1
00:00:00,040 --> 00:00:02,520
Speaker 1: Guess what will? What's that mango? So I know you've

2
00:00:02,520 --> 00:00:05,120
Speaker 1: gotten really unto running recently, and I think there's an

3
00:00:05,120 --> 00:00:08,200
Speaker 1: accessory from Japan you really need to invest in. It's

4
00:00:08,240 --> 00:00:13,040
Speaker 1: called tomaton. Tomaton that sounds cool. So so what's a toomaton? Well,

5
00:00:13,119 --> 00:00:16,160
Speaker 1: it's exactly what it sounds like. It's a robot backpack

6
00:00:16,239 --> 00:00:19,880
Speaker 1: that feeds you tomatoes on the run. That is not

7
00:00:19,920 --> 00:00:22,599
Speaker 1: what I expected to say that that is much less

8
00:00:22,600 --> 00:00:26,840
Speaker 1: cool than I expected. Basically, there's this giant cartoony tomato

9
00:00:26,880 --> 00:00:29,160
Speaker 1: that sits on your head and it's got this clear

10
00:00:29,240 --> 00:00:31,680
Speaker 1: tube full of tomatoes that kind of hangs down your spine.

11
00:00:32,120 --> 00:00:35,400
Speaker 1: And as you run, this robotic arm swoops back to

12
00:00:35,440 --> 00:00:37,960
Speaker 1: grab a tomato, and then it swoops down in front

13
00:00:38,000 --> 00:00:40,239
Speaker 1: of you with that juicy fruit and sticks it in

14
00:00:40,240 --> 00:00:43,720
Speaker 1: your face. And I'm kind of surprised more elite athletes

15
00:00:43,760 --> 00:00:46,959
Speaker 1: don't use it. Well, you know, my kids already make

16
00:00:47,000 --> 00:00:49,000
Speaker 1: fun of me for what I look like when I run,

17
00:00:49,080 --> 00:00:51,239
Speaker 1: and so I can only imagine this is just going

18
00:00:51,280 --> 00:00:53,959
Speaker 1: to give them that much more ammunition, which is probably

19
00:00:53,960 --> 00:00:56,240
Speaker 1: all the more reason I should do it. But I'm

20
00:00:56,320 --> 00:00:59,160
Speaker 1: curious why tomatoes, of all things I've never wanted a

21
00:00:59,200 --> 00:01:02,720
Speaker 1: tomato while I was running, Well you haven't before, but

22
00:01:02,840 --> 00:01:05,479
Speaker 1: now you will. So it was invented by this company

23
00:01:05,520 --> 00:01:09,119
Speaker 1: called Kagome, which is Japan's premiere tomato juice and catchup

24
00:01:09,120 --> 00:01:12,720
Speaker 1: manufacturer apparently. But I am not done talking about Tamaton

25
00:01:12,840 --> 00:01:16,440
Speaker 1: just yet. So while the original Tmaton is pretty good

26
00:01:16,440 --> 00:01:18,800
Speaker 1: for five k's, you might want to look into the

27
00:01:18,840 --> 00:01:23,800
Speaker 1: petite Tamaton for marathons. It's uh, it's lighter, smaller. Instead

28
00:01:23,840 --> 00:01:27,120
Speaker 1: of um that robot arm feeder, It's got a trigger

29
00:01:27,200 --> 00:01:29,480
Speaker 1: and a tube and you can actually shoot tiny tomatoes

30
00:01:29,480 --> 00:01:32,760
Speaker 1: in your mouth anytime you want. It does have a

31
00:01:32,800 --> 00:01:35,600
Speaker 1: tomato timer that prevents you from eating too many tomatoes

32
00:01:35,640 --> 00:01:38,720
Speaker 1: too quickly, so you do pace yourself with it. So

33
00:01:39,520 --> 00:01:41,560
Speaker 1: that's that's really impressive. I can't say that I've ever

34
00:01:41,600 --> 00:01:44,800
Speaker 1: been at risk of feeding myself too many of anything

35
00:01:44,840 --> 00:01:48,560
Speaker 1: while on a long run, but whatever, that's great. Yeah,

36
00:01:48,640 --> 00:01:50,919
Speaker 1: well it's just the first of nine facts about robots

37
00:01:51,000 --> 00:02:15,280
Speaker 1: we're covering for today's show, Celesidising Day their podcast listeners,

38
00:02:15,280 --> 00:02:17,639
Speaker 1: Welcome to Part Time Genius. I'm Will Pearson and is

39
00:02:17,720 --> 00:02:20,560
Speaker 1: always I'm joined by my good friend Mangesh Ticketer and

40
00:02:20,639 --> 00:02:23,400
Speaker 1: sitting behind the soundproof glass. Now this guy is working

41
00:02:23,520 --> 00:02:26,720
Speaker 1: on another movie reboot now Mango. We've seen him work

42
00:02:26,800 --> 00:02:30,040
Speaker 1: on so many ideas for movie reboots, and this one

43
00:02:30,120 --> 00:02:33,760
Speaker 1: is for a feminist reboot for Short Circuit too. So

44
00:02:34,000 --> 00:02:35,600
Speaker 1: I think this is gonna be a big one. This

45
00:02:35,720 --> 00:02:38,240
Speaker 1: may be his most brilliant one yet. But that's that's

46
00:02:38,240 --> 00:02:42,000
Speaker 1: our friend and producer Tristan McNeil over there. That's amazing Tristan.

