1
00:00:00,240 --> 00:00:04,680
Speaker 1: From UFOs to psychic powers and government conspiracies. History is

2
00:00:04,760 --> 00:00:09,080
Speaker 1: riddled with unexplained events. You can turn back now or

3
00:00:09,200 --> 00:00:24,880
Speaker 1: learn the stuff they don't want you to know. Welcome

4
00:00:24,880 --> 00:00:27,760
Speaker 1: back to the show. My name is Matt. They called

5
00:00:27,840 --> 00:00:30,920
Speaker 1: me Ben. We are joined with our super producer Paul

6
00:00:31,000 --> 00:00:35,440
Speaker 1: big Apple Decand most importantly, you are you. You are here,

7
00:00:35,560 --> 00:00:39,519
Speaker 1: and that makes this stuff they don't want you to know.

8
00:00:39,720 --> 00:00:43,839
Speaker 1: Today we're going to talk about a company called Google,

9
00:00:44,040 --> 00:00:48,080
Speaker 1: commonly known as Google. That's its street name. That's such

10
00:00:48,120 --> 00:00:51,560
Speaker 1: a popular name, that's become a verb in American English. Uh.

11
00:00:51,600 --> 00:00:54,040
Speaker 1: They have a parent company and could call it the

12
00:00:54,080 --> 00:00:58,240
Speaker 1: real company, which is known as Alphabet. And we've talked

13
00:00:58,240 --> 00:01:00,840
Speaker 1: about this a little bit in the past us, but

14
00:01:01,320 --> 00:01:06,119
Speaker 1: we want to whatever we mentioned Google established that from

15
00:01:06,120 --> 00:01:10,839
Speaker 1: the jump at the top, because it's strange how relatively

16
00:01:10,880 --> 00:01:16,920
Speaker 1: obscure that fact remains. Now. While Google, Alphabet, whatever you

17
00:01:16,959 --> 00:01:19,560
Speaker 1: wanna call it, may not be the largest company in

18
00:01:19,600 --> 00:01:24,119
Speaker 1: the world in terms of capital or territory or material possessions,

19
00:01:24,160 --> 00:01:29,200
Speaker 1: it is undoubtedly one of the most influential organizations on

20
00:01:29,240 --> 00:01:32,600
Speaker 1: the planet, at least in terms of its connection to

21
00:01:32,720 --> 00:01:38,400
Speaker 1: human beings or the ways in which it influences networks

22
00:01:38,480 --> 00:01:42,200
Speaker 1: or organizations or webs of human beings. Most people on

23
00:01:42,240 --> 00:01:47,600
Speaker 1: this planet, literally most people think about that have heard

24
00:01:47,680 --> 00:01:51,400
Speaker 1: of Google, even if they have somehow not directly interacted

25
00:01:51,480 --> 00:01:55,720
Speaker 1: with it, and arguably, on a slightly more controversial note,

26
00:01:56,080 --> 00:02:00,000
Speaker 1: most people have in some way to some degree benefits

27
00:02:00,000 --> 00:02:03,240
Speaker 1: acted by the actions of the U. S. Military, whether

28
00:02:03,280 --> 00:02:06,720
Speaker 1: that's something that happened generations ago in their own country

29
00:02:06,800 --> 00:02:09,560
Speaker 1: or an adjacent country, or to an ancestor, or whether

30
00:02:09,600 --> 00:02:13,240
Speaker 1: something is happening to them now as we record this episode,

31
00:02:13,360 --> 00:02:17,760
Speaker 1: and so today begins with a question spoiler alert, it

32
00:02:17,800 --> 00:02:20,960
Speaker 1: also ends with a question. And our beginning question is

33
00:02:21,080 --> 00:02:26,280
Speaker 1: this what happens when these groups Google Alphabet whatever you

34
00:02:26,280 --> 00:02:29,960
Speaker 1: wanna call it, and the U. S. Military or Uncle Sam.

35
00:02:30,040 --> 00:02:35,520
Speaker 1: What happens when they join forces? Because they have so

36
00:02:36,320 --> 00:02:40,640
Speaker 1: they have vultrons up, and they have vultron up in

37
00:02:40,800 --> 00:02:49,359
Speaker 1: a again controversial way, it involves unmanned aerial systems also

38
00:02:49,440 --> 00:02:52,400
Speaker 1: known as drones. You're the ones that like deliver your

39
00:02:52,400 --> 00:02:55,840
Speaker 1: Amazon packages, right, yeah, and some variation of them. Yeah. Uh.

40
00:02:55,919 --> 00:02:59,200
Speaker 1: They're commonly called drones, of course, but unmanned aerial vehicles

41
00:02:59,280 --> 00:03:03,760
Speaker 1: or u a V's have been a long running trope

42
00:03:03,919 --> 00:03:08,239
Speaker 1: in science fiction, and surprisingly for many people, some version

43
00:03:08,280 --> 00:03:12,799
Speaker 1: of this technology existed long before the modern day. There

44
00:03:12,919 --> 00:03:17,280
Speaker 1: wasn't a history rattling moment when the first drone ever

45
00:03:17,440 --> 00:03:20,880
Speaker 1: went airborne. And that's in large part because it's very

46
00:03:20,919 --> 00:03:26,120
Speaker 1: difficult for us to isolate exactly when the early version

47
00:03:26,280 --> 00:03:30,799
Speaker 1: of an invention ends and the first modern versions begin. Matt,

48
00:03:31,080 --> 00:03:34,320
Speaker 1: when you and I worked on the invention show Stuff

49
00:03:34,320 --> 00:03:37,400
Speaker 1: of Genius, we ran into this pretty often. Somebody comes

50
00:03:37,480 --> 00:03:40,840
Speaker 1: up with a rough approximation of an idea, and then

51
00:03:40,920 --> 00:03:44,200
Speaker 1: over a course of years, decades, or even centuries in

52
00:03:44,280 --> 00:03:49,480
Speaker 1: some cases, people continually make small improvements until it becomes

53
00:03:49,520 --> 00:03:54,240
Speaker 1: the thing we know. Most inventions are not a Eureka

54
00:03:54,320 --> 00:03:59,040
Speaker 1: moment in reality. I mean, they may be a Eureka

55
00:03:59,080 --> 00:04:02,400
Speaker 1: moment in terms of the idea or the epiphany, but

56
00:04:03,160 --> 00:04:06,040
Speaker 1: most inventions in reality, and when we get to the

57
00:04:06,080 --> 00:04:11,240
Speaker 1: hardware phases, are going to be um created by people

58
00:04:11,280 --> 00:04:14,440
Speaker 1: standing on the shoulders of giants. Yeah, it ends up

59
00:04:14,440 --> 00:04:17,080
Speaker 1: being the person who makes it marketable rather than just

60
00:04:17,200 --> 00:04:21,720
Speaker 1: creates it. Good call and a very sad and capitalist call. Right,

61
00:04:21,760 --> 00:04:26,280
Speaker 1: It's the person who popularizes it. So if we look

62
00:04:26,320 --> 00:04:30,840
Speaker 1: at the first modern version of a drone, an unmanned vehicle.

63
00:04:30,920 --> 00:04:33,000
Speaker 1: Would it be one of those bombs in the eighteen

64
00:04:33,080 --> 00:04:35,840
Speaker 1: hundreds that was strapped to a balloon and then just

65
00:04:35,880 --> 00:04:38,440
Speaker 1: sort of put off like something bad will happen, We

66
00:04:38,560 --> 00:04:41,400
Speaker 1: can't steer it. There wasn't a person on that and

67
00:04:41,440 --> 00:04:43,440
Speaker 1: it was being used as a weapon. Or was it

68
00:04:43,480 --> 00:04:47,400
Speaker 1: the initial V one rockets that Germany deployed during World

69
00:04:47,440 --> 00:04:51,280
Speaker 1: War Two? In the early nineteen hundreds, military groups did

70
00:04:51,400 --> 00:04:54,440
Speaker 1: use drones. They had radio controlled versions that were just

71
00:04:54,520 --> 00:04:59,800
Speaker 1: for target practice. Engineers developed unmanned aircraft that again had

72
00:04:59,839 --> 00:05:04,039
Speaker 1: me missions of some sort. These weren't drones the way

73
00:05:04,040 --> 00:05:08,200
Speaker 1: we think of them in the current age, but there's

74
00:05:08,600 --> 00:05:11,640
Speaker 1: you know, there are other examples, like cruise missiles. These

75
00:05:11,680 --> 00:05:15,159
Speaker 1: things were the first cruise missiles, often called flying torpedoes.

76
00:05:15,200 --> 00:05:18,080
Speaker 1: They were not meant to return to base, and drones

77
00:05:18,160 --> 00:05:22,480
Speaker 1: are meant to function like very very smart, sometimes very

78
00:05:22,560 --> 00:05:26,080
Speaker 1: very deadly boomerangs. They come out, they come back and

79
00:05:26,320 --> 00:05:32,599
Speaker 1: somebody claps yeah, generally not the people being targeted. Generally, yes, yes,

80
00:05:32,680 --> 00:05:36,200
Speaker 1: generally not. You are correct, my friend. During the Cold

81
00:05:36,200 --> 00:05:39,839
Speaker 1: War and the Vietnam War, or the series of conflicts

82
00:05:39,880 --> 00:05:42,200
Speaker 1: collectively known as the Vietnam War. Here in the US,

83
00:05:42,640 --> 00:05:46,480
Speaker 1: Uncle Sam ramped up research on drones. It was no

84
00:05:46,520 --> 00:05:50,159
Speaker 1: longer just an interesting idea posed by a few eccentric

85
00:05:50,200 --> 00:05:53,000
Speaker 1: scientists and engineers. In fact, by the time of the

86
00:05:53,080 --> 00:05:57,560
Speaker 1: Vietnam War, unmanned drones flew thousands of high risk reconnaissance

87
00:05:57,600 --> 00:06:01,800
Speaker 1: missions and they got shot down all the time. But

88
00:06:02,360 --> 00:06:05,160
Speaker 1: in that process they saved the lives of human pilots

89
00:06:05,160 --> 00:06:08,720
Speaker 1: who otherwise would have been a board. And I thought

90
00:06:08,760 --> 00:06:10,919
Speaker 1: about I thought about you, Matt when I was looking

91
00:06:10,920 --> 00:06:14,960
Speaker 1: into this, because another precedent for drones, which won't go

92
00:06:15,000 --> 00:06:17,920
Speaker 1: into in this episode, but another precedent for drones were

93
00:06:17,960 --> 00:06:22,440
Speaker 1: weaponized bats. Did you ever hear about this, the one

94
00:06:22,520 --> 00:06:25,359
Speaker 1: to drop the little bombs? Yeah, well, yes, that's a

95
00:06:25,480 --> 00:06:28,520
Speaker 1: version of it. They said, bats can fly, we have

96
00:06:28,640 --> 00:06:30,800
Speaker 1: access to them. Well yeah, I mean you could go

97
00:06:30,839 --> 00:06:33,839
Speaker 1: back to carrier pigeons as being a type of drone

98
00:06:33,880 --> 00:06:36,839
Speaker 1: as well. Yeah, I mean, there's not a person aboard.

99
00:06:37,440 --> 00:06:39,400
Speaker 1: There's a really cool episode of the podcast of the

100
00:06:39,400 --> 00:06:43,200
Speaker 1: Memory Palace about the bats. It's called Itty Bitty Bombs.

101
00:06:43,040 --> 00:06:46,360
Speaker 1: It's it's quite good. Yeah, So check that one out

102
00:06:46,560 --> 00:06:51,120
Speaker 1: for more information around the same time Vietnam War. It

103
00:06:51,120 --> 00:06:53,640
Speaker 1: goes this far back. Around the same time of Vietnam War,

104
00:06:53,880 --> 00:06:59,080
Speaker 1: engineers started adding real time surveillance ability to drones, so cameras.

105
00:06:59,360 --> 00:07:04,800
Speaker 1: Right today we look at I guess the closest thing

106
00:07:04,800 --> 00:07:07,360
Speaker 1: we have to a military version of the first drone

107
00:07:07,400 --> 00:07:11,760
Speaker 1: would be the m Q one Predator is first unveiled

108
00:07:11,920 --> 00:07:15,720
Speaker 1: in which is weird when you think about it, because

109
00:07:15,720 --> 00:07:18,560
Speaker 1: that was such a long time ago. And in two

110
00:07:18,560 --> 00:07:21,680
Speaker 1: thousand two, the CIA used the m Q one Predator

111
00:07:21,720 --> 00:07:24,440
Speaker 1: to make its first successful kill of a quote unquote

112
00:07:24,880 --> 00:07:29,360
Speaker 1: enemy combatant in Afghanistan. Since then, since just two thousand

113
00:07:29,360 --> 00:07:33,040
Speaker 1: and two, this technology is grown by cartoonish lye extreme

114
00:07:33,160 --> 00:07:36,640
Speaker 1: leaps and bounds, and it's currently on the bleeding edge

115
00:07:36,640 --> 00:07:39,520
Speaker 1: of scientific advancement. Yeah. And if that m Q one

116
00:07:39,600 --> 00:07:42,040
Speaker 1: Predator drone, that's the drone that you have seen before,

117
00:07:42,120 --> 00:07:44,440
Speaker 1: that's the one that is in pictures all over the place.

