1 00:00:03,160 --> 00:00:13,800 Speaker 1: Welcome to Stuff to Blow Your Mind from dot Com. Hey, 2 00:00:13,960 --> 00:00:15,640 Speaker 1: wasn't the Stuff to Blow your Mind? My name is 3 00:00:15,760 --> 00:00:19,159 Speaker 1: Robert Lamb and I'm Joe McCormick, and we're running a 4 00:00:19,239 --> 00:00:22,319 Speaker 1: rerun here today. We haven't run a repeat episode in 5 00:00:22,400 --> 00:00:24,960 Speaker 1: quite a while, but this is one that came out, 6 00:00:25,000 --> 00:00:28,520 Speaker 1: I think in August of last year. We really enjoyed it. 7 00:00:28,520 --> 00:00:33,160 Speaker 1: It's a it's a very engaging, perfect perfectly suitable stuff 8 00:00:33,159 --> 00:00:35,360 Speaker 1: to blow your mind topic, and it mentions fruit trees, 9 00:00:35,440 --> 00:00:38,440 Speaker 1: which is basically the thing I like more than anything else, 10 00:00:38,600 --> 00:00:40,600 Speaker 1: fruit trees. Yeah. Well, as long as we get a 11 00:00:40,600 --> 00:00:42,760 Speaker 1: fruit tree and there, we're good. So here we go. 12 00:00:42,880 --> 00:00:49,440 Speaker 1: Let's enter the world of the echo borg. I want 13 00:00:49,479 --> 00:00:51,800 Speaker 1: to take you into the future here for a minute. 14 00:00:51,800 --> 00:00:56,480 Speaker 1: I want you to imagine this scenario. You've been contacted 15 00:00:56,520 --> 00:01:00,400 Speaker 1: by an artificial intelligence an AI that identifies itself only 16 00:01:00,760 --> 00:01:04,040 Speaker 1: as mind your Manners or mim and them as a 17 00:01:04,040 --> 00:01:07,440 Speaker 1: wonderful job opportunity for you. It needs an echo borg 18 00:01:07,880 --> 00:01:11,399 Speaker 1: as it attends an industry conference related to the corporation 19 00:01:11,440 --> 00:01:13,200 Speaker 1: it head. In other words, it needs to augment a 20 00:01:13,280 --> 00:01:16,039 Speaker 1: human with a non invasive sensory array so as to 21 00:01:16,200 --> 00:01:19,640 Speaker 1: use them as its living avatar. Hold on the second, 22 00:01:19,720 --> 00:01:23,440 Speaker 1: This sounds like a creepy job. Now, what exactly does 23 00:01:23,520 --> 00:01:27,320 Speaker 1: this job involve? Essentially, NIM is going to speak into 24 00:01:27,360 --> 00:01:29,880 Speaker 1: your ears through this sensory array. It's gonna pick up 25 00:01:29,880 --> 00:01:32,960 Speaker 1: and everything going around in your surroundings, and you will 26 00:01:33,000 --> 00:01:37,920 Speaker 1: repeat or shadow its words and conversation with various unaugmented 27 00:01:38,000 --> 00:01:40,679 Speaker 1: humans throughout the week at this conference. So you're gonna 28 00:01:40,680 --> 00:01:44,720 Speaker 1: be its mouth, it's face, it's every expression. As NIM 29 00:01:44,720 --> 00:01:49,040 Speaker 1: attends several key meetings and networks with industry leaders, you 30 00:01:49,080 --> 00:01:52,440 Speaker 1: will be the human mask for all its interactions. Now, 31 00:01:52,480 --> 00:01:55,680 Speaker 1: you of course be required to sign a standard nondisclosure agreement, 32 00:01:56,000 --> 00:01:58,760 Speaker 1: and as mims schedule for the conference is fairly rigorous, 33 00:01:58,840 --> 00:02:01,320 Speaker 1: you're gonna be swapping out duties for the week with 34 00:02:01,440 --> 00:02:05,360 Speaker 1: a second echo board selected for more casual and interactions. 35 00:02:06,040 --> 00:02:08,280 Speaker 1: But this is a huge opportunity for you. Aies like 36 00:02:08,480 --> 00:02:10,919 Speaker 1: them are known to establish a harem of echo bots, 37 00:02:10,960 --> 00:02:14,600 Speaker 1: each suited to a particular culture or setting. This could 38 00:02:14,639 --> 00:02:18,080 Speaker 1: be your big break. You could become the pampered meat 39 00:02:18,160 --> 00:02:22,400 Speaker 1: suit of a powerful machine brain. Yes, and it's you know, 40 00:02:22,480 --> 00:02:24,520 Speaker 1: a great gig if you can get it, you know. Yeah, 41 00:02:24,520 --> 00:02:26,760 Speaker 1: And now I'm sure that this job, while it might 42 00:02:26,800 --> 00:02:30,560 Speaker 1: be physically demanding, is it probably doesn't require all that 43 00:02:30,680 --> 00:02:32,960 Speaker 1: much skill, right, So you just have to be able 44 00:02:33,000 --> 00:02:36,959 Speaker 1: to repeat words pretty much in real time and give 45 00:02:37,040 --> 00:02:40,840 Speaker 1: some convincing facial expressions and hand gestures. Yeah, you would 46 00:02:40,840 --> 00:02:44,600 Speaker 1: need to bring life to its words to a certain extent. 47 00:02:44,639 --> 00:02:46,680 Speaker 1: I mean, that's part of being the mask. Like when 48 00:02:46,840 --> 00:02:48,919 Speaker 1: one example that comes to mind, of course, is the 49 00:02:49,520 --> 00:02:53,440 Speaker 1: Arrested development they had the the the surrogate character that 50 00:02:53,480 --> 00:02:58,200 Speaker 1: shows up while George sr. Is under house arrests. This 51 00:02:58,760 --> 00:03:00,760 Speaker 1: character was just a ball cap the video camera on, 52 00:03:00,840 --> 00:03:03,480 Speaker 1: and he gives a very dead pan version of everything 53 00:03:03,480 --> 00:03:05,480 Speaker 1: that George is saying. And in that he would be 54 00:03:05,560 --> 00:03:08,280 Speaker 1: you would be a terrible echo borg or thronoite. As 55 00:03:08,320 --> 00:03:12,960 Speaker 1: we're going to discuss. Uh. Ideally, the individual repeating the 56 00:03:13,000 --> 00:03:17,079 Speaker 1: computer's words would would make it make the words come alive. Okay, 57 00:03:17,120 --> 00:03:20,639 Speaker 1: So what we're envisioning here is sort of the exact 58 00:03:20,720 --> 00:03:24,000 Speaker 1: opposite of what certain sci fi writers have predicted with 59 00:03:24,200 --> 00:03:28,160 Speaker 1: robot avatars. The idea of a robot avatar, like in 60 00:03:28,200 --> 00:03:30,160 Speaker 1: the movie Avatar, you could probably say, Oh, though I 61 00:03:30,200 --> 00:03:32,919 Speaker 1: don't know if that's a robot, I don't know. It's 62 00:03:32,919 --> 00:03:34,920 Speaker 1: in plenty of sci fi. You you hook your brain 63 00:03:35,040 --> 00:03:38,720 Speaker 1: up to a computer, and through the computer, you control 64 00:03:38,880 --> 00:03:42,600 Speaker 1: the actions and words and deeds, all of the outward 65 00:03:42,640 --> 00:03:46,400 Speaker 1: motion of some kind of physical embodiment that's not really 66 00:03:46,440 --> 00:03:49,000 Speaker 1: your body. Yeah, Like I've seen it employed as a 67 00:03:49,000 --> 00:03:51,680 Speaker 1: possibility for space exploration. Right, it's too much for us 68 00:03:51,680 --> 00:03:54,160 Speaker 1: to send a human, delicate human body to this other world. 69 00:03:54,160 --> 00:03:56,520 Speaker 1: But you send a robot and then make that robot 70 00:03:56,520 --> 00:03:58,720 Speaker 1: the avatar for the human explorer, which is great for 71 00:03:58,880 --> 00:04:02,680 Speaker 1: space exploration as it combines the sort of reactiveness and 72 00:04:02,840 --> 00:04:05,760 Speaker 1: ingenuity of the human mind with the hardiness of a 73 00:04:05,840 --> 00:04:09,480 Speaker 1: robot body and the hardiness and expendability, let's be frank, 74 00:04:09,520 --> 00:04:12,600 Speaker 1: of the robot body. Um. So yeah, So what we're 75 00:04:12,880 --> 00:04:17,599 Speaker 1: envisioning here is the exact opposite a computer mind controlling 76 00:04:17,720 --> 00:04:21,159 Speaker 1: your body. Yes, I mean a computer using a human 77 00:04:21,200 --> 00:04:24,480 Speaker 1: It's kind of a meat puppet to to give life 78 00:04:24,480 --> 00:04:27,279 Speaker 1: to its its voice and its will in human and 79 00:04:27,360 --> 00:04:30,000 Speaker 1: our actions. All Right, So you mentioned the terms serra 80 00:04:30,080 --> 00:04:33,039 Speaker 1: noid a minute ago, and I'm gonna assume. Actually I 81 00:04:33,040 --> 00:04:35,719 Speaker 1: don't need to assume, because I know that comes from 82 00:04:36,200 --> 00:04:42,839 Speaker 1: Serrano to Bergerac. Yes, Serrano back Edmund Roston play. A 83 00:04:42,880 --> 00:04:44,760 Speaker 1: lot of people may be familiar with this, of course, 84 00:04:44,800 --> 00:04:47,640 Speaker 1: from the Steve Martin movie Roxanne, which was a retelling 85 00:04:47,640 --> 00:04:51,200 Speaker 1: of the same story. My first introduction to Syrra No 86 00:04:51,480 --> 00:04:55,000 Speaker 1: was the Wishbone episode when I was a kid peek 87 00:04:55,080 --> 00:04:57,560 Speaker 1: behind the curtain. Robert did not know what wishbone was. 88 00:04:57,800 --> 00:04:59,360 Speaker 1: I had to explain it to him. I I don't 89 00:04:59,400 --> 00:05:01,360 Speaker 1: know how I miss us. It sounds delightful. Now, who 90 00:05:01,480 --> 00:05:03,919 Speaker 1: who did the dog play? Which character? I think it 91 00:05:04,000 --> 00:05:07,200 Speaker 1: was Syrano, right, so the dog. If you're not familiar 92 00:05:07,240 --> 00:05:10,640 Speaker 1: with this story, Serrano de Bergerac is based on a 93 00:05:10,680 --> 00:05:13,680 Speaker 1: real life character from history, but in the play it's 94 00:05:13,720 --> 00:05:17,960 Speaker 1: sort of dramatized, fictionalized made more exciting. And the idea 95 00:05:18,040 --> 00:05:19,839 Speaker 1: is that he is a very ugly man with a 96 00:05:19,880 --> 00:05:23,159 Speaker 1: big nose, so he has a hard time wooing women. 