1 00:00:00,200 --> 00:00:03,800 Speaker 1: Hi, and welcome back to Bloomberg Benchmark. It is October 2 00:00:03,920 --> 00:00:07,800 Speaker 1: twenty two, Wednesday. Oh God, why are dates so hard 3 00:00:07,840 --> 00:00:16,560 Speaker 1: for me? I mean, would never make that mistake. Hi, 4 00:00:16,680 --> 00:00:19,439 Speaker 1: and welcome back to Bloomberg Benchmark, a podcast about the 5 00:00:19,480 --> 00:00:24,079 Speaker 1: global economy. It is Thursday, October twenty two. I'm Tori Stillwell, 6 00:00:24,160 --> 00:00:26,920 Speaker 1: a US economics reporter in d C with Bloomberg News. 7 00:00:27,360 --> 00:00:30,280 Speaker 1: I am with my colleagues and go hosts. Dan Moss, 8 00:00:30,400 --> 00:00:34,080 Speaker 1: our executive editor for International Economics who just landed in Ottawa, 9 00:00:34,560 --> 00:00:38,640 Speaker 1: and Akiedo, our editor for Benchmark in San Francisco. Hey guys, 10 00:00:40,080 --> 00:00:43,320 Speaker 1: you guys are always traveling. I'm just stuck here in 11 00:00:43,440 --> 00:00:46,839 Speaker 1: d C. Dan, send me to Sydney. I'm ready to 12 00:00:47,159 --> 00:00:49,839 Speaker 1: beat it up there. Well, I'll have to give you 13 00:00:49,920 --> 00:00:54,160 Speaker 1: some great recommendations before you go. Okay, deal. So Aki, 14 00:00:54,680 --> 00:00:57,640 Speaker 1: you've been in Japan for part of the last week. 15 00:00:58,080 --> 00:01:01,320 Speaker 1: When did you get back and what perspectives on the 16 00:01:01,320 --> 00:01:04,200 Speaker 1: news did you bring back with you? Well, I got 17 00:01:04,240 --> 00:01:07,679 Speaker 1: back late Saturday night and had a wonderful time there. 18 00:01:07,880 --> 00:01:10,679 Speaker 1: I thought i'd talked about the Bank of Japan or 19 00:01:10,680 --> 00:01:13,800 Speaker 1: maybe Chinese economic statistics today because I was in Asia 20 00:01:13,880 --> 00:01:16,319 Speaker 1: all week. But then I just saw this piece of 21 00:01:16,400 --> 00:01:19,559 Speaker 1: news roll in this morning that was really interesting. It's 22 00:01:19,880 --> 00:01:22,560 Speaker 1: a little nerdy and a little niche, but there's this 23 00:01:22,680 --> 00:01:27,480 Speaker 1: government committee that oversees Sweden's central bank and apparently they're 24 00:01:27,480 --> 00:01:32,280 Speaker 1: getting together to potentially revisit the central bank's mandate there, 25 00:01:32,360 --> 00:01:35,600 Speaker 1: which sounds like really technical jargon and really boring, but 26 00:01:35,680 --> 00:01:37,880 Speaker 1: it could potentially be a really big deal if it 27 00:01:37,920 --> 00:01:42,160 Speaker 1: does end up leading to some legal changes to the 28 00:01:42,160 --> 00:01:46,360 Speaker 1: way Sweden's central bank operates. UM. And right now Sweden 29 00:01:46,400 --> 00:01:49,840 Speaker 1: has this inflation target of two. Some people are saying 30 00:01:49,880 --> 00:01:53,720 Speaker 1: maybe Sweden needs to raise the central bank inflation target 31 00:01:54,120 --> 00:01:56,960 Speaker 1: to create more of a cushion between the target and 32 00:01:57,560 --> 00:02:00,880 Speaker 1: zero percent UM. Some people we're saying, maybe you need 33 00:02:00,920 --> 00:02:04,920 Speaker 1: a lawer the inflation target because you shouldn't really have 34 00:02:04,960 --> 00:02:07,120 Speaker 1: the school that you can't achieve at the end of 35 00:02:07,120 --> 00:02:09,760 Speaker 1: the day. Um. But it comes down to this question 36 00:02:09,840 --> 00:02:13,160 Speaker 1: of what happens when you have the school but you're 37 00:02:13,200 --> 00:02:17,040 Speaker 1: not able to deliver on it for years and years 38 00:02:17,040 --> 00:02:20,880 Speaker 1: and years. So I'm gonna be