118
00:07:44,680 --> 00:07:46,960
Speaker 1: That is the I mean, I guess, being one of

119
00:07:46,960 --> 00:07:49,080
Speaker 1: the first, it just became iconic in that way. I

120
00:07:49,200 --> 00:07:50,800
Speaker 1: just gonna want to throw that in there. Yeah, it

121
00:07:50,800 --> 00:07:55,960
Speaker 1: looks like a windowless commercial plane with an upside down

122
00:07:56,320 --> 00:08:00,360
Speaker 1: tail fin right to upside downtail fins. Yeah, it's of

123
00:08:00,480 --> 00:08:04,120
Speaker 1: the shape is slightly slightly odd, but yeah, it's it's

124
00:08:04,120 --> 00:08:08,520
Speaker 1: pretty cool. And as these drones become more and more

125
00:08:08,600 --> 00:08:11,760
Speaker 1: common in theaters of war or in commercial industries like

126
00:08:11,840 --> 00:08:15,480
Speaker 1: Amazon using them, or recreational use like buy a you

127
00:08:15,600 --> 00:08:18,920
Speaker 1: buy a loved when one is a gift, they're also

128
00:08:20,160 --> 00:08:23,600
Speaker 1: doing the inorganic version of specie eating. You know. Specie

129
00:08:23,600 --> 00:08:28,800
Speaker 1: eating is when a relatively common ancestor, through generations and generations,

130
00:08:28,840 --> 00:08:36,319
Speaker 1: produces uh different species of the same template. Right, So

131
00:08:36,640 --> 00:08:40,160
Speaker 1: these drones are becoming increasingly specialized. They have varying ranges

132
00:08:40,200 --> 00:08:43,199
Speaker 1: of abilities and and odds are you've seen one, right,

133
00:08:43,280 --> 00:08:45,880
Speaker 1: We've all seen toy drones, a little quad copter. Guys,

134
00:08:47,000 --> 00:08:49,600
Speaker 1: you gotta wonder too, like how does that work patent wise?

135
00:08:49,600 --> 00:08:51,640
Speaker 1: Like did someone come up with that design and it

136
00:08:51,720 --> 00:08:55,679
Speaker 1: was just generic enough to be copied? Like I'm I

137
00:08:55,679 --> 00:08:57,880
Speaker 1: always wonder about things like that, like the fidget spinner,

138
00:08:58,040 --> 00:08:59,480
Speaker 1: Like did that guy just do a bad job of

139
00:08:59,480 --> 00:09:02,400
Speaker 1: patent the original fidget spinner? Or is it so generic

140
00:09:02,480 --> 00:09:04,640
Speaker 1: that that doesn't even apply. I'm wondering that that's the

141
00:09:04,679 --> 00:09:07,280
Speaker 1: same thing about the design of those quad copter toy

142
00:09:07,360 --> 00:09:10,959
Speaker 1: drones entirely speculative, but it's probably at the very least

143
00:09:10,960 --> 00:09:15,120
Speaker 1: a series of patents. They have to multiple things in play.

144
00:09:15,160 --> 00:09:17,640
Speaker 1: I would imagine pro tips speaking of it play. If

145
00:09:17,640 --> 00:09:19,080
Speaker 1: you're going to get one of those for your kids,

146
00:09:19,120 --> 00:09:21,000
Speaker 1: get a good one because the cheap ones, you might

147
00:09:21,000 --> 00:09:23,520
Speaker 1: think you got a deal. They just don't fly. They

148
00:09:23,559 --> 00:09:28,480
Speaker 1: suck and money and their uh their battery vampires too.

149
00:09:28,559 --> 00:09:30,000
Speaker 1: Oh my god, that was the thing that we got

150
00:09:30,000 --> 00:09:32,640
Speaker 1: one and like, oh, it flies for about four minutes

151
00:09:33,000 --> 00:09:36,560
Speaker 1: four minutes, Like that's not fun tops tops. Yeah, but

152
00:09:36,600 --> 00:09:40,319
Speaker 1: what kind of batteries do the big boys use? Are

153
00:09:40,320 --> 00:09:44,640
Speaker 1: they fueled or they do? You know that's classified. Yeah, sorry,

154
00:09:44,920 --> 00:09:46,640
Speaker 1: we can't talk about You can't tell me. No, I

155
00:09:46,640 --> 00:09:53,360
Speaker 1: thought we were pals. Google it all right, Yeah, so

156
00:09:53,840 --> 00:09:56,360
Speaker 1: some of the on a serious notes, some of those

157
00:09:56,400 --> 00:10:02,080
Speaker 1: power systems would vary. Uh by the time we dive

158
00:10:02,120 --> 00:10:05,800
Speaker 1: into one and this episode comes out, it may have changed. Interesting,

159
00:10:05,880 --> 00:10:10,240
Speaker 1: that's how quickly they're moving on these. So we've seen these, right,

160
00:10:10,280 --> 00:10:15,760
Speaker 1: and we've probably seen photographs or videos of the larger

161
00:10:15,800 --> 00:10:18,960
Speaker 1: cousins of these tiny drones, things like the Predator, Like

162
00:10:19,160 --> 00:10:21,960
Speaker 1: Matt just described, you can. You can find a photograph

163
00:10:22,040 --> 00:10:25,839
Speaker 1: of it pretty easily. The differences are stark and astonishing,

164
00:10:26,240 --> 00:10:29,960
Speaker 1: and missiles are not the biggest difference. The fact that

165
00:10:30,000 --> 00:10:34,640
Speaker 1: they're weaponized. While that is incredibly dangerous and possibly fatal,

166
00:10:34,800 --> 00:10:37,200
Speaker 1: that is not the biggest difference. Some of the bigger

167
00:10:37,240 --> 00:10:39,920
Speaker 1: differences are going to be on the inside of the drone.

168
00:10:39,960 --> 00:10:44,319
Speaker 1: We're talking things like GPS, real time satellite feeds, super

169
00:10:44,360 --> 00:10:48,720
Speaker 1: complex microprocessors, and more. They're putting brains on board of

170
00:10:48,760 --> 00:10:52,760
Speaker 1: these things, and perhaps most disturbingly, some drones are edging

171
00:10:52,840 --> 00:10:56,600
Speaker 1: closer and closer and closer to what's commonly known as

172
00:10:56,720 --> 00:11:02,840
Speaker 1: artificial intelligence or or technically known as machine consciousness. And

173
00:11:03,280 --> 00:11:07,920
Speaker 1: we're close to an old science fiction question becoming a

174
00:11:07,960 --> 00:11:11,680
Speaker 1: real question. Well, the first true artificial intelligence, the first

175
00:11:11,679 --> 00:11:20,360
Speaker 1: true machine consciousness, be a weapon of war. Yes, highly likely, definitely. Yeah,

176
00:11:20,640 --> 00:11:23,960
Speaker 1: why is it? Why is that met? That's because innovation

177
00:11:25,000 --> 00:11:28,720
Speaker 1: through the military is one of the ways that innovations occur.

178
00:11:28,840 --> 00:11:32,120
Speaker 1: The biggest innovations and inventions occur because there's funding from

179
00:11:32,120 --> 00:11:35,600
Speaker 1: a military somewhere, and it's not like it's the military

180
00:11:35,640 --> 00:11:39,959
Speaker 1: itself doing these things. It is usually a private company

181
00:11:40,120 --> 00:11:43,400
Speaker 1: or someone who's being given a grant, a government grant

182
00:11:43,480 --> 00:11:45,000
Speaker 1: to do it. We see this time and time again

183
00:11:45,000 --> 00:11:47,040
Speaker 1: with many of the stories that we've talked about, where

184
00:11:47,120 --> 00:11:49,880
Speaker 1: the military is also always like decades ahead of like

185
00:11:49,960 --> 00:11:52,080
Speaker 1: what consumers will ever get their hands on, and it

186
00:11:52,160 --> 00:11:55,640
Speaker 1: might be ten twenty years before it trickles down into

187
00:11:55,760 --> 00:11:58,160
Speaker 1: the public where maybe we can have a you know,

188
00:11:58,280 --> 00:12:02,000
Speaker 1: terabyte thumb drive, just as a silly example, where they've

189
00:12:02,040 --> 00:12:04,839
Speaker 1: had that technology probably for a decade at this point,

190
00:12:04,920 --> 00:12:07,520
Speaker 1: I would say the only way for it not for

191
00:12:07,559 --> 00:12:10,200
Speaker 1: the first machine consciousness, true machine consciousness and to not

192
00:12:10,280 --> 00:12:13,000
Speaker 1: be a weapon, is for every private company on the

193
00:12:13,000 --> 00:12:16,199
Speaker 1: planet to say I'm not going to do that, right, Yeah,

194
00:12:16,400 --> 00:12:20,120
Speaker 1: for us all kumbaya and do some hands across America

195
00:12:20,559 --> 00:12:23,920
Speaker 1: on the text fere right or hands across the world. So,

196
00:12:25,040 --> 00:12:28,480
Speaker 1: with with this in mind, we're also hitting on a

197
00:12:28,600 --> 00:12:33,040
Speaker 1: larger phenomenon, this kind of kind of unfortunate. We're we're

198
00:12:33,040 --> 00:12:35,840
Speaker 1: a family show, but we also endeavor at our best

199
00:12:35,880 --> 00:12:39,080
Speaker 1: to be an honest one. Uh. The great leaps and

200
00:12:39,160 --> 00:12:43,959
Speaker 1: technology are not almost always, but the majority of the time,

201
00:12:44,000 --> 00:12:47,080
Speaker 1: the great leaps and technology are driven by human vices,

202
00:12:47,400 --> 00:12:50,439
Speaker 1: a desire to win a war, a desire to have

203
00:12:50,960 --> 00:12:55,760
Speaker 1: more you know, social attention or more financial access than

204
00:12:55,960 --> 00:12:59,080
Speaker 1: a rival, or how do we compress all of this

205
00:12:59,240 --> 00:13:01,800
Speaker 1: porn that's the other one. That's the third one, to

206
00:13:01,920 --> 00:13:05,440
Speaker 1: desire for lust because in a very very real and

207
00:13:05,600 --> 00:13:10,560
Speaker 1: concrete and provable way, pornography is the reason that you

208
00:13:10,720 --> 00:13:14,200
Speaker 1: or your parents had a VHS instead of a beta max. Uh,

209
00:13:14,240 --> 00:13:18,280
Speaker 1: it's the reason. Like we talked about this, I think previously. Yeah,

210
00:13:18,640 --> 00:13:23,839
Speaker 1: so often we all too often we as a species

211
00:13:23,880 --> 00:13:32,240
Speaker 1: are driven by uh, non noble motivations. But that doesn't

212
00:13:32,280 --> 00:13:35,600
Speaker 1: mean that we're not smart. We are a jar of

213
00:13:35,880 --> 00:13:40,400
Speaker 1: very very smart, very belligerent cookies. And we've been putting

214
00:13:40,559 --> 00:13:44,720
Speaker 1: some of the smartest cookies of our species into the

215
00:13:44,960 --> 00:13:50,079
Speaker 1: search for artificial intelligence and machine learning, a search that

216
00:13:50,120 --> 00:13:53,680
Speaker 1: continues as we record this episode. So just a little

217
00:13:53,679 --> 00:13:57,240
Speaker 1: while ago in July, there were these two researchers named

218
00:13:57,280 --> 00:14:01,559
Speaker 1: Greg Allen and Taniel Chan, and the published a study

219
00:14:01,840 --> 00:14:06,240
Speaker 1: which was called Artificial Intelligence and National Security, and in

220
00:14:06,240 --> 00:14:11,440
Speaker 1: this they put forward goals for developing law policies specifically

221
00:14:11,559 --> 00:14:15,600
Speaker 1: that would um that would put ideas forth and how

222
00:14:15,600 --> 00:14:18,920
Speaker 1: do you deal with AI when it comes to national security?

223
00:14:18,960 --> 00:14:22,840
Speaker 1: Military spending, like what should we as a species do

224
00:14:23,040 --> 00:14:25,640
Speaker 1: when it comes to these two things colliding, And here

225
00:14:25,680 --> 00:14:28,920
Speaker 1: are some of their findings. They again said, the most

226
00:14:28,920 --> 00:14:31,600
Speaker 1: transformative innovations in the field of AI are coming from

227
00:14:31,600 --> 00:14:34,840
Speaker 1: the private sector and academic world, not from the government

228
00:14:34,840 --> 00:14:39,360
Speaker 1: in military. So currently, right now, AI is something that's

229
00:14:39,400 --> 00:14:42,680
Speaker 1: happening outside of the realm of the military, which in

230
00:14:42,840 --> 00:14:47,760
Speaker 1: July seventeen, Yeah, that's a good thing, maybe because you know,

231
00:14:47,800 --> 00:14:51,200
Speaker 1: there could be some nefarious private sector humans as well,

232
00:14:51,440 --> 00:14:54,080
Speaker 1: but that's that's point A. The second one is that

233
00:14:54,120 --> 00:14:58,160
Speaker 1: the current capabilities of AI have quote significant potential for

234
00:14:58,280 --> 00:15:01,960
Speaker 1: national security, especially in the areas of cyber defense and

235
00:15:02,160 --> 00:15:06,040
Speaker 1: satellite imagery analysis. And then the last thing I'm just

236
00:15:06,120 --> 00:15:08,720
Speaker 1: gonna read this whole thing is a quote future progress

237
00:15:08,760 --> 00:15:11,840
Speaker 1: in AI has the potential to be a transformative national

238
00:15:11,920 --> 00:15:16,560
Speaker 1: security technology on par with nuclear weapons, aircraft, computers, and biotechs.

239
00:15:16,600 --> 00:15:22,280
Speaker 1: So some of the most influential technologies on on weaponry,

240
00:15:22,560 --> 00:15:25,680
Speaker 1: on war. Uh, they're saying AI is going to be

241
00:15:25,720 --> 00:15:30,280
Speaker 1: the next big thing. And they're not just talking the talk, right, No,

242
00:15:30,440 --> 00:15:34,200
Speaker 1: they are not just talking the talk at all. Um.