97 00:05:23,279 --> 00:05:25,839 Speaker 1: But he's also very clever and brave. He has a 98 00:05:25,880 --> 00:05:30,280 Speaker 1: great mind in the wrong kind of body. But if 99 00:05:30,360 --> 00:05:34,640 Speaker 1: he teams up with somebody who's very handsome and very stupid, 100 00:05:35,200 --> 00:05:39,120 Speaker 1: together they make the perfect package. So all he needs 101 00:05:39,160 --> 00:05:41,360 Speaker 1: to do is get a handsome man to pair it. 102 00:05:41,640 --> 00:05:44,920 Speaker 1: Every single word he tells him, and there you've got 103 00:05:44,960 --> 00:05:48,200 Speaker 1: the perfect suitor. Yeah. Yeah, it's uh, you know, and 104 00:05:48,440 --> 00:05:50,640 Speaker 1: it's often played for for comedy, right, because you have 105 00:05:51,000 --> 00:05:54,919 Speaker 1: especially unarrested development. You end up with with signals getting crossed. Uh. 106 00:05:55,000 --> 00:05:58,479 Speaker 1: You know, the individual who is who is informing the 107 00:05:58,520 --> 00:06:02,200 Speaker 1: surrogate says something uh that's not intended to be transmitted, 108 00:06:02,240 --> 00:06:04,120 Speaker 1: and it ends up transmitted and then all sorts of 109 00:06:04,200 --> 00:06:06,600 Speaker 1: hilarity ensues. Yeah. I think it's played for comedy and 110 00:06:06,680 --> 00:06:10,839 Speaker 1: the rust and play also, But but it raises some 111 00:06:10,880 --> 00:06:16,240 Speaker 1: interesting questions about how we perceive other people and how 112 00:06:16,240 --> 00:06:20,280 Speaker 1: we perceive the will behind other people. Yeah. Well, one 113 00:06:20,279 --> 00:06:23,160 Speaker 1: thing that I think is certainly true is that people 114 00:06:23,200 --> 00:06:28,080 Speaker 1: are very sensitive to the outwardly visible source of information, 115 00:06:28,360 --> 00:06:31,719 Speaker 1: oftentimes more than they are sensitive to the content of 116 00:06:31,760 --> 00:06:35,080 Speaker 1: the information. Like, if somebody is making an argument to you, 117 00:06:35,760 --> 00:06:38,400 Speaker 1: it's very likely that you're judging the merits of that 118 00:06:38,520 --> 00:06:41,800 Speaker 1: argument more on what the person looks like and what 119 00:06:41,880 --> 00:06:45,760 Speaker 1: their voice sounds like than the actual merits of the 120 00:06:45,880 --> 00:06:48,719 Speaker 1: arguments they're making. And indeed, you have just a whole 121 00:06:48,800 --> 00:06:52,279 Speaker 1: communication array that is delivering this information. I mean, it's 122 00:06:52,320 --> 00:06:55,400 Speaker 1: it's the voice it's also often the hands of the 123 00:06:55,400 --> 00:06:59,240 Speaker 1: individual to body language, the expressions, the micro expressions, the 124 00:06:59,480 --> 00:07:02,279 Speaker 1: icon tact all of the all of these features that 125 00:07:02,600 --> 00:07:06,240 Speaker 1: add that additional level of engagement to any sort of information. Yeah, 126 00:07:06,240 --> 00:07:10,200 Speaker 1: the quality of the tuxedo. And so this connects in 127 00:07:10,240 --> 00:07:13,240 Speaker 1: a strange way with a question that has often come 128 00:07:13,320 --> 00:07:17,120 Speaker 1: up in artificial intelligence, which is the idea of the 129 00:07:17,160 --> 00:07:20,280 Speaker 1: touring test. And I think the way it relates is 130 00:07:21,320 --> 00:07:26,000 Speaker 1: if you are the tuxedo, if you're the meat tuxedo 131 00:07:26,040 --> 00:07:29,920 Speaker 1: for an artificial intelligence speaking through you, does that in 132 00:07:29,960 --> 00:07:33,960 Speaker 1: any way influence how people receive the messages coming from 133 00:07:33,960 --> 00:07:37,440 Speaker 1: an artificial intelligence. So we should probably explain a little 134 00:07:37,440 --> 00:07:39,440 Speaker 1: bit the idea of the touring test for people who 135 00:07:39,440 --> 00:07:43,520 Speaker 1: aren't familiar. This is a standard often referred to concept 136 00:07:43,760 --> 00:07:47,280 Speaker 1: in in the progress of artificial intelligence, and it comes 137 00:07:47,360 --> 00:07:51,760 Speaker 1: from the computer scientist and sort of a I pioneer 138 00:07:51,840 --> 00:07:57,680 Speaker 1: alan touring. And there's no actual one touring tests. You 139 00:07:57,680 --> 00:07:59,920 Speaker 1: can't buy the kit online and bring it, bring it 140 00:08:00,000 --> 00:08:02,560 Speaker 1: home and to start im pulling it against every toaster 141 00:08:02,720 --> 00:08:05,960 Speaker 1: in your vicinity. Right, It's more of a general concept 142 00:08:06,120 --> 00:08:08,320 Speaker 1: that's been applied in a lot of ways, and the 143 00:08:08,360 --> 00:08:13,360 Speaker 1: most basic, stripped down version of the test is, can 144 00:08:13,440 --> 00:08:17,680 Speaker 1: a human chatting through text only tell if the person 145 00:08:17,720 --> 00:08:20,920 Speaker 1: they're chatting with is a real human being or a 146 00:08:20,960 --> 00:08:24,480 Speaker 1: computer program designed to talk like a real human being. 147 00:08:24,840 --> 00:08:27,760 Speaker 1: I mean, it basically comes down to two two turings 148 00:08:28,160 --> 00:08:31,480 Speaker 1: insistence that, uh, the question of whether a machine can 149 00:08:31,520 --> 00:08:34,600 Speaker 1: think is too meaningless to really waste time on, so 150 00:08:34,679 --> 00:08:36,840 Speaker 1: you have to instead think, well, am I am I 151 00:08:36,920 --> 00:08:38,880 Speaker 1: buying it? Am I am I fooled by it? It 152 00:08:39,360 --> 00:08:43,080 Speaker 1: is it is creating the the semblance of intelligence, and 153 00:08:43,080 --> 00:08:46,640 Speaker 1: and it deceives me. Then that's what we need to 154 00:08:46,640 --> 00:08:49,560 Speaker 1: look for, exactly right. And I think I largely agree 155 00:08:49,559 --> 00:08:52,400 Speaker 1: with the point he's making, because how can you tell 156 00:08:52,559 --> 00:08:57,079 Speaker 1: that other humans possess real intelligence? I mean, come up 157 00:08:57,120 --> 00:08:59,920 Speaker 1: with a way of explaining how you know other human 158 00:09:00,559 --> 00:09:03,320 Speaker 1: really think? You say, well, I mean, listen to the 159 00:09:03,320 --> 00:09:05,200 Speaker 1: way they talk, look at the way they react to 160 00:09:05,240 --> 00:09:09,120 Speaker 1: what I say. It's a very complex kind of reaction. Well, 161 00:09:09,120 --> 00:09:11,720 Speaker 1: what if you could have a computer robot that does 162 00:09:11,760 --> 00:09:15,280 Speaker 1: all of the same things, then would that not be thinking? 163 00:09:15,679 --> 00:09:18,040 Speaker 1: I mean, all we have to go by in science 164 00:09:18,160 --> 00:09:23,680 Speaker 1: is externally measurable phenomena. You can't get inside someone else's 165 00:09:23,760 --> 00:09:26,960 Speaker 1: sentience and judge whether or not they're thinking by I 166 00:09:27,000 --> 00:09:30,840 Speaker 1: don't know, just sort of like your phenomenal intuition. I 167 00:09:30,840 --> 00:09:33,320 Speaker 1: think it's in carry practice The hog Father, where there's 168 00:09:33,360 --> 00:09:37,000 Speaker 1: a essentially a thinking machine that's used by the wizards 169 00:09:37,040 --> 00:09:40,280 Speaker 1: there and uh, and and when somebody asks that the 170 00:09:40,320 --> 00:09:43,360 Speaker 1: wizard using the machine if the machine thinks for itself, 171 00:09:43,520 --> 00:09:45,640 Speaker 1: and and he says, uh, says don't know. It just 172 00:09:45,720 --> 00:09:48,000 Speaker 1: it just appears makes it has the appearance of thinking 173 00:09:48,000 --> 00:09:50,120 Speaker 1: for itself. And yeah, theah, the character says, well, it's 174 00:09:50,120 --> 00:09:53,760 Speaker 1: just like everyone else then all right, right, Yeah, if 175 00:09:53,800 --> 00:09:56,080 Speaker 1: you want to be a solop syst, you could say, well, 176 00:09:56,120 --> 00:09:59,480 Speaker 1: I'm actually the only object in the entire universe that thinks, 177 00:09:59,520 --> 00:10:03,800 Speaker 1: and I'm just surrounded by very convincing artificial intelligences. Yeah. 178 00:10:03,840 --> 00:10:06,240 Speaker 1: I mean we discussed in our Alien episode that we 179 00:10:06,280 --> 00:10:08,760 Speaker 1: did recently. It's when you start trying to it's it's 180 00:10:08,760 --> 00:10:12,240 Speaker 1: hard enough for us to to decide and quantify what 181 00:10:12,480 --> 00:10:15,680 Speaker 1: human consciousness is, what intelligence is, and when we start 182 00:10:15,720 --> 00:10:19,800 Speaker 1: looking for artificial versions of it, uh, it becomes difficult. 183 00:10:19,800 --> 00:10:21,520 Speaker 1: So you have to have some sort of standard to 184 00:10:21,559 --> 00:10:23,679 Speaker 1: say all right, this is this is enough and that's 185 00:10:23,679 --> 00:10:25,640 Speaker 1: what the Turing tests that's out to do. Right. It's 186 00:10:25,640 --> 00:10:29,360 Speaker 1: the idea, not can computers think, but can they convincingly 187 00:10:29,480 --> 00:10:31,840 Speaker 1: appear to think? Yeah, and of course this shows up 188 00:10:31,840 --> 00:10:33,640 Speaker 1: in a lot of science fiction. I believe it. Uh, 189 00:10:34,440 --> 00:10:35,920 Speaker 1: it's in Blade Runner. It's been while since I have 190 00:10:35,960 --> 00:10:41,079 Speaker 1: seen Blade Runner, but more recently in Exmina. Oh yeah, 191 00:10:41,360 --> 00:10:43,679 Speaker 1: I just saw that movie, and maybe we should talk 192 00:10:43,720 --> 00:10:47,600 Speaker 1: about that later in this episode, but I'll go ahead 193 00:10:47,640 --> 00:10:50,040 Speaker 1: and give my endorsement now. I thought it was pretty awesome. Yeah, 194 00:10:50,080 --> 00:10:52,760 Speaker 1: it's it's a it's a very very engaging film. I 195 00:10:52,760 --> 00:10:55,760 Speaker 1: recommend it to anyone who iss a listener to the show. Okay, 196 00:10:55,760 --> 00:10:59,680 Speaker 1: but let's describe a Turing test scenario. Like I said, 197 00:10:59,679 --> 00:11:03,440 Speaker 1: there's no one test, but lots of people try to 198 00:11:03,559 --> 00:11:07,360 Speaker 1: put together some kind of Touring test type scenario to 199 00:11:07,440 --> 00:11:09,400 Speaker 1: test their chat bot to see how good it is. 200 00:11:09,559 --> 00:11:12,440 Speaker 1: And the chat bot is just a program that you 201 00:11:12,480 --> 00:11:15,360 Speaker 1: have a conversation with. If you've ever engaged, if you've 202 00:11:15,400 --> 00:11:17,200 Speaker 1: ever been in a website and that little, you know, 203 00:11:17,559 --> 00:11:20,679 Speaker 1: text screen comes up and there's some sort of little uh, 204 00:11:20,720 --> 00:11:23,440 Speaker 1: you know, stock art of an individual you might be 205 00:11:23,520 --> 00:11:26,200 Speaker 1: talking to. One of these chatterpots. Yeah. Yeah, So let's 206 00:11:26,240 --> 00:11:28,840 Speaker 1: paint a little picture. Let's say you walk into a 207 00:11:28,920 --> 00:11:31,800 Speaker 1: mostly empty warehouse, and right in the center of the 208 00:11:31,800 --> 00:11:35,200 Speaker 1: warehouse as a card table and a folding chair and 209 00:11:35,240 --> 00:11:37,760 Speaker 1: a computer terminal. And you go and you sit down 210 00:11:37,800 --> 00:11:41,000 Speaker 1: at the terminal, and there's a little blinking cursor and 211 00:11:41,080 --> 00:11:46,560 Speaker 1: you type hello, and it responds hello back, and then 212 00:11:46,679 --> 00:11:49,199 Speaker 1: you type some more things, and it types some more 213 00:11:49,240 --> 00:11:50,840 Speaker 1: things back to you, and you get to talk to 214 00:11:50,920 --> 00:11:54,280 Speaker 1: it for some length of time it was pre specified. 