watching this really closely. Yeah, 39 00:02:20,919 --> 00:02:24,600 Speaker 1: I agree with you, that's really quite sexy. That country 40 00:02:24,680 --> 00:02:27,400 Speaker 1: central bank, the ricks Bank was one of the first 41 00:02:28,720 --> 00:02:33,200 Speaker 1: to adopt the flesh inflation targeting, and more recently, they 42 00:02:33,280 --> 00:02:38,680 Speaker 1: found themselves in the crosshairs um principally but not only, 43 00:02:38,960 --> 00:02:43,560 Speaker 1: the cross hairs of Paul Krookman for raising rates quickly 44 00:02:44,400 --> 00:02:46,680 Speaker 1: once what appeared to be the worst of the global 45 00:02:46,760 --> 00:02:50,400 Speaker 1: recession was over, only to find themselves in a situation 46 00:02:50,440 --> 00:02:54,640 Speaker 1: where they had to reverse and not only cut but 47 00:02:54,760 --> 00:02:59,280 Speaker 1: do que So it's quite fascinating. There's almost a parable 48 00:02:59,360 --> 00:03:02,600 Speaker 1: of moderns in full banking day. Yeah, definitely, they're really 49 00:03:02,880 --> 00:03:06,560 Speaker 1: ahead of the pack. Um, Tori, what's your what's your 50 00:03:06,680 --> 00:03:10,239 Speaker 1: current event chatter of the week. Well, I am focused 51 00:03:10,240 --> 00:03:14,040 Speaker 1: as always on the US and right now we're in 52 00:03:14,080 --> 00:03:19,080 Speaker 1: the midst of this like monthly cycle of housing data 53 00:03:19,160 --> 00:03:21,880 Speaker 1: that we get, and it actually turns out things are 54 00:03:21,919 --> 00:03:25,840 Speaker 1: looking pretty good. Um. Home builder confidence is at a 55 00:03:25,919 --> 00:03:29,960 Speaker 1: decade high, which is great. We got housing starts data 56 00:03:30,200 --> 00:03:36,119 Speaker 1: on Tuesday that showed that construction of new homes rose 57 00:03:36,160 --> 00:03:38,840 Speaker 1: to the second highest level in eight years, So that's 58 00:03:38,880 --> 00:03:41,880 Speaker 1: great news. We're gonna get more data over the next week, 59 00:03:42,480 --> 00:03:45,560 Speaker 1: so I'm definitely keeping an eye on that. Cool. So, Tori, 60 00:03:45,680 --> 00:03:50,280 Speaker 1: the idea is housing will help underpin things while manufacturing 61 00:03:50,320 --> 00:03:54,320 Speaker 1: and exports of suffering. Precisely, consumers have really been doing 62 00:03:54,360 --> 00:03:57,200 Speaker 1: the heavy lifting for growth in the US, and this 63 00:03:58,160 --> 00:04:01,360 Speaker 1: shows that housing is no exception. Well, I want to 64 00:04:01,360 --> 00:04:04,800 Speaker 1: talk this week, Tory about a story that you published 65 00:04:04,800 --> 00:04:08,280 Speaker 1: on Monday called Social skills are the Last Line of 66 00:04:08,320 --> 00:04:12,840 Speaker 1: Defense for humans seeking work? And what better date? We're 67 00:04:12,840 --> 00:04:19,880 Speaker 1: recording this on Wednesday, October, the date Marty McFly arrived 68 00:04:19,920 --> 00:04:23,560 Speaker 1: in the future. Now, we've all written and read stories 69 00:04:23,640 --> 00:04:29,560 Speaker 1: about robotics and their increasing news in say, vehicle assembly lines, 70 00:04:30,200 --> 00:04:35,600 Speaker 1: but most of that commentary has also suggested that more 71 00:04:35,920 --> 00:04:41,880 Speaker 1: people focused softer emotional intelligence skills in the workplace. They're 72 00:04:41,920 --> 00:04:45,159 Speaker 1: still some years away for robots. But you had an 73 00:04:45,240 --> 00:04:50,560 Speaker 1: interesting adventure with somebody or rather something called Amy Ingram 74 00:04:50,600 --> 00:04:52,599 Speaker 1: when you were developing this story. Why don't you talk 75 00:04:52,640 --> 00:04:55,360 Speaker 1: about Amy and how she