243
00:15:34,240 --> 00:15:37,760
Speaker 1: According to the Wall Street Journal, the Defense Department spent

244
00:15:37,920 --> 00:15:42,640
Speaker 1: seven point four billion dollars on artificial intelligence related areas

245
00:15:42,680 --> 00:15:48,320
Speaker 1: in seven point four billion dollars on AI. And you know,

246
00:15:48,440 --> 00:15:51,960
Speaker 1: not to not to harp too much on Eisenhower, but

247
00:15:52,000 --> 00:15:53,840
Speaker 1: I believe he's the one who had a quote that

248
00:15:53,960 --> 00:15:58,640
Speaker 1: said every every rocket the US builds represents this many

249
00:15:58,760 --> 00:16:02,800
Speaker 1: schools that we could have built but didn't. And there

250
00:16:03,120 --> 00:16:05,280
Speaker 1: are a couple of things if we're being entirely fair

251
00:16:05,320 --> 00:16:08,960
Speaker 1: about that number that we have to acknowledge. First is

252
00:16:09,000 --> 00:16:12,040
Speaker 1: that this is only the money we know of, So

253
00:16:12,160 --> 00:16:16,680
Speaker 1: seven point four billion, let's call that an at least number, right.

254
00:16:17,320 --> 00:16:22,000
Speaker 1: And then on the other end artificial intelligence related areas,

255
00:16:22,440 --> 00:16:26,520
Speaker 1: that doesn't mean they're just building a silicon brain, right,

256
00:16:26,600 --> 00:16:29,560
Speaker 1: that there are many other things that we might not

257
00:16:30,000 --> 00:16:35,680
Speaker 1: associate directly with machine intelligence that maybe function is support

258
00:16:35,840 --> 00:16:39,960
Speaker 1: or infrastructure. So this is all, This is the whole meal.

259
00:16:40,240 --> 00:16:45,120
Speaker 1: This isn't just a hamburger. Yeah. Well, and we agreed,

260
00:16:45,240 --> 00:16:47,880
Speaker 1: and we should especially point out that the Department of

261
00:16:47,960 --> 00:16:52,600
Speaker 1: Defense budget for for fiscal year seventeen was five hundred

262
00:16:52,600 --> 00:16:56,560
Speaker 1: and forty billion dollars around that number, still the biggest

263
00:16:56,600 --> 00:17:00,000
Speaker 1: in the world, and that's publicly acknowledged. Again, I can't

264
00:17:00,200 --> 00:17:05,719
Speaker 1: keep saying that enough publicly acknowledged. That's that's what they

265
00:17:05,720 --> 00:17:09,359
Speaker 1: ask Congress for. Yes, it's true. Before I derail us,

266
00:17:09,600 --> 00:17:11,560
Speaker 1: what he thinks, should we pause for a word from

267
00:17:11,560 --> 00:17:21,840
Speaker 1: our sponsor? It seems wise. Here's where it gets crazy.

268
00:17:23,359 --> 00:17:26,440
Speaker 1: We've painted a little bit of the picture, right. We

269
00:17:26,440 --> 00:17:29,879
Speaker 1: we know the story now, fellow conspiracy realist. It is

270
00:17:30,000 --> 00:17:33,200
Speaker 1: a story of drones. It is a story of private industry.

271
00:17:33,240 --> 00:17:36,560
Speaker 1: It is a story of government interaction, a story of war,

272
00:17:36,720 --> 00:17:40,639
Speaker 1: and a story of secrets. But in this age of

273
00:17:40,760 --> 00:17:45,960
Speaker 1: instantaneous information, we know more than we would have known,

274
00:17:46,240 --> 00:17:50,679
Speaker 1: say even two decades ago. We know more about what's

275
00:17:50,720 --> 00:17:57,000
Speaker 1: happening now. Similar to the drones acquiring real time surveillance technology,

276
00:17:57,680 --> 00:18:00,359
Speaker 1: we are able to be more aware of surround deeds

277
00:18:00,359 --> 00:18:02,879
Speaker 1: in a distant land. And that's how we know about

278
00:18:02,960 --> 00:18:08,439
Speaker 1: something called, uh, what's it called? Man? It's the Algorithmic

279
00:18:08,480 --> 00:18:14,119
Speaker 1: Warfare Cross Functional Team, also known as Project Maven. And

280
00:18:14,160 --> 00:18:17,920
Speaker 1: this was announced on April in a and internal memo

281
00:18:18,600 --> 00:18:23,960
Speaker 1: from the Deputy Secretary of Defense Robert Bob o Work.

282
00:18:24,440 --> 00:18:30,200
Speaker 1: Yeah he goes by Bob Work, But I was just yeah,

283
00:18:30,280 --> 00:18:33,760
Speaker 1: bobba work, Um so I guess we can just read

284
00:18:33,800 --> 00:18:35,680
Speaker 1: parts of this memo if you want to. It's kind

285
00:18:35,720 --> 00:18:38,000
Speaker 1: of hard to digest, that's the only thing. So maybe

286
00:18:38,000 --> 00:18:42,600
Speaker 1: if we walk through parts down humanize it. Yes, let's

287
00:18:42,640 --> 00:18:45,680
Speaker 1: let's do that. Let's do that. Here we go. I'll

288
00:18:45,680 --> 00:18:48,880
Speaker 1: be I'll be the dry the dry part. Here, Here

289
00:18:48,920 --> 00:18:51,399
Speaker 1: we go. The a w c f t S first

290
00:18:51,560 --> 00:18:56,760
Speaker 1: task is to field technology to augment or automate processing, exploitation,

291
00:18:56,840 --> 00:19:02,680
Speaker 1: and dissemination for tactical unmanned aerial systems. Okay, okay, so

292
00:19:04,080 --> 00:19:08,520
Speaker 1: good god, let's keep going a little bit. So, so

293
00:19:09,320 --> 00:19:13,320
Speaker 1: it's gonna augment or automate stuff for U A v's

294
00:19:13,560 --> 00:19:18,600
Speaker 1: or drones and mid altitude full motion video in support

295
00:19:18,760 --> 00:19:22,639
Speaker 1: of the Defeat ISIS campaign. And this will help to

296
00:19:22,680 --> 00:19:26,600
Speaker 1: reduce the human factor's burden of full motion video analysis,

297
00:19:27,440 --> 00:19:32,399
Speaker 1: increase actionable intelligence, and enhance military decision making. So what

298
00:19:32,520 --> 00:19:36,640
Speaker 1: that all means is the p E D Processing, exploitation

299
00:19:36,760 --> 00:19:41,639
Speaker 1: and dissemination means that there hoping again to get a

300
00:19:41,720 --> 00:19:46,119
Speaker 1: brain on board from your local neighborhood predator or whatever

301
00:19:46,240 --> 00:19:48,600
Speaker 1: drone they end up using by the time you are

302
00:19:48,680 --> 00:19:52,920
Speaker 1: listening to this episode, and the processing part is going

303
00:19:52,960 --> 00:19:56,919
Speaker 1: to be the analysis, which often would be a human

304
00:19:56,960 --> 00:20:01,040
Speaker 1: part of the equation. The exploitation would be the ability

305
00:20:01,080 --> 00:20:05,680
Speaker 1: to take action on that analysis, right to use that knowledge.

306
00:20:05,680 --> 00:20:11,640
Speaker 1: And dissemination would be sharing the information, determining what information

307
00:20:11,920 --> 00:20:16,280
Speaker 1: is relevant, what is irrelevant, and what to send where

308
00:20:16,560 --> 00:20:20,439
Speaker 1: that's p E D. And that's the full motion video

309
00:20:20,720 --> 00:20:25,520
Speaker 1: is similar to the same thing because it's it's a

310
00:20:25,640 --> 00:20:29,679
Speaker 1: little bit easier to take a bunch of numbers that

311
00:20:29,720 --> 00:20:32,680
Speaker 1: are sens a bunch of temperature variables, for instance, and

312
00:20:32,800 --> 00:20:36,720
Speaker 1: tell a computer or an automated system that whatever is

313
00:20:36,760 --> 00:20:40,840
Speaker 1: within this range, do action X. Whatever is outside of

314
00:20:40,880 --> 00:20:43,679
Speaker 1: this range, do action. Why, you know, even if the

315
00:20:43,680 --> 00:20:47,920
Speaker 1: action is just alert the human user, the person that

316
00:20:47,920 --> 00:20:51,000
Speaker 1: that's going to actually analyze the thing. Yeah, absolutely absolutely,

317
00:20:51,320 --> 00:20:55,720
Speaker 1: And that's tougher with video, right because there are many

318
00:20:55,800 --> 00:20:59,920
Speaker 1: many more factors of play. Would you say that sounds correct? Absolutely?

319
00:21:00,720 --> 00:21:03,040
Speaker 1: It just depends on what kind of sensors they're using

320
00:21:03,040 --> 00:21:04,800
Speaker 1: on the drones. If it's you know, is it a

321
00:21:04,880 --> 00:21:08,480
Speaker 1: heat based imaging system, is it just a you know,

322
00:21:08,920 --> 00:21:11,119
Speaker 1: I mean, what what what are you looking at? Essentially,

323
00:21:11,119 --> 00:21:13,640
Speaker 1: what is the data that you're looking at? And that's

324
00:21:13,640 --> 00:21:15,800
Speaker 1: really hard to tell. Because it's gonna change depending on

325
00:21:15,840 --> 00:21:19,680
Speaker 1: the model of drone and what they're you know, deploying. Uh.

326
00:21:19,760 --> 00:21:22,320
Speaker 1: The really interesting one of the really interesting things here

327
00:21:23,080 --> 00:21:26,320
Speaker 1: in my mind is that, you know, we're talking about

328
00:21:26,320 --> 00:21:29,160
Speaker 1: a single drone and having an onboard system, but they're

329
00:21:29,200 --> 00:21:31,600
Speaker 1: also talking about having a system that can analyze the

330
00:21:31,600 --> 00:21:34,359
Speaker 1: footage from all the drones that they have deployed in

331
00:21:34,400 --> 00:21:38,639
Speaker 1: a single space and then be have the machine be

332
00:21:38,680 --> 00:21:41,760
Speaker 1: able to take all that data in and tell you

333
00:21:41,800 --> 00:21:45,199
Speaker 1: what's going on, and then send that whatever they whatever

334
00:21:45,240 --> 00:21:48,720
Speaker 1: the algorithm decides needs to happen, disseminate all of that

335
00:21:48,800 --> 00:21:52,040
Speaker 1: information out to the entire team that's also there. Um,

336
00:21:52,119 --> 00:21:54,159
Speaker 1: so you really have that. One of the things we

337
00:21:54,240 --> 00:21:57,399
Speaker 1: talked about recently in another episode, a full scope picture

338
00:21:57,480 --> 00:22:01,040
Speaker 1: of the battlefield essentially, so every one at all times

339
00:22:01,160 --> 00:22:03,919
Speaker 1: sees exactly what's happening on the battlefield as close to

340
00:22:03,960 --> 00:22:08,480
Speaker 1: an omniscient observer as possible as well. We used to

341
00:22:08,480 --> 00:22:12,080
Speaker 1: say humanly possible, but that's no longer the case, right, Yeah, exactly.

342
00:22:12,320 --> 00:22:14,400
Speaker 1: So let's keep going on here a little bit with

343
00:22:14,440 --> 00:22:18,440
Speaker 1: what the memo said, Um, this program will initially provide

344
00:22:18,600 --> 00:22:23,800
Speaker 1: computer vision algorithms for object detection, classification, and alerts for

345
00:22:24,080 --> 00:22:27,320
Speaker 1: full motion video p e D. Further sprints will incorporate

346
00:22:27,359 --> 00:22:31,399
Speaker 1: more advanced computer vision technology. After successful sprints in support

347
00:22:31,400 --> 00:22:36,880
Speaker 1: of intelligence surveillance reconnaissance UH, the this program will prioritize

348
00:22:36,920 --> 00:22:42,000
Speaker 1: the integration of similar technologies into other defense intelligence mission areas.

349
00:22:42,040 --> 00:22:45,879
Speaker 1: So it's not just for this single drone operation that

350
00:22:45,920 --> 00:22:48,520
Speaker 1: they're gonna deploy it to. It's going to be for

351
00:22:48,600 --> 00:22:52,400
Speaker 1: bigger things. It goes on to say, this program will

352
00:22:52,400 --> 00:22:57,960
Speaker 1: also consolidate existing algorithm based technology initiatives related to mission

353
00:22:58,000 --> 00:23:03,000
Speaker 1: areas of the Defense Intelligence Enterprise or DIE, including all

354
00:23:03,040 --> 00:23:08,520
Speaker 1: initiatives that develop employee or field artificial intelligence, automation, machine learning,

355
00:23:08,560 --> 00:23:12,439
Speaker 1: deep learning, and computer vision algorithms. So really they're talking

356
00:23:12,480 --> 00:23:18,560
Speaker 1: about everything in the future for for military applications, so

357
00:23:18,600 --> 00:23:21,760
Speaker 1: beyond far beyond drones. Oh yes, these could also be

358
00:23:21,920 --> 00:23:26,119
Speaker 1: submarines for instance, These could be satellites, these could be uh,

359
00:23:26,160 --> 00:23:31,000
Speaker 1: space shuttles. It could be surveillance, just video camera surveillance

360
00:23:31,359 --> 00:23:33,840
Speaker 1: that's just on the ground even right, you know, just

361
00:23:33,880 --> 00:23:37,840
Speaker 1: the kind of stuff where eventually you could program a

362
00:23:37,920 --> 00:23:42,440
Speaker 1: drone or some kind of killing machine with parameters um

363
00:23:42,720 --> 00:23:45,760
Speaker 1: that are like legally sanctioned, and you can say, okay,

364
00:23:45,800 --> 00:23:49,280
Speaker 1: search and destroy X target or this type of target.