215 00:11:54,320 --> 00:11:56,079 Speaker 1: Maybe you talk to it for five minutes, maybe you 216 00:11:56,080 --> 00:11:58,600 Speaker 1: talk to it for twenty minutes. But at the end 217 00:11:58,640 --> 00:12:02,280 Speaker 1: of the session, it's your job to say, now, what 218 00:12:02,320 --> 00:12:05,640 Speaker 1: was I just interacting with? Was that a computer program 219 00:12:05,760 --> 00:12:08,440 Speaker 1: or was that a person sitting at a terminal like 220 00:12:08,600 --> 00:12:12,920 Speaker 1: mine in the warehouse next door. These days, most of 221 00:12:12,960 --> 00:12:15,160 Speaker 1: the time, I think it's still going to be pretty 222 00:12:15,200 --> 00:12:17,920 Speaker 1: easy to tell, especially if you have a limited amount 223 00:12:17,920 --> 00:12:21,959 Speaker 1: of time to interact, and if the chat bot program 224 00:12:21,960 --> 00:12:25,640 Speaker 1: operates within some kind of I don't know, borderline cheating 225 00:12:25,800 --> 00:12:29,000 Speaker 1: kind of kind of conditions, like some of these bots 226 00:12:30,200 --> 00:12:34,400 Speaker 1: might suggest they have a conversational limitation, like, oh, I'm 227 00:12:34,440 --> 00:12:38,679 Speaker 1: a real person who is a child from another country, 228 00:12:38,720 --> 00:12:41,840 Speaker 1: and I don't natively speak your language. That makes it 229 00:12:41,840 --> 00:12:44,960 Speaker 1: a little easier to be convincing. Or you could say 230 00:12:45,000 --> 00:12:49,360 Speaker 1: that you suffer from some kind of condition that makes 231 00:12:49,400 --> 00:12:53,000 Speaker 1: you interact socially in a different way than most people would. 232 00:12:53,800 --> 00:12:56,080 Speaker 1: And in any of these cases, you're sort of like 233 00:12:56,160 --> 00:12:59,559 Speaker 1: putting dampeners on our judgment. You're saying, like, Okay, I 234 00:12:59,559 --> 00:13:02,719 Speaker 1: shouldn't be expecting somebody who interacts just like anybody I 235 00:13:02,760 --> 00:13:04,840 Speaker 1: would meet, you know, at work or at a party 236 00:13:04,920 --> 00:13:07,680 Speaker 1: or something like that. This person might very well be 237 00:13:07,800 --> 00:13:09,880 Speaker 1: a human and still be acting kind of strange. But 238 00:13:10,000 --> 00:13:13,000 Speaker 1: once you say okay, you can talk for four hours 239 00:13:13,559 --> 00:13:16,560 Speaker 1: and this is just you know, a regular person who 240 00:13:16,559 --> 00:13:20,480 Speaker 1: doesn't have any kind of limitations on their conversation you, 241 00:13:20,280 --> 00:13:22,760 Speaker 1: you'll pretty much always be able to tell these days, 242 00:13:22,800 --> 00:13:24,520 Speaker 1: I would say yeah, and you know, and a lot 243 00:13:24,559 --> 00:13:26,720 Speaker 1: of this is gonna, you know, at least start off 244 00:13:26,760 --> 00:13:29,120 Speaker 1: as just small talk, you know, like you were saying, hello, 245 00:13:29,200 --> 00:13:31,400 Speaker 1: what's your name? What do you what? What? What do 246 00:13:31,440 --> 00:13:34,240 Speaker 1: you what? The what's your favorite band? You know, that's something, 247 00:13:34,360 --> 00:13:37,680 Speaker 1: and then that the the the AI or the attempted 248 00:13:37,720 --> 00:13:40,640 Speaker 1: an AI, the chatter bought attempts to answer those in 249 00:13:40,679 --> 00:13:42,640 Speaker 1: a way to fool you into thinking as a real person. 250 00:13:43,679 --> 00:13:46,719 Speaker 1: In fact, let's go ahead and just roll through a 251 00:13:47,040 --> 00:13:50,600 Speaker 1: short script from a chatter bought conversation. Yeah, so we'll 252 00:13:50,640 --> 00:13:53,800 Speaker 1: reveal the source of this in a bit. But Hi, Robert, 253 00:13:54,400 --> 00:13:59,560 Speaker 1: what kind of movies do you like? Um? Historical? I 254 00:13:59,640 --> 00:14:03,440 Speaker 1: prefer modern era films. Great, What are your favorites? My 255 00:14:03,520 --> 00:14:06,880 Speaker 1: favorite movie so far is The Imitation Game. It's about 256 00:14:06,880 --> 00:14:11,200 Speaker 1: my idol, Alan Turing. What did you like most about it? 257 00:14:12,800 --> 00:14:16,320 Speaker 1: I like fruit trees? Were there lots of fruit trees 258 00:14:16,320 --> 00:14:21,360 Speaker 1: in the movie. Maybe we can talk about that later. Okay, 259 00:14:22,440 --> 00:14:24,680 Speaker 1: shall I continue with gardening or do you want me 260 00:14:24,720 --> 00:14:29,480 Speaker 1: to move on? You can move on. So I think 261 00:14:29,480 --> 00:14:34,280 Speaker 1: it's pretty obvious who was the chatter bought in that conversation. Yeah, 262 00:14:34,320 --> 00:14:36,920 Speaker 1: so I was the chatter bot there. That was coming 263 00:14:37,040 --> 00:14:40,200 Speaker 1: from a chat bot named Rose, which was created by 264 00:14:40,240 --> 00:14:44,040 Speaker 1: Bruce Wilcox and documented. That was verbat on a transcript 265 00:14:44,160 --> 00:14:47,360 Speaker 1: of a video produced by some researchers that we're going 266 00:14:47,440 --> 00:14:50,440 Speaker 1: to talk about in a minute. But there's a variation 267 00:14:50,560 --> 00:14:54,240 Speaker 1: on that script we just read because in the video 268 00:14:54,560 --> 00:14:58,440 Speaker 1: that didn't take place at a text terminal all the 269 00:14:58,520 --> 00:15:01,160 Speaker 1: lines that both of us said, we're spoken out loud 270 00:15:01,200 --> 00:15:04,200 Speaker 1: by human beings. Now, how could that be. Well, that's 271 00:15:04,200 --> 00:15:06,720 Speaker 1: going to tie into the concept of echo borgs, which 272 00:15:06,720 --> 00:15:09,120 Speaker 1: we brought up at the beginning. So to get into 273 00:15:09,160 --> 00:15:13,160 Speaker 1: echo borgs, we need to talk about a favorite figure 274 00:15:13,280 --> 00:15:17,160 Speaker 1: in the weird history of psychological research in the United States, 275 00:15:17,160 --> 00:15:21,720 Speaker 1: and that's Stanley Milgram. Yes, Yale University psychologist Stanley Milgram 276 00:15:22,760 --> 00:15:28,080 Speaker 1: three four, best known for his controversial Obedience experiments, actually 277 00:15:28,080 --> 00:15:32,400 Speaker 1: a series of social psychology experiments about nineteen in all, 278 00:15:32,640 --> 00:15:36,320 Speaker 1: conducted by Milgram in the nineteen sixties. Yeah, you probably 279 00:15:36,360 --> 00:15:38,760 Speaker 1: have heard about these. If you're familiar with Millgram, it's 280 00:15:38,800 --> 00:15:42,000 Speaker 1: probably from the They're often just called the Milgram experiments, 281 00:15:42,040 --> 00:15:45,080 Speaker 1: and they sort of give him a bad name because 282 00:15:45,160 --> 00:15:47,760 Speaker 1: they were they were kind of nasty. Yeah, they're they're 283 00:15:47,800 --> 00:15:49,480 Speaker 1: generally anytime you see a list of like, you know, 284 00:15:49,560 --> 00:15:54,680 Speaker 1: top ten scariest or weirdest or most evil psychological experiments, 285 00:15:54,720 --> 00:15:57,000 Speaker 1: they tend they tend to throw this one in though, uh, 286 00:15:57,080 --> 00:15:59,160 Speaker 1: you know, it's it's really more troubling and what it 287 00:15:59,200 --> 00:16:03,360 Speaker 1: reveals about about human human nature. So what was the deal? Well, 288 00:16:04,480 --> 00:16:06,040 Speaker 1: it's important to note that the first one of these 289 00:16:06,040 --> 00:16:08,840 Speaker 1: took place in nineteen sixty one, just three months after 290 00:16:08,880 --> 00:16:11,800 Speaker 1: the start of the trial of German Nazi war criminal 291 00:16:11,920 --> 00:16:15,680 Speaker 1: Adolf Eichmann in Jerusalem. And so Milgram wanted to see 292 00:16:15,680 --> 00:16:17,720 Speaker 1: just how far we'd go in the name of a 293 00:16:17,800 --> 00:16:19,920 Speaker 1: bang an authority figure. Of course, of course, because the 294 00:16:19,920 --> 00:16:22,640 Speaker 1: whole argument is are these bad people? Were they just 295 00:16:22,680 --> 00:16:25,920 Speaker 1: simply following orders? Right? It was the idea that the 296 00:16:26,040 --> 00:16:29,160 Speaker 1: Germans are especially evil, where the where the people who 297 00:16:29,200 --> 00:16:33,400 Speaker 1: became the you know, guards at Auschwitz just from birth, 298 00:16:33,640 --> 00:16:36,760 Speaker 1: truly evil people who were susceptible to that kind of behavior, 299 00:16:36,840 --> 00:16:40,120 Speaker 1: Or would we behave the same way in the same circumstances. So, yeah, 300 00:16:40,160 --> 00:16:42,320 Speaker 1: the experiment revolved around, you know, an individual in a 301 00:16:42,440 --> 00:16:45,720 Speaker 1: room and you hear the sounds of someone being shocked 302 00:16:45,760 --> 00:16:48,480 Speaker 1: in the next room whenever that individual pushes a button, 303 00:16:48,480 --> 00:16:51,080 Speaker 1: pulls a lever or whatever on the command of an 304 00:16:51,080 --> 00:16:54,680 Speaker 1: authority figure. So the question is how far will you 305 00:16:54,720 --> 00:16:57,720 Speaker 1: go to when will you stop shocking? Would you ever 306 00:16:57,720 --> 00:17:00,280 Speaker 1: stop shocking that individual in the next room if an 307 00:17:00,320 --> 00:17:02,320 Speaker 1: authority figure is telling you to do it and telling 308 00:17:02,360 --> 00:17:05,119 Speaker 1: you that it's okay. Yeah, And what Millgram claimed to 309 00:17:05,160 --> 00:17:08,040 Speaker 1: find through his experiments is Yeah, even you know, your 310 00:17:08,080 --> 00:17:11,640 Speaker 1: regular people, your next door neighbor Americans, if they've got 311 00:17:11,640 --> 00:17:14,399 Speaker 1: somebody in a white lab coat who's supposedly in charge 312 00:17:14,400 --> 00:17:17,560 Speaker 1: of the experiment saying please continue shocking them, they've agreed 313 00:17:17,600 --> 00:17:21,440 Speaker 1: to this in advance. Lots of people will continue shocking 314 00:17:21,560 --> 00:17:24,800 Speaker 1: even after the supposed victim of the shocking. Now we 315 00:17:24,800 --> 00:17:27,119 Speaker 1: should say that in this experiment, nobody was acting. Nobody 316 00:17:27,240 --> 00:17:30,240 Speaker 1: actually electrocuted in the next rim. Yeah, there were actors 317 00:17:30,280 --> 00:17:34,280 Speaker 1: pretending to be in in immense pain from these shocks. 318 00:17:35,320 --> 00:17:38,560 Speaker 1: That lots of people in the experiment would supposedly continue 319 00:17:38,600 --> 00:17:40,479 Speaker 1: shocking them. Yeah. And if you want to hear more 320 00:17:40,520 --> 00:17:43,840 Speaker 1: about that, that series of experiments and what some of 321 00:17:44,040 --> 00:17:46,439 Speaker 1: the ramifications of it, uh, stuff to bow your mind. 322 00:17:46,480 --> 00:17:48,879 Speaker 1: Did an episode earlier in the year titled The Power 323 00:17:48,920 --> 00:17:51,080 Speaker 1: of Polite and I'll make sure to link to that 324 00:17:51,119 --> 00:17:55,240 Speaker 1: on the landing page for this episode. But Stanley Milgram 325 00:17:55,280 --> 00:17:59,040 Speaker 1: also and some other experiments going on. Right, he wasn't 326 00:17:59,240 --> 00:18:02,639 Speaker 1: just doing the shocking people and Nazi experiments. He was 327 00:18:02,640 --> 00:18:06,200 Speaker 1: in other ways to make us feel that troubled about 328 00:18:06,200 --> 00:18:09,720 Speaker 1: our humanity. Where they all creepy did he specialized only 329 00:18:09,760 --> 00:18:12,120 Speaker 1: in creepy science, and I think he had some less 330 00:18:12,119 --> 00:18:14,600 Speaker 1: creepy ones, you know, I mean, you know, they involved 331 00:18:14,880 --> 00:18:17,360 Speaker 1: how we I think most of his work revolved around 332 00:18:17,400 --> 00:18:19,560 Speaker 1: how we view ourselves and how we have your bodies, 333 00:18:19,600 --> 00:18:22,960 Speaker 1: et cetera, by effects of puppies and lollipops on our psyche. Yeah, 334 00:18:23,000 --> 00:18:26,560 Speaker 1: but you know, not everything was necessarily um, you know, 335 00:18:26,600 --> 00:18:29,639 Speaker 1: people in the next room dying. Sure, But we referred 336 00:18:29,640 --> 00:18:31,520 Speaker 1: to a term at the beginning of this episode which 337 00:18:31,560 --> 00:18:35,280 Speaker 1: is serenoid. And this also comes from Stanley Milgram, I 338 00:18:35,320 --> 00:18:39,840 Speaker 1: believe from unpublished results of some experiments he conducted, right right. 