is essentially what this story 76 00:04:55,440 --> 00:04:59,240 Speaker 1: is about. Yeah, So, when I was doing research into 77 00:04:59,279 --> 00:05:03,600 Speaker 1: this story, I initially saw a great paper out by 78 00:05:03,760 --> 00:05:07,520 Speaker 1: David Demming over at Harvard University about social skills in 79 00:05:07,520 --> 00:05:11,400 Speaker 1: the job market, and he basically found that almost all 80 00:05:11,480 --> 00:05:13,839 Speaker 1: the job growth since nineteen eighty has been in work 81 00:05:13,920 --> 00:05:18,039 Speaker 1: that is social skill intensive. So um, while I was 82 00:05:18,240 --> 00:05:21,880 Speaker 1: doing a little more research. I wanted to reach out 83 00:05:22,000 --> 00:05:25,880 Speaker 1: to this founder of a technology startup. They do virtual 84 00:05:25,960 --> 00:05:29,120 Speaker 1: personal assistance. I send him an email. Can we set 85 00:05:29,160 --> 00:05:31,279 Speaker 1: up a time to chat over the phone. He's like, 86 00:05:31,320 --> 00:05:34,080 Speaker 1: no problem, I'm gonna have Amy set it up. Amy, 87 00:05:34,160 --> 00:05:37,200 Speaker 1: can you just take it from here. Um. So she 88 00:05:37,279 --> 00:05:44,280 Speaker 1: sent me preferred day and time. This is around seven PM, 89 00:05:44,360 --> 00:05:46,719 Speaker 1: and I actually had an event that night, so I 90 00:05:46,760 --> 00:05:51,760 Speaker 1: wasn't checking email. Then at three twenty one am, she 91 00:05:51,880 --> 00:05:55,600 Speaker 1: emails me again and it's like I wanted to follow up. 92 00:05:55,839 --> 00:05:57,640 Speaker 1: Of course I didn't see it because I was sound asleep. 93 00:05:57,880 --> 00:06:00,240 Speaker 1: And four hours later she sent me another mel This 94 00:06:00,360 --> 00:06:03,159 Speaker 1: is about seven twenty. At this point, it's like, I 95 00:06:03,240 --> 00:06:06,440 Speaker 1: haven't heard back from you about this meeting. So fortunately 96 00:06:06,440 --> 00:06:08,560 Speaker 1: for Amy, I'm up and getting ready for work and 97 00:06:08,560 --> 00:06:11,080 Speaker 1: I'm checking my emails. I'm already stressed about getting out 98 00:06:11,120 --> 00:06:14,360 Speaker 1: the door, and Amy keeps bugging me about this meeting 99 00:06:14,400 --> 00:06:16,240 Speaker 1: that I could handle like as soon as I got 100 00:06:16,279 --> 00:06:19,000 Speaker 1: to work, and so I think I got a little 101 00:06:19,040 --> 00:06:22,080 Speaker 1: irritated with her. I thought it was there's no way 102 00:06:22,080 --> 00:06:23,640 Speaker 1: that a humans going to email me at three am 103 00:06:23,640 --> 00:06:27,120 Speaker 1: about a meeting. Um. So it was at that point 104 00:06:27,200 --> 00:06:28,760 Speaker 1: that I was like, all right, this has got to 105 00:06:28,839 --> 00:06:31,680 Speaker 1: be a machine. And when I brought up this whole 106 00:06:31,720 --> 00:06:37,840 Speaker 1: experience to Dennis Mortenson, who is the founder of AMY, 107 00:06:37,960 --> 00:06:40,080 Speaker 1: I guess I should say. Um. He was like, this 108 00:06:40,160 --> 00:06:42,600 Speaker 1: is exactly the thing that we're looking to figure out. 109 00:06:42,640 --> 00:06:46,039 Speaker 1: You know, how many social skills do we need to 110 00:06:46,080 --> 00:06:50,000 Speaker 1: embed these agents, these machines with and how do we 111 00:06:50,120 --> 00:06:57,120 Speaker 1: do that? Okay, So just so we're clear, Amy Ingram, Um, 112 00:06:57,120 --> 00:07:01,360 Speaker 1: she's a she's a machine. She's a virtual person assistant. UM, 113 00:07:01,440 --> 00:07:04,560 Speaker 1: she answers emails, she sends emails as she sets up 114 00:07:04,560 --> 00:07:10,960 Speaker 1: your meetings. And for AMY, ingram stands for