fatal, that is not the biggest difference. Some

189
00:11:20,679 --> 00:11:22,719
Speaker 1: of the bigger differences are going to be on the

190
00:11:22,760 --> 00:11:26,280
Speaker 1: inside of the drone. We're talking things like GPS, real

191
00:11:26,320 --> 00:11:31,440
Speaker 1: time satellite feeds, super complex microprocessors, and more, they're putting

192
00:11:31,480 --> 00:11:35,199
Speaker 1: brains on board of these things. And perhaps most disturbingly,

193
00:11:35,480 --> 00:11:39,360
Speaker 1: some drones are edging closer and closer and closer to

194
00:11:39,360 --> 00:11:43,520
Speaker 1: what's commonly known as artificial intelligence or technically known as

195
00:11:43,600 --> 00:11:51,079
Speaker 1: machine consciousness. And we're close to an old science fiction

196
00:11:51,200 --> 00:11:55,319
Speaker 1: question becoming a real question. Well, the first true artificial intelligence,

197
00:11:55,360 --> 00:11:59,720
Speaker 1: the first true machine consciousness be a weapon of war.

198
00:12:00,480 --> 00:12:03,439
Speaker 2: Yes, highly likely, definitely.

199
00:12:04,000 --> 00:12:05,760
Speaker 1: Yeah, why is that met?

200
00:12:06,280 --> 00:12:11,199
Speaker 2: That's because innovation through the military is one of the

201
00:12:11,240 --> 00:12:14,839
Speaker 2: ways that innovations occur. The biggest innovations and inventions occur

202
00:12:14,920 --> 00:12:18,280
Speaker 2: because there's funding from a military somewhere, and it's not

203
00:12:18,600 --> 00:12:22,080
Speaker 2: like it's the military itself doing these things. It is

204
00:12:22,360 --> 00:12:26,400
Speaker 2: usually a private company or someone who's being given a grant,

205
00:12:26,559 --> 00:12:27,880
Speaker 2: a government grant to do it.

206
00:12:27,880 --> 00:12:29,360
Speaker 5: We see this time and time again with many of

207
00:12:29,400 --> 00:12:31,839
Speaker 5: the stories that we've talked about, where the military is

208
00:12:31,880 --> 00:12:34,959
Speaker 5: also always like decades ahead of what consumers will ever

209
00:12:34,960 --> 00:12:37,480
Speaker 5: get their hands on, and it might be ten twenty

210
00:12:37,559 --> 00:12:40,680
Speaker 5: years before it trickles down into the public where maybe

211
00:12:40,679 --> 00:12:43,680
Speaker 5: we can have a you know, terabyte thumb drive, just

212
00:12:44,280 --> 00:12:47,319
Speaker 5: as a silly example, where they've had that technology probably

213
00:12:47,360 --> 00:12:49,119
Speaker 5: for a decade at this point.

214
00:12:48,920 --> 00:12:51,520
Speaker 2: I would say the only way for it, not for

215
00:12:51,559 --> 00:12:54,360
Speaker 2: the first machine consciousness, true machine consciousness, to not be

216
00:12:54,440 --> 00:12:57,439
Speaker 2: a weapon, is for every private company on the planet

217
00:12:57,480 --> 00:12:58,679
Speaker 2: to say I'm not going to do.

218
00:12:58,640 --> 00:13:02,520
Speaker 1: That, right, Yeah, for us all kumbayah and do some

219
00:13:02,920 --> 00:13:05,960
Speaker 1: hands across America on the tech sphere, right, or hands

220
00:13:05,960 --> 00:13:11,320
Speaker 1: across the world. So with this in mind, we're also

221
00:13:11,480 --> 00:13:16,480
Speaker 1: hitting on a larger phenomenon, this kind of kind of unfortunate.

222
00:13:16,600 --> 00:13:19,640
Speaker 1: We're a family show, but we also endeavor at our

223
00:13:19,640 --> 00:13:23,120
Speaker 1: best to be an honest when the great leaps and

224
00:13:23,200 --> 00:13:28,000
Speaker 1: technology are not almost always, but the majority of the time,

225
00:13:28,040 --> 00:13:31,160
Speaker 1: the great leaps and technology are driven by human vices.

226
00:13:31,440 --> 00:13:34,520
Speaker 1: A desire to win a war, a desire to have

227
00:13:35,040 --> 00:13:39,880
Speaker 1: more you know, social attention or more financial access than

228
00:13:40,000 --> 00:13:41,079
Speaker 1: a rival.

229
00:13:41,320 --> 00:13:43,120
Speaker 2: Or how do we compress all of this?

230
00:13:43,240 --> 00:13:45,760
Speaker 1: Porn Right, that's the other one. That's the third one,

231
00:13:45,800 --> 00:13:49,199
Speaker 1: a desire for us because in a very very real

232
00:13:49,360 --> 00:13:53,680
Speaker 1: and concrete and provable way, pornography is the reason that

233
00:13:54,400 --> 00:13:56,840
Speaker 1: you or your parents had a VHS instead of a

234
00:13:56,880 --> 00:14:00,959
Speaker 1: Beta MAX. It's the reason, Like we talked about this

235
00:14:01,000 --> 00:14:06,040
Speaker 1: I think previously. Yes, yeah, so often we all too

236
00:14:06,080 --> 00:14:14,960
Speaker 1: often we as a species are driven by non noble motivations,

237
00:14:15,480 --> 00:14:18,560
Speaker 1: but that doesn't mean that we're not smart. We are

238
00:14:18,720 --> 00:14:23,400
Speaker 1: a jar of very very smart, very belligerent cookies, and

239
00:14:23,560 --> 00:14:26,720
Speaker 1: we've been putting some of the smartest cookies of our

240
00:14:26,800 --> 00:14:33,520
Speaker 1: species into the search for artificial intelligence and machine learning,

241
00:14:33,600 --> 00:14:36,720
Speaker 1: a search that continues as we record this episode.

242
00:14:37,080 --> 00:14:39,640
Speaker 2: So just a little while ago, in July of twenty seventeen,

243
00:14:39,680 --> 00:14:42,840
Speaker 2: there were these two researchers named Greg Allen and Taniel

244
00:14:43,000 --> 00:14:47,200
Speaker 2: Chan and they published a study which was called Artificial

245
00:14:47,240 --> 00:14:51,320
Speaker 2: Intelligence and National Security, and in this they put forward

246
00:14:51,400 --> 00:14:58,000
Speaker 2: goals for developing law policies specifically that would that would

247
00:14:58,000 --> 00:15:00,840
Speaker 2: put ideas forth in how do you with AI when

248
00:15:00,920 --> 00:15:04,800
Speaker 2: it comes to national security military spending? Like what should

249
00:15:04,800 --> 00:15:07,920
Speaker 2: we as a species do when it comes to these

250
00:15:07,960 --> 00:15:11,040
Speaker 2: two things colliding? And here are some of their findings.

251
00:15:11,400 --> 00:15:14,640
Speaker 2: They again said the most transformative innovations in the field

252
00:15:14,680 --> 00:15:17,880
Speaker 2: of AI are coming from the private sector and academic world,

253
00:15:18,000 --> 00:15:21,680
Speaker 2: not from the government in military. So currently, right now

254
00:15:22,120 --> 00:15:25,520
Speaker 2: AI is something that's happening outside of the realm of

255
00:15:25,560 --> 00:15:29,480
Speaker 2: the military, which in July twenty seventeen, yay, that's a

256
00:15:29,480 --> 00:15:32,400
Speaker 2: good thing, maybe because you know, there could be some

257
00:15:32,480 --> 00:15:36,800
Speaker 2: nefarious private sector humans as well, But that's point A.

258
00:15:37,240 --> 00:15:39,800
Speaker 2: The second one is that the current capabilities of AI

259
00:15:40,000 --> 00:15:44,200
Speaker 2: have quote significant potential for national security, especially in the

260
00:15:44,240 --> 00:15:49,160
Speaker 2: areas of cyber defense and satellite imagery analysis. And then

261
00:15:49,160 --> 00:15:50,960
Speaker 2: the last thing, I'm just going to read this whole

262
00:15:50,960 --> 00:15:53,800
Speaker 2: thing as a quote. Future progress in AI has the

263
00:15:53,840 --> 00:15:57,880
Speaker 2: potential to be a transformative national security technology on par

264
00:15:58,000 --> 00:16:01,040
Speaker 2: with nuclear weapons, aircraft, computers, and biotech, so some of

265
00:16:01,080 --> 00:16:08,720
Speaker 2: the most influential technologies on weaponry on war. They're saying

266
00:16:08,760 --> 00:16:10,720
Speaker 2: AI is going to be the next big thing.

267
00:16:12,160 --> 00:16:14,040
Speaker 1: And they're not just talking the talk, right.

268
00:16:14,240 --> 00:16:16,800
Speaker 2: No, they are not just talking to the talk at all.

269
00:16:18,280 --> 00:16:21,760
Speaker 2: According to the Wall Street Journal, the Defense Department spent

270
00:16:21,960 --> 00:16:26,680
Speaker 2: seven point four billion dollars on artificial intelligence related areas

271
00:16:26,720 --> 00:16:31,000
Speaker 2: in twenty seventeen. Seven point four billion dollars on AI.

272
00:16:31,440 --> 00:16:34,600
Speaker 1: And you know, not to not to harp too much

273
00:16:34,720 --> 00:16:37,320
Speaker 1: on Eisenhower, but I believe he's the one who had

274
00:16:37,320 --> 00:16:42,160
Speaker 1: a quote that said every rocket the US builds represents

275
00:16:42,200 --> 00:16:44,920
Speaker 1: this many schools that we could have built but didn't.

276
00:16:46,080 --> 00:16:48,480
Speaker 1: And there are a couple of things, if we're being

277
00:16:48,600 --> 00:16:51,680
Speaker 1: entirely fair about that number, that we have to acknowledge.

278
00:16:51,760 --> 00:16:55,240
Speaker 1: First is that this is only the money we know of,

279
00:16:55,960 --> 00:17:00,720
Speaker 1: so seven point four billion, let's call that an at least, right,

280
00:17:01,360 --> 00:17:06,040
Speaker 1: And then on the other end artificial intelligence related areas.

281
00:17:06,480 --> 00:17:10,560
Speaker 1: That doesn't mean they're just building a silicon brain, right,

282
00:17:10,640 --> 00:17:13,600
Speaker 1: that there are many other things that we might not

283
00:17:14,040 --> 00:17:19,680
Speaker 1: associate directly with machine intelligence, that maybe function is support

284
00:17:19,880 --> 00:17:24,000
Speaker 1: or infrastructure. So this is all, This is the whole meal.

285
00:17:24,280 --> 00:17:26,959
Speaker 1: This isn't just the hamburger.

286
00:17:27,359 --> 00:17:30,840
Speaker 2: Yeah, well, and we agreed. We should especially point out

287
00:17:30,920 --> 00:17:34,800
Speaker 2: that the Department of Defense budget for fiscal year twenty

288
00:17:34,840 --> 00:17:39,960
Speaker 2: seventeen was five hundred and forty billion dollars around that number.

289
00:17:39,880 --> 00:17:43,480
Speaker 1: Still the biggest in the world. Yeah, and that's publicly acknowledged. Again,

290
00:17:43,680 --> 00:17:47,480
Speaker 1: I can't keep saying that enough. Publicly acknowledged.

291
00:17:49,320 --> 00:17:50,639
Speaker 2: That's what they ask Congress for.

292
00:17:50,880 --> 00:17:53,919
Speaker 1: Yes, it's true. Before I derail us, what do you

293
00:17:53,920 --> 00:17:56,160
Speaker 1: think should pause for a word from our sponsor?

294
00:17:56,640 --> 00:17:57,399
Speaker 4: That seems wise.

295
00:18:03,960 --> 00:18:09,520
Speaker 1: Here's where it gets crazy. We've painted a little bit

296
00:18:09,560 --> 00:18:12,639
Speaker 1: of the picture, right. We know this story now, fellow

297
00:18:12,720 --> 00:18:16,199
Speaker 1: conspiracy realist. It is a story of drones. It is

298
00:18:16,240 --> 00:18:18,800
Speaker 1: a story of private industry. It is a story of

299
00:18:19,080 --> 00:18:23,159
Speaker 1: government interaction, a story of war and a story of secrets.