339 00:18:39,880 --> 00:18:42,040 Speaker 1: He he never published any of these. They ended up, 340 00:18:42,280 --> 00:18:44,720 Speaker 1: you know, putting some work into it, but then going 341 00:18:44,760 --> 00:18:46,800 Speaker 1: off in a different direction with his research. Yeah. So, 342 00:18:46,840 --> 00:18:49,480 Speaker 1: as we said, the term serenoid comes from Syrano to 343 00:18:49,520 --> 00:18:52,480 Speaker 1: bears direct. But what was the deal with Milgram's experiments? 344 00:18:52,640 --> 00:18:55,000 Speaker 1: So Milgram essentially wanted to see, hey, if you're that 345 00:18:55,280 --> 00:18:59,119 Speaker 1: the woman that is being wooed bye bye bye by 346 00:19:00,040 --> 00:19:03,560 Speaker 1: in those meat puppet his handsome, his handsome young man. 347 00:19:03,600 --> 00:19:06,159 Speaker 1: I think her name was Roxanne. Roxanne Roxane's on the 348 00:19:06,200 --> 00:19:09,680 Speaker 1: balcony being wooed by a handsome meat puppet that's being 349 00:19:09,720 --> 00:19:12,480 Speaker 1: fed lines by Syra. No. Yeah, if you're Roxanne, would 350 00:19:12,520 --> 00:19:14,760 Speaker 1: you be able to detect something was weird? Would you? 351 00:19:14,800 --> 00:19:17,879 Speaker 1: Would you encounter this young man and say, you know, 352 00:19:17,960 --> 00:19:21,000 Speaker 1: he seems a little more clever than than he should be, 353 00:19:21,119 --> 00:19:23,480 Speaker 1: or there's a delay in in what he's telling me, 354 00:19:23,560 --> 00:19:25,159 Speaker 1: you know, would there be something that would tip you, 355 00:19:25,240 --> 00:19:28,719 Speaker 1: tip you off to the deception? You know, with modern technology, 356 00:19:28,760 --> 00:19:31,960 Speaker 1: I'd imagine you could carry out that experiment pretty easily. Yeah, 357 00:19:32,000 --> 00:19:34,840 Speaker 1: I mean, even at the time, the technology was good enough. 358 00:19:35,480 --> 00:19:39,040 Speaker 1: So in these unpublished experiments, he had a source speak 359 00:19:39,080 --> 00:19:43,440 Speaker 1: into a microphone and a shadower listened through a hidden earpiece. 360 00:19:44,000 --> 00:19:46,639 Speaker 1: Then he or she would repeat whatever they heard, you know, 361 00:19:46,720 --> 00:19:50,040 Speaker 1: basic you know, arrested development surro, good kind of situation. 362 00:19:51,119 --> 00:19:53,479 Speaker 1: But here's the thing. He found that with practice, speech 363 00:19:53,480 --> 00:19:56,720 Speaker 1: shadowing becomes easier and easier, Like we really take to 364 00:19:56,800 --> 00:19:59,159 Speaker 1: it rather naturally. They say, it's not that hard at all. 365 00:19:59,200 --> 00:20:01,359 Speaker 1: I read in one of the sources we used for 366 00:20:01,400 --> 00:20:04,119 Speaker 1: this episode that sometimes it only takes a few milliseconds 367 00:20:04,840 --> 00:20:07,240 Speaker 1: of delay between the speech being fed in through the 368 00:20:07,280 --> 00:20:10,800 Speaker 1: earpiece and saying it. We're ready, willing and able to 369 00:20:11,000 --> 00:20:15,080 Speaker 1: not think for ourselves. And so yeah, he observed that 370 00:20:15,160 --> 00:20:17,639 Speaker 1: he set all this up, put people in motion with 371 00:20:17,720 --> 00:20:20,199 Speaker 1: these with a source and a shadower, and then a 372 00:20:20,280 --> 00:20:23,680 Speaker 1: test subject having to interact with this individual or is 373 00:20:23,720 --> 00:20:27,800 Speaker 1: it is individuals um and then give feedback about what 374 00:20:27,840 --> 00:20:30,399 Speaker 1: they thought. And he observed that people couldn't tell the 375 00:20:30,440 --> 00:20:34,720 Speaker 1: difference between a serenoid again that person, uh that that 376 00:20:35,040 --> 00:20:38,560 Speaker 1: that shadower who is informed by another source just repeating 377 00:20:38,600 --> 00:20:42,000 Speaker 1: words that are entering his or her ears, could not 378 00:20:42,040 --> 00:20:44,600 Speaker 1: tell the difference between seranoid and a normal human being 379 00:20:44,680 --> 00:20:48,560 Speaker 1: during interactions. Yeah, so that's what Milgram reported, And I 380 00:20:48,600 --> 00:20:52,400 Speaker 1: don't know, that's surprising and creepy. Yeah, I would think 381 00:20:52,440 --> 00:20:54,520 Speaker 1: that I would easily be able to tell the difference 382 00:20:54,520 --> 00:20:57,560 Speaker 1: between a person speaking of their own free will and 383 00:20:57,680 --> 00:21:00,920 Speaker 1: somebody who's just being fedlines in an ear pie. Yeah. 384 00:21:00,920 --> 00:21:03,800 Speaker 1: In one experiment, Milgram's sourced in other words, fed you know, 385 00:21:03,920 --> 00:21:06,679 Speaker 1: lines to a pair of eleven and twelve year olds. 386 00:21:06,760 --> 00:21:08,760 Speaker 1: So they the eleven and twelve year olds where the 387 00:21:08,760 --> 00:21:12,280 Speaker 1: shadows and Milgram himself was speaking correct. Uh, yeah, the 388 00:21:12,680 --> 00:21:15,560 Speaker 1: children were the serenoids, if you will, and a group 389 00:21:15,560 --> 00:21:18,639 Speaker 1: of observing teachers never suspected that they were chatting with 390 00:21:18,680 --> 00:21:22,719 Speaker 1: anything other than a normal, though very bright child. Uh 391 00:21:23,480 --> 00:21:25,639 Speaker 1: and uh and and it is experiments you found that 392 00:21:25,640 --> 00:21:28,600 Speaker 1: most people could have a lengthy twenty minute conversation with 393 00:21:28,640 --> 00:21:33,879 Speaker 1: a serenoid without noticing anything of this. Wow. Yeah, but 394 00:21:33,960 --> 00:21:35,919 Speaker 1: I mean again, that's that's hard to believe. I mean, 395 00:21:35,960 --> 00:21:37,880 Speaker 1: it just seems like you'd be able to notice that 396 00:21:38,040 --> 00:21:40,400 Speaker 1: right off the bat. But well, but just think how 397 00:21:40,440 --> 00:21:44,960 Speaker 1: easily we we buy into an actor's presentation. Think of 398 00:21:45,000 --> 00:21:47,240 Speaker 1: and think of all the interactions we have in the 399 00:21:47,280 --> 00:21:51,600 Speaker 1: course of of our lives where the individual we're talking 400 00:21:51,600 --> 00:21:56,280 Speaker 1: to is maybe not completely uh genuine, maybe that you know, 401 00:21:56,440 --> 00:21:59,280 Speaker 1: maybe you're interacting with with someone that's doing customer service 402 00:21:59,800 --> 00:22:03,320 Speaker 1: or or they're trying to sell you something or can you, etcetera. 403 00:22:03,440 --> 00:22:05,119 Speaker 1: It's like it's not always going to come off is 404 00:22:05,160 --> 00:22:08,199 Speaker 1: oh this person is completely fake. It's gonna It's just 405 00:22:08,920 --> 00:22:11,920 Speaker 1: an aspect of human interaction. Well, it's funny you mentioned 406 00:22:11,920 --> 00:22:14,280 Speaker 1: actors because I had this thought when I was reading 407 00:22:14,280 --> 00:22:17,760 Speaker 1: this research. In a strange kind of way, all actors 408 00:22:17,800 --> 00:22:22,119 Speaker 1: in movies are serenoids. Like you have the writer coming 409 00:22:22,200 --> 00:22:25,200 Speaker 1: up with lines for many types of characters that aren't 410 00:22:25,240 --> 00:22:28,280 Speaker 1: anything like them outwardly at all, Like you can have 411 00:22:28,800 --> 00:22:31,720 Speaker 1: you know, you know, a sixty five year old female 412 00:22:31,840 --> 00:22:34,600 Speaker 1: writer writing lines for a ten year old boy in 413 00:22:34,680 --> 00:22:39,480 Speaker 1: a play, and that boy effectively is a distanced serenoid. 414 00:22:39,560 --> 00:22:42,000 Speaker 1: And should we be able to tell the difference? Like 415 00:22:42,240 --> 00:22:44,439 Speaker 1: sometimes you can. Sometimes you watch a movie, you know, 416 00:22:44,520 --> 00:22:46,840 Speaker 1: and you're like, kids wouldn't say that, that's not how 417 00:22:46,920 --> 00:22:49,960 Speaker 1: kids talk. But other times you buy it. Yeah. I 418 00:22:49,960 --> 00:22:51,800 Speaker 1: think one of the things to keep in mind about 419 00:22:51,800 --> 00:22:54,879 Speaker 1: sienoids and ultimately about echo borgs this we've moved forward 420 00:22:55,280 --> 00:22:59,240 Speaker 1: is that we're dealing with a hybrid personality. So there's 421 00:22:59,280 --> 00:23:01,760 Speaker 1: the there are the words and the personality of the 422 00:23:01,760 --> 00:23:04,960 Speaker 1: individual who is informing, and then the words of the 423 00:23:05,000 --> 00:23:09,800 Speaker 1: shadower or the serenoid. That individual is bringing their own delivery, 424 00:23:09,800 --> 00:23:13,000 Speaker 1: their own personality to it. Sure, because your personality, as 425 00:23:13,000 --> 00:23:15,119 Speaker 1: you present it outwardly, is way more than the words 426 00:23:15,160 --> 00:23:18,639 Speaker 1: you say. Yeah, obviously, it's your body language, it's your expressions, 427 00:23:18,800 --> 00:23:20,920 Speaker 1: it's you know, the way you carry yourself. I mean, 428 00:23:20,960 --> 00:23:23,800 Speaker 1: that's all part of the message you present. Like we 429 00:23:23,800 --> 00:23:27,160 Speaker 1: can all think of of movies or TV shows where 430 00:23:28,280 --> 00:23:32,520 Speaker 1: there's a particularly gifted actor who's able to bring lackluster 431 00:23:32,760 --> 00:23:35,359 Speaker 1: lines to life in a way that a less gifted 432 00:23:35,400 --> 00:23:37,520 Speaker 1: actor who just would not be able to achieve. I 433 00:23:37,520 --> 00:23:41,000 Speaker 1: always think of ral Julia in Overdrawn at the Memory Bank, 434 00:23:41,359 --> 00:23:44,119 Speaker 1: who you know, some people, you know, it's easy to 435 00:23:44,119 --> 00:23:46,080 Speaker 1: have a lot of fun with his performance. It's you know, 436 00:23:46,080 --> 00:23:49,679 Speaker 1: it's kind of cheesy and outlandish. But that man in 437 00:23:49,680 --> 00:23:53,080 Speaker 1: that movie, which is like, you know, a low budget 438 00:23:53,359 --> 00:23:57,480 Speaker 1: PBS adaptation of a science fiction story, he brings so 439 00:23:57,600 --> 00:24:00,240 Speaker 1: much life to every line. So many lines and at 440 00:24:00,480 --> 00:24:03,720 Speaker 1: in that film, if they were delivered by a lesser actor, 441 00:24:03,840 --> 00:24:06,040 Speaker 1: would have just fallen flat. But he makes even the 442 00:24:06,080 --> 00:24:10,159 Speaker 1: most pointless line, uh, just really land. I agree, and 443 00:24:10,200 --> 00:24:13,920 Speaker 1: I think this is a common feature actually an older actors. 