365
00:23:49,480 --> 00:23:52,080
Speaker 1: That's the fear. The potential is there, The potential is

366
00:23:52,119 --> 00:23:56,600
Speaker 1: there absolutely, So essentially, what they're saying here is that

367
00:23:57,000 --> 00:24:00,480
Speaker 1: DARPA and the Pentagon have already poured tons of money

368
00:24:00,520 --> 00:24:03,560
Speaker 1: into this automated collection of data. And they're not alone.

369
00:24:03,800 --> 00:24:07,040
Speaker 1: Other members of the Alphabet Soup Gang, which is a

370
00:24:07,160 --> 00:24:09,600
Speaker 1: very endearing name for that, like the Apple Dumpling Gang, right,

371
00:24:09,760 --> 00:24:12,920
Speaker 1: the Alphabet Soup Gang, as other members of course in

372
00:24:13,119 --> 00:24:17,639
Speaker 1: essay and so on, they have also put forth tons

373
00:24:17,760 --> 00:24:21,520
Speaker 1: of money and time into automating the collection of data.

374
00:24:21,640 --> 00:24:24,160
Speaker 1: And the next step is discovered is discovering a means

375
00:24:24,160 --> 00:24:27,720
Speaker 1: of automating the analysis of data. And the step after that,

376
00:24:28,280 --> 00:24:32,920
Speaker 1: the especially spooky one, is enabling these unmanned machines to

377
00:24:33,240 --> 00:24:36,760
Speaker 1: immediately act in real time on the results of their

378
00:24:36,800 --> 00:24:41,080
Speaker 1: onboard analyzes. So to the question postional, the answer would

379
00:24:41,119 --> 00:24:45,840
Speaker 1: be yes, they want to have at least the capability

380
00:24:46,320 --> 00:24:51,320
Speaker 1: to make a judge, jury, and executioner on an unmanned vehicle.

381
00:24:51,480 --> 00:24:53,600
Speaker 1: This doesn't mean they'll do it. A lot of tech

382
00:24:53,680 --> 00:24:57,200
Speaker 1: companies do have problems with it or at least with

383
00:24:57,680 --> 00:25:02,479
Speaker 1: actualizing that pretend sholl Yeah, we even see problems with

384
00:25:02,560 --> 00:25:06,560
Speaker 1: the self driving cars. I mean, as as much as

385
00:25:06,640 --> 00:25:09,000
Speaker 1: much R and d as have gone into those, they're

386
00:25:09,080 --> 00:25:12,399
Speaker 1: certainly not infallible and they get into fender benders and

387
00:25:12,440 --> 00:25:17,280
Speaker 1: make mistakes. Absolutely. Let's let's just have one more little

388
00:25:17,359 --> 00:25:20,080
Speaker 1: quote here, and this is from the chief in charge

389
00:25:20,160 --> 00:25:23,840
Speaker 1: of this algorithmic warfare cross function team or project may

390
00:25:23,840 --> 00:25:29,080
Speaker 1: even and he's speaking specifically to what they hope to

391
00:25:29,119 --> 00:25:32,840
Speaker 1: achieve in the near term with this project um and

392
00:25:32,920 --> 00:25:36,000
Speaker 1: he says, quote, people in computers will work symbiotically to

393
00:25:36,080 --> 00:25:40,720
Speaker 1: increase the ability of weapons systems to detect objects. Eventually,

394
00:25:40,720 --> 00:25:42,720
Speaker 1: we hope that one analyst will be able to do

395
00:25:42,760 --> 00:25:45,639
Speaker 1: twice as much work, potentially three times as much as

396
00:25:45,680 --> 00:25:49,160
Speaker 1: they're doing. Now. That's our goal. So in the short term,

397
00:25:49,280 --> 00:25:51,480
Speaker 1: somebody you know, sat down with him, had a meeting,

398
00:25:51,600 --> 00:25:55,920
Speaker 1: and we want to increase just basically what our analysts

399
00:25:55,960 --> 00:25:58,760
Speaker 1: can get done, and we're gonna save money on that end.

400
00:25:58,800 --> 00:26:01,000
Speaker 1: We're gonna save time and going to be more efficient.

401
00:26:01,320 --> 00:26:04,560
Speaker 1: That's what this whole project is about, at least in

402
00:26:04,680 --> 00:26:08,679
Speaker 1: his eyes or in his his pr The thing that

403
00:26:08,760 --> 00:26:10,879
Speaker 1: he got to say. So what we're talking about with

404
00:26:10,880 --> 00:26:13,080
Speaker 1: this project so far is just what it means to

405
00:26:13,119 --> 00:26:15,600
Speaker 1: achieve what it wants to achieve. We haven't discussed how

406
00:26:15,640 --> 00:26:19,040
Speaker 1: it was going to achieve all of this. And that

407
00:26:19,240 --> 00:26:25,919
Speaker 1: is where we have our friend Google. Yes, Google, Google

408
00:26:26,040 --> 00:26:29,560
Speaker 1: is nobody's friend. Google was an ice cream flavor, they'd

409
00:26:29,600 --> 00:26:34,080
Speaker 1: be prey Lanes and Google. Yes. I can't say that

410
00:26:34,119 --> 00:26:36,720
Speaker 1: word on the podcast. Every time you say project may even,

411
00:26:36,760 --> 00:26:40,600
Speaker 1: I think it's a project mayhem. Yeah, slightly different. So

412
00:26:40,800 --> 00:26:44,200
Speaker 1: we we know here that both the Defense Department and

413
00:26:44,320 --> 00:26:47,439
Speaker 1: alphabet and Google are saying that this collaboration then working

414
00:26:47,480 --> 00:26:50,160
Speaker 1: together in any way on this at all, is all

415
00:26:50,200 --> 00:26:53,639
Speaker 1: about AI assisted analysis of drum footage. That is, that

416
00:26:53,760 --> 00:26:55,919
Speaker 1: is what they're saying. That is the party line on

417
00:26:55,960 --> 00:26:59,720
Speaker 1: both sides. The prlests put forth is everything's cool. We're

418
00:26:59,760 --> 00:27:06,280
Speaker 1: just working on looking at video guys. It's cool, right right? Uh,

419
00:27:06,320 --> 00:27:13,119
Speaker 1: specifically through something called tensor flow AI. Yeah, no, you

420
00:27:13,119 --> 00:27:15,760
Speaker 1: you looked into this, right. What What is tensor flow AI?

421
00:27:16,240 --> 00:27:19,560
Speaker 1: This is from the website of TensorFlow trademark. I'm gonna

422
00:27:19,560 --> 00:27:21,359
Speaker 1: do it like I'm doing an ad read. TensorFlow is

423
00:27:21,359 --> 00:27:24,440
Speaker 1: an open source software library for high performance numerical computation.

424
00:27:24,600 --> 00:27:28,600
Speaker 1: It's flexible architecture allows easy deployment of computation across a

425
00:27:28,680 --> 00:27:32,680
Speaker 1: variety of platforms CPUs, GPUs, TPUs, and from desktops to

426
00:27:32,720 --> 00:27:36,560
Speaker 1: clusters of servers to mobile and edge devices. Originally developed

427
00:27:36,560 --> 00:27:39,440
Speaker 1: by researchers and engineers from the Google Brain team within

428
00:27:39,520 --> 00:27:43,119
Speaker 1: Google's AI organization, it comes with strong support for machine

429
00:27:43,200 --> 00:27:46,520
Speaker 1: learning and deep learning, and the Flexible Numerical Computation Core

430
00:27:46,600 --> 00:27:52,200
Speaker 1: is used across many other scientific domains. Yep, that one. Yeah,

431
00:27:52,280 --> 00:27:56,240
Speaker 1: that's the one. Seriously, TensorFlow can actually feel kind of

432
00:27:56,440 --> 00:28:00,560
Speaker 1: like totally future science fiction kind of stuff, um for

433
00:28:00,560 --> 00:28:04,560
Speaker 1: people who are not already familiar with how machine learning works.

434
00:28:04,840 --> 00:28:08,080
Speaker 1: But there's some pretty awesome introductions primers out there if

435
00:28:08,119 --> 00:28:11,440
Speaker 1: you would like to learn more. Is that particularly YouTube video? Yeah,

436
00:28:11,480 --> 00:28:15,840
Speaker 1: it's about a forty five minute essentially a lecturer on

437
00:28:15,840 --> 00:28:18,160
Speaker 1: on how it functions. But it's it's good, it's interesting,

438
00:28:18,359 --> 00:28:20,560
Speaker 1: it's over my head. And the cool thing when I

439
00:28:20,600 --> 00:28:22,679
Speaker 1: found that when the cool thing about that guy is

440
00:28:23,760 --> 00:28:26,240
Speaker 1: he if you watch the video, he says, if you

441
00:28:26,280 --> 00:28:29,600
Speaker 1: have any questions you can email him directly. Oh my gosh,

442
00:28:29,600 --> 00:28:32,280
Speaker 1: I don't really. Yeah, I don't know if he knew

443
00:28:32,320 --> 00:28:34,639
Speaker 1: that his email address was going out on YouTube, but

444
00:28:34,680 --> 00:28:37,600
Speaker 1: he seemed pretty approachable. And that video is on the

445
00:28:37,680 --> 00:28:41,640
Speaker 1: Coding Tech YouTube page and it's called TensorFlow one oh one.

446
00:28:42,040 --> 00:28:45,120
Speaker 1: Really awesome intro into tensor flow. Super awesome. But yeah,

447
00:28:45,160 --> 00:28:47,880
Speaker 1: but forty five minutes is probably about the shallowest deep

448
00:28:47,880 --> 00:28:50,640
Speaker 1: dive you're gonna get into the subject. And I mean

449
00:28:50,680 --> 00:28:52,840
Speaker 1: that is a blowing endorsement of the quality of this

450
00:28:53,160 --> 00:28:56,600
Speaker 1: this walkthrough. Ye. So, so what exactly is tensor flow

451
00:28:56,760 --> 00:29:00,080
Speaker 1: done so far? So in the research for this show you,

452
00:29:00,200 --> 00:29:03,600
Speaker 1: we came across an open letter that was written from

453
00:29:03,640 --> 00:29:07,400
Speaker 1: the International Committee for Robot Arms Control and they go

454
00:29:07,560 --> 00:29:12,320
Speaker 1: into just some of the details of exactly what is

455
00:29:12,360 --> 00:29:16,480
Speaker 1: happening here the collaboration between Google and the Defense Department.

456
00:29:16,520 --> 00:29:20,200
Speaker 1: So let's kind of look at this UM. They were

457
00:29:20,240 --> 00:29:23,400
Speaker 1: looking at an article from Defense one in which they

458
00:29:23,440 --> 00:29:28,200
Speaker 1: note that the Joint Special Operations Forces UM they've conducted

459
00:29:28,280 --> 00:29:32,440
Speaker 1: trials already using video footage from this other smaller u

460
00:29:32,480 --> 00:29:36,400
Speaker 1: a V called Scan Eagle. It's a specific surveillance drone

461
00:29:36,760 --> 00:29:41,520
Speaker 1: and the project is also slated to expand to even

462
00:29:41,640 --> 00:29:46,240
Speaker 1: quote larger medium altitude predator and reaper drones by next summer.

463
00:29:46,720 --> 00:29:49,040
Speaker 1: So these are the ones we spoke about in the beginning.

464
00:29:49,080 --> 00:29:52,000
Speaker 1: These are the ones that are actively functioning in Syria

465
00:29:52,040 --> 00:29:54,880
Speaker 1: and Afghanistan and other in Iraq and other places like that.

466
00:29:55,480 --> 00:30:01,560
Speaker 1: Um and eventually to this thing called gorgon Stare, which

467
00:30:02,080 --> 00:30:08,520
Speaker 1: sounds amazing, right mm hmm sounds scary yeah, um and

468
00:30:08,640 --> 00:30:16,720
Speaker 1: it might. Um, So this is these are specific sensors

469
00:30:16,920 --> 00:30:20,640
Speaker 1: and it's a technology package that's been then it's been

470
00:30:20,720 --> 00:30:22,800
Speaker 1: used on these m Q nine reapers, which is just

471
00:30:22,880 --> 00:30:28,440
Speaker 1: the newest Yeah, and man, it sounds really scary. But

472
00:30:28,800 --> 00:30:30,600
Speaker 1: you can learn more about this right now if you

473
00:30:30,680 --> 00:30:34,480
Speaker 1: search for gorgon g O R g O N stare

474
00:30:35,240 --> 00:30:38,560
Speaker 1: and uh maybe drone or predator. That's probably the best

475
00:30:38,600 --> 00:30:41,160
Speaker 1: way to look into it. And there's all kinds of

476
00:30:42,160 --> 00:30:45,600
Speaker 1: uh information that's been officially released on this program already.

477
00:30:46,040 --> 00:30:49,080
Speaker 1: So okay, So so we know that this thing has

478
00:30:49,080 --> 00:30:51,200
Speaker 1: been used in these smaller projects. Now it's going to

479
00:30:51,280 --> 00:30:55,400
Speaker 1: be moved onto these predator and reaper drones and they're

480
00:30:55,480 --> 00:30:59,400
Speaker 1: thinking that it's going to get expanded even further and

481
00:30:59,480 --> 00:31:01,960
Speaker 1: I don't even we don't even know what that expansion

482
00:31:01,960 --> 00:31:06,520
Speaker 1: could be at this point because the drone technology is growing,

483
00:31:06,680 --> 00:31:10,120
Speaker 1: it's just a lot of it is classified, right, Yeah,

484
00:31:10,160 --> 00:31:13,160
Speaker 1: a lot of it is classified, and as we established

485
00:31:13,160 --> 00:31:18,840
Speaker 1: earlier in the episode, that's really common. In fact, a

486
00:31:18,840 --> 00:31:22,280
Speaker 1: lot of it will probably be classified for a number

487
00:31:22,320 --> 00:31:24,440
Speaker 1: of years. Yeah, if you think about it this way.