300
00:18:24,119 --> 00:18:29,159
Speaker 1: But in this age of instantaneous information, we know more

301
00:18:29,440 --> 00:18:33,120
Speaker 1: than we would have known, say even two decades ago.

302
00:18:33,800 --> 00:18:37,040
Speaker 1: We know more about what's happening now. Similar to the

303
00:18:37,119 --> 00:18:43,240
Speaker 1: drones acquiring real time surveillance technology, we are able to

304
00:18:43,280 --> 00:18:46,479
Speaker 1: be more aware of surroundings in a distant land. And

305
00:18:46,520 --> 00:18:51,040
Speaker 1: that's how we know about something called what's it called METT.

306
00:18:51,720 --> 00:18:56,280
Speaker 2: It's the Algorithmic Warfare Cross Functional Team, also known as

307
00:18:56,320 --> 00:18:57,240
Speaker 2: Project Maven.

308
00:18:58,680 --> 00:19:00,680
Speaker 5: And this was announced on April twenty six of twenty

309
00:19:00,680 --> 00:19:05,680
Speaker 5: seventeen in a and internal memo from the Deputy Secretary

310
00:19:05,680 --> 00:19:08,600
Speaker 5: of Defense Robert Bob o work.

311
00:19:09,080 --> 00:19:14,840
Speaker 2: Yeahobo, he goes by bob work. But I just thought, yeah,

312
00:19:14,880 --> 00:19:18,680
Speaker 2: Boba work. So I guess we can just read parts

313
00:19:18,680 --> 00:19:20,439
Speaker 2: of this memo if you want to. It's kind of

314
00:19:20,440 --> 00:19:23,119
Speaker 2: hard to digest, that's the only thing. So maybe if we.

315
00:19:23,280 --> 00:19:26,560
Speaker 1: Anyway walk through parts, get down, humanize it.

316
00:19:26,720 --> 00:19:29,280
Speaker 2: Yes, let's let's do that. Let's do that. Here we go.

317
00:19:30,080 --> 00:19:33,160
Speaker 2: I'll be I'll be the dry and the dry part. Here,

318
00:19:33,320 --> 00:19:37,159
Speaker 2: Here we go. The aw CFT's first task is to

319
00:19:37,160 --> 00:19:42,320
Speaker 2: field technology to augment or automate processing, exploitation, and dissemination

320
00:19:42,840 --> 00:19:49,240
Speaker 2: for tactical unmanned aerial systems. Okay, okay, so good god,

321
00:19:50,240 --> 00:19:54,160
Speaker 2: Well let's keep going a little bit. So, so it's

322
00:19:54,160 --> 00:19:59,680
Speaker 2: gonna augment or automate stuff for UAVs or drones and

323
00:20:00,280 --> 00:20:04,320
Speaker 2: mid altitude full motion video in support of the Defeat

324
00:20:04,359 --> 00:20:08,119
Speaker 2: ISIS campaign, and this will help to reduce the human

325
00:20:08,160 --> 00:20:13,800
Speaker 2: factor's burden of full motion video analysis, increase actionable intelligence,

326
00:20:14,240 --> 00:20:16,280
Speaker 2: and enhance military decision making.

327
00:20:16,760 --> 00:20:21,160
Speaker 1: So what that all means is the PED processing, exploitation,

328
00:20:21,400 --> 00:20:26,320
Speaker 1: and dissemination means that they're hoping again to get a

329
00:20:26,320 --> 00:20:31,160
Speaker 1: brain on board your local neighborhood predator or whatever drone

330
00:20:31,200 --> 00:20:33,679
Speaker 1: they end up using by the time you are listening

331
00:20:33,720 --> 00:20:37,719
Speaker 1: to this episode. And the processing part is going to

332
00:20:37,720 --> 00:20:41,800
Speaker 1: be the analysis, which often would be the human part

333
00:20:41,800 --> 00:20:45,840
Speaker 1: of the equation. The exploitation would be the ability to

334
00:20:45,920 --> 00:20:50,320
Speaker 1: take action on that analysis, right to use that knowledge,

335
00:20:50,320 --> 00:20:56,199
Speaker 1: and dissemination would be sharing the information, determining what information

336
00:20:56,520 --> 00:21:01,400
Speaker 1: is relevant, what is irrelevant, and to send where that's

337
00:21:01,480 --> 00:21:06,080
Speaker 1: PED And that's the full motion video is similar to

338
00:21:06,119 --> 00:21:11,359
Speaker 1: the same thing because it's it's a little bit easier

339
00:21:11,480 --> 00:21:14,960
Speaker 1: to take a bunch of numbers that are sensed a

340
00:21:14,960 --> 00:21:18,640
Speaker 1: bunch of temperature variables, for instance, and tell a computer

341
00:21:19,160 --> 00:21:22,520
Speaker 1: or an automated system that whatever is within this range,

342
00:21:23,119 --> 00:21:26,680
Speaker 1: do action X. Whatever is outside of this range, do action.

343
00:21:26,880 --> 00:21:30,160
Speaker 2: Why, you know, even if the action is just alert

344
00:21:30,560 --> 00:21:33,680
Speaker 2: the human user, the person that's going to actually analyze

345
00:21:33,680 --> 00:21:33,960
Speaker 2: the thing.

346
00:21:34,080 --> 00:21:39,720
Speaker 1: Yeah, absolutely absolutely, And that's tougher with video, right because

347
00:21:39,720 --> 00:21:42,440
Speaker 1: there are many many more factors of play. Would you

348
00:21:42,480 --> 00:21:43,399
Speaker 1: say that sounds correct?

349
00:21:43,920 --> 00:21:47,360
Speaker 2: Absolutely? It just depends on what kind of sensors they're

350
00:21:47,440 --> 00:21:49,320
Speaker 2: using on the drones. If it's you know, is it

351
00:21:49,359 --> 00:21:53,119
Speaker 2: a heat based imaging system, is it just a you know,

352
00:21:53,560 --> 00:21:55,960
Speaker 2: I mean, what are you looking at? Essentially? What is

353
00:21:56,000 --> 00:21:58,640
Speaker 2: the data that you're looking at? And that's really hard

354
00:21:58,680 --> 00:22:00,399
Speaker 2: to tell because it's going to change the depending on

355
00:22:00,480 --> 00:22:03,000
Speaker 2: the model of drone and what they're you know, deploying.

356
00:22:04,359 --> 00:22:06,960
Speaker 2: The really interesting one of the really interesting things here

357
00:22:07,720 --> 00:22:10,920
Speaker 2: in my mind is that, you know, we're talking about

358
00:22:10,960 --> 00:22:13,760
Speaker 2: a single drone and having an onboard system, but they're

359
00:22:13,800 --> 00:22:16,240
Speaker 2: also talking about having a system that can analyze the

360
00:22:16,240 --> 00:22:19,000
Speaker 2: footage from all the drones that they have deployed in

361
00:22:19,040 --> 00:22:23,560
Speaker 2: a single space and then have the machine be able

362
00:22:23,600 --> 00:22:26,640
Speaker 2: to take all that data in and tell you what's

363
00:22:26,680 --> 00:22:31,440
Speaker 2: going on, and then send that whatever the algorithm decides

364
00:22:31,600 --> 00:22:34,159
Speaker 2: needs to happen, disseminate all of that information out to

365
00:22:34,200 --> 00:22:37,600
Speaker 2: the entire team that's also there, so you really have that.

366
00:22:38,200 --> 00:22:40,560
Speaker 2: One of the things we talked about recently in another episode,

367
00:22:40,600 --> 00:22:44,880
Speaker 2: a full scope picture of the battlefield essentially, so everyone

368
00:22:44,960 --> 00:22:48,400
Speaker 2: at all times sees exactly what's happening on the battlefield.

369
00:22:47,920 --> 00:22:52,560
Speaker 1: As close to an omniscient observer as possible. Yeah, as well,

370
00:22:52,640 --> 00:22:55,080
Speaker 1: we used to say humanly possible, but that's no longer

371
00:22:55,119 --> 00:22:55,960
Speaker 1: the case, right.

372
00:22:55,840 --> 00:22:58,640
Speaker 2: Yeah, exactly, So let's keep going on here a little

373
00:22:58,640 --> 00:23:02,520
Speaker 2: bit with what the memo said. This program will initially

374
00:23:02,560 --> 00:23:08,080
Speaker 2: provide computer vision algorithms for object detection, classification and alerts

375
00:23:08,160 --> 00:23:12,119
Speaker 2: for full motion video ped Further sprints will incorporate more

376
00:23:12,160 --> 00:23:16,119
Speaker 2: advanced computer vision technology. After successful sprints in support of

377
00:23:16,119 --> 00:23:22,440
Speaker 2: intelligence surveillance reconnaissance, this program will prioritize the integration of

378
00:23:22,480 --> 00:23:27,040
Speaker 2: similar technologies into other defense intelligence mission areas, so it's

379
00:23:27,119 --> 00:23:31,040
Speaker 2: not just for this single drone operation that they're gonna

380
00:23:31,680 --> 00:23:33,919
Speaker 2: deploy it to. It's going to be for bigger things.

381
00:23:35,040 --> 00:23:38,240
Speaker 2: It goes on to say, this program will also consolidate

382
00:23:38,800 --> 00:23:43,000
Speaker 2: existing algorithm based technology initiatives related to mission areas of

383
00:23:43,040 --> 00:23:48,560
Speaker 2: the Defense Intelligence Enterprise or DIE, including all initiatives that

384
00:23:48,640 --> 00:23:53,760
Speaker 2: develop employ or field artificial intelligence, automation, machine learning, deep learning,

385
00:23:53,840 --> 00:23:59,720
Speaker 2: and computer vision algorithms. So really they're talking about everything

386
00:24:00,440 --> 00:24:03,760
Speaker 2: in the future for military applications.

387
00:24:03,080 --> 00:24:06,240
Speaker 1: So beyond far beyond drones. Oh yes, these could also

388
00:24:06,280 --> 00:24:09,680
Speaker 1: be submarines for instance. These could be satellites, these could

389
00:24:09,720 --> 00:24:11,600
Speaker 1: be space shuttles.

390
00:24:11,800 --> 00:24:16,320
Speaker 2: It could be surveillance, just video camera surveillance that's just

391
00:24:16,400 --> 00:24:18,480
Speaker 2: on the ground either, right, you know, is this.

392
00:24:18,520 --> 00:24:22,480
Speaker 5: The kind of stuff where eventually you could program a

393
00:24:22,560 --> 00:24:27,439
Speaker 5: drone or some kind of killing machine with parameters that

394
00:24:27,480 --> 00:24:30,760
Speaker 5: are like legally sanctioned and you can say, okay, search

395
00:24:30,800 --> 00:24:33,879
Speaker 5: and destroy X target or this type of target.

396
00:24:34,080 --> 00:24:37,760
Speaker 1: That's the fear. The potential is there, The potential is there. Absolutely,

397
00:24:38,240 --> 00:24:42,360
Speaker 1: So essentially, what they're saying here is that DARPA and

398
00:24:42,400 --> 00:24:45,520
Speaker 1: the Pentagon have already poured tons of money into this

399
00:24:45,600 --> 00:24:49,040
Speaker 1: automated collection of data. And they're not alone. Other members

400
00:24:49,040 --> 00:24:52,359
Speaker 1: of the Alphabet Soup Gang, which is a very endearing

401
00:24:52,440 --> 00:24:54,480
Speaker 1: name for them, like the Apple Dumpling Gang, right, the

402
00:24:54,520 --> 00:24:57,920
Speaker 1: Alphabet Soup Gang as other members of course n SA

403
00:24:58,000 --> 00:25:02,560
Speaker 1: and so on. They have also put forth tons of

404
00:25:02,680 --> 00:25:06,320
Speaker 1: money and time into automating the collection of data, and

405
00:25:06,400 --> 00:25:08,879
Speaker 1: the next step is discover is discovering a means of

406
00:25:08,960 --> 00:25:12,280
Speaker 1: automating the analysis of data. And the step after that,

407
00:25:12,920 --> 00:25:17,600
Speaker 1: the especially spooky one, is enabling these unmanned machines to

408
00:25:17,880 --> 00:25:21,360
Speaker 1: immediately act in real time on the results of their

409
00:25:21,400 --> 00:25:25,720
Speaker 1: onboard analysis. So to the questioning Posonal, the answer would

410
00:25:25,760 --> 00:25:30,480
Speaker 1: be yes, they want to have at least the capability

411
00:25:31,000 --> 00:25:35,959
Speaker 1: to make a judge, jury, and executioner on an unmanned vehicle.