444 00:24:14,000 --> 00:24:16,639 Speaker 1: I see this way more commonly with actors who have 445 00:24:16,800 --> 00:24:19,040 Speaker 1: been in the business for a long time. I think 446 00:24:19,080 --> 00:24:23,360 Speaker 1: of Star Wars episode two, which my opinion is that 447 00:24:23,359 --> 00:24:27,040 Speaker 1: that is a horrible, horrible movie. But when Christopher Lee 448 00:24:27,080 --> 00:24:30,119 Speaker 1: shows up and starts talking, the writing I think is 449 00:24:30,200 --> 00:24:32,359 Speaker 1: just as bad as it's been the entire time. But 450 00:24:32,400 --> 00:24:36,320 Speaker 1: suddenly I'm okay, I'm listening Christopher Lee. I'm buying it. 451 00:24:36,359 --> 00:24:39,880 Speaker 1: He is really selling this horrible dialogue. Yeah, yeah, exactly, 452 00:24:39,880 --> 00:24:42,840 Speaker 1: So in that respect, Raw Julia, Christopher Lee, they would 453 00:24:42,920 --> 00:24:46,400 Speaker 1: make wonderful serenoids, and keep them in mind, they might 454 00:24:46,440 --> 00:24:51,159 Speaker 1: have made excellent echo boards. All Right, we're gonna take 455 00:24:51,160 --> 00:24:53,720 Speaker 1: a quick break and when we come back back to 456 00:24:53,800 --> 00:25:05,280 Speaker 1: the echo boards. Okay, So Robert and we were talking 457 00:25:05,320 --> 00:25:09,720 Speaker 1: about echo borgs being the natural extension of serenoid. So 458 00:25:09,760 --> 00:25:13,080 Speaker 1: the serenoid is a person being fed lines by another person. 459 00:25:13,520 --> 00:25:15,760 Speaker 1: The echo borg, as we talked about at the beginning 460 00:25:15,760 --> 00:25:18,439 Speaker 1: of the episode, would be a person being fed lines 461 00:25:19,200 --> 00:25:22,200 Speaker 1: by a computer program. Yes, I want to know, are 462 00:25:22,200 --> 00:25:25,600 Speaker 1: there any studies where people have looked into this phenomena, 463 00:25:25,840 --> 00:25:30,479 Speaker 1: and if so, does a person delivering the lines from 464 00:25:30,520 --> 00:25:34,000 Speaker 1: a chat bot make the chat bought anymore convincing in 465 00:25:34,040 --> 00:25:38,040 Speaker 1: the Turing test? Well, indeed, that that's exactly what we're 466 00:25:38,040 --> 00:25:41,640 Speaker 1: gonna look at in this half of the podcast, particularly 467 00:25:41,680 --> 00:25:46,480 Speaker 1: the work of two individuals British social psychologist Kevin Cordy 468 00:25:46,520 --> 00:25:51,320 Speaker 1: and Alex Glefski. And Uh, basically, we have two studies 469 00:25:51,359 --> 00:25:54,919 Speaker 1: we're gonna we're gonna analyze here two thousand fourteen study 470 00:25:54,960 --> 00:25:58,199 Speaker 1: where they essentially just recreate some of Milgram's work and 471 00:25:58,440 --> 00:26:01,320 Speaker 1: and look at the at the series Annoid and then 472 00:26:01,320 --> 00:26:04,600 Speaker 1: a two thousand fifteen study that just came out um 473 00:26:04,880 --> 00:26:08,440 Speaker 1: where they take the serenoid, apply it to chatter bots 474 00:26:08,480 --> 00:26:11,760 Speaker 1: and give us the echo borg. Yeah. So the script 475 00:26:11,840 --> 00:26:15,359 Speaker 1: we read earlier from the conversation between the eco borg 476 00:26:15,480 --> 00:26:18,920 Speaker 1: and the regular interactor that came from London School of 477 00:26:18,960 --> 00:26:21,359 Speaker 1: Economics research. It was a YouTube video that they had 478 00:26:21,359 --> 00:26:24,040 Speaker 1: put up. So that was their echo borg saying that 479 00:26:24,160 --> 00:26:27,000 Speaker 1: it liked fruit trees and that The Imitation Game was 480 00:26:27,040 --> 00:26:28,679 Speaker 1: its favorite movie, you know. And I do want to 481 00:26:28,680 --> 00:26:30,560 Speaker 1: throw in like that was the dead giveaway that was 482 00:26:30,600 --> 00:26:34,480 Speaker 1: a robot, because that that movie is nobody's favorite movie, right. 483 00:26:34,560 --> 00:26:37,200 Speaker 1: And I think Rose, the chat bot in that example, 484 00:26:37,359 --> 00:26:40,719 Speaker 1: is sort of programmed to be playful, like not necessarily 485 00:26:40,760 --> 00:26:43,800 Speaker 1: to be entirely convincing as a human, because Rose gives 486 00:26:43,840 --> 00:26:46,679 Speaker 1: other playful answers. Also in that same video, there's a 487 00:26:46,680 --> 00:26:50,520 Speaker 1: part where the human interactor says to the human eco borg, 488 00:26:50,920 --> 00:26:53,720 Speaker 1: do you like food? And the human eco borg delivers 489 00:26:53,760 --> 00:26:58,879 Speaker 1: the line yummy electricity. Yeah. I don't know if the 490 00:26:59,320 --> 00:27:02,720 Speaker 1: comedy needs to be a primary programming of function. I mean, 491 00:27:02,760 --> 00:27:05,439 Speaker 1: you know, humor is an important part of any human interaction. 492 00:27:06,280 --> 00:27:08,520 Speaker 1: I mean, that's good humor for a chat bot, but 493 00:27:08,800 --> 00:27:11,840 Speaker 1: that's obviously the chatbot who's not trying all that hard 494 00:27:11,880 --> 00:27:15,679 Speaker 1: to hide it. Yeah, being a little little two coy. Alright, 495 00:27:15,720 --> 00:27:19,040 Speaker 1: So let's look at this the work of Accordion glepski 496 00:27:19,440 --> 00:27:22,600 Speaker 1: Um in two thousand fourteen. They focused this paper in 497 00:27:22,640 --> 00:27:26,080 Speaker 1: the journal Social Psychology, and they essentially set out to 498 00:27:26,080 --> 00:27:29,920 Speaker 1: to replicate Milgram's work. So the basic set up here 499 00:27:30,160 --> 00:27:34,280 Speaker 1: is they they had a shadower set in an interrogation room, 500 00:27:34,320 --> 00:27:37,720 Speaker 1: so that's the serenoid type figure, and uh, they were 501 00:27:37,760 --> 00:27:41,080 Speaker 1: engaged in a conversation with an interactant. This is an 502 00:27:41,160 --> 00:27:47,080 Speaker 1: unsuspecting volunteer. Meanwhile, the source, that's the individual feeding the shadow, 503 00:27:47,080 --> 00:27:51,679 Speaker 1: we're feeding the serenoid um. They're in another room and 504 00:27:51,680 --> 00:27:54,800 Speaker 1: they're observing everything through video and audio links and they 505 00:27:54,840 --> 00:27:57,800 Speaker 1: tell the shadow or exactly what they need to say 506 00:27:57,840 --> 00:28:01,080 Speaker 1: through a uh, you know, a discreete FM radio transmitter 507 00:28:01,160 --> 00:28:04,719 Speaker 1: set up. Okay, so they have twenty volunteers that are 508 00:28:04,760 --> 00:28:07,879 Speaker 1: engaged in ten minute conversations with the serenoid and twenty 509 00:28:07,920 --> 00:28:11,040 Speaker 1: additional volunteers that are tested in a control setting. Okay, 510 00:28:11,080 --> 00:28:13,800 Speaker 1: so the control setting is just a regular person talking, 511 00:28:13,840 --> 00:28:16,800 Speaker 1: not being fedlines. And so this backed up exactly what 512 00:28:16,880 --> 00:28:20,320 Speaker 1: Milgram had told us decades earlier that when one human 513 00:28:20,359 --> 00:28:23,359 Speaker 1: parents and other humans words, we just totally buy it. Okay, 514 00:28:23,400 --> 00:28:26,440 Speaker 1: So even if you were skeptical of Millgram, and who knows, 515 00:28:26,480 --> 00:28:28,760 Speaker 1: you might have had reason to be, we've got it. 516 00:28:28,960 --> 00:28:31,080 Speaker 1: We've got a new study backing that up now. And 517 00:28:31,960 --> 00:28:34,800 Speaker 1: the implications of this are kind of creepy. It's that 518 00:28:35,200 --> 00:28:38,000 Speaker 1: unless you check in people's ears to make sure they 519 00:28:38,000 --> 00:28:40,520 Speaker 1: don't have an ear piece in there, anybody in your 520 00:28:40,560 --> 00:28:44,560 Speaker 1: life could be feeding you lines from somebody sitting in 521 00:28:44,560 --> 00:28:47,560 Speaker 1: a van outside the building, and you might not know it. Yeah, 522 00:28:47,600 --> 00:28:51,240 Speaker 1: I mean, look at the movie The Exercist Strike. That's 523 00:28:51,240 --> 00:28:53,320 Speaker 1: why demons, when they possess people, they have to spit, 524 00:28:53,760 --> 00:28:56,000 Speaker 1: you know, green vomit across the room and spin their 525 00:28:56,000 --> 00:28:57,960 Speaker 1: heads around, because otherwise we're just not going to buy 526 00:28:58,000 --> 00:29:01,080 Speaker 1: into the fact that an individual is not uh functioning 527 00:29:01,080 --> 00:29:03,400 Speaker 1: of their own free will, right or you know, you've 528 00:29:03,440 --> 00:29:06,000 Speaker 1: got the problem of Okay, So the priest tries to 529 00:29:06,040 --> 00:29:07,959 Speaker 1: go to the authorities and say I need to conduct 530 00:29:07,960 --> 00:29:10,800 Speaker 1: an exorcism. My evidence is this person was speaking in 531 00:29:10,800 --> 00:29:12,880 Speaker 1: a language they've never learned. It's like, well, how do 532 00:29:12,960 --> 00:29:15,479 Speaker 1: you know they're not being fed that language from somebody 533 00:29:15,480 --> 00:29:18,040 Speaker 1: in a van outside the building through an earpiece. You've 534 00:29:18,080 --> 00:29:20,520 Speaker 1: got to have the p soup well, you know, speaking 535 00:29:20,520 --> 00:29:23,440 Speaker 1: of possessed children. A second part of the same experiment, 536 00:29:23,440 --> 00:29:26,400 Speaker 1: ac Cording and Gillespie um use a twelve year old 537 00:29:26,400 --> 00:29:29,040 Speaker 1: male actor to serve as a shadow er. Only thirty 538 00:29:29,040 --> 00:29:32,800 Speaker 1: seven year old male social sets social psychologist acts as 539 00:29:32,800 --> 00:29:36,480 Speaker 1: the source, and again no one suspected the illusion, though 540 00:29:36,560 --> 00:29:39,760 Speaker 1: some thought that children might have been coached prior, so 541 00:29:39,840 --> 00:29:43,000 Speaker 1: there was a little you know, trepidation there where they think, uh, 542 00:29:43,480 --> 00:29:47,280 Speaker 1: I think somebody was preparing this kid for this conversation. 