47
00:02:42,160 --> 00:02:45,480
Speaker 1: So I know we both love robots, and I certainly

48
00:02:45,520 --> 00:02:49,040
Speaker 1: was a Jetsons fan. I had kept Sella as a kid. Uh,

49
00:02:49,360 --> 00:02:52,560
Speaker 1: lots of radio shot kits, and you know, I loved

50
00:02:52,600 --> 00:02:55,920
Speaker 1: building weird sort of automated inventions. But there's actually another

51
00:02:55,919 --> 00:02:59,080
Speaker 1: reason we're doing a show on robots today, right, that's right.

52
00:02:59,120 --> 00:03:01,440
Speaker 1: So one of our four her guest Simone Yatch, and

53
00:03:01,480 --> 00:03:04,480
Speaker 1: you probably know her for her amazing crappy robots that

54
00:03:04,520 --> 00:03:07,880
Speaker 1: she builds and tests on YouTube. And we are huge

55
00:03:07,919 --> 00:03:09,799
Speaker 1: fans of Simone and if you think about some of

56
00:03:09,840 --> 00:03:12,639
Speaker 1: the examples of things that she's built, she built the

57
00:03:12,639 --> 00:03:15,200
Speaker 1: the robot that shampoo's your hair for you just got

58
00:03:15,200 --> 00:03:18,280
Speaker 1: a hilarious video of this, or the alarm clock that

59
00:03:18,280 --> 00:03:20,280
Speaker 1: wakes you up just by slapping you in the face,

60
00:03:20,320 --> 00:03:24,400
Speaker 1: which was also incredibly amusing. But you know, she she

61
00:03:24,480 --> 00:03:27,560
Speaker 1: got some not so great news, though she um feels

62
00:03:27,560 --> 00:03:29,600
Speaker 1: like she's gonna be okay. So she found out that

63
00:03:29,639 --> 00:03:32,919
Speaker 1: she has a brain tumor, is going to undergo surgery

64
00:03:32,960 --> 00:03:36,200
Speaker 1: this month, and they're optimistic about the prospects for this.

65
00:03:36,320 --> 00:03:38,520
Speaker 1: But we still we really wanted to wish her the

66
00:03:38,680 --> 00:03:41,000
Speaker 1: very best of luck, so we made this little nine

67
00:03:41,040 --> 00:03:44,360
Speaker 1: things in her honor. That's right, So get well soon, simone.

68
00:03:44,920 --> 00:03:47,480
Speaker 1: So back to our love of robots for a second.

69
00:03:47,560 --> 00:03:50,280
Speaker 1: So people always ask me about my favorite stories and

70
00:03:50,280 --> 00:03:53,240
Speaker 1: facts for Mental Floss, and they're a handful I always

71
00:03:53,240 --> 00:03:55,800
Speaker 1: talked about. But my favorite invention has to be this

72
00:03:55,960 --> 00:03:58,840
Speaker 1: robotic fish that gets other fish to follow it, because

73
00:03:58,880 --> 00:04:02,320
Speaker 1: it sounds like a small idea, like why do you

74
00:04:02,440 --> 00:04:05,400
Speaker 1: care that you can create a leader among fish? And

75
00:04:05,440 --> 00:04:08,240
Speaker 1: the answer is so much more interesting. It's so that,

76
00:04:08,520 --> 00:04:11,360
Speaker 1: like if there's an oil spill or a disaster or

77
00:04:11,400 --> 00:04:14,520
Speaker 1: something devastating, you can actually lead schools of fish away

78
00:04:14,560 --> 00:04:17,279
Speaker 1: from that disaster. And I just think that's amazing. It

79
00:04:17,400 --> 00:04:19,640
Speaker 1: is pretty cool, and I would say it's I mean

80
00:04:19,680 --> 00:04:22,880
Speaker 1: it's not as important as the backpack Tomato robot, but

81
00:04:23,080 --> 00:04:25,480
Speaker 1: it feels like it's almost as important as the backpack

82
00:04:25,560 --> 00:04:29,919
Speaker 1: to make no picking on thomasons. What robot factor are

83
00:04:29,920 --> 00:04:31,720
Speaker 1: you gonna start with? Well, all right, well, here's one

84
00:04:31,760 --> 00:04:34,120
Speaker 1: that I hadn't really thought about before, and it's what

85
00:04:34,279 --> 00:04:37,760
Speaker 1: personality your vacuum cleaners should have? And you know, I

86
00:04:37,800 --> 00:04:39,520
Speaker 1: know you've probably thought about this. You come up with

87
00:04:39,600 --> 00:04:42,120
Speaker 1: lots of great questions that I had never thought about it,

88
00:04:42,160 --> 00:04:45,000
Speaker 1: and apparently a lot of scientists hadn't thought about it either.