412
00:25:36,080 --> 00:25:38,280
Speaker 1: This doesn't mean they'll do it. A lot of tech

413
00:25:38,320 --> 00:25:41,720
Speaker 1: companies do have problems with it, or at least with

414
00:25:42,359 --> 00:25:44,840
Speaker 1: actualizing that potential.

415
00:25:45,400 --> 00:25:48,879
Speaker 5: Yeah, you even see problems with the self driving cars.

416
00:25:49,600 --> 00:25:52,000
Speaker 5: I mean, as much as much R and d as

417
00:25:52,040 --> 00:25:55,679
Speaker 5: have gone into those, they're certainly not infallible.

418
00:25:55,280 --> 00:25:58,200
Speaker 4: And they get into fender benders and make mistakes.

419
00:25:58,119 --> 00:26:02,480
Speaker 2: Absolutely of Let's just have one more little quote here,

420
00:26:02,480 --> 00:26:05,040
Speaker 2: and this is from the chief in charge of this

421
00:26:05,160 --> 00:26:09,960
Speaker 2: Algorithmic Warfare Cross Function Team or Project MAVEN, and he's

422
00:26:10,040 --> 00:26:14,520
Speaker 2: speaking specifically to what they hope to achieve in the

423
00:26:14,560 --> 00:26:18,800
Speaker 2: near term with this project. And he says, quote people

424
00:26:18,800 --> 00:26:21,800
Speaker 2: in computers will work symbiotically to increase the ability of

425
00:26:21,840 --> 00:26:26,120
Speaker 2: weapons systems to detect objects. Eventually, we hope that one

426
00:26:26,160 --> 00:26:28,399
Speaker 2: analyst will be able to do twice as much work,

427
00:26:28,560 --> 00:26:31,640
Speaker 2: potentially three times as much as they're doing now. That's

428
00:26:31,680 --> 00:26:35,120
Speaker 2: our goal. So in the short term, somebody sat down

429
00:26:35,119 --> 00:26:37,760
Speaker 2: with him, had a meeting, and we want to increase

430
00:26:39,200 --> 00:26:42,160
Speaker 2: just basically what our analysts can get done, and we're

431
00:26:42,160 --> 00:26:44,320
Speaker 2: gonna save money on that end, We're gonna save time

432
00:26:44,359 --> 00:26:47,040
Speaker 2: and we're gonna be more efficient. That's what this whole

433
00:26:47,080 --> 00:26:50,080
Speaker 2: project is about, at least in his eyes or his

434
00:26:51,760 --> 00:26:54,880
Speaker 2: pr thing that he got to say. So what we're

435
00:26:54,880 --> 00:26:57,080
Speaker 2: talking about with this project so far is just what

436
00:26:57,119 --> 00:26:59,159
Speaker 2: it means to achieve what it wants to achieve. We

437
00:26:59,160 --> 00:27:02,080
Speaker 2: haven't discussed how it was going to achieve all of this,

438
00:27:03,200 --> 00:27:07,240
Speaker 2: and that is where we have our friend Google.

439
00:27:08,280 --> 00:27:12,720
Speaker 5: Yes, Google, Google is nobody's friend. If Google was an

440
00:27:12,760 --> 00:27:16,840
Speaker 5: ice cream flavor, they'd be pre Lanes and Google.

441
00:27:17,320 --> 00:27:17,560
Speaker 2: Yes.

442
00:27:18,000 --> 00:27:20,240
Speaker 4: And I can't say that word on the podcast. Every

443
00:27:20,280 --> 00:27:22,640
Speaker 4: time you say project Mayven, I think it's a project Mayhem.

444
00:27:22,920 --> 00:27:27,439
Speaker 2: Yeah, slightly different. So we know here that both the

445
00:27:27,480 --> 00:27:31,040
Speaker 2: Defense Department and Alphabet and Google are saying that this collaboration,

446
00:27:31,520 --> 00:27:34,080
Speaker 2: them working together in any way on this at all,

447
00:27:34,400 --> 00:27:38,000
Speaker 2: is all about AI assisted analysis of drum footage. That is,

448
00:27:38,160 --> 00:27:40,440
Speaker 2: that is what they're saying. That is the party line

449
00:27:40,440 --> 00:27:44,040
Speaker 2: on both sides. The p rtists put forth is everything's cool.

450
00:27:44,119 --> 00:27:46,160
Speaker 2: We're just working on looking at video guys.

451
00:27:46,160 --> 00:27:54,359
Speaker 1: It's cool, right, right, specifically through something called tensor flow AI.

452
00:27:56,000 --> 00:27:59,960
Speaker 2: Yeah, Noel, you looked into this, right. What is TensorFlow?

453
00:28:00,880 --> 00:28:04,200
Speaker 5: This is from the website of TensorFlow trademark. I'm gonna

454
00:28:04,200 --> 00:28:05,960
Speaker 5: do it like I'm doing an ad read. TensorFlow is

455
00:28:05,960 --> 00:28:09,000
Speaker 5: an open source software library for high performance numerical computation.

456
00:28:09,240 --> 00:28:13,240
Speaker 5: It's flexible architecture allows easy deployment of computation across a

457
00:28:13,320 --> 00:28:17,320
Speaker 5: variety of platforms CPUs, GPUs, TPUs, and from desktops to

458
00:28:17,359 --> 00:28:21,160
Speaker 5: clusters of servers to mobile and edge devices. Originally developed

459
00:28:21,200 --> 00:28:24,040
Speaker 5: by researchers and engineers from the Google Brain team within

460
00:28:24,160 --> 00:28:27,760
Speaker 5: Google's AI organization, it comes with strong support from machine

461
00:28:27,800 --> 00:28:31,160
Speaker 5: learning and deep learning, and the flexible numerical computation core

462
00:28:31,240 --> 00:28:33,440
Speaker 5: is used across many other scientific domains.

463
00:28:34,880 --> 00:28:36,320
Speaker 4: Yep, that one.

464
00:28:36,640 --> 00:28:37,240
Speaker 2: Yeah, that's the what.

465
00:28:37,960 --> 00:28:42,200
Speaker 5: No, seriously, TensorFlow can actually feel kind of like totally

466
00:28:42,280 --> 00:28:46,320
Speaker 5: future science fiction kind of stuff for people who are

467
00:28:46,360 --> 00:28:49,880
Speaker 5: not already familiar with how machine learning works. But there's

468
00:28:49,880 --> 00:28:53,080
Speaker 5: some pretty awesome introductions primers out there if you would

469
00:28:53,120 --> 00:28:55,400
Speaker 5: like to learn more, particularly a YouTube video.

470
00:28:55,880 --> 00:29:00,760
Speaker 2: Yeah, it's about a forty five minute essentially on how

471
00:29:00,800 --> 00:29:03,800
Speaker 2: it functions. But it's good, it's interesting, it's over my head.

472
00:29:03,880 --> 00:29:05,840
Speaker 1: And the cool thing when I found that one the

473
00:29:05,880 --> 00:29:09,760
Speaker 1: cool thing about that guy is he's If you watch

474
00:29:09,840 --> 00:29:11,800
Speaker 1: the video, he says, if you have any questions, you

475
00:29:11,840 --> 00:29:13,200
Speaker 1: can email him directly.

476
00:29:13,560 --> 00:29:14,800
Speaker 2: Oh my gosh, I don't really.

477
00:29:14,880 --> 00:29:17,520
Speaker 1: Yeah, I don't know if he knew that his email

478
00:29:17,560 --> 00:29:20,600
Speaker 1: address was going out on YouTube, but he seemed pretty approachable.

479
00:29:20,760 --> 00:29:24,360
Speaker 2: And that video is on the Coding Tech YouTube page

480
00:29:24,400 --> 00:29:27,600
Speaker 2: and it's called TensorFlow one oh one. Really awesome intro

481
00:29:27,760 --> 00:29:28,600
Speaker 2: into TensorFlow.

482
00:29:28,760 --> 00:29:29,400
Speaker 4: Super awesome.

483
00:29:29,440 --> 00:29:31,680
Speaker 5: But yeah, but forty five minutes is probably about the

484
00:29:31,720 --> 00:29:34,120
Speaker 5: shallowest deep dive you're going to get into the subject.

485
00:29:34,720 --> 00:29:36,760
Speaker 5: And I mean that is a glowing endorsement of the

486
00:29:36,840 --> 00:29:38,400
Speaker 5: quality of this this walkthrough.

487
00:29:38,560 --> 00:29:42,320
Speaker 1: Yeah, so what exactly has TensorFlow done so far?

488
00:29:42,640 --> 00:29:45,640
Speaker 2: So in the research for this show, we came across

489
00:29:45,680 --> 00:29:49,480
Speaker 2: an open letter that was written from the International Committee

490
00:29:49,560 --> 00:29:53,880
Speaker 2: for Robot Arms Control, and they go into just some

491
00:29:53,920 --> 00:29:57,720
Speaker 2: of the details of exactly what is happening here the

492
00:29:57,960 --> 00:30:02,440
Speaker 2: collaboration between Google and the Department. So let's kind of

493
00:30:02,440 --> 00:30:06,080
Speaker 2: look at this. They were looking at an article from

494
00:30:06,120 --> 00:30:09,560
Speaker 2: Defense one in which they note that the Joint Special

495
00:30:09,560 --> 00:30:15,120
Speaker 2: Operations Forces they've conducted trials already using video footage from

496
00:30:15,160 --> 00:30:20,040
Speaker 2: this other smaller UAV called scan Eagle. It's a specific

497
00:30:20,080 --> 00:30:25,240
Speaker 2: surveillance drone, and the project is also slated to expand

498
00:30:25,640 --> 00:30:29,719
Speaker 2: to even quote larger medium altitude predator and reaper drones

499
00:30:29,800 --> 00:30:33,040
Speaker 2: by next summer. So these are the ones we spoke

500
00:30:33,080 --> 00:30:34,800
Speaker 2: about in the beginning. These are the ones that are

501
00:30:34,880 --> 00:30:38,360
Speaker 2: actively functioning in Syria and Afghanistan and other in Iraq

502
00:30:38,400 --> 00:30:42,640
Speaker 2: and other places like that, and eventually to this thing

503
00:30:42,800 --> 00:30:48,200
Speaker 2: called gorgon Stare, which sounds amazing, right.

504
00:30:50,440 --> 00:30:53,320
Speaker 4: Sounds scary, yeah, and it.

505
00:30:53,440 --> 00:31:01,240
Speaker 2: Turns it might. So this is these are specific sensors

506
00:31:01,560 --> 00:31:05,560
Speaker 2: and it's a technology package that's been it's been used

507
00:31:05,560 --> 00:31:09,080
Speaker 2: on these MQ nine reapers, which is just the newest relation.

508
00:31:09,240 --> 00:31:13,520
Speaker 2: Yeah yeah, and man, it sounds really scary. But you

509
00:31:13,560 --> 00:31:15,600
Speaker 2: can learn more about this right now. If you search

510
00:31:15,720 --> 00:31:22,280
Speaker 2: for gorgan go r gon Stare and maybe drone or predator,

511
00:31:22,320 --> 00:31:24,360
Speaker 2: that's probably the best way to look into it. And

512
00:31:24,400 --> 00:31:29,200
Speaker 2: there's all kinds of information that's been officially released on

513
00:31:29,240 --> 00:31:33,360
Speaker 2: this program already. So okay, so we know that this

514
00:31:33,440 --> 00:31:35,760
Speaker 2: thing's been used in these smaller projects. Now it's going

515
00:31:35,840 --> 00:31:38,360
Speaker 2: to be moved on to these Predator and Reaper drones,

516
00:31:38,840 --> 00:31:43,640
Speaker 2: and they're thinking that it's going to get expanded even further.