artificial intelligence. Yeah, 115 00:07:10,960 --> 00:07:14,000 Speaker 1: they share the share of the same initials UM and 116 00:07:14,360 --> 00:07:18,360 Speaker 1: ingram is actually, God, this is like so over my 117 00:07:18,400 --> 00:07:20,440 Speaker 1: head he was trying to explain it TV. But it's 118 00:07:20,480 --> 00:07:24,880 Speaker 1: like a model used in natural language processing that helps 119 00:07:25,000 --> 00:07:29,440 Speaker 1: machines understand human speech. UM. So you know, for all 120 00:07:29,640 --> 00:07:33,280 Speaker 1: like the suber Nerds out there, they're probably oh, yeah, 121 00:07:33,280 --> 00:07:37,440 Speaker 1: that's so cool. But but yeah, she's she's a robot 122 00:07:37,520 --> 00:07:40,120 Speaker 1: for sure. And every any listeners who are interested in this, 123 00:07:40,280 --> 00:07:43,720 Speaker 1: it's uh. The website is x dot ai. If you 124 00:07:43,760 --> 00:07:45,520 Speaker 1: guys want to check it out. You know, one of 125 00:07:45,520 --> 00:07:48,600 Speaker 1: the things that intrigued me about this story is it 126 00:07:48,800 --> 00:07:53,400 Speaker 1: suggests that because Amy is sort of groping for some 127 00:07:53,880 --> 00:07:58,160 Speaker 1: emotional and empathetic characteristics, that it's not that far off 128 00:07:58,160 --> 00:08:02,880 Speaker 1: in the future. Yeah, I mean, I think computer scientists 129 00:08:02,880 --> 00:08:06,560 Speaker 1: will quickly tell you that they are trying to figure 130 00:08:06,600 --> 00:08:11,200 Speaker 1: out how to at least get robots to mimic social skills. Well, actually, 131 00:08:11,240 --> 00:08:14,360 Speaker 1: you're in San Francisco, You're in the heart of all this, 132 00:08:14,600 --> 00:08:17,040 Speaker 1: and you've spent a lot of time writing about the 133 00:08:17,120 --> 00:08:21,320 Speaker 1: impact of technological advances on the economy, not just the 134 00:08:21,360 --> 00:08:24,440 Speaker 1: macro economy, at what's happening in the micro economy at 135 00:08:24,440 --> 00:08:29,440 Speaker 1: an enterprise level. Did tory story mesh with your experience? 136 00:08:29,480 --> 00:08:33,080 Speaker 1: Didn't resonate with you? Yeah, definitely. I mean, so I 137 00:08:33,120 --> 00:08:37,120 Speaker 1: wrote this big story about artificial intelligence UM about a 138 00:08:37,200 --> 00:08:40,160 Speaker 1: year and a half ago, and even since then, even 139 00:08:40,200 --> 00:08:42,800 Speaker 1: in that year and a half, there's been a remarkable 140 00:08:42,880 --> 00:08:46,240 Speaker 1: amount of progress in that field. It's amazing how quickly 141 00:08:46,559 --> 00:08:49,240 Speaker 1: scientists are making progress in this field. But at the 142 00:08:49,280 --> 00:08:51,680 Speaker 1: same time, you know, one thing that I learned that 143 00:08:51,800 --> 00:08:55,040 Speaker 1: was really important is that there are a lot of 144 00:08:55,040 --> 00:08:58,200 Speaker 1: things that computers can do, but there are many, many, 145 00:08:58,240 --> 00:09:02,160 Speaker 1: many other things that computers cannot do. Um Our human 146 00:09:02,200 --> 00:09:06,040 Speaker 1: brains are amazing. They're so complex in ways that you 147 00:09:06,200 --> 00:09:09,319 Speaker 1: never would have thought before, and are capable of doing 148 00:09:09,400 --> 00:09:12,840 Speaker 1: so many things that robots aren't able to do. Daniel said, Like, 149 00:09:12,920 --> 00:09:16,760 Speaker 1: it feels like we're not that far away from computers 150 00:09:17,160 --> 00:09:20,440 Speaker 1: being able to replicate the full social skills of a 151 00:09:20,559 --> 00:09:23,960 Speaker 1: human being. UM. From the conversations that I had with 152 00:09:24,080 --> 00:09:27,920 Speaker 1: artificial intelligence experts, most people felt that this was something 153 00:09:28,000 --> 00:09:31,679 Speaker 1: that was decades and decades away. We should probably define 154 00:09:31,720 --> 00:09:35,120 Speaker 1: what where, what social skills aren't, what qualifies as social skills? 