89
00:04:45,560 --> 00:04:48,880
Speaker 1: So robots are mostly supposed to be functional, right, but

90
00:04:49,120 --> 00:04:51,400
Speaker 1: you know, we're obviously going to have preferences for one

91
00:04:51,520 --> 00:04:54,839
Speaker 1: robot vacuum cleaner over another if they exhibit these certain

92
00:04:54,880 --> 00:04:57,320
Speaker 1: behaviors with their sounds. You know, if you think about

93
00:04:57,640 --> 00:04:59,640
Speaker 1: the sounds they make, or the light they put out,

94
00:04:59,720 --> 00:05:02,120
Speaker 1: or the motions they go through, you know we will

95
00:05:02,160 --> 00:05:05,359
Speaker 1: have preferences among these. So kay, can you talk me

96
00:05:05,360 --> 00:05:08,280
Speaker 1: through a few of these. Well, let's say you programmed

97
00:05:08,279 --> 00:05:11,000
Speaker 1: a robot vacuum cleaner to not leave a certain area

98
00:05:11,040 --> 00:05:14,080
Speaker 1: until it removed a spot. So just imagine it's just

99
00:05:14,120 --> 00:05:16,800
Speaker 1: going over and over the same area, and these kind

100
00:05:16,800 --> 00:05:19,760
Speaker 1: of short jarring motions and if it does that, it

101
00:05:19,839 --> 00:05:22,560
Speaker 1: might come across as obsessive, you know, just going back

102
00:05:22,600 --> 00:05:25,159
Speaker 1: and forth over that one unimportant spot while there's this

103
00:05:25,240 --> 00:05:28,640
Speaker 1: whole room it needs to vacuum. So scientists at delfta

104
00:05:28,680 --> 00:05:31,600
Speaker 1: University in the Netherlands decided to determine what types of

105
00:05:31,600 --> 00:05:35,440
Speaker 1: personalities people want from their vacuum cleaners. So they rounded

106
00:05:35,520 --> 00:05:38,760
Speaker 1: up fifteen early adopters. These are people who were predisposed

107
00:05:38,800 --> 00:05:42,080
Speaker 1: to buying robot vacuum cleaners. And then things got a

108
00:05:42,120 --> 00:05:45,320
Speaker 1: little bit weird, so they hired a bunch of actors

109
00:05:45,360 --> 00:05:49,039
Speaker 1: to either crawl around or walk around pretending to be

110
00:05:49,120 --> 00:05:52,760
Speaker 1: vacuum cleaners with different personality traits. Now, at first I

111
00:05:52,760 --> 00:05:55,200
Speaker 1: thought this was just made up, but this was an

112
00:05:55,240 --> 00:05:58,920
Speaker 1: actual study. And the only noises they could make were

113
00:05:59,000 --> 00:06:02,480
Speaker 1: vacuum cleaner noise, so you think worrying and sucking sounds.

114
00:06:02,520 --> 00:06:05,799
Speaker 1: But they could change patterns and ways that they moved,

115
00:06:06,200 --> 00:06:08,640
Speaker 1: and they were actually asked to act out these thirty

116
00:06:08,680 --> 00:06:12,080
Speaker 1: different personality traits for vacuum cleaners. How in the world

117
00:06:12,160 --> 00:06:14,680
Speaker 1: they could have these thirty different possible traits, I don't know,

118
00:06:14,760 --> 00:06:17,359
Speaker 1: but you know, just as examples. They could be bold

119
00:06:17,520 --> 00:06:20,840
Speaker 1: or calm, or talkative or several other things. But it's

120
00:06:20,880 --> 00:06:23,680
Speaker 1: just crazy to imagine this. That's funny. It sounds like

121
00:06:23,720 --> 00:06:27,960
Speaker 1: a terrible improv activity. But as it does, so when

122
00:06:27,960 --> 00:06:31,039
Speaker 1: did the scientists actually learn from this? Well, I mean

123
00:06:31,320 --> 00:06:33,920
Speaker 1: most of it was honestly pretty straightforward, like you don't

124
00:06:33,920 --> 00:06:37,080
Speaker 1: want to chatterbox for a vacuum cleaner, and most people

125
00:06:37,120 --> 00:06:40,160
Speaker 1: also don't want something that feels whimsical, like it might

126
00:06:40,200 --> 00:06:42,400
Speaker 1: make more sense for a vacuum cleaner to move in

127
00:06:42,520 --> 00:06:46,080
Speaker 1: circles or you know, maybe these unpredictable patterns once it's

128
00:06:46,160 --> 00:06:48,520
Speaker 1: mapped out all of the dirt that's in the room,

129
00:06:49,000 --> 00:06:52,560
Speaker 1: but people actually want to see straight lines and routine.

130
00:06:52,760 --> 00:06:56,119
Speaker 1: So whether that's actually, you know, the most efficient way

131
00:06:56,160 --> 00:06:59,159
Speaker 1: to vacuum a room, people perceive that to be the

132
00:06:59,200 --> 00:07:02,560
Speaker 1: most efficient. But yeah, actually, honestly, I just like that.