517
00:31:43,880 --> 00:31:46,000
Speaker 2: And I don't even we don't even know what that

518
00:31:46,080 --> 00:31:49,920
Speaker 2: expansion could be at this point, because the drone technology

519
00:31:50,600 --> 00:31:53,880
Speaker 2: is growing. It's just a lot of it is classified.

520
00:31:54,000 --> 00:31:56,800
Speaker 1: Right, Yeah, a lot of it is classified, and as

521
00:31:56,840 --> 00:32:02,240
Speaker 1: we've established earlier in the episode ode, that's really common.

522
00:32:02,520 --> 00:32:06,000
Speaker 1: In fact, a lot of it will probably be classified

523
00:32:06,120 --> 00:32:07,360
Speaker 1: for a number of years.

524
00:32:07,560 --> 00:32:09,280
Speaker 2: Well yeah, if you think about it this way. So

525
00:32:09,320 --> 00:32:13,360
Speaker 2: the gorgon Stare, it can allegedly and I just say

526
00:32:13,360 --> 00:32:15,840
Speaker 2: this because I haven't actually seen it, but it uses

527
00:32:15,880 --> 00:32:18,280
Speaker 2: a high tech series of cameras that can quote view

528
00:32:18,680 --> 00:32:20,320
Speaker 2: entire towns.

529
00:32:20,440 --> 00:32:23,800
Speaker 5: Like what like Google Earth style like yeah, that kind

530
00:32:23,800 --> 00:32:25,000
Speaker 5: of zoom, yeah, okay.

531
00:32:24,880 --> 00:32:27,680
Speaker 2: Active running looking at an entire town. And then if

532
00:32:27,720 --> 00:32:30,400
Speaker 2: you apply this machine learning to it and you can

533
00:32:30,440 --> 00:32:34,040
Speaker 2: see every human being walking around, and then the technology

534
00:32:34,040 --> 00:32:36,280
Speaker 2: gets more and more powerful, then you could really just

535
00:32:36,360 --> 00:32:38,680
Speaker 2: have a complete surveillance state from above.

536
00:32:38,840 --> 00:32:41,200
Speaker 5: Yeah, that's that's pretty scary, Like you can see them

537
00:32:41,200 --> 00:32:42,880
Speaker 5: in their homes. Is it have some kind of like

538
00:32:43,480 --> 00:32:45,840
Speaker 5: heat vision that works from afar?

539
00:32:46,080 --> 00:32:48,520
Speaker 2: That's a great question. Then I don't know the answer too.

540
00:32:49,440 --> 00:32:51,960
Speaker 2: I certainly hope not. Maybe it would be good for

541
00:32:52,000 --> 00:32:54,800
Speaker 2: a military application, but that's terrifying.

542
00:32:54,960 --> 00:32:57,080
Speaker 5: That's the problem, right, Like you can't put this stuff

543
00:32:57,120 --> 00:32:59,080
Speaker 5: back in the box. Like, even if it starts off

544
00:32:59,080 --> 00:33:02,120
Speaker 5: as a military application, it's going to eventually get in

545
00:33:02,160 --> 00:33:04,800
Speaker 5: the hands of Not to say that the military has

546
00:33:04,880 --> 00:33:08,640
Speaker 5: our best interests at all times, but it could make

547
00:33:08,680 --> 00:33:11,600
Speaker 5: its way into a more nefarious use.

548
00:33:12,200 --> 00:33:14,720
Speaker 1: If we were to put a list of all the

549
00:33:14,800 --> 00:33:18,400
Speaker 1: technologies that expanded past their intended use on one side,

550
00:33:18,720 --> 00:33:21,280
Speaker 1: and a list of all the technologies that people agreed

551
00:33:21,360 --> 00:33:25,320
Speaker 1: not to pursue on the other side, one list would

552
00:33:25,360 --> 00:33:30,000
Speaker 1: be very very long. I mean, we for a species

553
00:33:30,040 --> 00:33:34,160
Speaker 1: that loses cities and even entire civilizations, we're pretty good

554
00:33:34,160 --> 00:33:37,200
Speaker 1: at finding and replicating technology, and we don't like to

555
00:33:37,280 --> 00:33:40,800
Speaker 1: lose it. We did an earlier episode on things like

556
00:33:41,440 --> 00:33:45,240
Speaker 1: so called Damascus steel or Greek fire. But looking in

557
00:33:45,280 --> 00:33:48,640
Speaker 1: that it was very difficult for us to find examples

558
00:33:48,760 --> 00:33:52,040
Speaker 1: of technology that was truly lost in the modern day.

559
00:33:52,760 --> 00:33:58,240
Speaker 1: Once Pandora's jar is unscrewed, then the it's kadie bar

560
00:33:58,400 --> 00:34:01,240
Speaker 1: the door. Yeaistic statement, I Ardi.

561
00:34:01,840 --> 00:34:03,400
Speaker 4: Those badgers get out of that bag.

562
00:34:04,000 --> 00:34:06,560
Speaker 1: They will eat you alive, and you can't get them

563
00:34:06,600 --> 00:34:06,880
Speaker 1: back in.

564
00:34:07,040 --> 00:34:10,520
Speaker 2: No, man, Well, here's the thing. The badgers with this

565
00:34:10,680 --> 00:34:13,919
Speaker 2: project badgers in the bag. They don't even know where

566
00:34:13,920 --> 00:34:16,320
Speaker 2: the other one is right now. They're so far gone

567
00:34:16,320 --> 00:34:18,919
Speaker 2: from each other. Because this is where you get into

568
00:34:18,920 --> 00:34:22,640
Speaker 2: the whole targeted killing thing that was really big in

569
00:34:22,840 --> 00:34:25,440
Speaker 2: what two thousand and seven, two thousand and eight as

570
00:34:25,480 --> 00:34:30,000
Speaker 2: we got on through the Obama administration, where drones were

571
00:34:30,040 --> 00:34:33,839
Speaker 2: being used more and more to target mostly males, and

572
00:34:34,040 --> 00:34:36,640
Speaker 2: males were considered if you're of a certain age, you

573
00:34:36,680 --> 00:34:38,799
Speaker 2: could kind of prove that this male was of a

574
00:34:38,800 --> 00:34:41,680
Speaker 2: certain age even from a drone, that person would be

575
00:34:41,719 --> 00:34:45,920
Speaker 2: considered an enemy combatant in certain aspects or in certain ways.

576
00:34:46,480 --> 00:34:50,560
Speaker 2: And if you're using these algorithms to just figure out

577
00:34:50,640 --> 00:34:52,160
Speaker 2: you know, if you can just figure out it's a

578
00:34:52,160 --> 00:34:57,000
Speaker 2: male in in you know, above a certain age. Then

579
00:34:57,040 --> 00:34:59,319
Speaker 2: that's what Google is helping them do is figure out

580
00:34:59,360 --> 00:35:03,080
Speaker 2: how to kill people from Afar that are males of

581
00:35:03,120 --> 00:35:04,040
Speaker 2: a certain age. Right.

582
00:35:04,120 --> 00:35:08,240
Speaker 1: And this goes into a quotation that you found, Matt,

583
00:35:08,320 --> 00:35:10,800
Speaker 1: that gave me some chills.

584
00:35:11,560 --> 00:35:14,480
Speaker 2: Oh yeah. A quote from Eric Schmidt, who was who

585
00:35:14,760 --> 00:35:17,319
Speaker 2: up until last year or the end of last year,

586
00:35:17,360 --> 00:35:20,920
Speaker 2: he was the executive chairman at Google in alphabet, and

587
00:35:21,920 --> 00:35:25,600
Speaker 2: he says, there's a general concern in the tech community

588
00:35:25,719 --> 00:35:29,480
Speaker 2: of somehow the military industrial complex using their stuff to

589
00:35:29,640 --> 00:35:30,919
Speaker 2: kill people incorrectly.

590
00:35:32,200 --> 00:35:34,640
Speaker 4: Hold on there, how do you kill people incorrectly?

591
00:35:34,800 --> 00:35:37,399
Speaker 1: That's that's the way of doing Yeah, that's what gave

592
00:35:37,440 --> 00:35:40,600
Speaker 1: me chills, Matt, when you were originally talking about this,

593
00:35:40,719 --> 00:35:44,200
Speaker 1: the idea of killing people incorrectly. The implication there is

594
00:35:44,440 --> 00:35:52,800
Speaker 1: frightening because Schmidt is jumping straight across and completely sidelining

595
00:35:53,160 --> 00:35:57,080
Speaker 1: the recognition that maybe the tech community doesn't want to

596
00:35:57,160 --> 00:35:58,200
Speaker 1: kill people in general.

597
00:35:58,320 --> 00:36:01,359
Speaker 4: Is that just linguistic jiu jitsu kind of it's a bit.

598
00:36:01,520 --> 00:36:03,359
Speaker 1: Yeah, absolutely, yeah, it's a bit.

599
00:36:03,640 --> 00:36:05,360
Speaker 2: And that just so you know, that was a quote

600
00:36:05,760 --> 00:36:08,279
Speaker 2: from a keynote address he was giving at the New

601
00:36:08,280 --> 00:36:11,960
Speaker 2: American Security, Artificial Intelligence and Global Security.

602
00:36:11,560 --> 00:36:15,360
Speaker 1: Summit, right before the panel on how to correctly kill people.

603
00:36:15,960 --> 00:36:18,360
Speaker 2: Yeah, it was in November of twenty seventeen. Yeah, it

604
00:36:18,360 --> 00:36:19,360
Speaker 2: was right before that panel.

605
00:36:19,800 --> 00:36:22,640
Speaker 4: You shoot him in the face, that's how you do it,

606
00:36:22,920 --> 00:36:24,040
Speaker 4: and make eye contacts.

607
00:36:24,080 --> 00:36:24,720
Speaker 1: Huh yeah.

608
00:36:24,800 --> 00:36:25,280
Speaker 4: Yeah.

609
00:36:25,280 --> 00:36:32,960
Speaker 1: So the situation then seems on the cusp of something

610
00:36:33,040 --> 00:36:36,360
Speaker 1: a rapid evolution, perhaps an evolution that's already occurring. And

611
00:36:36,400 --> 00:36:39,279
Speaker 1: we've talked about the motivations of the military, we've talked

612
00:36:39,280 --> 00:36:43,440
Speaker 1: about the technical aspects, we've talked about what the suits

613
00:36:43,520 --> 00:36:48,359
Speaker 1: and the generals want. But what about the employees, What

614
00:36:48,400 --> 00:36:51,520
Speaker 1: about the people who are actually doing the work. As

615
00:36:51,560 --> 00:36:53,640
Speaker 1: we know, often those are the people who are the most.

616
00:36:53,480 --> 00:36:56,000
Speaker 2: Ignored, a lot of them bucked right, yes, and we'll

617
00:36:56,040 --> 00:36:57,960
Speaker 2: learn about that right after a quick break. Oo.

618
00:37:04,560 --> 00:37:07,919
Speaker 1: So to bring the focus on the employees, the ones

619
00:37:07,920 --> 00:37:11,200
Speaker 1: who are actually out there writing the code that you know,

620
00:37:11,280 --> 00:37:15,040
Speaker 1: many times they probably won't get credit for, we need

621
00:37:15,080 --> 00:37:19,880
Speaker 1: to look at their own internal activism. See Google and

622
00:37:19,920 --> 00:37:24,520
Speaker 1: Alphabet employees were well aware of these moves far before

623
00:37:24,600 --> 00:37:27,879
Speaker 1: the official announcement, way before twenty seventeen. Of course, they

624
00:37:27,880 --> 00:37:30,520
Speaker 1: know this stuff is happening because they're working on things

625
00:37:30,600 --> 00:37:32,320
Speaker 1: like this. Behind the scenes.

626
00:37:32,320 --> 00:37:35,400
Speaker 2: Might be compartmentalized, but they're working on things. Sure they

627
00:37:35,440 --> 00:37:37,239
Speaker 2: can you know, employees talk.