155 00:09:35,120 --> 00:09:38,040 Speaker 1: What you guys think, Yeah, definitely, I mean, let's let's 156 00:09:38,040 --> 00:09:40,400 Speaker 1: start with this. So, like, you know, imagine this kind 157 00:09:40,400 --> 00:09:45,720 Speaker 1: of spectrum between of tasks. Some's tasks are really repetitive. 158 00:09:45,760 --> 00:09:47,920 Speaker 1: You're doing the same thing over and over again, and 159 00:09:47,960 --> 00:09:50,560 Speaker 1: it's easy to write out a full instruction manual of 160 00:09:50,559 --> 00:09:53,240 Speaker 1: what you're supposed to do. There's this, you know, on 161 00:09:53,280 --> 00:09:56,400 Speaker 1: the other end of the spectrum, there are tasks that 162 00:09:56,520 --> 00:10:00,800 Speaker 1: are really complex, really diverse. You're doing something different every 163 00:10:00,840 --> 00:10:03,719 Speaker 1: single day. It's really creative. You're you're coming up with 164 00:10:03,760 --> 00:10:06,920 Speaker 1: novel solutions all the time. The kind of tasks that 165 00:10:06,960 --> 00:10:10,160 Speaker 1: are on the repetitive end of the spectrum are really 166 00:10:10,240 --> 00:10:14,440 Speaker 1: easy to replicate, and those jobs are already gone. You 167 00:10:14,480 --> 00:10:16,880 Speaker 1: can think of something you would do in a factory 168 00:10:16,880 --> 00:10:19,280 Speaker 1: and like an auto factory, for example, where you're putting 169 00:10:19,280 --> 00:10:22,120 Speaker 1: on the same part over and over again. Those kind 170 00:10:22,120 --> 00:10:25,319 Speaker 1: of jobs are gone. But you think of something really complicated, 171 00:10:25,800 --> 00:10:29,760 Speaker 1: like um being an executive editor like Dan, where you're 172 00:10:29,760 --> 00:10:33,840 Speaker 1: making all these different decisions every single day, you're providing oversight, 173 00:10:33,960 --> 00:10:36,520 Speaker 1: You're coming up with new things. That's the kind of 174 00:10:36,520 --> 00:10:40,000 Speaker 1: thing a computer hasn't been able to do yet. Right, 175 00:10:40,040 --> 00:10:45,040 Speaker 1: So it's a sense of collaboration, reading people like, kind 176 00:10:45,080 --> 00:10:49,040 Speaker 1: of working off of social cues. All those things qualify 177 00:10:49,280 --> 00:10:52,320 Speaker 1: social skills, and at least one expert that I talked 178 00:10:52,400 --> 00:10:57,760 Speaker 1: to said that upward of the workforce needs at least 179 00:10:57,840 --> 00:11:02,319 Speaker 1: some sort of collaboration to get their job done. So 180 00:11:02,360 --> 00:11:05,920 Speaker 1: it's a lot. If we've advanced this quickly, what makes 181 00:11:06,000 --> 00:11:10,760 Speaker 1: us think that that final stage is still some steps away? 182 00:11:10,960 --> 00:11:13,920 Speaker 1: Because zacky, this made me recall a story you wrote 183 00:11:14,320 --> 00:11:19,280 Speaker 1: in twenty fourteen about robots being deployed in lawyers offices 184 00:11:19,840 --> 00:11:22,920 Speaker 1: doing legal work. You know, that's a whole other step 185 00:11:22,960 --> 00:11:26,160 Speaker 1: away from the generic shot of a robot putting an 186 00:11:26,200 --> 00:11:30,360 Speaker 1: engine head into an suv. Yeah, definitely. So this was 187 00:11:30,760 --> 00:11:34,800 Speaker 1: you know, a very specific task in that lawyers UM 188 00:11:35,080 --> 00:11:38,200 Speaker 1: currently are performing called you