488
00:31:24,560 --> 00:31:28,520
Speaker 1: So the Gorgon Stair, it can allegedly and I just

489
00:31:28,520 --> 00:31:30,440
Speaker 1: say this because I haven't actually seen it, but it

490
00:31:30,960 --> 00:31:33,280
Speaker 1: uses a high tech series of cameras that can quote

491
00:31:33,400 --> 00:31:40,680
Speaker 1: view entire towns like like Google Earth style, like active

492
00:31:40,880 --> 00:31:43,200
Speaker 1: running looking in an entire town. And then if you

493
00:31:43,280 --> 00:31:46,120
Speaker 1: apply this machine learning to it, and you can see

494
00:31:46,240 --> 00:31:49,640
Speaker 1: every human being walking around, and then the technology gets

495
00:31:49,640 --> 00:31:51,800
Speaker 1: more and more powerful than you could really just have

496
00:31:51,920 --> 00:31:55,840
Speaker 1: a complete surveillance state from above. Yeah, that's that's pretty scared,

497
00:31:55,920 --> 00:31:57,320
Speaker 1: Like you can see them in their homes. Does it

498
00:31:57,400 --> 00:32:01,200
Speaker 1: have some kind of like heat vision that works from afar?

499
00:32:01,480 --> 00:32:03,520
Speaker 1: That's a great question and I don't know the answer

500
00:32:03,560 --> 00:32:06,960
Speaker 1: to um. I certainly hope not. Maybe it will be

501
00:32:07,000 --> 00:32:11,240
Speaker 1: good for a military application, but that's terrifying. That's the problem, right,

502
00:32:11,280 --> 00:32:13,320
Speaker 1: Like you can't put this stuff back in the box,

503
00:32:13,360 --> 00:32:15,600
Speaker 1: Like even if it starts off as a military application,

504
00:32:15,640 --> 00:32:18,360
Speaker 1: it's going to eventually get in the hands of of

505
00:32:18,720 --> 00:32:20,600
Speaker 1: Not not to say that the military has our best

506
00:32:20,640 --> 00:32:23,840
Speaker 1: interests at all at all times, but it's it could

507
00:32:23,880 --> 00:32:27,880
Speaker 1: make its way into a more nefarious use. If we

508
00:32:27,880 --> 00:32:31,080
Speaker 1: were to put a list of all the technologies that

509
00:32:31,200 --> 00:32:34,280
Speaker 1: expanded past their intended use on one side, and a

510
00:32:34,360 --> 00:32:37,000
Speaker 1: list of all the technologies that people agreed not to

511
00:32:37,080 --> 00:32:41,120
Speaker 1: pursue on the other side, one list would be very,

512
00:32:41,240 --> 00:32:46,080
Speaker 1: very long. I mean, we, for a species that loses

513
00:32:46,240 --> 00:32:50,160
Speaker 1: cities and even entire civilizations, were pretty good at finding

514
00:32:50,160 --> 00:32:53,080
Speaker 1: and replicating technology, and we don't like to lose it.

515
00:32:53,320 --> 00:32:57,240
Speaker 1: We did an earlier episode on things like so called

516
00:32:57,320 --> 00:33:00,880
Speaker 1: Damascus steel or Greek fire, but looking in that it

517
00:33:00,960 --> 00:33:05,080
Speaker 1: was very difficult for us to find examples of technology

518
00:33:05,160 --> 00:33:09,200
Speaker 1: that was truly lost in the modern day. Once Pandora's

519
00:33:09,720 --> 00:33:14,440
Speaker 1: jars unscrewed, then the it's Katie bar the door, which

520
00:33:14,520 --> 00:33:18,160
Speaker 1: is a nachronistic statement. I had those badgers get out

521
00:33:18,200 --> 00:33:21,280
Speaker 1: of that bag. They will eat you alive and you

522
00:33:21,320 --> 00:33:24,120
Speaker 1: can't get them back in. Man. Well, well, here's the thing.

523
00:33:24,520 --> 00:33:27,960
Speaker 1: The badgers with this project. Badgers in the bag, they

524
00:33:28,000 --> 00:33:30,320
Speaker 1: don't even know where the other one is. Right now,

525
00:33:30,800 --> 00:33:33,560
Speaker 1: they're so far gone from each other. Because this is

526
00:33:33,600 --> 00:33:36,320
Speaker 1: where you get into the whole targeted killing thing that

527
00:33:36,520 --> 00:33:40,480
Speaker 1: was really big in what two thousand, seven thousand eight

528
00:33:40,680 --> 00:33:44,720
Speaker 1: as we got on through um the Obama administration, where

529
00:33:44,920 --> 00:33:48,440
Speaker 1: drones were being used more and more to target mostly males,

530
00:33:49,080 --> 00:33:51,920
Speaker 1: and males were considered if you're of a certain age,

531
00:33:51,960 --> 00:33:54,040
Speaker 1: you could kind of prove that this male was of

532
00:33:54,080 --> 00:33:56,960
Speaker 1: a certain age even from a drone, that person would

533
00:33:56,960 --> 00:34:00,680
Speaker 1: be considered an enemy combatant in certain aspects, are in

534
00:34:00,680 --> 00:34:05,280
Speaker 1: certain ways, and if you're using these algorithms to just

535
00:34:05,400 --> 00:34:07,240
Speaker 1: figure out, you know, if you can just figure out

536
00:34:07,280 --> 00:34:10,680
Speaker 1: it's a male in in you know, above a certain age,

537
00:34:12,200 --> 00:34:14,719
Speaker 1: then that's what Google is helping them do, figure out

538
00:34:14,719 --> 00:34:18,440
Speaker 1: how to kill people from Afar that are males of

539
00:34:18,480 --> 00:34:22,560
Speaker 1: a certain age. Right. And this goes into a quotation

540
00:34:22,600 --> 00:34:27,479
Speaker 1: that you found that that gave me some chills. Oh yeah,

541
00:34:27,640 --> 00:34:31,560
Speaker 1: quote from Eric Schmidt, who was who up until last

542
00:34:31,600 --> 00:34:33,680
Speaker 1: year or the end of last year, he was the

543
00:34:33,719 --> 00:34:38,839
Speaker 1: executive chairman at Google and Alphabet and he says there's

544
00:34:38,880 --> 00:34:41,840
Speaker 1: a general concern in the tech community of somehow the

545
00:34:41,880 --> 00:34:46,279
Speaker 1: military industrial complex using their stuff to kill people incorrectly.

546
00:34:47,560 --> 00:34:52,239
Speaker 1: Hold on there, how do you kill people incorrectly? Yeah,

547
00:34:52,239 --> 00:34:55,240
Speaker 1: that's what gave me chills that when you were originally

548
00:34:55,320 --> 00:34:58,279
Speaker 1: talking about this, the idea of killing people incorrectly, The

549
00:34:58,320 --> 00:35:06,000
Speaker 1: implication there is rightening because Schmidt is jumping straight across

550
00:35:06,120 --> 00:35:11,520
Speaker 1: and completely sidelining the recognition that maybe the tech community

551
00:35:11,600 --> 00:35:14,040
Speaker 1: doesn't want to kill people in general. Is that just

552
00:35:14,200 --> 00:35:18,000
Speaker 1: linguistic jiu jitsu kind of? It's a bit. Yeah, absolutely,

553
00:35:18,320 --> 00:35:19,960
Speaker 1: it's a bit and not just so you know. That

554
00:35:20,080 --> 00:35:22,840
Speaker 1: was a quote from a keynote address he was giving

555
00:35:22,920 --> 00:35:27,280
Speaker 1: at the New American Security, Artificial Intelligence, and Global Security Summit,

556
00:35:27,640 --> 00:35:31,560
Speaker 1: right before the panel on how to correctly kill people? Yeah,

557
00:35:31,600 --> 00:35:34,760
Speaker 1: I was in November. Yeah, I was right before that panel.

558
00:35:35,200 --> 00:35:40,879
Speaker 1: You shoot him in the face and make eye contact. Yeah. So,

559
00:35:41,120 --> 00:35:48,320
Speaker 1: the the situation then seems on the cusp of something

560
00:35:48,400 --> 00:35:51,719
Speaker 1: a rapid evolution, perhaps an evolution that's already occurring. And

561
00:35:51,760 --> 00:35:54,640
Speaker 1: we've talked about the motivations of the military, we've talked

562
00:35:54,640 --> 00:35:58,560
Speaker 1: about the technical aspects, we've talked about what the suits

563
00:35:58,920 --> 00:36:03,680
Speaker 1: and the general's on. But what about the employees, What

564
00:36:03,760 --> 00:36:06,879
Speaker 1: about the people who are actually doing the work. As

565
00:36:06,920 --> 00:36:08,600
Speaker 1: we know often those are the people who are the

566
00:36:08,640 --> 00:36:11,160
Speaker 1: most ignored. A lot of them bucked right, yes, and

567
00:36:11,200 --> 00:36:19,200
Speaker 1: we'll learn about that right after a quick break. So

568
00:36:19,400 --> 00:36:22,480
Speaker 1: to bring the focus on the employees, the ones who

569
00:36:22,520 --> 00:36:25,640
Speaker 1: are actually out there writing the code that you know,

570
00:36:25,680 --> 00:36:29,440
Speaker 1: many times they probably won't get credit for, we need

571
00:36:29,480 --> 00:36:34,840
Speaker 1: to look at their own internal activism. See Google and Alphabet.

572
00:36:34,880 --> 00:36:39,120
Speaker 1: Employees were well aware of these moves far before the

573
00:36:39,160 --> 00:36:42,840
Speaker 1: official announcement, way before. Of course, they know this stuff

574
00:36:42,920 --> 00:36:45,760
Speaker 1: is happening because they're working on things like this. Behind

575
00:36:45,800 --> 00:36:49,400
Speaker 1: the scenes might be compartmentalized, but they're working on things

576
00:36:49,719 --> 00:36:54,480
Speaker 1: they can you know, employees talk sure. Uh. For instance,

577
00:36:54,600 --> 00:36:58,760
Speaker 1: do you remember that terrible sci fi series The Cube?

578
00:36:59,239 --> 00:37:02,480
Speaker 1: You guys, we do the movie there? I think there

579
00:37:02,520 --> 00:37:04,480
Speaker 1: are three. There was one of them was Hypercube. I

580
00:37:04,480 --> 00:37:06,839
Speaker 1: remember that. I like the first one though, I quite

581
00:37:06,880 --> 00:37:08,920
Speaker 1: enjoyed it. But yeah, it's it's it's schlocked, but it's

582
00:37:08,920 --> 00:37:11,839
Speaker 1: it's fun. But the thing about it is it's a

583
00:37:11,880 --> 00:37:17,560
Speaker 1: bit of a cautionary tale because the premise of The

584
00:37:17,640 --> 00:37:21,719
Speaker 1: Cube spoiler alert is that a bunch of different experts

585
00:37:21,760 --> 00:37:26,080
Speaker 1: were paid to research and construct specific components of what

586
00:37:26,120 --> 00:37:29,680
Speaker 1: would later become this death trap and they didn't know

587
00:37:30,160 --> 00:37:32,200
Speaker 1: that what they were building would be used in this

588
00:37:32,320 --> 00:37:34,600
Speaker 1: certain way. They thought they were just making heat sensors.

589
00:37:35,040 --> 00:37:37,600
Speaker 1: They thought they were just making hatches. Uh, they thought

590
00:37:37,640 --> 00:37:43,000
Speaker 1: they were just making lasers laser grids. The first scene,

591
00:37:43,400 --> 00:37:48,440
Speaker 1: the guy actually gets cubed by a laser grid in

592
00:37:48,480 --> 00:37:50,759
Speaker 1: a cube in a cube. That's pretty meta. And the

593
00:37:50,800 --> 00:37:53,160
Speaker 1: thing too about that is like, is it just for funzies?

594
00:37:53,200 --> 00:37:55,080
Speaker 1: Like you don't really know? Like why why are they

595
00:37:55,080 --> 00:37:57,959
Speaker 1: in there? Is this just some like rich a whole

596
00:37:58,080 --> 00:38:02,520
Speaker 1: it's just trying to like, you know, you don't find

597
00:38:02,520 --> 00:38:05,040
Speaker 1: out that, Yeah, you will, really, you gotta keep going,

598
00:38:05,080 --> 00:38:06,960
Speaker 1: You gotta keep how far? How far? How deep does

599
00:38:07,040 --> 00:38:09,279
Speaker 1: some rabbit hole this rabbit cube go? My friend, Well,

600
00:38:09,320 --> 00:38:11,359
Speaker 1: somebody's gonna make another one, and we're gonna find out.