628
00:37:37,520 --> 00:37:43,160
Speaker 1: Sure. For instance, do you remember that terrible sci fi

629
00:37:43,400 --> 00:37:45,319
Speaker 1: series The Cube guys we.

630
00:37:45,280 --> 00:37:46,320
Speaker 2: Do the movie?

631
00:37:47,680 --> 00:37:48,399
Speaker 1: I think there are three?

632
00:37:48,520 --> 00:37:50,480
Speaker 4: There was one of them was Hypercube. I remember that.

633
00:37:51,040 --> 00:37:52,799
Speaker 5: I like the first one though, I quite enjoyed it.

634
00:37:52,840 --> 00:37:54,879
Speaker 5: But yeah, it's it's it's shlock, but it's fun.

635
00:37:55,160 --> 00:37:57,920
Speaker 1: But the thing about it is it's a bit of

636
00:37:58,120 --> 00:38:03,960
Speaker 1: a cautionary tale because the premise of The Cube spoiler

637
00:38:04,040 --> 00:38:07,920
Speaker 1: alert is that a bunch of different experts were paid

638
00:38:08,000 --> 00:38:12,120
Speaker 1: to research and construct specific components of what would later

639
00:38:12,200 --> 00:38:16,040
Speaker 1: become this death trap, and they didn't know that what

640
00:38:16,080 --> 00:38:18,439
Speaker 1: they were building would be used in this certain way.

641
00:38:18,520 --> 00:38:20,719
Speaker 1: They thought they were just making heat sensors, right, they

642
00:38:20,719 --> 00:38:23,880
Speaker 1: thought they were just making hatches. They thought they were

643
00:38:23,920 --> 00:38:26,280
Speaker 1: just making lasers, Yeah, laser grids.

644
00:38:27,880 --> 00:38:31,880
Speaker 5: The first scene the guy actually gets cubed by a

645
00:38:31,960 --> 00:38:36,000
Speaker 5: laser grid in a cube in a cube. That's pretty mad.

646
00:38:36,160 --> 00:38:37,880
Speaker 5: And the thing too about that is like, is it

647
00:38:37,920 --> 00:38:40,120
Speaker 5: just for funzies? Like you don't really know, Like why

648
00:38:40,160 --> 00:38:42,160
Speaker 5: why are they in there? Is this just some like

649
00:38:42,719 --> 00:38:45,560
Speaker 5: rich a hole. It's just trying to like, you know,

650
00:38:47,560 --> 00:38:48,480
Speaker 5: you don't find out.

651
00:38:48,280 --> 00:38:50,600
Speaker 2: That, yeah, you will, really, you gotta keep going.

652
00:38:50,640 --> 00:38:52,319
Speaker 4: We got a cue. How far? How far? How deep

653
00:38:52,360 --> 00:38:54,200
Speaker 4: does this rabbit hole, this rabbit cube go?

654
00:38:54,320 --> 00:38:56,200
Speaker 2: My friend, Well, somebody's going to make another one, and

655
00:38:56,200 --> 00:38:56,920
Speaker 2: we're gonna find out.

656
00:38:56,920 --> 00:38:58,160
Speaker 4: You think some of the definitive one.

657
00:38:58,280 --> 00:39:00,640
Speaker 1: Sure, it'll probably be a prequel or something that's cool.

658
00:39:00,760 --> 00:39:03,920
Speaker 1: You can also, you know, I'm sure read spoilers on

659
00:39:03,960 --> 00:39:06,880
Speaker 1: the wiki or on the online in a form somewhere.

660
00:39:07,239 --> 00:39:12,840
Speaker 1: But luckily, at least for their own ethical well being

661
00:39:12,960 --> 00:39:16,400
Speaker 1: or their ability to sleep at night, the many of

662
00:39:16,440 --> 00:39:20,040
Speaker 1: the Google and Alphabet employees were able to communicate with

663
00:39:20,120 --> 00:39:24,080
Speaker 1: each other. They were aware of the implications of what

664
00:39:24,120 --> 00:39:27,319
Speaker 1: would happen if all this stuff was assembled together like

665
00:39:27,360 --> 00:39:33,560
Speaker 1: the Avengers or Captain Planet and the Planeteers, so they

666
00:39:33,560 --> 00:39:37,320
Speaker 1: took action. First, they gathered to write an internal petition

667
00:39:37,480 --> 00:39:42,600
Speaker 1: requesting the organization pull away from this agreement with Uncle Sam. Petitions,

668
00:39:42,680 --> 00:39:45,960
Speaker 1: it should be said, are not inherently unusual at Google,

669
00:39:46,280 --> 00:39:49,600
Speaker 1: but this one was particularly important and particularly crucial to

670
00:39:49,640 --> 00:39:52,880
Speaker 1: the future of the company because first, it was created

671
00:39:52,920 --> 00:39:56,880
Speaker 1: with the knowledge that this program, this cooperation with Project

672
00:39:56,920 --> 00:40:01,839
Speaker 1: Maven is only a very small baby step in what

673
00:40:02,239 --> 00:40:07,239
Speaker 1: management hopes will be a long, elaborate staircase, a much

674
00:40:07,360 --> 00:40:11,000
Speaker 1: larger series of cooperations. You see, Maven is not a

675
00:40:11,120 --> 00:40:14,879
Speaker 1: one off thing. It's not its own it's not its

676
00:40:14,880 --> 00:40:18,680
Speaker 1: own centerpiece. It is a pilot program. It is taking

677
00:40:19,400 --> 00:40:21,839
Speaker 1: a concept and running it around the block to see

678
00:40:21,880 --> 00:40:26,280
Speaker 1: how it drives. Secondly, Google employees believe that this project

679
00:40:26,320 --> 00:40:28,759
Speaker 1: is symptomatic of what they see as a larger and

680
00:40:28,840 --> 00:40:32,520
Speaker 1: growing problem in Google. They say the company is becoming

681
00:40:32,680 --> 00:40:38,480
Speaker 1: less and less transparent across all all facets, public facing,

682
00:40:38,760 --> 00:40:45,480
Speaker 1: government facing, employee facing, user facing, and they've even shed

683
00:40:45,719 --> 00:40:48,680
Speaker 1: their old famous motto don't be evil.

684
00:40:49,640 --> 00:40:52,759
Speaker 2: Yeah, but now if you go to let's see the

685
00:40:53,560 --> 00:40:58,720
Speaker 2: Google dot com slash about our company slash our dash company,

686
00:40:58,800 --> 00:41:01,759
Speaker 2: it says this is what their mission statement is now

687
00:41:02,040 --> 00:41:05,440
Speaker 2: quote organize the world's information and make it universally accessible

688
00:41:05,560 --> 00:41:08,760
Speaker 2: and useful, and maybe be a little evil if needed.

689
00:41:08,880 --> 00:41:10,600
Speaker 2: I don't see that written anywhere where.

690
00:41:10,680 --> 00:41:13,120
Speaker 5: I think it's implied. I think getting rid of that

691
00:41:13,440 --> 00:41:16,320
Speaker 5: original mission statement and getting into these kind of deals

692
00:41:16,760 --> 00:41:18,480
Speaker 5: implies such things.

693
00:41:19,360 --> 00:41:23,399
Speaker 1: They do have a on their Code of Conduct they

694
00:41:23,440 --> 00:41:29,200
Speaker 1: made some changes that have a wonky timeline about them too,

695
00:41:30,080 --> 00:41:36,240
Speaker 1: Because there's an article on Gizmoto by Kate Conger wherein

696
00:41:36,600 --> 00:41:40,160
Speaker 1: the author finds the section of the old Code of

697
00:41:40,200 --> 00:41:42,960
Speaker 1: Conduct that was archived by the way back Machine on

698
00:41:43,120 --> 00:41:46,320
Speaker 1: and listening to the State carefully April twenty first, twenty

699
00:41:46,400 --> 00:41:51,279
Speaker 1: eighteen that still has all that don't be evil mentioned. Yeah, right,

700
00:41:51,760 --> 00:41:56,160
Speaker 1: And then the updated version that was first archived on

701
00:41:56,280 --> 00:42:01,239
Speaker 1: May fourth, twenty eighteen took out all but one of

702
00:42:01,280 --> 00:42:04,360
Speaker 1: those mentions of don't be evil, and the Code of

703
00:42:04,440 --> 00:42:07,440
Speaker 1: Conduct itself says it has not been updated since April

704
00:42:07,480 --> 00:42:10,440
Speaker 1: fifth of twenty eighteen, So the timeline doesn't match up.

705
00:42:10,600 --> 00:42:14,440
Speaker 1: User error. Possibly hmm, possibly.

706
00:42:14,239 --> 00:42:16,279
Speaker 2: Wayback machine, what you're doing over there?

707
00:42:16,400 --> 00:42:20,640
Speaker 1: Yeah? Did they get to you, wayback machine? Regardless of

708
00:42:20,840 --> 00:42:24,040
Speaker 1: you know, whether you feel that this is just a

709
00:42:24,200 --> 00:42:29,640
Speaker 1: series of unconnected dots being forced into a larger perceived pattern,

710
00:42:29,920 --> 00:42:33,120
Speaker 1: whether you think people are practicing confirmation bias, or whether

711
00:42:33,600 --> 00:42:37,879
Speaker 1: there really is something rotten there in Silicon Valley, the

712
00:42:37,920 --> 00:42:39,600
Speaker 1: fact of the matter is you can read the full

713
00:42:39,640 --> 00:42:42,759
Speaker 1: text of the petition online and find out what the

714
00:42:42,960 --> 00:42:46,839
Speaker 1: Google employees themselves thought. We have a few highlights here.

715
00:42:47,120 --> 00:42:49,960
Speaker 1: We cannot, they say in the petition outsource the moral

716
00:42:50,000 --> 00:42:54,840
Speaker 1: responsibility of our technologies to third parties. Google stated values

717
00:42:54,880 --> 00:42:58,600
Speaker 1: make this clear. Every one of our users is trusting us,

718
00:42:58,719 --> 00:43:03,400
Speaker 1: never jeopardize that. Ever, this contract referring to MAVEN puts

719
00:43:03,440 --> 00:43:07,200
Speaker 1: Google's reputation at risk and stands in direct opposition to

720
00:43:07,320 --> 00:43:11,600
Speaker 1: our core values. Building this technology to assist the US

721
00:43:11,719 --> 00:43:16,839
Speaker 1: government in military surveillance and potentially lethal outcomes is not acceptable.

722
00:43:17,239 --> 00:43:21,080
Speaker 1: Recognizing Google's moral and ethical responsibility and the threat to

723
00:43:21,120 --> 00:43:27,120
Speaker 1: Google's reputation, we request that you cancel this project immediately, draft, publicize,

724
00:43:27,160 --> 00:43:31,319
Speaker 1: and enforce a clear policy stating that neither Google nor

725
00:43:31,360 --> 00:43:35,000
Speaker 1: its contractors will ever build warfare technology.

726
00:43:35,280 --> 00:43:37,360
Speaker 5: And a writer for en Gadget actually reached out to

727
00:43:37,400 --> 00:43:41,200
Speaker 5: Google and they were provided with the following statement.

728
00:43:42,360 --> 00:43:42,600
Speaker 4: Quote.

729
00:43:42,640 --> 00:43:45,120
Speaker 5: An important part of our culture is having employees who

730
00:43:45,120 --> 00:43:47,759
Speaker 5: are actively engaged in the work that we do. We

731
00:43:47,840 --> 00:43:50,080
Speaker 5: know that there are many open questions involved in the

732
00:43:50,160 --> 00:43:52,880
Speaker 5: use of new technologies, so these conversations with employees and

733
00:43:52,880 --> 00:43:56,319
Speaker 5: outside experts are hugely important and beneficial. MAVEN is a

734
00:43:56,360 --> 00:43:59,920
Speaker 5: well publicized DoD project, and Google is working on one

735
00:44:00,160 --> 00:44:03,400
Speaker 5: part of it, specifically scope to be for non offensive

736
00:44:03,480 --> 00:44:07,600
Speaker 5: purposes and using open source object recognition software available to

737
00:44:07,719 --> 00:44:11,600
Speaker 5: any Google Cloud customer. The models are based on unclassified

738
00:44:11,640 --> 00:44:14,319
Speaker 5: data only. The technology is used to flag images for

739
00:44:14,400 --> 00:44:17,279
Speaker 5: human review and is intended to save lives and save

740
00:44:17,320 --> 00:44:19,920
Speaker 5: people from having to do highly tedious work.