know, it's document reading in 189 00:11:38,280 --> 00:11:42,120 Speaker 1: this initial stage of litigation where you have to decide 190 00:11:42,160 --> 00:11:45,400 Speaker 1: which documents are relevant and which documents aren't relevant to 191 00:11:45,440 --> 00:11:48,960 Speaker 1: your case. Before, it was impossible to have computers do 192 00:11:49,040 --> 00:11:51,880 Speaker 1: that because it was just too complicated of a task. 193 00:11:52,400 --> 00:11:56,040 Speaker 1: Every case of litigation was too different in order to 194 00:11:56,120 --> 00:11:58,720 Speaker 1: kind of come up with this like master set of rules. 195 00:11:59,280 --> 00:12:01,760 Speaker 1: But they found a way to do that do that 196 00:12:01,920 --> 00:12:06,120 Speaker 1: by um giving a small group of lawyers the small 197 00:12:06,240 --> 00:12:10,000 Speaker 1: subset of documents and having them say whether something is 198 00:12:10,040 --> 00:12:13,080 Speaker 1: relevant or not relevant, and the computer is watching those 199 00:12:13,160 --> 00:12:16,640 Speaker 1: humans make those decisions, and then the computer learns from 200 00:12:16,679 --> 00:12:21,040 Speaker 1: that experience and is able to amplify that experience across 201 00:12:21,040 --> 00:12:25,240 Speaker 1: a much broader set of documents for that specific case. UM. 202 00:12:25,240 --> 00:12:29,360 Speaker 1: This leads to a lot of savings and dollars you know, 203 00:12:29,400 --> 00:12:32,360 Speaker 1: in terms of like the labor that you have to employ, 204 00:12:32,400 --> 00:12:35,920 Speaker 1: and it also makes the litigation process go a lot more, 205 00:12:36,200 --> 00:12:40,320 Speaker 1: um go much faster. But I feel like to really 206 00:12:40,360 --> 00:12:44,520 Speaker 1: illustrate how far away robots are in terms of just 207 00:12:44,679 --> 00:12:48,480 Speaker 1: really being able to moot very well, maybe we should 208 00:12:48,520 --> 00:12:50,800 Speaker 1: try to have a little bit of a conversation with 209 00:12:50,920 --> 00:12:54,160 Speaker 1: Siri and see how that goes. Yeah, let's do it. 210 00:12:55,000 --> 00:13:01,240 Speaker 1: Let's see, Siri, I'm feeling really sad today. I would 211 00:13:01,240 --> 00:13:03,840 Speaker 1: give you a shoulder to cry on, Victoria if I 212 00:13:03,920 --> 00:13:08,480 Speaker 1: had one. Thanks, that's really nice, but it didn't really 213 00:13:08,520 --> 00:13:14,720 Speaker 1: make me feel better. You're welcome, okay, all right, Well, 214 00:13:15,080 --> 00:13:17,800 Speaker 1: I'm not sure that has quite the same effect as 215 00:13:17,840 --> 00:13:21,679 Speaker 1: if you know, Aki was asking me exactly why I 216 00:13:21,840 --> 00:13:24,680 Speaker 1: was sad. I feel like if if I told you 217 00:13:24,760 --> 00:13:27,160 Speaker 1: I was sad and you didn't ask me why or 218 00:13:27,559 --> 00:13:29,959 Speaker 1: try to make me feel a little bit better about it, 219 00:13:30,280 --> 00:13:34,760 Speaker 1: I might have to unfriend you. So that's that's a 220 00:13:35,120 --> 00:13:38,200 Speaker 1: you know, a kind of social skill that Siri has 221 00:13:38,240 --> 00:13:43,840 Speaker 1: been unable to learn yet. Can we follow or um 222 00:13:44,280 --> 00:13:47,960 Speaker 1: framed Amy? I mean, let's talk a bit about Amy. 223 00:13:48,120 --> 00:13:53,560 Speaker 1: Does she or it exists inside a computer, inside a 224 00:13:53,640 --> 00:13:58,800 Speaker 1: micro ship or in this chap's office tory? Is there 225 00:13:58,840 --> 00:14:03,679 Speaker 1: a humanoid look machine typing away to keyboard and her 226 00:14:04,040 --> 00:14:07,720 Speaker 1: or it the name happens to be Amy. What are 227 00:14:07,720 --> 00:14:11,200 Speaker 1: the dimensions that we're dealing with here? Amy is a machine. 