133
00:07:02,600 --> 00:07:05,039
Speaker 1: We had to hire a bunch of Shakespearean trained actors

134
00:07:05,080 --> 00:07:07,760
Speaker 1: to play the role of like vacuum number three, just

135
00:07:07,800 --> 00:07:12,440
Speaker 1: to figure all of this out. That's horrible, But I

136
00:07:12,480 --> 00:07:14,560
Speaker 1: do think it's an interesting idea about how much should

137
00:07:14,560 --> 00:07:17,680
Speaker 1: we care for and about our robots. And I was

138
00:07:17,680 --> 00:07:20,000
Speaker 1: looking at something to that same effect. I found this

139
00:07:20,120 --> 00:07:23,800
Speaker 1: article from two thousand and thirteen that you know, these

140
00:07:23,880 --> 00:07:27,400
Speaker 1: soldiers who were using bomb disposal robots on battlefields were

141
00:07:27,400 --> 00:07:30,160
Speaker 1: actually becoming too emotionally attached to them, or at least

142
00:07:30,200 --> 00:07:33,040
Speaker 1: that was the concern. And the robots weren't particularly cute,

143
00:07:33,040 --> 00:07:35,680
Speaker 1: like they didn't have like sweet faces or anything, but

144
00:07:35,680 --> 00:07:39,440
Speaker 1: but the soldiers actually started treating them as mascots or pets,

145
00:07:40,000 --> 00:07:42,000
Speaker 1: and then when they were destroyed, they actually have funerals

146
00:07:42,040 --> 00:07:44,480
Speaker 1: for them. But one of the big questions for robot

147
00:07:44,520 --> 00:07:47,480
Speaker 1: makers at the time was just how anthropomorphic, like do

148
00:07:47,480 --> 00:07:50,720
Speaker 1: you design these instruments, because you know, you do want

149
00:07:50,800 --> 00:07:53,600
Speaker 1: soldiers to care for the equipment and maintain it properly,

150
00:07:53,680 --> 00:07:55,720
Speaker 1: but you don't want them to care so much and

151
00:07:55,800 --> 00:07:58,800
Speaker 1: be so gingerly with these things that they refused to

152
00:07:58,800 --> 00:08:01,360
Speaker 1: put them in harm's way. You know, with too much

153
00:08:01,360 --> 00:08:03,440
Speaker 1: emotional attachment, you might not send your robot into a

154
00:08:03,440 --> 00:08:06,440
Speaker 1: space that you think dangerous, and that's precisely what they're

155
00:08:06,480 --> 00:08:08,760
Speaker 1: built for. Yeah, that is pretty interesting. It's not something

156
00:08:08,800 --> 00:08:11,200
Speaker 1: i'd really thought about before. All Right, well, here's a

157
00:08:11,320 --> 00:08:14,160
Speaker 1: robot that you might have a little bit less sympathy for,

158
00:08:14,320 --> 00:08:17,000
Speaker 1: so you remember the carnivorous furniture we read about And

159
00:08:17,040 --> 00:08:20,000
Speaker 1: this was a while back, but you remember this, right, Yeah.

160
00:08:20,360 --> 00:08:22,520
Speaker 1: I feel like it was some European designer was something,

161
00:08:22,520 --> 00:08:25,600
Speaker 1: and they'd created these porch lamps that were, um, they

162
00:08:25,600 --> 00:08:28,840
Speaker 1: were like powered by catching and digesting insects and flies.

163
00:08:29,000 --> 00:08:32,480
Speaker 1: It was a little gruesome for my taste. Well, if

164
00:08:32,520 --> 00:08:34,760
Speaker 1: that was gruesome for you, this is definitely not going

165
00:08:34,800 --> 00:08:38,080
Speaker 1: to ease your anxieties. And this comes from two thousand nine.

166
00:08:38,120 --> 00:08:40,640
Speaker 1: There was a Wired article and they reported that the

167
00:08:40,640 --> 00:08:44,320
Speaker 1: Department of Defense was funding battlefield robots that could power

168
00:08:44,360 --> 00:08:49,520
Speaker 1: themselves by collecting organic matter. Now that organic matter included

169
00:08:50,000 --> 00:08:54,600
Speaker 1: human corpses, So that's just great. We're gonna have these

170
00:08:54,640 --> 00:08:57,120
Speaker 1: flesh powered battle bots on the fields now, I mean,

171
00:08:57,160 --> 00:08:59,120
Speaker 1: it's true, that's exactly what they are. And it's like

172
00:08:59,160 --> 00:09:02,160
Speaker 1: these scientists have never watched a sci fi horror movie.