741
00:44:20,320 --> 00:44:22,600
Speaker 4: I'm just gonna finish. I think it's worth it. Any

742
00:44:22,680 --> 00:44:26,640
Speaker 4: military use of machine learning naturally raises valid concerns. We're

743
00:44:26,680 --> 00:44:30,799
Speaker 4: actively engaged across the company in a comprehensive discussion of

744
00:44:30,840 --> 00:44:33,400
Speaker 4: this important topic and also with outside experts as we

745
00:44:33,440 --> 00:44:36,600
Speaker 4: continue to develop our policies around the development and use

746
00:44:36,640 --> 00:44:38,879
Speaker 4: of our machine learning technologies.

747
00:44:39,200 --> 00:44:43,520
Speaker 2: So basically, we're gonna keep making this project. Sorry, guys.

748
00:44:43,680 --> 00:44:45,600
Speaker 4: Yeah, it's cool, but it's okay.

749
00:44:45,440 --> 00:44:48,160
Speaker 1: Because we're going to have a quote unquote healthy conversation

750
00:44:48,280 --> 00:44:50,879
Speaker 1: about it. Here's what happens next. As you said, mat,

751
00:44:50,920 --> 00:44:53,360
Speaker 1: Google refuses to backwave from this project. They're still going

752
00:44:53,440 --> 00:44:56,719
Speaker 1: to go through Noel. As as we can clean from

753
00:44:56,760 --> 00:45:00,120
Speaker 1: the statement you read, they say it is going to

754
00:45:00,120 --> 00:45:04,560
Speaker 1: be nonviolent and there's no spooky stuff going on. It's

755
00:45:04,600 --> 00:45:10,600
Speaker 1: all unclassified. Over three thousand engineers signed this petition and

756
00:45:11,520 --> 00:45:16,319
Speaker 1: a dozen resigned. More may resign in the future. This

757
00:45:16,400 --> 00:45:19,680
Speaker 1: resignation wave is happening as we record this, and it's

758
00:45:19,800 --> 00:45:25,600
Speaker 1: due to a coldly fascinating moral quandary. It's this, how

759
00:45:25,680 --> 00:45:29,879
Speaker 1: responsible are these engineers for the applications of their inventions?

760
00:45:30,360 --> 00:45:33,440
Speaker 1: Is for example, the creator of the Winchester repeating rifle

761
00:45:33,520 --> 00:45:36,319
Speaker 1: guilty for the deaths caused by that weapon. His wife

762
00:45:36,400 --> 00:45:38,760
Speaker 1: thought so, and that's why she built a crazy mansion

763
00:45:38,800 --> 00:45:41,719
Speaker 1: out west, which we still, believe it or not, have

764
00:45:41,880 --> 00:45:42,480
Speaker 1: never visited.

765
00:45:42,600 --> 00:45:43,600
Speaker 4: Oh man, I really want to go.

766
00:45:43,800 --> 00:45:46,160
Speaker 1: Yeah, I think we should absolutely go. Paul, would you

767
00:45:46,200 --> 00:45:47,719
Speaker 1: go with us if we If we go to the

768
00:45:47,760 --> 00:45:50,719
Speaker 1: Winchester mansion, we got we got an ardent thumbs up

769
00:45:50,719 --> 00:45:51,600
Speaker 1: from him, they both.

770
00:45:51,920 --> 00:45:53,960
Speaker 4: Paul's got some some beefy thumbs Yeah.

771
00:45:54,000 --> 00:45:55,560
Speaker 2: You go into all the rooms first, Paul.

772
00:45:55,760 --> 00:45:57,880
Speaker 4: Yeah, it's just a test for traps.

773
00:45:58,000 --> 00:45:58,920
Speaker 2: Yeah. No.

774
00:45:59,040 --> 00:46:02,879
Speaker 5: But it also like Oppenheimer, right, he was famously had

775
00:46:03,120 --> 00:46:08,000
Speaker 5: serious qualms about the way his technology was used in

776
00:46:08,120 --> 00:46:10,240
Speaker 5: terms of being weapons of mass destruction.

777
00:46:10,400 --> 00:46:14,960
Speaker 1: And Einstein also tortured by the He didn't build a

778
00:46:14,960 --> 00:46:18,239
Speaker 1: physical thing, but he was tortured by the idea that

779
00:46:18,320 --> 00:46:22,440
Speaker 1: his own realizations had led to this. So with this

780
00:46:22,520 --> 00:46:25,919
Speaker 1: in mind, if an engineer builds software for the purpose of, say,

781
00:46:26,360 --> 00:46:31,200
Speaker 1: more efficiently delivering payloads to locations in near Earth orbit,

782
00:46:31,520 --> 00:46:35,279
Speaker 1: and that same software is later used to deliver payloads

783
00:46:35,280 --> 00:46:39,960
Speaker 1: of bombs to people that the drones owners don't care for.

784
00:46:40,520 --> 00:46:44,160
Speaker 1: Is that engineer then responsible for the ensuing damage and

785
00:46:44,400 --> 00:46:48,720
Speaker 1: or death. It's a quandary that a lot of people

786
00:46:48,840 --> 00:46:51,279
Speaker 1: are having a hard time answering to go back and

787
00:46:51,320 --> 00:46:54,840
Speaker 1: noll to something you mentioned earlier about the dilemmas of

788
00:46:55,040 --> 00:47:02,160
Speaker 1: autonomous vehicles. One of the current nie unanswerable questions, or

789
00:47:02,160 --> 00:47:04,879
Speaker 1: at least when that hasn't been answered yet, is what

790
00:47:05,560 --> 00:47:09,520
Speaker 1: happens if there is an imminent accident that's going to

791
00:47:09,560 --> 00:47:13,400
Speaker 1: occur and you're in an autonomous vehicle. Does the vehicle

792
00:47:14,280 --> 00:47:19,280
Speaker 1: swerve and hit a pedestrian? Does it prioritize the person

793
00:47:19,360 --> 00:47:22,080
Speaker 1: in the vehicle their life over the life of a

794
00:47:22,600 --> 00:47:25,120
Speaker 1: person who deserves to live just as much but happens

795
00:47:25,120 --> 00:47:28,560
Speaker 1: not to be in the vehicle, And if so, who

796
00:47:28,640 --> 00:47:31,960
Speaker 1: is responsible? Is it the person who is in the

797
00:47:32,040 --> 00:47:34,799
Speaker 1: vehicle for not taking control if they could. Is it

798
00:47:34,880 --> 00:47:37,600
Speaker 1: the company that manufactured the vehicle? Is it the engineer

799
00:47:37,640 --> 00:47:40,840
Speaker 1: who made the software that made the vehicle make that choice.

800
00:47:41,120 --> 00:47:46,000
Speaker 1: The trolley problem of the old thought exercise becomes increasingly concrete,

801
00:47:46,360 --> 00:47:51,440
Speaker 1: increasingly dangerous, and increasingly complicated, and frankly, we're not equipped

802
00:47:51,440 --> 00:47:52,520
Speaker 1: to answer it at this point.

803
00:47:52,680 --> 00:47:56,839
Speaker 5: No, it's it's it's it's an utter conundrum because what

804
00:47:57,160 --> 00:47:59,799
Speaker 5: you know to be that programmer who gets to make

805
00:47:59,800 --> 00:48:02,480
Speaker 5: that decision, you're essentially playing god, aren't you.

806
00:48:02,560 --> 00:48:05,800
Speaker 4: It's like what life is more valuable? You know? Is

807
00:48:05,840 --> 00:48:08,520
Speaker 4: it our customer or is it the pedestrian.

808
00:48:08,920 --> 00:48:11,320
Speaker 5: There's really no way of properly answering that without making

809
00:48:11,320 --> 00:48:15,080
Speaker 5: a serious judgment, call a hard and fast choice.

810
00:48:14,880 --> 00:48:17,480
Speaker 1: Right, Yeah, you could bake in something where it's just

811
00:48:17,560 --> 00:48:22,320
Speaker 1: an absolute calculation of the number of lives. So, for instance,

812
00:48:22,360 --> 00:48:26,080
Speaker 1: in that case, if there is a crowd of five

813
00:48:26,200 --> 00:48:31,120
Speaker 1: people who are gathered for their Paul Decant fan club meeting,

814
00:48:31,360 --> 00:48:34,799
Speaker 1: which inexplicably takes place on the sidewalk by a busy road.

815
00:48:35,360 --> 00:48:37,880
Speaker 1: If those five people are gathered for that meeting and

816
00:48:37,920 --> 00:48:41,440
Speaker 1: there's only one person in the car, does the car say, Okay,

817
00:48:41,600 --> 00:48:45,120
Speaker 1: we'll just pop the airbags and hope this one person survives,

818
00:48:45,160 --> 00:48:47,000
Speaker 1: but we're not going to kill five people for one

819
00:48:47,560 --> 00:48:50,360
Speaker 1: And then what if there are two people in that

820
00:48:50,440 --> 00:48:52,640
Speaker 1: car and one of them is pregnant. And what if

821
00:48:52,680 --> 00:48:55,839
Speaker 1: the five people on the sidewalk celebrating with the Paul

822
00:48:55,880 --> 00:48:58,880
Speaker 1: Decant fan club, what if they're all in their seventies

823
00:48:58,960 --> 00:49:02,799
Speaker 1: right and pass report of age generally speaking? You know

824
00:49:02,840 --> 00:49:05,120
Speaker 1: what I mean? This this is the kind of stuff

825
00:49:05,200 --> 00:49:08,000
Speaker 1: that we have not yet, as far as we know,

826
00:49:09,360 --> 00:49:10,520
Speaker 1: learned to program for.

827
00:49:10,800 --> 00:49:13,080
Speaker 5: Then before you know it, you've got you know, power

828
00:49:13,160 --> 00:49:16,520
Speaker 5: hungry cars mowing down old ladies because they've you know,

829
00:49:16,600 --> 00:49:17,840
Speaker 5: they've they've lived their life.

830
00:49:18,200 --> 00:49:19,000
Speaker 4: They're expendable.

831
00:49:19,120 --> 00:49:21,360
Speaker 1: But like, what what about a school bus? Does a

832
00:49:21,400 --> 00:49:25,239
Speaker 1: school bus, based on the number of kids it has,

833
00:49:25,480 --> 00:49:28,720
Speaker 1: does that have just immunity to always prioritize the children

834
00:49:28,800 --> 00:49:32,279
Speaker 1: on the bus. It's very politically difficult to argue against.

835
00:49:31,920 --> 00:49:35,480
Speaker 2: That, oh man man.

836
00:49:35,360 --> 00:49:36,480
Speaker 1: Matt leaned back for that.

837
00:49:36,560 --> 00:49:39,480
Speaker 2: Oof, Yeah, because of what happens when a public bus

838
00:49:39,520 --> 00:49:43,000
Speaker 2: and then a school bus are both operating and there's

839
00:49:43,000 --> 00:49:44,480
Speaker 2: a crash a minute.

840
00:49:44,200 --> 00:49:46,080
Speaker 1: I don't know, because I'm marta. Most of the buses

841
00:49:46,080 --> 00:49:50,480
Speaker 1: are empty, Oh shade, It's true, It's true.

842
00:49:50,560 --> 00:49:58,400
Speaker 5: What about the street cars? The street car even emptier empties.

843
00:49:57,360 --> 00:49:59,240
Speaker 1: There's there's not even a driver.

844
00:49:59,520 --> 00:50:02,000
Speaker 2: Yeah. At this point, the Marta buses are just hopping

845
00:50:02,080 --> 00:50:02,760
Speaker 2: on the trolleys.

846
00:50:02,760 --> 00:50:05,320
Speaker 4: This is some serious Atlanta inside baseball.