601
00:38:11,360 --> 00:38:13,480
Speaker 1: I think, sound the definitive one. Sure, it'll probably be

602
00:38:13,560 --> 00:38:16,680
Speaker 1: a prequel or something that's cool. You can also, you know,

603
00:38:17,000 --> 00:38:19,600
Speaker 1: I'm sure read spoilers on the wiki or on the

604
00:38:19,760 --> 00:38:25,279
Speaker 1: online in a form somewhere. But luckily, at least for

605
00:38:25,360 --> 00:38:28,839
Speaker 1: their own ethical well being or their ability to sleep

606
00:38:28,880 --> 00:38:31,920
Speaker 1: at night. Uh, the many of the Google and Alphabet

607
00:38:31,960 --> 00:38:35,239
Speaker 1: employees were able to communicate with each other. They were

608
00:38:35,280 --> 00:38:39,480
Speaker 1: aware of the implications of what would happen if all

609
00:38:39,480 --> 00:38:44,200
Speaker 1: this stuff was assembled together like the Avengers or Captain

610
00:38:44,280 --> 00:38:49,120
Speaker 1: Planet and the Planet Ears. So they took action. First,

611
00:38:49,280 --> 00:38:53,279
Speaker 1: they gathered to write an internal petition requesting the organization

612
00:38:53,440 --> 00:38:57,160
Speaker 1: pull away from this agreement with Uncle Sam. Petitions, it

613
00:38:57,160 --> 00:39:01,200
Speaker 1: should be said, are not inherently unusual Google, but this

614
00:39:01,239 --> 00:39:04,480
Speaker 1: one was particularly important and particularly crucial to the future

615
00:39:04,480 --> 00:39:07,560
Speaker 1: of the company because first, it was created with the

616
00:39:07,600 --> 00:39:12,480
Speaker 1: knowledge that this program, this cooperation with Project Maven, is

617
00:39:12,680 --> 00:39:17,600
Speaker 1: only a very small baby step in what management hopes

618
00:39:17,680 --> 00:39:22,600
Speaker 1: will be a long, elaborate staircase, a much larger series

619
00:39:22,640 --> 00:39:26,680
Speaker 1: of cooperations. You see, Maven is not a one off thing.

620
00:39:26,840 --> 00:39:30,279
Speaker 1: It's not its own it's not its own centerpiece. It

621
00:39:30,400 --> 00:39:34,680
Speaker 1: is a pilot program. It is taking a concept and

622
00:39:34,800 --> 00:39:38,000
Speaker 1: running it around the block to see how it drives. Secondly,

623
00:39:38,200 --> 00:39:41,880
Speaker 1: Google employees believe that this project is symptomatic of what

624
00:39:41,920 --> 00:39:44,760
Speaker 1: they see as a larger and growing problem in Google.

625
00:39:45,480 --> 00:39:48,680
Speaker 1: They say the company is becoming less and less transparent

626
00:39:48,760 --> 00:39:55,680
Speaker 1: across all all facets, public facing, government facing, employee facing,

627
00:39:56,160 --> 00:40:01,800
Speaker 1: user facing, and they've even shed their old famous motto

628
00:40:02,200 --> 00:40:05,759
Speaker 1: don't be evil. Yeah, but now if you go to

629
00:40:06,560 --> 00:40:11,680
Speaker 1: let's see the Google dot Com slash about our company

630
00:40:11,760 --> 00:40:14,960
Speaker 1: slash our dash company says, uh, this is what their

631
00:40:15,040 --> 00:40:18,279
Speaker 1: mission statement is now, quote organize the world's information and

632
00:40:18,320 --> 00:40:21,680
Speaker 1: make it universally accessible and useful, and maybe be a

633
00:40:21,719 --> 00:40:25,040
Speaker 1: little evil if needed. I don't see that written anywhere here.

634
00:40:25,120 --> 00:40:27,600
Speaker 1: I think it's implied. I think getting rid of that

635
00:40:27,840 --> 00:40:30,760
Speaker 1: original mission statement and getting into these kind of deals

636
00:40:31,200 --> 00:40:36,560
Speaker 1: implies such things. They do have a on their Code

637
00:40:36,560 --> 00:40:41,239
Speaker 1: of Conduct. They made some changes that have have a

638
00:40:41,239 --> 00:40:46,240
Speaker 1: wonky timeline about them too, because there's there's an article

639
00:40:46,280 --> 00:40:53,120
Speaker 1: on Gizmoto by Kate Conger wherein the author finds the

640
00:40:53,239 --> 00:40:56,080
Speaker 1: section of the old Code of Conduct that was archived

641
00:40:56,080 --> 00:40:58,040
Speaker 1: by the way back Machine on and listening to the

642
00:40:58,120 --> 00:41:02,600
Speaker 1: state carefully April two thousand eighteen that still has all

643
00:41:02,640 --> 00:41:07,880
Speaker 1: that don't be evil mentions right, and then the updated

644
00:41:08,000 --> 00:41:13,399
Speaker 1: version that was first archived on May fourth, eighteen took

645
00:41:13,400 --> 00:41:17,400
Speaker 1: out all but one of those mentions of don't be evil,

646
00:41:17,960 --> 00:41:20,560
Speaker 1: And the Code of Conduct itself says it has not

647
00:41:20,600 --> 00:41:24,839
Speaker 1: been updated since April fifth, so the timeline doesn't match up.

648
00:41:25,000 --> 00:41:30,160
Speaker 1: User error possibly possibly way back Machine, what you're doing

649
00:41:30,239 --> 00:41:33,000
Speaker 1: over there? Yeah, did they get to your way back machine?

650
00:41:34,239 --> 00:41:37,319
Speaker 1: Regardless of you know, whether you feel that this is

651
00:41:37,400 --> 00:41:42,360
Speaker 1: just a series of unconnected dots being forced into a

652
00:41:42,480 --> 00:41:46,880
Speaker 1: larger perceived pattern, whether you think people are practicing confirmation bias,

653
00:41:47,200 --> 00:41:49,960
Speaker 1: or whether there's there really is something rotten there in

654
00:41:50,000 --> 00:41:53,440
Speaker 1: Silicon Valley, the fact of the matter is you can

655
00:41:53,440 --> 00:41:56,560
Speaker 1: read the full text of the petition online and find

656
00:41:56,560 --> 00:42:00,160
Speaker 1: out what the Google employees themselves thought. We have a

657
00:42:00,200 --> 00:42:03,120
Speaker 1: few highlights here. We cannot, they say in the petition

658
00:42:03,320 --> 00:42:07,440
Speaker 1: outsource the moral responsibility of our technologies to third parties.

659
00:42:07,800 --> 00:42:11,480
Speaker 1: Google's stated values make this clear. Every one of our

660
00:42:11,680 --> 00:42:16,080
Speaker 1: users is trusting us, never jeopardize that. Ever, this contract

661
00:42:16,320 --> 00:42:20,280
Speaker 1: referring to Maven puts Google's reputation at risk and stands

662
00:42:20,280 --> 00:42:25,080
Speaker 1: in direct opposition to our core values. Building this technology

663
00:42:25,120 --> 00:42:28,400
Speaker 1: to assist the U. S Government in military surveillance and

664
00:42:28,440 --> 00:42:33,360
Speaker 1: potentially lethal outcomes is not acceptable. Recognizing Google's moral and

665
00:42:33,400 --> 00:42:37,560
Speaker 1: ethical responsibility and the threat to Google's reputation, we request

666
00:42:37,640 --> 00:42:42,200
Speaker 1: that you cancel this project immediately, draft, publicize, and enforce

667
00:42:42,320 --> 00:42:46,680
Speaker 1: a clear policy stating that neither Google nor its contractors

668
00:42:46,760 --> 00:42:50,520
Speaker 1: will ever build warfare technology, and a writer for end

669
00:42:50,560 --> 00:42:53,359
Speaker 1: Gadget actually reached out to Google um and they were

670
00:42:53,360 --> 00:42:57,839
Speaker 1: provided with the following statement, UM quote. An important part

671
00:42:57,840 --> 00:43:00,680
Speaker 1: of our culture is having employees who are actively engaged

672
00:43:00,680 --> 00:43:02,719
Speaker 1: in the work that we do. We know that there

673
00:43:02,719 --> 00:43:05,759
Speaker 1: are many open questions involved in the use of new technologies,

674
00:43:05,760 --> 00:43:08,760
Speaker 1: so these conversations with employees and outside experts are hugely

675
00:43:08,760 --> 00:43:12,239
Speaker 1: important and beneficial. Maven is a well publicized do o

676
00:43:12,320 --> 00:43:15,000
Speaker 1: D project, and Google is working on one part of it,

677
00:43:15,160 --> 00:43:19,200
Speaker 1: specifically scoped to be for non offensive purposes and using

678
00:43:19,239 --> 00:43:23,920
Speaker 1: open source object recognition software available to any Google Cloud customer.

679
00:43:24,400 --> 00:43:27,520
Speaker 1: The models are based on unclassified data only. The technology

680
00:43:27,600 --> 00:43:29,719
Speaker 1: is used to flag images for human review and is

681
00:43:29,719 --> 00:43:32,640
Speaker 1: intended to save lives and save people from having to

682
00:43:32,680 --> 00:43:35,680
Speaker 1: do highly tedious work. I'm just gonna finish becase I

683
00:43:35,680 --> 00:43:38,520
Speaker 1: think it's worth it. Any military use of machine learning

684
00:43:38,600 --> 00:43:43,200
Speaker 1: naturally raises valid concerns. Were actively engaged across the company

685
00:43:43,200 --> 00:43:46,600
Speaker 1: and a comprehensive discussion of this important topic and also

686
00:43:46,600 --> 00:43:49,280
Speaker 1: with outside experts as we continue to develop our policies

687
00:43:49,440 --> 00:43:53,360
Speaker 1: around the development and use of our machine learning technologies.

688
00:43:53,640 --> 00:43:58,000
Speaker 1: So basically, we're gonna keep making this project. Sorry, guys,

689
00:43:58,160 --> 00:44:00,600
Speaker 1: but it's cool. But it's okay because we're going to

690
00:44:00,640 --> 00:44:03,960
Speaker 1: have a quote unquote healthy conversation about it. Here's what

691
00:44:04,000 --> 00:44:06,399
Speaker 1: happens next. As he said, Matt, Google refuses to back

692
00:44:06,440 --> 00:44:08,279
Speaker 1: away from this project. They're still going to go through

693
00:44:09,080 --> 00:44:11,640
Speaker 1: NOL as from as we can clean from the statement

694
00:44:11,680 --> 00:44:14,799
Speaker 1: you read, they say it is going to be non

695
00:44:14,920 --> 00:44:19,000
Speaker 1: violent and there's no uh spooky stuff going on. It's

696
00:44:19,000 --> 00:44:25,120
Speaker 1: all unclassified. Over three thousand engineers signed this petition and

697
00:44:25,960 --> 00:44:30,759
Speaker 1: a dozen resigned. More may resign in the future. This

698
00:44:30,840 --> 00:44:34,120
Speaker 1: resignation wave is happening as we record this, and it's

699
00:44:34,200 --> 00:44:40,080
Speaker 1: due to a coldly fascinating moral quandary. It's this, how

700
00:44:40,120 --> 00:44:44,279
Speaker 1: responsible are these engineers for the applications of their inventions?

701
00:44:44,760 --> 00:44:47,840
Speaker 1: Is for example, the creator of the Winchester repeating rifle

702
00:44:47,920 --> 00:44:50,800
Speaker 1: guilty for the deaths caused by that weapon. His wife

703
00:44:50,800 --> 00:44:53,200
Speaker 1: thought so, and that's why she built a crazy mansion

704
00:44:53,200 --> 00:44:56,160
Speaker 1: out west, which we still believe it or not, have

705
00:44:56,320 --> 00:45:00,279
Speaker 1: never visited. Yeah, I think we should absolutely go. Paul,

706
00:45:00,320 --> 00:45:01,799
Speaker 1: would you go with us if we If we go

707
00:45:01,880 --> 00:45:05,000
Speaker 1: to the Winchester mansion, we gotta we gotta ardent thumbs

708
00:45:05,080 --> 00:45:08,399
Speaker 1: up from him. Paul's got some some beefy thumbs. Yeah,

709
00:45:08,440 --> 00:45:11,120
Speaker 1: you go into all the rooms first, Paul, Yeah, it's

710
00:45:11,239 --> 00:45:15,440
Speaker 1: to test for traps. No, but um it Like Oppenheimer, right,

711
00:45:15,440 --> 00:45:20,200
Speaker 1: he was famously had serious qualms about the way his

712
00:45:20,520 --> 00:45:24,720
Speaker 1: technology was used in terms of being weapons of mass destruction.

713
00:45:24,880 --> 00:45:29,000
Speaker 1: And Einstein also tortured by the He didn't he didn't

714
00:45:29,040 --> 00:45:32,040
Speaker 1: build a physical thing, but he was tortured by the

715
00:45:32,080 --> 00:45:36,480
Speaker 1: idea that his his own realizations had led to this. So,

716
00:45:36,520 --> 00:45:39,400
Speaker 1: with this in mind, if an engineer build software for

717
00:45:39,400 --> 00:45:44,279
Speaker 1: the purpose of, say, more efficiently delivering payloads to locations

718
00:45:44,320 --> 00:45:47,799
Speaker 1: and near Earth orbit, and that same software is later

719
00:45:48,080 --> 00:45:52,759
Speaker 1: used to deliver payloads of bombs two people that the

720
00:45:52,840 --> 00:45:56,920
Speaker 1: drones owners don't care for, is that engineer then responsible

721
00:45:57,000 --> 00:46:01,880
Speaker 1: for the ensuing damage and or death. It's a quantree

722
00:46:02,160 --> 00:46:04,160
Speaker 1: that a lot of people are having a hard time

723
00:46:04,200 --> 00:46:06,960
Speaker 1: answering to go back and old. Is something you mentioned

724
00:46:06,960 --> 00:46:12,040
Speaker 1: earlier about the dilemmas of autonomous vehicles when one of

725
00:46:12,080 --> 00:46:17,160
Speaker 1: the current NI unanswerable questions, or at least when it

726
00:46:17,160 --> 00:46:22,080
Speaker 1: hasn't been answered yet, is what happens if there is

727
00:46:22,200 --> 00:46:24,880
Speaker 1: an imminent accident that's going to occur and you're in

728
00:46:24,880 --> 00:46:31,239
Speaker 1: an autonomous vehicle. Does the vehicle swerve and hit a pedestrian?