173
00:09:02,200 --> 00:09:05,920
Speaker 1: And it feels pretty predictable how your war might end

174
00:09:05,960 --> 00:09:08,600
Speaker 1: in all of this, And and honestly, the worst part

175
00:09:08,640 --> 00:09:11,440
Speaker 1: is that the project is called let's see, it's called

176
00:09:12,040 --> 00:09:17,360
Speaker 1: energetically Autonomous Tactical Robot or EATER for short. Yeah, I'm

177
00:09:17,400 --> 00:09:19,280
Speaker 1: not sure how much I love that fact, I feel

178
00:09:19,280 --> 00:09:23,400
Speaker 1: like you're just viewing my nightmares. But speaking of human

179
00:09:23,520 --> 00:09:26,240
Speaker 1: robot relations, I'd like you to know that we've got

180
00:09:26,240 --> 00:09:28,720
Speaker 1: work to do on both sides of that equation. And

181
00:09:29,080 --> 00:09:31,959
Speaker 1: this comes from the story of a little robot called Hitchbot.

182
00:09:32,440 --> 00:09:35,440
Speaker 1: So Hitchbot was built around two thousand fifteen, and it

183
00:09:35,520 --> 00:09:39,120
Speaker 1: was this ultra q hitchhiking robot that's only real mission

184
00:09:39,160 --> 00:09:41,439
Speaker 1: was to see the world and find some good humanity.

185
00:09:41,640 --> 00:09:43,560
Speaker 1: I mean, that's a nice idea. I'm not sure I

186
00:09:43,640 --> 00:09:47,680
Speaker 1: understand exactly what that means. So how did hitch Bot work. Well,

187
00:09:47,800 --> 00:09:51,400
Speaker 1: it was a Canadian creation and it was solar powered. Um,

188
00:09:51,640 --> 00:09:54,480
Speaker 1: it had led lights, it wore a smile all the time,

189
00:09:54,559 --> 00:09:57,640
Speaker 1: and had these rain boots and uh, hitchpot was about

190
00:09:57,679 --> 00:10:00,360
Speaker 1: the size of a small child. So it was pretty

191
00:10:00,400 --> 00:10:02,959
Speaker 1: hard not to love this cute robot. And it was

192
00:10:03,000 --> 00:10:04,640
Speaker 1: also built to be light enough that you could just

193
00:10:04,679 --> 00:10:06,400
Speaker 1: pick it up and buckle it into your back seat.

194
00:10:06,960 --> 00:10:08,800
Speaker 1: The whole idea was it would take photos along the

195
00:10:08,800 --> 00:10:11,000
Speaker 1: way to document the trip, and it had a few

196
00:10:11,000 --> 00:10:13,080
Speaker 1: small benefits for the people who picked it up, like

197
00:10:13,400 --> 00:10:16,360
Speaker 1: it came with a cigarette lighter and adapter. So I

198
00:10:16,440 --> 00:10:18,440
Speaker 1: guess you could charge your phone from it or whatever.

199
00:10:18,559 --> 00:10:20,840
Speaker 1: But the idea was like the creators would put a

200
00:10:20,840 --> 00:10:23,520
Speaker 1: destination on hitchpot. So one of the destinations with like

201
00:10:23,559 --> 00:10:25,920
Speaker 1: San Francisco, and they tell people to pick him up

202
00:10:25,960 --> 00:10:27,800
Speaker 1: and then drop them off somewhere you thought someone else

203
00:10:27,840 --> 00:10:31,800
Speaker 1: might pick him up. Really easy directions, right, So hitchbot

204
00:10:31,840 --> 00:10:35,760
Speaker 1: traveled across Canada, he took a trip across Europe, and

205
00:10:35,800 --> 00:10:39,160
Speaker 1: then he came to the US. That sounds like a

206
00:10:39,160 --> 00:10:43,720
Speaker 1: setup for something. So what happened exactly? So hitchpot had

207
00:10:43,760 --> 00:10:46,280
Speaker 1: this great bucket list he wanted to see before you

208
00:10:46,320 --> 00:10:48,120
Speaker 1: got to San Francisco. You want to see like jazz

209
00:10:48,120 --> 00:10:49,960
Speaker 1: in New Orleans. You wanted to see the magic of

210
00:10:49,960 --> 00:10:53,160
Speaker 1: Disney World, various other landmarks. And he started his trip

211
00:10:53,200 --> 00:10:56,040
Speaker 1: well from Massachusetts. But then two weeks later he was

212
00:10:56,080 --> 00:11:01,719
Speaker 1: found dismembered in Philadelphia. Oh no, that's terrible. Yeah, I mean,

213
00:11:01,760 --> 00:11:04,400
Speaker 1: it's doubly sad. When you read his goal on his

214
00:11:04,440 --> 00:11:08,160
Speaker 1: home page, it reads, quote, I hope that my hitchhiking

215
00:11:08,200 --> 00:11:10,920
Speaker 1: trip will allow me to meet many interesting people, see

216
00:11:10,960 --> 00:11:14,920
Speaker 1: beautiful places, and learn more about humanity. So I mean,

217
00:11:15,080 --> 00:11:16,480
Speaker 1: I guess the lesson is you're not supposed to go

218
00:11:16,559 --> 00:11:22,040
Speaker 1: to Philly to learn about humanity, but apparently not well.