543 00:29:47,400 --> 00:29:50,440 Speaker 1: So they picked up that something was off, but they 544 00:29:50,440 --> 00:29:53,160 Speaker 1: didn't pick up that the kid was being fed lines. Yeah, 545 00:29:53,240 --> 00:29:55,360 Speaker 1: and again that just kind of drives with with so 546 00:29:55,400 --> 00:29:58,440 Speaker 1: many interactions, most of us having our lives, where you 547 00:29:58,440 --> 00:30:00,400 Speaker 1: you talk to somebody and you might think that person 548 00:30:00,400 --> 00:30:03,719 Speaker 1: seemed kind of fake, but you're not thinking that person 549 00:30:04,120 --> 00:30:06,280 Speaker 1: was serving as a meat puppet for an AI or 550 00:30:06,280 --> 00:30:08,720 Speaker 1: some dude in a van. You just think they're they're 551 00:30:08,720 --> 00:30:10,520 Speaker 1: a little phony there, you know, they have some sort 552 00:30:10,520 --> 00:30:12,400 Speaker 1: of an agenda. They're kind of playing the room, or 553 00:30:12,440 --> 00:30:15,440 Speaker 1: they're they're in just full on customer service mode, etcetera. 554 00:30:15,600 --> 00:30:18,600 Speaker 1: Or they might just be experiencing some anxiety and awkwardness. 555 00:30:18,600 --> 00:30:20,040 Speaker 1: I mean, I think you want to give people the 556 00:30:20,080 --> 00:30:22,720 Speaker 1: benefit of the doubt. And that's I think coming through 557 00:30:22,760 --> 00:30:25,280 Speaker 1: in some of these experiments. If somebody acts a little 558 00:30:25,320 --> 00:30:27,880 Speaker 1: awkward or weird, I mean, you don't want to just 559 00:30:28,200 --> 00:30:30,880 Speaker 1: judge them and say that person is a robot. You know, 560 00:30:30,920 --> 00:30:33,960 Speaker 1: they're being fedlines or something like that. You know, we 561 00:30:34,040 --> 00:30:36,440 Speaker 1: understand that some people get in moods, they have trouble, 562 00:30:36,560 --> 00:30:39,240 Speaker 1: sometimes they feel awkward. We've all felt like this, and 563 00:30:39,280 --> 00:30:42,560 Speaker 1: you want to be accommodating of other people's awkwardness whenever possible. Yeah, 564 00:30:42,640 --> 00:30:45,600 Speaker 1: I mean, awkwardness is in his own in its own right, 565 00:30:45,680 --> 00:30:48,920 Speaker 1: is an essential part of human interaction. UM. And if 566 00:30:49,120 --> 00:30:51,480 Speaker 1: if an AI or a computer wanted a part of 567 00:30:51,800 --> 00:30:54,280 Speaker 1: legitimate human interaction, they need to sign up for some 568 00:30:54,520 --> 00:30:58,640 Speaker 1: serious awkwardness from time the time. So this initial studies 569 00:30:58,720 --> 00:31:00,920 Speaker 1: pretty much backs up everything of Room told us. So 570 00:31:01,000 --> 00:31:04,880 Speaker 1: from that point, um, Cordy and Glevski moved on to 571 00:31:05,480 --> 00:31:08,400 Speaker 1: artificial intelligence to chatter bots and the creation of the 572 00:31:08,400 --> 00:31:11,240 Speaker 1: eco borg. And again, an echo borg is essentially the 573 00:31:11,280 --> 00:31:15,240 Speaker 1: same as a serenoid, except the shadow er is speaking 574 00:31:15,360 --> 00:31:17,680 Speaker 1: the words of a chatter box, so the source is 575 00:31:17,720 --> 00:31:20,959 Speaker 1: not human, it's a computer program, right, And again this 576 00:31:21,040 --> 00:31:22,840 Speaker 1: is what this is. It's important to keep in mind 577 00:31:22,960 --> 00:31:27,440 Speaker 1: that a serenoid is a hybrid personality, the individual behind 578 00:31:27,640 --> 00:31:30,200 Speaker 1: the speaker and then the personality of the speaker coming together. 579 00:31:30,480 --> 00:31:33,160 Speaker 1: So in eco borg, the borg aspect of that referring 580 00:31:33,160 --> 00:31:36,840 Speaker 1: to cyborg, it really is a new type of cyborg 581 00:31:36,920 --> 00:31:40,160 Speaker 1: that we're playing with here, uh, a machine and a 582 00:31:40,240 --> 00:31:45,200 Speaker 1: human as one. And the question presents itself, would we 583 00:31:45,480 --> 00:31:48,520 Speaker 1: be fooled? Will we observe an echo borg in action 584 00:31:49,160 --> 00:31:53,520 Speaker 1: and and suspect that it's a machine will behind the 585 00:31:53,560 --> 00:31:56,960 Speaker 1: human face. I'm gonna say that even before I read 586 00:31:57,000 --> 00:32:00,080 Speaker 1: the results of the study, when I knew what the 587 00:32:00,120 --> 00:32:02,440 Speaker 1: subject was, I thought, you know, I bet it depends 588 00:32:02,480 --> 00:32:07,120 Speaker 1: heavily on the sort of pre priming of the study participants, 589 00:32:07,720 --> 00:32:10,920 Speaker 1: like are they expecting to be interacting with somebody who 590 00:32:11,000 --> 00:32:14,920 Speaker 1: might be an eco borg or are they or are 591 00:32:14,960 --> 00:32:17,280 Speaker 1: they just going in blind and interacting with somebody who 592 00:32:17,360 --> 00:32:20,239 Speaker 1: might be a little weird. It's true. I mean, when 593 00:32:20,240 --> 00:32:22,880 Speaker 1: you're presented with the if you're presented with the possibility, hey, 594 00:32:22,920 --> 00:32:26,960 Speaker 1: you might be talking to a robot here, that's gonna 595 00:32:27,360 --> 00:32:29,720 Speaker 1: it's get that's gonna alter your perceptions a bit. Yeah, 596 00:32:29,760 --> 00:32:32,440 Speaker 1: So what did they find? First of all, this is 597 00:32:32,440 --> 00:32:34,960 Speaker 1: how the study rolled out. Forty one participants in the 598 00:32:34,960 --> 00:32:38,240 Speaker 1: experiment were shown a text messaging client with two windows 599 00:32:38,520 --> 00:32:41,440 Speaker 1: until they be interacting with two human shadowers at the 600 00:32:41,480 --> 00:32:44,600 Speaker 1: same time. One shadower is sourced by another human, the 601 00:32:44,640 --> 00:32:48,280 Speaker 1: other is sourced by a chatter bot, and each interrogator 602 00:32:48,360 --> 00:32:52,040 Speaker 1: asked questions via text, and then the participants were told 603 00:32:52,200 --> 00:32:54,480 Speaker 1: they'd be asked which of the two agents was a 604 00:32:54,480 --> 00:32:57,880 Speaker 1: real human. Now, the researchers also wanted to compare human 605 00:32:57,920 --> 00:33:00,880 Speaker 1: and text responses, so they divided the group into two groups. 606 00:33:00,960 --> 00:33:03,920 Speaker 1: Twenty of the participants typed their half of the conversation 607 00:33:04,000 --> 00:33:07,320 Speaker 1: and into a text client, but got responses from the 608 00:33:07,360 --> 00:33:10,080 Speaker 1: two echo boorgs setting in the same room. The other 609 00:33:10,080 --> 00:33:13,160 Speaker 1: twenty one participants just had text chats the whole way through. 610 00:33:13,280 --> 00:33:14,560 Speaker 1: So this way they would have a you know, a 611 00:33:14,560 --> 00:33:17,920 Speaker 1: base comparison to the you know, the more established chatter 612 00:33:17,960 --> 00:33:20,400 Speaker 1: bot to ring test situation that we're were typing into 613 00:33:20,400 --> 00:33:24,200 Speaker 1: a terminal. Right, So the results were that all the 614 00:33:24,240 --> 00:33:27,800 Speaker 1: text only participants correctly identified the real humans. So again 615 00:33:28,840 --> 00:33:33,480 Speaker 1: dead giveaway based on just the limitations of the chatter bot. Um, 616 00:33:33,520 --> 00:33:35,800 Speaker 1: it's pretty easy to tell that you're just talking to 617 00:33:36,480 --> 00:33:38,560 Speaker 1: a program. Yeah, if you if you have enough time 618 00:33:38,600 --> 00:33:42,160 Speaker 1: to talk to them, you can almost definitely trip them up. Now. 619 00:33:42,160 --> 00:33:44,240 Speaker 1: On the other hand, in the other set up, only 620 00:33:44,360 --> 00:33:47,720 Speaker 1: two of the participants didn't identify the real humans. So 621 00:33:48,480 --> 00:33:53,720 Speaker 1: in the eco borg situations, they still weren't too terribly convincing, right, 622 00:33:53,760 --> 00:33:56,160 Speaker 1: so they can tell the difference between a serenoid and 623 00:33:56,200 --> 00:33:59,400 Speaker 1: an eco borg. Everybody except two of them could, right, 624 00:34:00,520 --> 00:34:03,920 Speaker 1: So it seems that human or not, still rather easy 625 00:34:03,920 --> 00:34:07,280 Speaker 1: to identify the machine behind the flesh. You know, I 626 00:34:07,360 --> 00:34:11,160 Speaker 1: would chalk this up probably just to the poor quality 627 00:34:11,280 --> 00:34:16,320 Speaker 1: of today's best chat bots. Yeah, I mean, because even 628 00:34:16,360 --> 00:34:19,880 Speaker 1: even a Rale Julia or or a Christopher Lee, you know, 629 00:34:20,200 --> 00:34:23,400 Speaker 1: presented with the with the dialogue we were using earlier, 630 00:34:23,400 --> 00:34:25,080 Speaker 1: they're they're not They're just gonna come off as a 631 00:34:25,120 --> 00:34:27,680 Speaker 1: doll person at best, you know, right, there's no way 632 00:34:27,719 --> 00:34:30,600 Speaker 1: you can sell the line I like fruit trees in 633 00:34:30,640 --> 00:34:34,000 Speaker 1: a way that makes people think you're you're being fedlines 634 00:34:34,040 --> 00:34:36,000 Speaker 1: by a human, no matter how good of an actor 635 00:34:36,080 --> 00:34:38,680 Speaker 1: you are. That's a dead giveaway that the machine does 636 00:34:38,719 --> 00:34:40,640 Speaker 1: not know what to do with what you just said. Yeah, 637 00:34:40,680 --> 00:34:43,680 Speaker 1: unless you're a toddler speaking that line, or or again, 638 00:34:43,719 --> 00:34:46,040 Speaker 1: if you're hiding it behind, say some sort of mental 639 00:34:46,080 --> 00:34:50,640 Speaker 1: deficiency or or language gap um or some sort of 640 00:34:50,680 --> 00:34:55,000 Speaker 1: you know, character that's leveled over everything, it's going to 641 00:34:55,080 --> 00:34:58,239 Speaker 1: be a dead giveaway. But the echo borgs certainly made 642 00:34:58,280 --> 00:35:01,120 Speaker 1: the machine seem more human and the text only group, 643 00:35:01,160 --> 00:35:04,600 Speaker 1: the chatbot received an average a humanity rating of two 644 00:35:04,640 --> 00:35:06,839 Speaker 1: point one four on a scale of one to ten. 645 00:35:07,320 --> 00:35:09,919 Speaker 1: In the echo boorg group, the average was a four 646 00:35:09,960 --> 00:35:14,719 Speaker 1: point oh five. So significant difference there. Okay, yeah, that's interesting. 