847
00:50:05,400 --> 00:50:10,000
Speaker 1: It is true. Learn about our public infrastructure debacles. We're

848
00:50:10,040 --> 00:50:11,880
Speaker 1: working on it. We're doing our best things.

849
00:50:11,960 --> 00:50:15,319
Speaker 4: It's gonna be called it's gonna be rebranded to the atl.

850
00:50:15,160 --> 00:50:16,560
Speaker 2: WHOA, that's cool.

851
00:50:16,920 --> 00:50:20,799
Speaker 1: Yeah, nothing weird about that, nothing forced about that. Right.

852
00:50:21,080 --> 00:50:25,080
Speaker 1: So Google is unfazed. They're not only going to continue

853
00:50:25,080 --> 00:50:27,440
Speaker 1: this project, which we should also say is not super

854
00:50:27,480 --> 00:50:31,080
Speaker 1: super big. It's publicly described as being worth a minimum

855
00:50:31,120 --> 00:50:33,840
Speaker 1: of nine million dollars. That's a lot of money to

856
00:50:34,120 --> 00:50:37,120
Speaker 1: normal people. That's not a lot of money to Google.

857
00:50:38,320 --> 00:50:42,560
Speaker 1: But they are vying for additional projects within this military space,

858
00:50:42,600 --> 00:50:46,480
Speaker 1: and this is not unprecedented. IBM has a long standing

859
00:50:46,520 --> 00:50:50,720
Speaker 1: relationship with the US military on a number of fronts

860
00:50:50,960 --> 00:50:53,760
Speaker 1: and also cough cough, Nazi Germany cough cough.

861
00:50:54,160 --> 00:50:56,600
Speaker 4: Yes, did you guys know the Hugo boss designed the

862
00:50:56,680 --> 00:50:57,480
Speaker 4: Nazi uniform?

863
00:50:57,640 --> 00:50:57,839
Speaker 1: Yes?

864
00:50:57,960 --> 00:51:00,480
Speaker 4: I did not know that until yesterday. Shock.

865
00:51:01,560 --> 00:51:03,799
Speaker 5: What a great pr team they must have to still

866
00:51:03,840 --> 00:51:05,239
Speaker 5: be around after that.

867
00:51:05,400 --> 00:51:08,680
Speaker 1: Seriously, well, I mean there were a lot of German

868
00:51:08,760 --> 00:51:09,600
Speaker 1: companies that.

869
00:51:09,560 --> 00:51:12,760
Speaker 4: Were for sure Volkswagen, even I think was interesting.

870
00:51:13,560 --> 00:51:17,000
Speaker 1: On an additional note, IBM itself may be drawing an

871
00:51:17,000 --> 00:51:21,200
Speaker 1: ethical line. They have signaled that they will not create

872
00:51:21,239 --> 00:51:24,440
Speaker 1: intelligent drones for warfare, and a blog post they had

873
00:51:24,440 --> 00:51:28,200
Speaker 1: recently IBM said they're committed to the ethical and responsible

874
00:51:28,239 --> 00:51:32,400
Speaker 1: advancement of AI technology and the AI or machine consciousness

875
00:51:32,520 --> 00:51:35,759
Speaker 1: should be used to augment human decision making rather than

876
00:51:35,840 --> 00:51:40,480
Speaker 1: replace it. So that's similar to the argument that automated

877
00:51:41,280 --> 00:51:44,640
Speaker 1: vehicle systems for consumers should be doing a lot of

878
00:51:44,760 --> 00:51:47,760
Speaker 1: assisted driving, but there should always be a human behind

879
00:51:47,800 --> 00:51:48,240
Speaker 1: the wheel.

880
00:51:48,840 --> 00:51:51,360
Speaker 2: That's probably a good call International Business Machines.

881
00:51:52,719 --> 00:51:57,200
Speaker 1: That's true. Oh, that's true. And of course shout out

882
00:51:57,239 --> 00:51:59,640
Speaker 1: to everybody who listened to our other episodes on the

883
00:51:59,680 --> 00:52:06,200
Speaker 1: evil of machine consciousness, And personally I prefer that term

884
00:52:06,239 --> 00:52:09,040
Speaker 1: to AI because what why is it artificial?

885
00:52:09,440 --> 00:52:09,680
Speaker 2: Right?

886
00:52:09,719 --> 00:52:10,920
Speaker 1: Why are we so special?

887
00:52:11,719 --> 00:52:13,919
Speaker 2: Yeah, And at this point I'm just going to urge

888
00:52:13,960 --> 00:52:16,279
Speaker 2: everybody to go through and read that open letter that

889
00:52:16,400 --> 00:52:19,480
Speaker 2: was written by the International Committee for Robot Arms Control.

890
00:52:19,840 --> 00:52:23,919
Speaker 2: There are some pretty terrifying things that they go into there,

891
00:52:23,960 --> 00:52:28,480
Speaker 2: specifically about the slippery slope that Project Maven represents. So

892
00:52:28,680 --> 00:52:30,719
Speaker 2: I would go and read that. You can find it

893
00:52:31,800 --> 00:52:33,719
Speaker 2: Robot Arms Control Open Letter.

894
00:52:34,280 --> 00:52:37,319
Speaker 1: I when you first showed this to me, Matt, I

895
00:52:37,480 --> 00:52:41,560
Speaker 1: was cartoonishly tickled because I originally read the title as

896
00:52:41,920 --> 00:52:44,240
Speaker 1: the International Committee for Robot Arms.

897
00:52:44,480 --> 00:52:44,760
Speaker 2: Yeah.

898
00:52:44,960 --> 00:52:50,439
Speaker 1: Yeah, and I didn't read the control part. At the end, Man,

899
00:52:50,480 --> 00:52:53,919
Speaker 1: it was a long day. So what do they say, though,

900
00:52:53,920 --> 00:52:56,680
Speaker 1: what's one thing that really stood out to you in

901
00:52:56,719 --> 00:52:57,160
Speaker 1: this letter?

902
00:52:57,560 --> 00:53:00,200
Speaker 2: Well, here's the quote. We are just a short up

903
00:53:00,239 --> 00:53:04,840
Speaker 2: away from authorizing autonomous drones to kill automatically without human

904
00:53:04,880 --> 00:53:09,719
Speaker 2: supervision or meaningful human control. Tight that alone. Just you know,

905
00:53:09,800 --> 00:53:14,480
Speaker 2: if you arm these weapons, these flying autonomous weapons, with

906
00:53:14,560 --> 00:53:17,840
Speaker 2: the ability to make those decisions, yeah, that's a target

907
00:53:18,040 --> 00:53:20,399
Speaker 2: that's not a target, that is definitely a target kill.

908
00:53:20,719 --> 00:53:22,719
Speaker 1: What happens if they get hacked as well.

909
00:53:23,880 --> 00:53:28,440
Speaker 2: Or if they just are just operating someone loses control

910
00:53:28,480 --> 00:53:30,080
Speaker 2: somehow just however.

911
00:53:30,000 --> 00:53:34,520
Speaker 1: Or if they're operating independently somewhere and they have some

912
00:53:34,560 --> 00:53:37,560
Speaker 1: sort of cognitive leap, and they say, you know, the

913
00:53:37,600 --> 00:53:40,759
Speaker 1: best way to achieve the objective is to get rid

914
00:53:40,800 --> 00:53:44,160
Speaker 1: of everyone. That would almost never happen. We're talking the

915
00:53:44,160 --> 00:53:46,359
Speaker 1: odds of winning the lottery eight times in a row.

916
00:53:46,400 --> 00:53:48,760
Speaker 4: We're talking about like a terminator scenario, right.

917
00:53:48,640 --> 00:53:53,200
Speaker 1: Some Skynet stuff for now that is still relegated to

918
00:53:53,280 --> 00:53:55,320
Speaker 1: the realm of dystopian science fiction.

919
00:53:56,080 --> 00:53:58,120
Speaker 2: But they do make another really great point. I just

920
00:53:58,120 --> 00:54:01,560
Speaker 2: want to say here. They are deeply concerned about the

921
00:54:01,560 --> 00:54:05,560
Speaker 2: possible integration of Google's data, the data that they have

922
00:54:05,680 --> 00:54:09,719
Speaker 2: on everyone on you, probably massive data set, integrating that

923
00:54:10,320 --> 00:54:15,880
Speaker 2: with this military surveillance data and combining that and applying

924
00:54:15,880 --> 00:54:17,480
Speaker 2: it to this targeted killing.

925
00:54:17,239 --> 00:54:18,760
Speaker 4: Notion so you can cross reference.

926
00:54:19,440 --> 00:54:21,960
Speaker 2: Yeah, and this is a you know, these aren't just

927
00:54:22,000 --> 00:54:24,200
Speaker 2: some Joe Schmoe's getting together and saying, hey, we need

928
00:54:24,239 --> 00:54:27,960
Speaker 2: to write this letter, you guys. These are some serious

929
00:54:28,560 --> 00:54:32,640
Speaker 2: engineers and academics and people who are working in these

930
00:54:32,680 --> 00:54:35,640
Speaker 2: fields saying like raising a flag and saying no, no, no, no, no,

931
00:54:35,680 --> 00:54:36,560
Speaker 2: we got to watch this.

932
00:54:36,840 --> 00:54:41,680
Speaker 1: We know where this could go. Yeah, and now we

933
00:54:42,480 --> 00:54:47,080
Speaker 1: draw this episode to a close, and we don't have

934
00:54:47,960 --> 00:54:53,640
Speaker 1: a solid answer or prediction. We're right here with you, folks,

935
00:54:53,800 --> 00:54:58,160
Speaker 1: unless you are working directly for Project Mavin, right, we

936
00:54:58,320 --> 00:55:02,360
Speaker 1: don't know. No one knows where exactly we are going

937
00:55:02,920 --> 00:55:05,480
Speaker 1: to paraphrase Jim Wilder and Willie walk up.

938
00:55:05,560 --> 00:55:08,760
Speaker 5: Unless we forget to mention that Google is also obviously

939
00:55:08,760 --> 00:55:12,759
Speaker 5: doing some super innovative, fascinating things that are quite good

940
00:55:12,800 --> 00:55:17,760
Speaker 5: for humanity. Potentially, they're using machine learning to develop testing

941
00:55:17,840 --> 00:55:23,319
Speaker 5: software that can actually find different complications related to diabetes

942
00:55:23,880 --> 00:55:29,040
Speaker 5: and also early signs of breast cancer, and the FDA,

943
00:55:29,160 --> 00:55:34,040
Speaker 5: according to this article from Wired, is already in early

944
00:55:34,080 --> 00:55:39,600
Speaker 5: stages of approving AI software that can help doctors make

945
00:55:39,960 --> 00:55:43,160
Speaker 5: very important life or death medical decisions. And this is

946
00:55:43,160 --> 00:55:45,560
Speaker 5: from an article called Google's new AI head is so

947
00:55:45,600 --> 00:55:48,799
Speaker 5: smart he doesn't need AI. About this guy, Jeff Dean,

948
00:55:49,200 --> 00:55:52,040
Speaker 5: who is a big part of a lot of Google's

949
00:55:52,040 --> 00:55:53,160
Speaker 5: AI innovations.

950
00:55:53,480 --> 00:55:57,000
Speaker 1: That's a great point because we can't we cannot forget

951
00:55:57,239 --> 00:56:02,520
Speaker 1: how large and varied Google and Alphabetter's organizations are. There's

952
00:56:02,560 --> 00:56:05,120
Speaker 1: Google dot org as well, where you can learn about

953
00:56:05,400 --> 00:56:09,440
Speaker 1: how they're using data to uncover racial injustice or building

954
00:56:09,520 --> 00:56:14,120
Speaker 1: open source platforms to translate books for disadvantaged kids. It's

955
00:56:14,160 --> 00:56:16,480
Speaker 1: absolutely right. I really appreciate bringing it up because they're

956
00:56:16,520 --> 00:56:21,919
Speaker 1: not just this solely evil thing, right, And sometimes those

957
00:56:21,960 --> 00:56:26,040
Speaker 1: aims of these these large departments within Google may even

958
00:56:26,080 --> 00:56:27,120
Speaker 1: contradict one another.