219
00:11:22,120 --> 00:11:24,959
Speaker 1: Speaking about robots learning, I found this great article and

220
00:11:25,040 --> 00:11:28,040
Speaker 1: news scientists, and it was called eight hilarious ways AI

221
00:11:28,120 --> 00:11:31,040
Speaker 1: has outsmarted us to get the job done. It's a

222
00:11:31,120 --> 00:11:33,880
Speaker 1: really fun piece because you know, we mostly only think

223
00:11:33,920 --> 00:11:37,120
Speaker 1: that humans and animals can come up with creative solutions,

224
00:11:37,120 --> 00:11:39,480
Speaker 1: and I really don't think about robots being able to

225
00:11:39,520 --> 00:11:43,000
Speaker 1: do that. But these evolutionary algorithms do the same thing.

226
00:11:43,080 --> 00:11:45,920
Speaker 1: And as the computers are trying to optimize tasks based

227
00:11:45,960 --> 00:11:48,920
Speaker 1: on the rules you give them, it's pretty funny because

228
00:11:48,960 --> 00:11:51,000
Speaker 1: they actually end up failing in some of the most

229
00:11:51,120 --> 00:11:53,760
Speaker 1: ridiculous ways that are that are great to read about.

230
00:11:54,040 --> 00:11:57,360
Speaker 1: So I already like this, but I'm curious about the specifics.

231
00:11:57,600 --> 00:11:59,000
Speaker 1: Can you give me a couple Well, one of the

232
00:11:59,040 --> 00:12:01,640
Speaker 1: scientists who has interview put it this way quote, if

233
00:12:01,679 --> 00:12:05,560
Speaker 1: you close one loophole, the artificial intelligence finds another. To

234
00:12:05,640 --> 00:12:08,920
Speaker 1: some extent, it feels like you're wily coyote with the roadrunner.

235
00:12:09,040 --> 00:12:11,800
Speaker 1: You set up these wildly complex systems that fail in

236
00:12:11,840 --> 00:12:15,320
Speaker 1: ways you didn't expect. So you know, for instance, he

237
00:12:15,320 --> 00:12:18,160
Speaker 1: has the computer to create a creature that jumps really high,

238
00:12:18,280 --> 00:12:20,800
Speaker 1: and the way it measured how high the jump was

239
00:12:20,800 --> 00:12:23,640
Speaker 1: was by how high the creature's foot was off the ground.

240
00:12:24,000 --> 00:12:26,400
Speaker 1: So instead of trying to train the creature to jump

241
00:12:26,480 --> 00:12:29,720
Speaker 1: this wiley AI just had this creature do a somersault

242
00:12:29,760 --> 00:12:31,679
Speaker 1: on the ground and then kick its feet up mid

243
00:12:31,800 --> 00:12:34,040
Speaker 1: role and so it's just you imagine, this is pretty

244
00:12:34,040 --> 00:12:36,960
Speaker 1: fun to see. But there was also a different simulation

245
00:12:37,000 --> 00:12:40,600
Speaker 1: where the AI was supposed to create the fastest possible creature,

246
00:12:41,160 --> 00:12:44,000
Speaker 1: but instead of designing a snake or like some aerodynamic

247
00:12:44,080 --> 00:12:47,040
Speaker 1: hawk or something like that, it basically spent all its

248
00:12:47,040 --> 00:12:50,800
Speaker 1: time building this giant creature and then all they did

249
00:12:51,160 --> 00:12:53,600
Speaker 1: was push it over because you know, the velocity the

250
00:12:53,720 --> 00:12:56,640
Speaker 1: fall was faster than anything else that it could possibly

251
00:12:57,400 --> 00:13:00,520
Speaker 1: That's really interesting. Yeah, yeah, I was trying to even

252
00:13:00,520 --> 00:13:02,679
Speaker 1: think like how they would screw that one up. But honestly,

253
00:13:02,720 --> 00:13:05,920
Speaker 1: I think my favorite way that AI failed was when

254
00:13:05,920 --> 00:13:09,160
Speaker 1: it was tasked with limiting a computer's processor usage. So

255
00:13:09,280 --> 00:13:13,480
Speaker 1: the solution the robot just turned the computer off. I

256
00:13:13,559 --> 00:13:17,600
Speaker 1: love that actually, So here's a click fact. It's in

257
00:13:17,640 --> 00:13:20,120
Speaker 1: a similar vein. I I was looking up a cutting

258
00:13:20,200 --> 00:13:22,320
Speaker 1: edge story on how far robots had come in two

259
00:13:22,320 --> 00:13:26,160
Speaker 1: thousand seven and uh scientists in NYC in Japan designed

260
00:13:26,160 --> 00:13:29,040
Speaker 1: this really cute wine bot and the idea was that

261
00:13:29,080 --> 00:13:31,440
Speaker 1: you could feed it wine, cheese, or meat, those three

262
00:13:31,440 --> 00:13:33,640
Speaker 1: things and it would identify the food for you, kind

263
00:13:33,640 --> 00:13:37,760
Speaker 1: of like a personal somalia or a metal gastronomist, that's

264
00:13:37,760 --> 00:13:40,160
Speaker 1: what they called it. And it actually do it without