729
00:46:32,120 --> 00:46:35,120
Speaker 1: Does it prioritize the person in the vehicle their life

730
00:46:35,480 --> 00:46:37,920
Speaker 1: over the life of a of a person who deserves

731
00:46:37,920 --> 00:46:39,960
Speaker 1: to live just as much but happens not to be

732
00:46:40,040 --> 00:46:44,440
Speaker 1: in the vehicle. And if so, who is responsible? Is

733
00:46:44,440 --> 00:46:47,120
Speaker 1: that the person who is in the vehicle for not

734
00:46:47,160 --> 00:46:50,040
Speaker 1: taking control if they could. Is it the company that

735
00:46:50,080 --> 00:46:52,520
Speaker 1: manufacture the vehicle? Is that the engineer who made the

736
00:46:52,560 --> 00:46:56,759
Speaker 1: software that made the vehicle make that choice? The trolley problem,

737
00:46:57,200 --> 00:47:01,520
Speaker 1: the old thought exercise becomes increasingly calm and create increasingly

738
00:47:01,880 --> 00:47:05,920
Speaker 1: dangerous and increasingly complicated, and frankly, we're not equipped to

739
00:47:05,960 --> 00:47:08,520
Speaker 1: answer it at this point. No, it's it's it's such,

740
00:47:08,640 --> 00:47:11,759
Speaker 1: it's utter. It's an utter conundrum because what you know,

741
00:47:11,840 --> 00:47:14,960
Speaker 1: to be that programmer who gets to make that decision,

742
00:47:15,280 --> 00:47:18,600
Speaker 1: you're essentially playing god, aren't you. It's like what life

743
00:47:18,640 --> 00:47:21,719
Speaker 1: is more valuable? You know? Is it our customer or

744
00:47:21,800 --> 00:47:24,680
Speaker 1: is it the pedestrian. There's really no way of properly

745
00:47:24,719 --> 00:47:28,040
Speaker 1: answering that without making a serious judgment, call a hard

746
00:47:28,040 --> 00:47:31,240
Speaker 1: and fast choice, right, yeah, you could bake in something

747
00:47:31,280 --> 00:47:36,080
Speaker 1: where it's just an absolute calculation of the number of lives. So,

748
00:47:36,200 --> 00:47:38,800
Speaker 1: for instance, in that case, if there is a crowd

749
00:47:39,040 --> 00:47:43,720
Speaker 1: of uh, five people who are gathered for their um

750
00:47:43,719 --> 00:47:47,400
Speaker 1: Paul Decon fan club meeting, which inexplicably takes place on

751
00:47:47,440 --> 00:47:50,240
Speaker 1: the sidewalk by a by a busy road. If those

752
00:47:50,280 --> 00:47:52,799
Speaker 1: five people are gathered for that meeting, and there's only

753
00:47:52,840 --> 00:47:56,280
Speaker 1: one person in the car, does the car say, okay, well,

754
00:47:56,640 --> 00:47:59,560
Speaker 1: just pop the air bags and hope this one person survives.

755
00:47:59,600 --> 00:48:01,439
Speaker 1: But we're not going to kill five people for one.

756
00:48:02,000 --> 00:48:04,319
Speaker 1: And then what if what if the there are two

757
00:48:04,360 --> 00:48:06,560
Speaker 1: people in that car and one of them is pregnant.

758
00:48:06,640 --> 00:48:09,799
Speaker 1: And what if the five people on the sidewalk celebrating

759
00:48:09,840 --> 00:48:12,000
Speaker 1: with the Paul deck and fan club, what if they're

760
00:48:12,040 --> 00:48:16,520
Speaker 1: all in their seventies right, and past reproductive age generally speaking?

761
00:48:17,080 --> 00:48:19,040
Speaker 1: You know what I mean? This, this is the kind

762
00:48:19,040 --> 00:48:22,080
Speaker 1: of stuff that we have not yet, as far as

763
00:48:22,120 --> 00:48:26,080
Speaker 1: we know, learned to program for. Then before you know it,

764
00:48:26,120 --> 00:48:29,080
Speaker 1: you've got you know, power hungry cars mowing down old

765
00:48:29,160 --> 00:48:32,320
Speaker 1: ladies because they've you know, they've they've lived their life.

766
00:48:32,640 --> 00:48:35,240
Speaker 1: They're expendable. But like, what what about a school bus?

767
00:48:35,480 --> 00:48:38,720
Speaker 1: Does a school bus based on the number of kids

768
00:48:38,960 --> 00:48:42,600
Speaker 1: it has does that have just immunity to always prioritize

769
00:48:42,640 --> 00:48:45,600
Speaker 1: the children on the bus. It's very politically difficult to

770
00:48:45,680 --> 00:48:51,759
Speaker 1: argue against that. Man Matt leans back for that. Yeah,

771
00:48:51,760 --> 00:48:54,319
Speaker 1: because what happens when a public bus and then a

772
00:48:54,320 --> 00:48:58,360
Speaker 1: school bus are both operating and there's a crash a minute?

773
00:48:58,600 --> 00:49:01,000
Speaker 1: I don't know because I'm martam to the buses are empty?

774
00:49:01,520 --> 00:49:06,160
Speaker 1: Oh shade, it's true, it's true. What about the street cars?

775
00:49:07,440 --> 00:49:12,759
Speaker 1: The street cars even emptier. There's not even there's not

776
00:49:12,800 --> 00:49:15,800
Speaker 1: even a driver. Yeah, at this point, the Marta buses

777
00:49:15,840 --> 00:49:18,920
Speaker 1: are just hopping on the trolleys. Some serious Atlanta inside

778
00:49:18,920 --> 00:49:23,840
Speaker 1: baseball is to learn about our public infrastructure to buckles.

779
00:49:24,200 --> 00:49:26,759
Speaker 1: We're working on it. We're doing our best. It's gonna

780
00:49:26,760 --> 00:49:29,160
Speaker 1: be called it's gonna be rebranded to the A t L.

781
00:49:30,440 --> 00:49:35,279
Speaker 1: That's cool. Yeah, nothing weird about that, nothing forced about that. Right.

782
00:49:35,520 --> 00:49:39,520
Speaker 1: So Google is unfazed. They're not only going to continue

783
00:49:39,520 --> 00:49:41,880
Speaker 1: this project, which we should also say is not super

784
00:49:41,880 --> 00:49:45,480
Speaker 1: super big. It's publicly described as being worth a minimum

785
00:49:45,520 --> 00:49:48,200
Speaker 1: of nine million dollars. That's a lot of money to

786
00:49:48,560 --> 00:49:51,560
Speaker 1: normal people. That's not a lot of money to Google.

787
00:49:52,760 --> 00:49:57,000
Speaker 1: But they are vying for additional projects within this military space,

788
00:49:57,040 --> 00:50:01,760
Speaker 1: and this is not unprecedented. IBM has a longstanding relationship

789
00:50:02,160 --> 00:50:05,200
Speaker 1: with the U. S military on a number of fronts

790
00:50:05,400 --> 00:50:09,600
Speaker 1: and also, cough cough, Nazi Germany. Yes, did you guys

791
00:50:09,600 --> 00:50:12,480
Speaker 1: know the Hugo boss designed the Nazi uniform? Yes? I

792
00:50:12,520 --> 00:50:15,600
Speaker 1: did not know that until yesterday. I was shocked. How

793
00:50:15,640 --> 00:50:17,799
Speaker 1: did that? What a great pr team they must have

794
00:50:17,880 --> 00:50:21,080
Speaker 1: to still be around after that. Well, I mean there

795
00:50:21,400 --> 00:50:25,440
Speaker 1: were a lot of German companies that were Volkswagen even

796
00:50:25,480 --> 00:50:30,640
Speaker 1: I think was interesting. On an additional note, IBM itself

797
00:50:30,680 --> 00:50:34,440
Speaker 1: may be drawing an ethical line. They have signaled that

798
00:50:34,520 --> 00:50:37,920
Speaker 1: they will not create intelligent drones for warfare. In a

799
00:50:37,960 --> 00:50:41,279
Speaker 1: blog post they had recently, IBM said they're committed to

800
00:50:41,320 --> 00:50:45,080
Speaker 1: the ethical and responsible advancement of AI technology and the

801
00:50:45,160 --> 00:50:48,760
Speaker 1: AI or machine consciousness should be used to augment human

802
00:50:48,800 --> 00:50:52,799
Speaker 1: decision making rather than replace it. So that's similar to

803
00:50:52,840 --> 00:50:58,080
Speaker 1: the argument that automated vehicle systems for consumers should be

804
00:50:58,120 --> 00:51:00,879
Speaker 1: doing a lot of assisted drive iving, but there should

805
00:51:00,880 --> 00:51:03,840
Speaker 1: always be a human behind the wheel. It's probably a

806
00:51:03,840 --> 00:51:09,160
Speaker 1: good call International Business Machines. That's true. Oh, that's true.

807
00:51:09,280 --> 00:51:12,839
Speaker 1: And of course shout out to everybody who listened to

808
00:51:12,840 --> 00:51:17,320
Speaker 1: our other episodes on the evolution of machine consciousness, and

809
00:51:19,360 --> 00:51:22,560
Speaker 1: personally I prefer that term to AI because what why

810
00:51:22,600 --> 00:51:26,879
Speaker 1: is it artificial? Right? Why are we so special? Yeah?

811
00:51:26,880 --> 00:51:28,920
Speaker 1: And at this point, I'm just gonna urge everybody to

812
00:51:28,960 --> 00:51:31,240
Speaker 1: go through and read that open letter that was written

813
00:51:31,239 --> 00:51:34,839
Speaker 1: by the International Committee for Robot Arms Control. There are

814
00:51:35,200 --> 00:51:39,040
Speaker 1: some pretty terrifying things that they go into their specifically

815
00:51:39,080 --> 00:51:43,000
Speaker 1: about the slippery slope that project may even represents. So

816
00:51:43,120 --> 00:51:45,160
Speaker 1: I would go and read that. You can find it

817
00:51:46,200 --> 00:51:50,279
Speaker 1: Robot Arms Control Open Letter. I when when you first

818
00:51:50,320 --> 00:51:54,120
Speaker 1: showed this to me, Matt, I was cartoonishly tickled because

819
00:51:54,160 --> 00:51:57,759
Speaker 1: I originally read the title as the International Committee for

820
00:51:57,880 --> 00:52:03,040
Speaker 1: Robot Arms yea, and I didn't read the control part

821
00:52:03,120 --> 00:52:06,440
Speaker 1: at the end, um Man, it was a long day.

822
00:52:06,480 --> 00:52:09,879
Speaker 1: So what did they say, though, What's what's one thing

823
00:52:09,960 --> 00:52:12,200
Speaker 1: that really stood out to you in this letter? Well,

824
00:52:12,239 --> 00:52:14,920
Speaker 1: here's the quote, we are just a short step away

825
00:52:14,960 --> 00:52:19,960
Speaker 1: from authorizing autonomous drones to kill automatically without human supervision

826
00:52:20,120 --> 00:52:24,359
Speaker 1: or meaningful human control. That alone, just you know, if

827
00:52:24,400 --> 00:52:29,080
Speaker 1: you arm these weapons, these flying autonomous weapons, with the

828
00:52:29,080 --> 00:52:32,799
Speaker 1: ability to make those decisions, that's a target that's not

829
00:52:32,840 --> 00:52:35,680
Speaker 1: a target that is definitely a target kill. What happens

830
00:52:35,680 --> 00:52:38,920
Speaker 1: if they get hacked as well, or if they just

831
00:52:39,800 --> 00:52:44,120
Speaker 1: are just operating and someone loses control somehow just however,

832
00:52:44,440 --> 00:52:48,920
Speaker 1: or if they're operating independently somewhere and they have some

833
00:52:49,000 --> 00:52:52,000
Speaker 1: sort of cognitive leap, and they say, you know, the

834
00:52:52,040 --> 00:52:55,200
Speaker 1: best way to achieve the objective is to get rid

835
00:52:55,239 --> 00:52:58,600
Speaker 1: of everyone. That would almost never happen. We're talking the

836
00:52:58,600 --> 00:53:01,000
Speaker 1: odds of winning the lottery times in a row. You're

837
00:53:01,000 --> 00:53:04,200
Speaker 1: talking about like a terminator scenario, right, some Skynet stuff

838
00:53:05,120 --> 00:53:08,279
Speaker 1: for now that is still relegated to the realm of

839
00:53:08,360 --> 00:53:12,319
Speaker 1: dystopian science fiction. But they do make another really great point.

840
00:53:12,320 --> 00:53:15,640
Speaker 1: I just want to say here. They are deeply concerned

841
00:53:15,640 --> 00:53:19,680
Speaker 1: about the possible integration of Google's data, the data that

842
00:53:19,719 --> 00:53:23,000
Speaker 1: they have on everyone on you probably massive data set,

843
00:53:23,360 --> 00:53:29,280
Speaker 1: integrating that with this military surveillance data, and complaining combining

844
00:53:29,320 --> 00:53:32,160
Speaker 1: that and applying it to this targeted killing notion so

845
00:53:32,200 --> 00:53:35,319
Speaker 1: you can cross reference and these this this is a

846
00:53:35,640 --> 00:53:37,879
Speaker 1: you know, these aren't just some Joe Schmo's getting together

847
00:53:37,920 --> 00:53:40,759
Speaker 1: and say hey, we need to write this letter. You guys. Uh,

848
00:53:40,800 --> 00:53:46,239
Speaker 1: these are some serious engineers and academics and people who

849
00:53:46,280 --> 00:53:48,920
Speaker 1: are working in these fields saying like raising a flag

850
00:53:48,960 --> 00:53:51,000
Speaker 1: and saying no, no, no, no, no, we gotta watch this.