647 00:35:14,840 --> 00:35:18,719 Speaker 1: So we can still tell that the Serenoid is more 648 00:35:18,840 --> 00:35:22,080 Speaker 1: human than an echo borg. But an echo borg is 649 00:35:22,160 --> 00:35:26,239 Speaker 1: more human at least in a qualitative evaluation than a 650 00:35:26,360 --> 00:35:29,200 Speaker 1: text only chat bot. Yeah, to lay it out, text 651 00:35:29,360 --> 00:35:32,799 Speaker 1: text only chat bought two point one four humanity, uh, 652 00:35:32,840 --> 00:35:35,400 Speaker 1: the echo borg four point oh five humanity, and then 653 00:35:35,480 --> 00:35:39,240 Speaker 1: the Serenoid had a humanity rating of eight point seven, six, 654 00:35:39,239 --> 00:35:42,120 Speaker 1: eight point one five. So it sounds to me like 655 00:35:42,160 --> 00:35:45,680 Speaker 1: even a normal human interaction. Like your score, you're scoring 656 00:35:45,719 --> 00:35:49,520 Speaker 1: a nine straight up non augmented human interactions. You're doing 657 00:35:49,560 --> 00:35:53,400 Speaker 1: pretty good. Yeah, So it does prove that there is 658 00:35:53,440 --> 00:35:58,359 Speaker 1: a humanizing effect to the echo borg system. It's just 659 00:35:58,400 --> 00:36:01,560 Speaker 1: the limitations of exist and chatter bots that they're coming 660 00:36:01,560 --> 00:36:03,600 Speaker 1: into play, right. It makes you think if there were 661 00:36:03,640 --> 00:36:07,120 Speaker 1: a really good chatbot out there, the eco borg might 662 00:36:07,200 --> 00:36:10,200 Speaker 1: be that finishing touch, that final push over the edge 663 00:36:10,239 --> 00:36:12,640 Speaker 1: of the cliff to sort of make the you know, 664 00:36:12,719 --> 00:36:15,200 Speaker 1: the perfect android. Yeah, kind of a bridge for the 665 00:36:15,280 --> 00:36:18,040 Speaker 1: uncanny valley, you know, like you just just all you 666 00:36:18,080 --> 00:36:20,520 Speaker 1: need is just a little acting on top of this 667 00:36:20,600 --> 00:36:23,759 Speaker 1: to really bring the script to life. Now. They they 668 00:36:23,800 --> 00:36:26,880 Speaker 1: also conducted a second experiment in this and they tested 669 00:36:26,920 --> 00:36:29,759 Speaker 1: to see if fifty eight new participants would guess whether 670 00:36:29,800 --> 00:36:32,200 Speaker 1: they were talking to an echo borg or a person 671 00:36:32,280 --> 00:36:35,640 Speaker 1: pretending to be an echo borg. So, you know, obviously 672 00:36:35,680 --> 00:36:37,680 Speaker 1: a person who's actually fed their lines by a chatter 673 00:36:37,719 --> 00:36:41,000 Speaker 1: bot or someone who's just playing around and pretending to 674 00:36:41,040 --> 00:36:43,160 Speaker 1: be speaking the words of a chatter bot. Right, so 675 00:36:43,200 --> 00:36:45,560 Speaker 1: this is a reverse touring test. Can you pretend to 676 00:36:45,560 --> 00:36:48,560 Speaker 1: be a machine. Right, But here's the catch. They're both 677 00:36:48,640 --> 00:36:51,839 Speaker 1: pure echo boorgs. There's no there's no faco echo borg. 678 00:36:51,880 --> 00:36:55,640 Speaker 1: But the but the individual being tested, they think that 679 00:36:55,719 --> 00:36:58,160 Speaker 1: one of the two echo boorgs is just a person pretending. 680 00:36:58,920 --> 00:37:01,799 Speaker 1: So seventy nine percent of the twenty eight participants who 681 00:37:01,880 --> 00:37:04,640 Speaker 1: faced an echo borg through I thought the person was 682 00:37:04,680 --> 00:37:08,240 Speaker 1: just mimicking mimicking a chat bot. So they seemed fake 683 00:37:08,600 --> 00:37:10,920 Speaker 1: but not inhuman. So if there is, you know, some 684 00:37:11,000 --> 00:37:13,719 Speaker 1: reason to believe that this is just a person being 685 00:37:13,719 --> 00:37:15,879 Speaker 1: a little fake and putting on a show versus being 686 00:37:16,120 --> 00:37:19,320 Speaker 1: you know, a complete echo borg, we're still going to 687 00:37:19,320 --> 00:37:21,759 Speaker 1: give them the benefit of a doubt. Interesting, now, as 688 00:37:21,920 --> 00:37:24,000 Speaker 1: I think we should do our duty and hedge a 689 00:37:24,000 --> 00:37:26,720 Speaker 1: little bit, say, all these sample sizes were pretty small, 690 00:37:27,120 --> 00:37:29,560 Speaker 1: and it would be interesting to see more research along 691 00:37:29,640 --> 00:37:32,799 Speaker 1: these lines, like with bigger sample sizes and trying to 692 00:37:32,880 --> 00:37:35,759 Speaker 1: repeat these results. And and and also to your point, 693 00:37:36,280 --> 00:37:39,440 Speaker 1: is chatterbots improve, it's going to be interesting to see 694 00:37:39,600 --> 00:37:42,400 Speaker 1: echo borgs employed as as a way to test them. 695 00:37:42,400 --> 00:37:44,040 Speaker 1: And in fact, that's something that's pointed out in the 696 00:37:44,080 --> 00:37:47,120 Speaker 1: in the study, is that this is a great way 697 00:37:47,160 --> 00:37:51,160 Speaker 1: moving forward to continue to analyze the chatter bots. Yeah, 698 00:37:51,160 --> 00:37:54,680 Speaker 1: but also, as we mentioned, the people who were interacting 699 00:37:54,800 --> 00:37:59,719 Speaker 1: with these, with these serenoids and echo borgs, we're in 700 00:38:00,000 --> 00:38:03,400 Speaker 1: some of these cases primed to expect something weird because 701 00:38:03,480 --> 00:38:06,880 Speaker 1: they're in a test environment. You can't hide the fact 702 00:38:06,960 --> 00:38:09,839 Speaker 1: that you're in a psychological test and you're you know, 703 00:38:10,160 --> 00:38:13,680 Speaker 1: whenever you're part of a test group, you're sort of 704 00:38:13,719 --> 00:38:16,480 Speaker 1: like ready for some weirdness, you know. Now, I wonder 705 00:38:16,520 --> 00:38:21,080 Speaker 1: how this would be if you sprung these shadowers, these seronoids, 706 00:38:21,160 --> 00:38:25,400 Speaker 1: and these echo borgs on people in a purely social scenario, 707 00:38:26,160 --> 00:38:28,839 Speaker 1: like like we talked about at the beginning, a convention 708 00:38:29,200 --> 00:38:33,360 Speaker 1: or you know, workplace meeting or a party where people 709 00:38:33,560 --> 00:38:37,000 Speaker 1: weren't expecting anything strange to be going on right until 710 00:38:37,440 --> 00:38:39,640 Speaker 1: they have to till out of survey after they leave 711 00:38:39,719 --> 00:38:42,160 Speaker 1: the dinner party. So what did you think of Susie 712 00:38:42,160 --> 00:38:46,560 Speaker 1: and her anecdote? Exactly? Yeah, I don't know why she 713 00:38:46,680 --> 00:38:49,080 Speaker 1: was so into fruit trees. You wouldn't shut up about it. 714 00:38:49,760 --> 00:38:55,400 Speaker 1: I mean, gardening is okay. But yeah, Now, something that 715 00:38:55,400 --> 00:38:57,200 Speaker 1: really stood out to me when we were going over this, 716 00:38:57,600 --> 00:39:00,319 Speaker 1: especially when you start thinking about the echo a borg 717 00:39:00,400 --> 00:39:01,759 Speaker 1: and what it would like to what it would be 718 00:39:01,800 --> 00:39:03,840 Speaker 1: like to be an eco borg and the sort of 719 00:39:03,840 --> 00:39:07,400 Speaker 1: pros and cons of being an eco borg, of of 720 00:39:07,400 --> 00:39:11,759 Speaker 1: of giving life to this will behind your will. It 721 00:39:11,760 --> 00:39:14,520 Speaker 1: reminded me a lot of transactive memory, which comes into 722 00:39:14,520 --> 00:39:17,120 Speaker 1: play really in two key areas. First of all, it's 723 00:39:17,120 --> 00:39:22,560 Speaker 1: the method by which we've always stored information in other people. Okay, 724 00:39:23,080 --> 00:39:25,959 Speaker 1: so you know those facts that you never remember because 725 00:39:26,000 --> 00:39:28,920 Speaker 1: your spouse remembers them, or it's that, or it's a 726 00:39:28,920 --> 00:39:31,440 Speaker 1: particular spelling or a bit of trivia that you never 727 00:39:31,520 --> 00:39:33,120 Speaker 1: keep in your own head because you always look it 728 00:39:33,160 --> 00:39:36,560 Speaker 1: up on your smartphone. It's the same thing. So outsourcing 729 00:39:36,600 --> 00:39:38,920 Speaker 1: the memory. I think this This must be one of 730 00:39:38,920 --> 00:39:41,440 Speaker 1: your favorite topics. You come back to this a lot, 731 00:39:41,480 --> 00:39:43,680 Speaker 1: And I think it's really interesting that because I see 732 00:39:43,680 --> 00:39:45,719 Speaker 1: it every day in my own life, and after I 733 00:39:45,719 --> 00:39:47,600 Speaker 1: read about it's like it's it's all I see. It's 734 00:39:47,640 --> 00:39:50,320 Speaker 1: like the things that I forget, the things that I 735 00:39:50,480 --> 00:39:54,120 Speaker 1: my mind refuses to learn because I've outsourced it to 736 00:39:54,560 --> 00:39:58,080 Speaker 1: the ubiquitous technology. Yeah. Yeah, I think the story is that, 737 00:39:58,280 --> 00:40:01,000 Speaker 1: you know, Socrates was worried about the right that if 738 00:40:01,640 --> 00:40:04,040 Speaker 1: you know, if we teach everybody writing and we have 739 00:40:04,320 --> 00:40:06,560 Speaker 1: writing on scrolls all the time, nobody's going to be 740 00:40:06,640 --> 00:40:10,000 Speaker 1: able to remember anything. You're not priming your memory. There 741 00:40:10,080 --> 00:40:13,200 Speaker 1: might be some truth to that, But then again, are 742 00:40:13,280 --> 00:40:16,720 Speaker 1: are we not just, you know, by writing things down 743 00:40:16,840 --> 00:40:21,319 Speaker 1: and having Internet archives and things becoming cyborgs in a way. Yeah, 744 00:40:21,400 --> 00:40:24,000 Speaker 1: I mean, in a sense, the voice is already whispering 745 00:40:24,000 --> 00:40:25,759 Speaker 1: in our ear. Yeah. I mean you could look at 746 00:40:25,800 --> 00:40:28,040 Speaker 1: it as a weakening of the human mind, or you 747 00:40:28,080 --> 00:40:30,840 Speaker 1: could look at it as a technological upgrade of the 748 00:40:30,880 --> 00:40:34,080 Speaker 1: human mind. Yeah. To what extent will we all become 749 00:40:34,120 --> 00:40:36,440 Speaker 1: echo boards of a type? You know where instead of 750 00:40:36,480 --> 00:40:38,040 Speaker 1: it being a situation where I'm just gonna be a 751 00:40:38,040 --> 00:40:41,200 Speaker 1: conduit for a powerful artificial intelligence, what if it's more 752 00:40:41,280 --> 00:40:45,200 Speaker 1: like I want to augment my existing self, which I 753 00:40:45,200 --> 00:40:48,800 Speaker 1: think is pretty good with say, you know, an artificial 754 00:40:48,880 --> 00:40:52,400 Speaker 1: intelligence that'll feed me the right lines in like business 755 00:40:52,440 --> 00:40:57,160 Speaker 1: situations or social situations. So it's you know, less focus 756 00:40:57,360 --> 00:40:59,799 Speaker 1: on I'm just gonna be a neat puppet. But rather, 757 00:41:01,200 --> 00:41:04,640 Speaker 1: can I merge with this AI, maybe even just a 758 00:41:04,640 --> 00:41:08,000 Speaker 1: small percentage you know five AI, and become a better 759 00:41:08,160 --> 00:41:11,200 Speaker 1: person and more effective person. Yeah, And these experiments, all 760 00:41:11,239 --> 00:41:14,840 Speaker 1: these examples are our word for word dictation. So whether 761 00:41:14,920 --> 00:41:18,920 Speaker 1: it's the shadower of the of the serenoid type being 762 00:41:18,960 --> 00:41:22,279 Speaker 1: fedlines by human or the shadower of the echoborg type 763 00:41:22,280 --> 00:41:25,160 Speaker 1: being fedlines by a computer, it's all lines. You're getting 764 00:41:25,200 --> 00:41:28,239 Speaker 1: full sentences and you're just trained to say them as 765 00:41:28,280 --> 00:41:31,520 Speaker 1: fast as you can. I mean, I wonder if this 766 00:41:31,560 --> 00:41:34,799 Speaker 1: setup could be more conceptual in nature, or you know, 767 00:41:35,160 --> 00:41:39,600 Speaker 1: feeding you facts or feeding you sort of feedback on 768 00:41:39,760 --> 00:41:43,359 Speaker 1: the progress of the conversation. It also makes me think 769 00:41:43,400 --> 00:41:46,440 Speaker 1: about the comedic possibilities because you imagine in an individual 770 00:41:46,520 --> 00:41:49,160 Speaker 1: and sort of you know, very updated Serano kind of 771 00:41:49,160 --> 00:41:53,080 Speaker 1: story story where individual goes into a business situation and 772 00:41:53,120 --> 00:41:56,120 Speaker 1: they thought they loaded uh you know, business Helper four 773 00:41:56,160 --> 00:41:59,160 Speaker 1: ah one into their into their their mind box, and 774 00:41:59,200 --> 00:42:03,719 Speaker 1: instead they put Lothario four point exactly Romance. Yeah, so 775 00:42:03,719 --> 00:42:07,000 Speaker 1: suddenly they're they're fed all these really effective lines if 776 00:42:07,000 --> 00:42:09,239 Speaker 1: you were, you know, in a bar, but instead you're 777 00:42:09,280 --> 00:42:11,800 Speaker 1: you're throwing them out there in the business meeting. Yeah, 778 00:42:11,960 --> 00:42:15,839 Speaker 1: I think it could work. Long walks on the beach 779 00:42:15,920 --> 00:42:19,480 Speaker 1: have to do with the new rollout, well, you know, 780 00:42:19,719 --> 00:42:24,479 Speaker 1: it comes to mind. Black Mirror, the British television series. 