265
00:13:40,200 --> 00:13:43,240
Speaker 1: tasting the food too, So you can actually put a

266
00:13:43,240 --> 00:13:45,560
Speaker 1: wine bottle in front of it and the robot could

267
00:13:45,559 --> 00:13:48,240
Speaker 1: tell if the wine was authentic and the beverage matched

268
00:13:48,240 --> 00:13:51,560
Speaker 1: the label outside the bottle. That would have been good

269
00:13:51,600 --> 00:13:54,800
Speaker 1: for our old wine frauds episode. Yeah, But you know

270
00:13:54,840 --> 00:13:57,880
Speaker 1: what's interesting is that after impressing journalists with the demonstration,

271
00:13:58,000 --> 00:14:00,160
Speaker 1: this cameraman put his hand up to the robot up

272
00:14:00,480 --> 00:14:03,800
Speaker 1: and the robot identified the human hand is bacon. So

273
00:14:04,000 --> 00:14:06,600
Speaker 1: apparently the tech fact then still had a little ways

274
00:14:06,640 --> 00:14:09,960
Speaker 1: to go. Wow, that would be unfortunate if your hand

275
00:14:10,000 --> 00:14:12,640
Speaker 1: tasted like or smelled like bacon. Rather, I think that

276
00:14:12,679 --> 00:14:15,200
Speaker 1: would just be be torture. But you know, also I'm

277
00:14:15,240 --> 00:14:17,920
Speaker 1: a little curious because cannibals have always said humans taste

278
00:14:17,920 --> 00:14:20,360
Speaker 1: like chickens, so maybe they're a little off. I know

279
00:14:20,960 --> 00:14:23,760
Speaker 1: my feeling is that, like maybe bacon So delicious and

280
00:14:23,840 --> 00:14:26,000
Speaker 1: cannibals are just trying to protect their interests and keep

281
00:14:26,080 --> 00:14:28,560
Speaker 1: hipsters away from human meat. I I don't know, But

282
00:14:29,200 --> 00:14:31,240
Speaker 1: what I do know is we've got two more robot

283
00:14:31,240 --> 00:14:47,200
Speaker 1: facts to give you right after this break. Welcome back

284
00:14:47,200 --> 00:14:50,240
Speaker 1: to Part Time Genius. We're talking robots or mango. I

285
00:14:50,240 --> 00:14:52,760
Speaker 1: don't know if you knew this. Robots as people used

286
00:14:52,800 --> 00:14:55,000
Speaker 1: to call them back in the fifties. Is that true?

287
00:14:55,040 --> 00:14:57,680
Speaker 1: That's how it was pronounced. Yeah. I was actually watching

288
00:14:57,680 --> 00:15:00,160
Speaker 1: this old documentary and it was about how scientists we're

289
00:15:00,240 --> 00:15:04,120
Speaker 1: working on robot butlers of the future, and the narrator

290
00:15:04,200 --> 00:15:07,200
Speaker 1: and even the people on the screen kept calling them robots.

291
00:15:07,200 --> 00:15:10,840
Speaker 1: For somebody, that's super weird. So I feel like we

292
00:15:10,840 --> 00:15:14,480
Speaker 1: should bring that fact. But what's your last fact about robots? Well,

293
00:15:14,520 --> 00:15:17,640
Speaker 1: I know we've talked about humans hijacking animals in the past,

294
00:15:17,680 --> 00:15:19,600
Speaker 1: and we did this story a while back on how

295
00:15:19,640 --> 00:15:22,480
Speaker 1: scientists could manipulate cockroaches, you know, to get them to

296
00:15:22,520 --> 00:15:24,840
Speaker 1: move right or left with these I guess it was

297
00:15:24,880 --> 00:15:28,000
Speaker 1: with electrical pulses and and we can actually do the

298
00:15:28,080 --> 00:15:31,360
Speaker 1: same thing with beatles as well, yeah, I've seen that

299
00:15:31,400 --> 00:15:33,720
Speaker 1: you can actually put a um tiny camera on a

300
00:15:33,720 --> 00:15:36,560
Speaker 1: beetle now, which is pretty amazing, that's right, And so

301
00:15:36,680 --> 00:15:38,880
Speaker 1: to think about how that might be used. You know,

302
00:15:38,920 --> 00:15:41,920
Speaker 1: the advantage of a swarm of beetles oversay, like you

303
00:15:41,920 --> 00:15:44,360
Speaker 1: know this army of drones is, is that the beatles

304
00:15:44,360 --> 00:15:47,280
Speaker 1: are used to flying in bad weather, so they naturally

305
00:15:47,320 --> 00:15:50,280
Speaker 1: know where to feed themselves, how to rest, and so

306
00:15:50,320 --> 00:15:52,280
Speaker 1: because of that, you don't have to worry about solar

307
00:15:52,320 --> 00:15:55,040
Speaker 1: panels not working or a battery that might run out,

308
00:15:55,080 --> 00:15:57,720
Speaker 1: and you know, they're obviously super nimble, and they're just

309
00:15:57,960 --> 00:16:01,480
Speaker 1: way more cost effective too. But the weirdest animal I've

310
00:16:01,480 --> 00:16:04,480
Speaker 1: read about humans using to their advantage are turtles, which

311
00:16:04,600 --> 00:16:08,320
Speaker 1: obviously nothing like beetles, and apparently scientists from Korea have

312
00:16:08,440 --> 00:16:12,280
Speaker 1: created these parasitic robots that live on top of turtles

313
00:16:12,840 --> 00:16:15,680
Speaker 1: and kind of like your tomato robot, it will actually

314
00:16:15,800 --> 00:16:18,480
Speaker 1: feed the turtle snacks when they move in the right direction.