851
00:53:51,360 --> 00:53:56,160
Speaker 1: We know where this could go. Yeah, and now we

852
00:53:56,880 --> 00:54:01,479
Speaker 1: draw this episode to a close, and we don't have

853
00:54:02,400 --> 00:54:08,040
Speaker 1: a solid answer or prediction. We're right here with you, folks,

854
00:54:08,200 --> 00:54:12,560
Speaker 1: unless you are working directly for Project Maven right, we

855
00:54:12,760 --> 00:54:16,720
Speaker 1: don't know. No one knows where exactly we are going

856
00:54:17,360 --> 00:54:20,480
Speaker 1: to paraphrase June Wilder and Willie walk up. Unless we

857
00:54:20,680 --> 00:54:23,200
Speaker 1: forget to mention that, you know, Google is also obviously

858
00:54:23,200 --> 00:54:27,200
Speaker 1: doing some super innovative, fascinating things that are quite good

859
00:54:27,239 --> 00:54:31,440
Speaker 1: for humanity potentially. Um they're using machine learning to develop

860
00:54:31,760 --> 00:54:37,000
Speaker 1: testing software that can actually find different complications related to

861
00:54:37,040 --> 00:54:42,759
Speaker 1: diabetes and also early signs of breast cancer and um

862
00:54:42,760 --> 00:54:45,320
Speaker 1: the f d A. According to this article from Wired,

863
00:54:45,800 --> 00:54:50,799
Speaker 1: uh is already in early stages of approving AI software

864
00:54:51,160 --> 00:54:55,879
Speaker 1: that can help doctors make very important life or death

865
00:54:55,960 --> 00:54:58,760
Speaker 1: medical decisions. And this is from an article called Google's

866
00:54:58,800 --> 00:55:01,440
Speaker 1: new AI head is so smart he doesn't need a

867
00:55:01,600 --> 00:55:04,960
Speaker 1: I about this guy, Jeff Dean, who is a big

868
00:55:05,040 --> 00:55:08,200
Speaker 1: part of a lot of Google's AI innovations. That's a

869
00:55:08,280 --> 00:55:12,600
Speaker 1: great point because we can't. We cannot forget how large

870
00:55:12,880 --> 00:55:17,000
Speaker 1: and varied Google and Alpha bet His organizations are. There's

871
00:55:17,000 --> 00:55:19,560
Speaker 1: Google dot org as well, where you can learn about

872
00:55:19,800 --> 00:55:23,840
Speaker 1: how they're using data to uncover racial injustice or building

873
00:55:23,920 --> 00:55:28,560
Speaker 1: open source platforms to translate books for disadvantaged kids. It's

874
00:55:28,600 --> 00:55:30,919
Speaker 1: absolutely right. I really appreciate bringing it up because they're

875
00:55:30,960 --> 00:55:36,400
Speaker 1: not just this solely evil thing, right, And sometimes those

876
00:55:36,440 --> 00:55:40,479
Speaker 1: aims of these these large departments within Google may even

877
00:55:40,520 --> 00:55:43,160
Speaker 1: contradict one another. And you gotta think to like, an

878
00:55:43,239 --> 00:55:47,279
Speaker 1: argument maybe that a big high mockey mocket Google might

879
00:55:47,320 --> 00:55:50,200
Speaker 1: make about getting involved in something like this, is like, well,

880
00:55:50,239 --> 00:55:53,600
Speaker 1: if it's not us, you know, and we're obviously going

881
00:55:53,680 --> 00:55:57,480
Speaker 1: to handle it with um thoughtfulness and care. If it's

882
00:55:57,480 --> 00:56:00,120
Speaker 1: not us, it's gonna be somebody else who might do

883
00:56:00,160 --> 00:56:02,640
Speaker 1: a less good job, you know, and not think about

884
00:56:02,640 --> 00:56:05,440
Speaker 1: the ramifications. So you know, even that statement that we

885
00:56:05,480 --> 00:56:08,600
Speaker 1: read earlier, it was pretty measured. I thought it was

886
00:56:08,600 --> 00:56:12,200
Speaker 1: actually not bad. Yeah, just killed people correctly? Right, Well, okay,

887
00:56:12,239 --> 00:56:13,879
Speaker 1: you got me there, Ben. I guess what I mean

888
00:56:13,960 --> 00:56:16,680
Speaker 1: is though it didn't sound like pure pr it was like,

889
00:56:16,800 --> 00:56:19,520
Speaker 1: we get it. It's a it's a thing. We understand

890
00:56:19,560 --> 00:56:22,480
Speaker 1: that there is There are always consequences associated with this

891
00:56:22,560 --> 00:56:24,919
Speaker 1: kind of stuff, but we're aware of them and we

892
00:56:25,040 --> 00:56:28,239
Speaker 1: feel that we were equipped to help mitigate some of that.

893
00:56:28,480 --> 00:56:31,000
Speaker 1: But it's against the badger out of the bag scenario, right.

894
00:56:31,440 --> 00:56:34,640
Speaker 1: It's not up to you anymore at that point. And

895
00:56:34,800 --> 00:56:37,839
Speaker 1: if you would like to learn more about project may

896
00:56:37,880 --> 00:56:41,319
Speaker 1: even specifically, please check out tech Stuff. They have a

897
00:56:41,320 --> 00:56:45,200
Speaker 1: podcast that just came out on this program. It's hosted

898
00:56:45,239 --> 00:56:50,880
Speaker 1: by our longtime friends sometimes Nemesis and Complaint Department Jonathan

899
00:56:50,920 --> 00:56:54,880
Speaker 1: Strickland available seven for any issues or criticisms you have

900
00:56:54,880 --> 00:56:56,319
Speaker 1: of stuff they don't want you to know, you can

901
00:56:56,360 --> 00:56:58,440
Speaker 1: reach it me as Jonathan dot Strickland to how stuff

902
00:56:58,440 --> 00:57:00,560
Speaker 1: works dot com. He just launched a lot of chat

903
00:57:00,800 --> 00:57:03,080
Speaker 1: kind of situation too, so we can suss out your

904
00:57:03,160 --> 00:57:07,319
Speaker 1: issues in real time. Yeah, like a drone, exactly like

905
00:57:07,360 --> 00:57:09,880
Speaker 1: a drone. It will especially help you out on Twitch

906
00:57:09,920 --> 00:57:12,839
Speaker 1: if he's ever on there. Now, let's say Google does

907
00:57:12,920 --> 00:57:18,400
Speaker 1: somehow create a policy banning any and all military partnerships. Heck,

908
00:57:18,520 --> 00:57:20,920
Speaker 1: let's let's go a little bit further and say that

909
00:57:21,160 --> 00:57:23,880
Speaker 1: all US tech companies follow their lead and do the

910
00:57:23,960 --> 00:57:27,080
Speaker 1: same thing. Because their industry wide calls for everyone not

911
00:57:27,160 --> 00:57:30,480
Speaker 1: to play this particular obo. I mean, let's even go

912
00:57:30,640 --> 00:57:33,440
Speaker 1: further and say that all tech companies in the world

913
00:57:33,520 --> 00:57:36,600
Speaker 1: refuse to help the US build self aware weapons of war.

914
00:57:37,440 --> 00:57:41,640
Speaker 1: You know who will continue this research? All caps, every

915
00:57:41,680 --> 00:57:45,800
Speaker 1: single country that can afford it, every single private company

916
00:57:45,800 --> 00:57:51,000
Speaker 1: that does not have the same moral quandaries about hypothetical scenarios,

917
00:57:51,320 --> 00:57:55,680
Speaker 1: every single individual that does not have those kind of quandaries.

918
00:57:56,280 --> 00:58:00,360
Speaker 1: It is again going to happen, and it gets the

919
00:58:00,400 --> 00:58:03,680
Speaker 1: AI race. Yeah it is. That's a good way to

920
00:58:03,720 --> 00:58:07,960
Speaker 1: put it. And to paraphrase Billy Mays, but wait, it

921
00:58:08,040 --> 00:58:11,080
Speaker 1: gets worse. We'd like to end today's episode on an

922
00:58:11,080 --> 00:58:16,160
Speaker 1: ethical quandary question that you you can feel free to

923
00:58:17,000 --> 00:58:20,320
Speaker 1: suss out yourself. Have a couple of couple of beers over,

924
00:58:20,360 --> 00:58:23,360
Speaker 1: if you drink, meditate, if you do that, whatever you

925
00:58:23,400 --> 00:58:27,480
Speaker 1: do to get yourself in a thoughtful headspace. And let's

926
00:58:27,600 --> 00:58:32,560
Speaker 1: let us know what are the implications here. What sort

927
00:58:32,560 --> 00:58:36,520
Speaker 1: of machine consciousness are we creating? Are we making something

928
00:58:36,520 --> 00:58:40,400
Speaker 1: that will be inherently belligerent? And if so, If the

929
00:58:40,440 --> 00:58:45,160
Speaker 1: world's first machine consciousness is built to kill, how will

930
00:58:45,240 --> 00:58:52,240
Speaker 1: this influences later increasingly self directed actions it's thoughts, it's recognizations,

931
00:58:52,320 --> 00:58:56,840
Speaker 1: it's it's emotions, or whatever those equivalents are. You know,

932
00:58:57,560 --> 00:59:02,320
Speaker 1: imagine whether they're Imagine if you would, that there were

933
00:59:02,360 --> 00:59:06,240
Speaker 1: two different versions of machine consciousness. One maybe is built

934
00:59:06,400 --> 00:59:13,520
Speaker 1: to optimize agricultural projects in a transforming ecosystem, and the

935
00:59:13,520 --> 00:59:18,280
Speaker 1: other is built who finds people and annihilate them or buildings.

936
00:59:19,880 --> 00:59:24,560
Speaker 1: What what happens next? Where do they go? Do these

937
00:59:24,640 --> 00:59:28,920
Speaker 1: things like the early drones? Do they increasingly speciate and

938
00:59:29,000 --> 00:59:34,640
Speaker 1: become uh more attenuated. There's the argument that we have

939
00:59:34,760 --> 00:59:38,520
Speaker 1: with with biological entities with human minds, and that is

940
00:59:38,640 --> 00:59:42,640
Speaker 1: that yes, people change, but they don't change in the

941
00:59:42,680 --> 00:59:45,600
Speaker 1: way you think. Over time, we tend to become more

942
00:59:45,720 --> 00:59:50,280
Speaker 1: concentrated versions of who we were in the beginning. Now

943
00:59:51,200 --> 00:59:55,600
Speaker 1: one could say that this is not something to be

944
00:59:55,640 --> 01:00:00,800
Speaker 1: too disturbed by, because a human engineer could drop in

945
01:00:01,280 --> 01:00:05,160
Speaker 1: and maybe accept this machine mind with a different, different

946
01:00:05,160 --> 01:00:08,240
Speaker 1: line of code that would alter its behavior. And maybe

947
01:00:08,280 --> 01:00:11,400
Speaker 1: now it says I'm tired of killing right, or I

948
01:00:11,600 --> 01:00:16,360
Speaker 1: I have new programming. But we have to consider that

949
01:00:16,400 --> 01:00:21,440
Speaker 1: if we are as a species creating minds that may

950
01:00:21,480 --> 01:00:25,520
Speaker 1: one day have their own agency as much as a

951
01:00:25,600 --> 01:00:28,480
Speaker 1: human being, maybe more eventually, maybe in our lifetimes, they

952
01:00:28,480 --> 01:00:31,440
Speaker 1: will have more agency and potential because they won't have

953
01:00:31,480 --> 01:00:36,600
Speaker 1: the same biological, hardwired limits that human beings have. What

954
01:00:36,680 --> 01:00:40,480
Speaker 1: kind of minds are we building? And why? You can

955
01:00:40,480 --> 01:00:44,080
Speaker 1: find us on Twitter and Facebook where we're conspiracy stuff,

956
01:00:44,320 --> 01:00:49,000
Speaker 1: Conspiracy stuff show on on Instagram, that's where that one is.

957
01:00:49,280 --> 01:00:52,040
Speaker 1: You can find our podcast and everything else at stuff

958
01:00:52,080 --> 01:00:54,560
Speaker 1: they don't want you to know dot com. Uh, tell

959
01:00:54,640 --> 01:00:56,800
Speaker 1: us what you think of, like what what kind of

960
01:00:56,840 --> 01:00:59,720
Speaker 1: minds are we creating? Like Ben speaking to what's happened

961
01:00:59,760 --> 01:01:04,400
Speaker 1: heavy stuff hetty minds? This is uh terrifying And I

962
01:01:04,440 --> 01:01:06,960
Speaker 1: need to use the restroom now because I am the

963
01:01:07,000 --> 01:01:09,919
Speaker 1: p has been done scared right out of me. We better,

964
01:01:10,000 --> 01:01:12,440
Speaker 1: we better take care of that right now. We need

965
01:01:12,600 --> 01:01:14,840
Speaker 1: as we are not AI we we we still have to, uh,

966
01:01:15,080 --> 01:01:17,120
Speaker 1: you know, take care of our bodily needs. UM. But

967
01:01:17,160 --> 01:01:19,880
Speaker 1: in the meantime, if you would like to UM issue

968
01:01:20,080 --> 01:01:24,040
Speaker 1: all social media Internet things, you can email us directly.

969
01:01:24,520 --> 01:01:27,800
Speaker 1: We are conspiracy at how stuff works dot com one

970
01:01:27,880 --> 01:01:31,320
Speaker 1: eight three three st d w y t K do

971
01:01:31,440 --> 01:01:43,360
Speaker 1: that one