959
00:56:27,360 --> 00:56:29,640
Speaker 5: And you got to think, too, like an argument, maybe

960
00:56:29,680 --> 00:56:33,279
Speaker 5: that a big high mucky mucket Google might make about

961
00:56:33,320 --> 00:56:35,920
Speaker 5: getting involved in something like this is like, well, if

962
00:56:35,960 --> 00:56:39,320
Speaker 5: it's not us, you know, and we're obviously going to

963
00:56:39,360 --> 00:56:43,360
Speaker 5: handle it with thoughtfulness and care. If it's not us,

964
00:56:43,400 --> 00:56:45,759
Speaker 5: it's going to be somebody else who might do a

965
00:56:45,880 --> 00:56:49,160
Speaker 5: less good job, you know, and not think about the ramifications.

966
00:56:49,320 --> 00:56:51,680
Speaker 5: So you know, even that statement that we read earlier,

967
00:56:52,000 --> 00:56:53,480
Speaker 5: it was pretty measured.

968
00:56:53,640 --> 00:56:55,040
Speaker 4: I thought it was actually not bad.

969
00:56:55,200 --> 00:56:57,040
Speaker 1: Yeah, just kill people correctly, right.

970
00:56:57,120 --> 00:56:59,160
Speaker 5: Well, okay, you got me there, Ben, I guess what

971
00:56:59,200 --> 00:57:01,440
Speaker 5: I mean is though it did sound like pure pr

972
00:57:01,760 --> 00:57:03,520
Speaker 5: it was like, we get it.

973
00:57:03,160 --> 00:57:04,120
Speaker 4: It's a thing.

974
00:57:04,360 --> 00:57:07,840
Speaker 5: We understand that there is there are always consequences associated

975
00:57:07,880 --> 00:57:09,799
Speaker 5: with this kind of stuff, but we're aware of them

976
00:57:10,080 --> 00:57:13,400
Speaker 5: and we feel that we're equipped to help mitigate some

977
00:57:13,480 --> 00:57:15,560
Speaker 5: of that. But it's against the badger're out of the

978
00:57:15,560 --> 00:57:18,160
Speaker 5: bag scenario, right, It's not up to you anymore at

979
00:57:18,160 --> 00:57:18,520
Speaker 5: that point.

980
00:57:20,040 --> 00:57:23,240
Speaker 1: And if you would like to learn more about Project

981
00:57:23,320 --> 00:57:26,840
Speaker 1: Maven specifically, please check out tech Stuff. They have a

982
00:57:26,880 --> 00:57:30,720
Speaker 1: podcast that just came out on this program. It's hosted

983
00:57:30,760 --> 00:57:36,920
Speaker 1: by our longtime friends sometimes Nemesis and Complaint Department Jonathan Strickland.

984
00:57:36,960 --> 00:57:40,160
Speaker 1: Available twenty four to seven for any issues or criticisms

985
00:57:40,240 --> 00:57:41,560
Speaker 1: you have of stuff they don't want you to know,

986
00:57:41,640 --> 00:57:43,520
Speaker 1: you can reach of me is Jonathan dot Strickland to

987
00:57:43,600 --> 00:57:44,680
Speaker 1: HowStuffWorks dot com.

988
00:57:44,720 --> 00:57:47,439
Speaker 5: He just launched a live chat kind of situation too,

989
00:57:47,480 --> 00:57:50,120
Speaker 5: so we can suss out your issues in real time.

990
00:57:50,280 --> 00:57:53,560
Speaker 2: Yeah, yeah, yeah, like a drone, exactly like a drone.

991
00:57:53,960 --> 00:57:55,920
Speaker 2: It'll especially help you out on Twitch if he's ever

992
00:57:56,000 --> 00:57:56,320
Speaker 2: on there.

993
00:57:57,000 --> 00:58:00,840
Speaker 1: No, let's say Google does somehow create a policy banning

994
00:58:00,920 --> 00:58:05,160
Speaker 1: any and all military partnerships. He let's go a little

995
00:58:05,200 --> 00:58:08,640
Speaker 1: bit further and say that all US tech companies follow

996
00:58:08,680 --> 00:58:10,439
Speaker 1: their lead and do the same thing, because there are

997
00:58:10,640 --> 00:58:14,600
Speaker 1: industry wide calls for everyone not to play this particular obo.

998
00:58:15,080 --> 00:58:17,640
Speaker 1: I mean, let's even go further and say that all

999
00:58:17,680 --> 00:58:20,320
Speaker 1: tech companies in the world refuse to help the US

1000
00:58:20,400 --> 00:58:23,600
Speaker 1: build self aware weapons of war. You know who will

1001
00:58:23,600 --> 00:58:28,840
Speaker 1: continue this research? All caps, every single country that can

1002
00:58:28,920 --> 00:58:32,120
Speaker 1: afford it, every single private company that does not have

1003
00:58:32,200 --> 00:58:39,000
Speaker 1: the same moral quandaries about hypothetical scenarios, every single individual

1004
00:58:39,280 --> 00:58:42,320
Speaker 1: that does not have those kind of quandaries. It is

1005
00:58:42,840 --> 00:58:45,280
Speaker 1: again going to happen, and it.

1006
00:58:45,240 --> 00:58:46,600
Speaker 4: Gets worse the AI race.

1007
00:58:47,640 --> 00:58:49,600
Speaker 1: Yeah, it is. That's a good way to put it

1008
00:58:49,960 --> 00:58:54,120
Speaker 1: and to paraphrase Billy Mays, but wait, it gets worse.

1009
00:58:54,560 --> 00:58:57,800
Speaker 1: We'd like to end today's episode on an ethical quandary

1010
00:58:58,760 --> 00:59:03,640
Speaker 1: question that you could feel free to suss out yourself.

1011
00:59:03,880 --> 00:59:07,680
Speaker 1: Have a couple of couple of beers over, if you drink, meditate,

1012
00:59:07,760 --> 00:59:09,800
Speaker 1: if you do that, whatever you do to get yourself

1013
00:59:09,840 --> 00:59:15,880
Speaker 1: in a thoughtful headspace, and let us know what are

1014
00:59:15,880 --> 00:59:20,400
Speaker 1: the implications here. What sort of machine consciousness are we creating?

1015
00:59:21,080 --> 00:59:25,000
Speaker 1: Are we making something that will be inherently belligerent? And

1016
00:59:25,080 --> 00:59:28,680
Speaker 1: if so, If the world's first machine consciousness is built

1017
00:59:28,720 --> 00:59:33,520
Speaker 1: to kill, how will this influence is later increasingly self

1018
00:59:33,600 --> 00:59:39,400
Speaker 1: directed actions, its thoughts, it's recognizations, it's emotions, or whatever

1019
00:59:39,440 --> 00:59:47,479
Speaker 1: those equivalents are. Imagine whether they're imagine if you would

1020
00:59:47,520 --> 00:59:50,640
Speaker 1: that there were two different versions of machine consciousness. One

1021
00:59:50,760 --> 00:59:58,520
Speaker 1: maybe is built to optimize agricultural projects in a transforming ecosystem,

1022
00:59:58,880 --> 01:00:02,280
Speaker 1: and the other is built to finds people and annihilate

1023
01:00:02,320 --> 01:00:09,080
Speaker 1: them or buildings. What happens next? Where do they go?

1024
01:00:09,560 --> 01:00:12,960
Speaker 1: Do these things like the early drones? Do they increasingly

1025
01:00:13,200 --> 01:00:19,960
Speaker 1: speciate and become more attenuated. There's the argument that we

1026
01:00:20,080 --> 01:00:24,800
Speaker 1: have with biological entities, with human minds, and that is that, yes,

1027
01:00:25,200 --> 01:00:28,880
Speaker 1: people change, but they don't change in the way you think.

1028
01:00:28,960 --> 01:00:32,880
Speaker 1: Over time, we tend to become more concentrated versions of

1029
01:00:32,920 --> 01:00:37,880
Speaker 1: who we were in the beginning. Now one could say

1030
01:00:37,920 --> 01:00:42,360
Speaker 1: that this is not something to be too disturbed by,

1031
01:00:42,520 --> 01:00:47,880
Speaker 1: because a human engineer could drop in and maybe incept

1032
01:00:48,080 --> 01:00:51,400
Speaker 1: this machine mind with a different, different line of code

1033
01:00:51,560 --> 01:00:54,800
Speaker 1: that would alter its behavior. And maybe now it says

1034
01:00:54,840 --> 01:00:58,760
Speaker 1: I'm tired of killing right, or I have new programming.

1035
01:01:00,440 --> 01:01:03,840
Speaker 1: But we have to consider that if we are as

1036
01:01:03,880 --> 01:01:07,880
Speaker 1: a species creating minds that may one day have their

1037
01:01:07,880 --> 01:01:12,000
Speaker 1: own agency as much as as a human being, maybe

1038
01:01:12,000 --> 01:01:14,880
Speaker 1: more eventually, maybe in our lifetimes, they will have more

1039
01:01:15,200 --> 01:01:18,160
Speaker 1: agency and potential because they won't have the same biological,

1040
01:01:18,480 --> 01:01:22,840
Speaker 1: hardwired limits that human beings have. What kind of minds

1041
01:01:22,840 --> 01:01:25,160
Speaker 1: are we building? And why?

1042
01:01:25,800 --> 01:01:28,440
Speaker 2: You can find us on Twitter and Facebook where we're

1043
01:01:28,520 --> 01:01:33,760
Speaker 2: conspiracy stuff, conspiracy stuff show on on uh Instagram that's

1044
01:01:33,800 --> 01:01:36,640
Speaker 2: the where that one is. You can find our podcast

1045
01:01:36,720 --> 01:01:38,320
Speaker 2: and everything else that's stuff they don't want you to

1046
01:01:38,360 --> 01:01:41,560
Speaker 2: know dot com. UH tell us what you think of,

1047
01:01:41,600 --> 01:01:43,480
Speaker 2: like what kind of minds are we creating? Like Ben

1048
01:01:43,680 --> 01:01:47,000
Speaker 2: speaking to what's happened Heady stuff heady mind. Yes, this

1049
01:01:47,160 --> 01:01:51,040
Speaker 2: is uh terrifying and I need to use the restroom

1050
01:01:51,040 --> 01:01:53,800
Speaker 2: now because I am the p has been done scared

1051
01:01:53,880 --> 01:01:54,360
Speaker 2: right out of me.

1052
01:01:54,920 --> 01:01:57,440
Speaker 4: We better, we better take care of that right now.

1053
01:01:57,480 --> 01:01:59,760
Speaker 5: Oh we need As we are not AI, we still

1054
01:01:59,800 --> 01:02:02,320
Speaker 5: have to, you know, take care of our bodily needs.

1055
01:02:02,600 --> 01:02:03,560
Speaker 4: But in the meantime, if you.

1056
01:02:03,560 --> 01:02:06,640
Speaker 2: Would like it, and that's the end of this classic episode.

1057
01:02:06,680 --> 01:02:10,520
Speaker 2: If you have any thoughts or questions about this episode,

1058
01:02:10,880 --> 01:02:12,919
Speaker 2: you can get into contact with us in a number

1059
01:02:12,960 --> 01:02:14,960
Speaker 2: of different ways. One of the best is to give

1060
01:02:15,040 --> 01:02:17,560
Speaker 2: us a call. Our number is one eight three three

1061
01:02:17,880 --> 01:02:21,120
Speaker 2: st d WYTK. If you don't want to do that,

1062
01:02:21,400 --> 01:02:23,400
Speaker 2: you can send us a good old fashioned email.

1063
01:02:23,680 --> 01:02:27,840
Speaker 1: We are conspiracy at iHeartRadio dot com.

1064
01:02:27,960 --> 01:02:30,040
Speaker 2: Stuff they Don't want you to Know is a production

1065
01:02:30,160 --> 01:02:34,680
Speaker 2: of iHeartRadio. For more podcasts from iHeartRadio, visit the iHeartRadio app,

1066
01:02:34,760 --> 01:02:37,640
Speaker 2: Apple Podcasts, or wherever you listen to your favorite shows.