315
00:16:19,160 --> 00:16:21,680
Speaker 1: So I totally get insects. It's creepy that you can

316
00:16:21,680 --> 00:16:24,360
Speaker 1: turn them into cyborgs and control them, but you know,

317
00:16:24,520 --> 00:16:26,960
Speaker 1: they can get anywhere. It makes sense. They can like

318
00:16:27,000 --> 00:16:29,720
Speaker 1: fly and crawl wherever. But why do you want to

319
00:16:29,760 --> 00:16:33,040
Speaker 1: control a turtle? Well, I mean the whole ideas that

320
00:16:33,080 --> 00:16:35,840
Speaker 1: there are going to be disasters that happen from time

321
00:16:35,880 --> 00:16:39,920
Speaker 1: to time. So but this like floods or spills or earthquakes,

322
00:16:39,960 --> 00:16:42,560
Speaker 1: and we just can't get to these spots very fast.

323
00:16:42,600 --> 00:16:45,560
Speaker 1: But if you hijack a turtle population in the area,

324
00:16:45,720 --> 00:16:48,840
Speaker 1: or any animal really, and then you survey the land

325
00:16:48,880 --> 00:16:52,479
Speaker 1: by using them as your rover, it actually becomes really beneficial.

326
00:16:52,920 --> 00:16:55,640
Speaker 1: And unlike the bugs, which you have to outfit with cameras,

327
00:16:55,880 --> 00:16:58,720
Speaker 1: if you're just dropping these parasite robots on land and

328
00:16:58,760 --> 00:17:00,960
Speaker 1: they each find a turtle, right, it ends up being

329
00:17:01,040 --> 00:17:04,600
Speaker 1: much more efficient. That's pretty crazy, Okay, So here's my

330
00:17:04,680 --> 00:17:07,760
Speaker 1: last robot fact. In two thousand eight, a post STOC

331
00:17:07,840 --> 00:17:11,000
Speaker 1: fellow at Hakkaido University wanted to figure out if robot

332
00:17:11,040 --> 00:17:13,960
Speaker 1: comedians can make people laugh. So he had people read

333
00:17:14,000 --> 00:17:16,720
Speaker 1: a bunch of text message jokes and then he had

334
00:17:16,880 --> 00:17:20,879
Speaker 1: robots deliver those same jokes. So what happened? Where the

335
00:17:21,000 --> 00:17:23,720
Speaker 1: where the robots actually funny? Yeah, I mean they were

336
00:17:23,800 --> 00:17:27,080
Speaker 1: marginally funnier. On a scale of one to five. The

337
00:17:27,080 --> 00:17:29,720
Speaker 1: text jokes scored a two point three, while the average

338
00:17:29,800 --> 00:17:32,720
Speaker 1: robot comedian landed a two point eight. Oh that's not bad.

339
00:17:32,760 --> 00:17:35,840
Speaker 1: I mean that's half a point more. It's pretty good. Actually, yeah,

340
00:17:35,960 --> 00:17:38,359
Speaker 1: But then he evolved the robots to make them even

341
00:17:38,480 --> 00:17:43,520
Speaker 1: funnier by adding fart noises. Of course, the combination of

342
00:17:43,640 --> 00:17:47,960
Speaker 1: robots making jokes and fart noises drove the numbers even higher,

343
00:17:47,960 --> 00:17:50,760
Speaker 1: and the moral was pretty easy to figure out, like,

344
00:17:50,880 --> 00:17:54,320
Speaker 1: robot comedians can be funny, but gassy robot comedians are

345
00:17:54,440 --> 00:17:59,360
Speaker 1: even funnier. That's some pretty good science, you know. I

346
00:17:59,400 --> 00:18:01,840
Speaker 1: feel like we had some good facts here, but I

347
00:18:01,840 --> 00:18:04,159
Speaker 1: feel like what you just closed with with some pretty

348
00:18:04,320 --> 00:18:07,480
Speaker 1: solid science and the fact that even robots can figure

349
00:18:07,480 --> 00:18:09,880
Speaker 1: out that fart noises are funny, I feel like I'm

350
00:18:09,880 --> 00:18:12,000
Speaker 1: gonna have to give you the trophy today, Mango. Thank

351
00:18:12,040 --> 00:18:14,399
Speaker 1: you so much, well, and thank you guys for listening.

352
00:18:14,440 --> 00:18:16,440
Speaker 1: We'll be back with a full length episode tomorrow