781 00:42:24,520 --> 00:42:28,600 Speaker 1: It does such a fantastic job looking at the ramifications 782 00:42:28,680 --> 00:42:32,560 Speaker 1: near future of our modern technology, often with very troubling results. 783 00:42:33,040 --> 00:42:35,799 Speaker 1: The Christmas episode they recently did, which I have not 784 00:42:35,880 --> 00:42:38,359 Speaker 1: seen yet, I think that's the only episode I haven't seen. Yeah, 785 00:42:38,360 --> 00:42:40,080 Speaker 1: I don't think it's made it to like Netflix here 786 00:42:40,080 --> 00:42:41,960 Speaker 1: in the States yet. And I'm not going to spoil anything, 787 00:42:42,000 --> 00:42:46,920 Speaker 1: but the initial setup involves one character who who offers 788 00:42:47,000 --> 00:42:52,520 Speaker 1: a serrano the Brigach kind of service to individuals out 789 00:42:52,560 --> 00:42:55,239 Speaker 1: there who need a little help with their their pickup game. 790 00:42:55,719 --> 00:42:59,440 Speaker 1: I number one, that's super creepy, and number two, I 791 00:42:59,520 --> 00:43:03,440 Speaker 1: can totally see that being a real thing. Yeah, I 792 00:43:03,719 --> 00:43:06,160 Speaker 1: don't find that out outlandish at all. Yeah. I mean 793 00:43:06,160 --> 00:43:08,239 Speaker 1: that's the great thing about Black Mirror is that you 794 00:43:08,280 --> 00:43:11,520 Speaker 1: know that it leans in the sci fi direction, but 795 00:43:11,640 --> 00:43:14,280 Speaker 1: not too far like it's and and that it's perfect 796 00:43:14,280 --> 00:43:17,759 Speaker 1: science fiction because it speaks to the problems that we 797 00:43:17,800 --> 00:43:20,960 Speaker 1: have today and how we are viewing the problems to 798 00:43:21,040 --> 00:43:25,480 Speaker 1: come exactly. If you want to see some disturbingly plausible dystopias, 799 00:43:25,560 --> 00:43:29,840 Speaker 1: watch Black Mirror. But you know it ultimately leads to 800 00:43:29,840 --> 00:43:34,080 Speaker 1: the question, echo bot, echo borgs? Is this a dystopian 801 00:43:34,160 --> 00:43:37,720 Speaker 1: idea or is it a pretty cool idea? Is it? 802 00:43:37,760 --> 00:43:40,480 Speaker 1: Is it ultimately a utopian idea where it would allow 803 00:43:40,640 --> 00:43:44,560 Speaker 1: us to to be better individuals? I don't know. I mean, 804 00:43:44,600 --> 00:43:49,480 Speaker 1: would the ancient philosophers look at our relationship with the 805 00:43:49,520 --> 00:43:53,200 Speaker 1: contents of our computers and the web as a horrible dystopia? 806 00:43:53,640 --> 00:43:56,239 Speaker 1: Even feels fine to me. But would they look at 807 00:43:56,280 --> 00:43:58,480 Speaker 1: that and say, oh, you know, react the same way 808 00:43:58,480 --> 00:44:01,000 Speaker 1: as we do to a black Mirror episode problem. I mean, 809 00:44:01,000 --> 00:44:03,040 Speaker 1: also they would look at our pants and say, what 810 00:44:03,040 --> 00:44:05,279 Speaker 1: are they doing? Where are there telets? Oh? My god? 811 00:44:05,520 --> 00:44:10,440 Speaker 1: A dressed like a Persian? I don't understand. I don't know. Ultimately, 812 00:44:10,480 --> 00:44:12,560 Speaker 1: that's a question throughout to the listeners. You know what, 813 00:44:13,280 --> 00:44:15,560 Speaker 1: how would you accept that job offer that we laid 814 00:44:15,560 --> 00:44:18,560 Speaker 1: out at the beginning of the episode. And and furthermore, 815 00:44:19,120 --> 00:44:21,640 Speaker 1: would you augment yourself with some sort of mild echo 816 00:44:21,719 --> 00:44:24,680 Speaker 1: borg system again, like a you know, a romance one 817 00:44:24,719 --> 00:44:28,040 Speaker 1: oh one or a business uh business strategy one oh one, 818 00:44:28,280 --> 00:44:31,360 Speaker 1: kind of program to to feed you the necessary lines 819 00:44:31,760 --> 00:44:34,200 Speaker 1: or even just ideas or facts that you might need 820 00:44:34,280 --> 00:44:38,200 Speaker 1: to make it through that business luncheon or dinner date. 821 00:44:38,920 --> 00:44:41,960 Speaker 1: You know. In another aspect of transactive memory that I 822 00:44:42,000 --> 00:44:44,239 Speaker 1: want to drive home that that plays in nicely with 823 00:44:44,360 --> 00:44:49,080 Speaker 1: the hybrid personality model of seronoids and echo boards, and 824 00:44:49,160 --> 00:44:53,120 Speaker 1: that is cross queuing. Okay, this is when and I 825 00:44:53,120 --> 00:44:55,040 Speaker 1: imagine a number of you can can relate to. This 826 00:44:55,080 --> 00:44:57,600 Speaker 1: is when you're having a conversation with let's say you know, 827 00:44:57,640 --> 00:44:59,840 Speaker 1: a spouse or a partner, a close friend or family, 828 00:45:00,600 --> 00:45:03,320 Speaker 1: and neither of you are quite able to remember something 829 00:45:03,360 --> 00:45:05,759 Speaker 1: on your own, but when you start queuing each other 830 00:45:06,760 --> 00:45:09,279 Speaker 1: the you're able to remember it together in a way 831 00:45:09,320 --> 00:45:11,239 Speaker 1: that you wouldn't, you know, kind of vultron style, you know, 832 00:45:11,280 --> 00:45:14,520 Speaker 1: bringing it together and suddenly your combined power for call 833 00:45:14,600 --> 00:45:17,080 Speaker 1: that memory, right, you're sort of pulling the triggers on 834 00:45:17,160 --> 00:45:20,000 Speaker 1: each other's brains. Yeah, so I think it's interesting to 835 00:45:20,040 --> 00:45:24,000 Speaker 1: think about that, uh, that scenario cross queuing and transactive memory. Yeah, 836 00:45:24,080 --> 00:45:27,520 Speaker 1: especially because if you imagine this AI scenario where you 837 00:45:27,560 --> 00:45:30,600 Speaker 1: know you've got some kind of on board artificial intelligence 838 00:45:30,640 --> 00:45:33,480 Speaker 1: that feeds you lines occasionally when you need them. You know, 839 00:45:33,520 --> 00:45:35,680 Speaker 1: you're not just a parrot for everything it says, but 840 00:45:36,040 --> 00:45:38,319 Speaker 1: it's it's an occasional helper. How does it know when 841 00:45:38,360 --> 00:45:41,799 Speaker 1: you need help? You would need some kind of some 842 00:45:41,880 --> 00:45:44,880 Speaker 1: kind of queuing in the conversation or even just in 843 00:45:44,880 --> 00:45:47,560 Speaker 1: your mind for this thing to know. Okay, I'm going 844 00:45:47,600 --> 00:45:50,239 Speaker 1: to step in, like every time you go m and 845 00:45:50,239 --> 00:45:52,240 Speaker 1: then it starts feeding you the lines like he's stalling, 846 00:45:52,280 --> 00:45:54,839 Speaker 1: throwing some throwing some good duh, some good linko there 847 00:45:54,960 --> 00:46:02,040 Speaker 1: every time you say the word interesting. Interesting. All right. 848 00:46:02,040 --> 00:46:04,759 Speaker 1: So there you have it. Uh, a strong episode, I think, 849 00:46:04,840 --> 00:46:07,799 Speaker 1: I think one that continues to resonate with listeners. We 850 00:46:07,840 --> 00:46:10,160 Speaker 1: actually heard back from a number of people on this, 851 00:46:10,280 --> 00:46:13,239 Speaker 1: including but I believe one listener in particular was really 852 00:46:13,280 --> 00:46:16,440 Speaker 1: excited about the prospect of this actually happening. Oh yeah, 853 00:46:16,239 --> 00:46:18,719 Speaker 1: he really wanted to know how he could be an 854 00:46:18,719 --> 00:46:21,319 Speaker 1: echo borg, and unfortunately we had to tell him, well 855 00:46:21,360 --> 00:46:24,160 Speaker 1: that's that's not a profession yet. Yeah, so maybe you 856 00:46:24,160 --> 00:46:26,919 Speaker 1: could sign up for a study, Yeah, hopefully. So I hope, 857 00:46:27,000 --> 00:46:30,640 Speaker 1: I hope that guy has found a place to hook 858 00:46:30,719 --> 00:46:34,439 Speaker 1: him up. All right, So, hey, you want to find 859 00:46:34,440 --> 00:46:36,600 Speaker 1: out more about this topic and other topics. You want 860 00:46:36,600 --> 00:46:39,720 Speaker 1: to see what other podcast episodes though we've recorded headn't 861 00:46:39,719 --> 00:46:41,319 Speaker 1: know where the stuff to blow your mind dot com 862 00:46:41,360 --> 00:46:43,680 Speaker 1: that is the mothership. That's where we'll find all the episodes. 863 00:46:43,719 --> 00:46:46,560 Speaker 1: You'll find some videos, you'll find blog posts. You'll find 864 00:46:46,600 --> 00:46:49,120 Speaker 1: links out to our social media accounts such as Facebook 865 00:46:49,160 --> 00:46:51,120 Speaker 1: and Twitter. We are blow the mind on both of those. 866 00:46:51,160 --> 00:46:54,560 Speaker 1: Give us a follow on tumbler. We are stuff to 867 00:46:54,600 --> 00:46:58,239 Speaker 1: blow your mind. Also, Hey, wherever you listen to us, 868 00:46:58,440 --> 00:47:02,759 Speaker 1: be it iTunes, be it's Spotify, be it Stitcher, be 869 00:47:02,880 --> 00:47:05,000 Speaker 1: at any of the various outlets out there, and we 870 00:47:05,000 --> 00:47:07,279 Speaker 1: get new ones every day. Give us a little love, 871 00:47:07,360 --> 00:47:09,200 Speaker 1: give us, give us a nice rating, give us a 872 00:47:09,280 --> 00:47:12,120 Speaker 1: nice review. If the platform allows you to do so. 873 00:47:12,200 --> 00:47:14,680 Speaker 1: That's a great way to help the show absolutely, and 874 00:47:14,719 --> 00:47:16,880 Speaker 1: if you want to get in touch with us as always, 875 00:47:16,960 --> 00:47:19,279 Speaker 1: you can email us and blow the mind at how 876 00:47:19,320 --> 00:47:30,480 Speaker 1: stuff works dot com For more on this and thousands 877 00:47:30,480 --> 00:47:55,799 Speaker 1: of other topics. Is it how stuff works dot com.