228 00:14:11,200 --> 00:14:13,480 Speaker 1: I mean she's not. She's not a robot sitting at 229 00:14:13,480 --> 00:14:17,280 Speaker 1: a computer typing. You know, she she recognized. The machine 230 00:14:17,360 --> 00:14:21,840 Speaker 1: recognizes patterns in the emails, you know, today, tomorrow, a time, 231 00:14:22,560 --> 00:14:26,360 Speaker 1: and they use that very similar to how Google does. 232 00:14:26,480 --> 00:14:28,800 Speaker 1: If if any of the listeners out there have Gmail, 233 00:14:28,920 --> 00:14:31,560 Speaker 1: I'm sure they see when when someone sends an email 234 00:14:31,600 --> 00:14:33,520 Speaker 1: with a time and a date, they'll see these dash 235 00:14:33,600 --> 00:14:37,120 Speaker 1: lines under it and you can add it to your calendar. Um. 236 00:14:37,160 --> 00:14:40,520 Speaker 1: You know, Amy is analyzing these these dates and these 237 00:14:40,600 --> 00:14:44,600 Speaker 1: times and these words, um and being able to respond 238 00:14:44,640 --> 00:14:48,240 Speaker 1: to that and and added to your calendar. So UM, 239 00:14:48,280 --> 00:14:49,680 Speaker 1: I don't think you should think of it as like 240 00:14:49,880 --> 00:14:52,760 Speaker 1: a robot sitting at a at a computer, like slaving 241 00:14:52,760 --> 00:15:01,520 Speaker 1: away at a keyboard. Should we be scared of Amy? Well, 242 00:15:01,560 --> 00:15:04,320 Speaker 1: I don't think we should be scared of Amy. As 243 00:15:04,360 --> 00:15:08,920 Speaker 1: a Dennis Mortinson again, the creator Um said he's just 244 00:15:09,000 --> 00:15:13,840 Speaker 1: trying to sort of eliminate this email back and forth 245 00:15:13,960 --> 00:15:17,520 Speaker 1: that any human has to do. Whether you're lucky enough 246 00:15:17,520 --> 00:15:20,360 Speaker 1: to have a personal assistant and your personal assistant does it, 247 00:15:20,640 --> 00:15:23,600 Speaker 1: or like Aggie and I have to do it. Um, 248 00:15:23,640 --> 00:15:26,160 Speaker 1: he just wants to eliminate that, so it frees you 249 00:15:26,280 --> 00:15:29,040 Speaker 1: up to do things that are more productive and more 250 00:15:29,160 --> 00:15:32,040 Speaker 1: worth a humans time. So I don't think we should 251 00:15:32,080 --> 00:15:34,320 Speaker 1: be scared of amy. It's really interesting because I was 252 00:15:34,360 --> 00:15:38,160 Speaker 1: doing research for this story, I was able to talk 253 00:15:38,200 --> 00:15:40,600 Speaker 1: to people about what they think the future looks like 254 00:15:40,640 --> 00:15:44,120 Speaker 1: in terms of how much work robots will be taking over, 255 00:15:44,200 --> 00:15:50,160 Speaker 1: et cetera. And Paedro Domingos over at the University of Washington. 256 00:15:50,520 --> 00:15:52,480 Speaker 1: He's he's the author of a new book on machine 257 00:15:52,560 --> 00:15:55,160 Speaker 1: learning called The Master Algorithm, if any of you guys 258 00:15:55,200 --> 00:15:58,560 Speaker 1: want to check it out. Um, but he sort of 259 00:15:58,680 --> 00:16:03,080 Speaker 1: envisioned this world are robots will be able to do 260 00:16:03,160 --> 00:16:07,000 Speaker 1: basically everything that humans currently do now in their work, 261 00:16:07,600 --> 00:16:09,960 Speaker 1: but they're going to be certain things that will want 262 00:16:10,000 --> 00:16:13,000 Speaker 1: a human touch for. So, you know, you don't want 263 00:16:13,080 --> 00:16:17,040 Speaker 1: to go to the bar and like pour out like 264 00:16:17,120 --> 00:16:20,360 Speaker 1: your feelings about how your girlfriend dumped you to a 265 00:16:20,520 --> 00:16:24,000 Speaker 1: robot bartender. You want a real bartender who can take 266 00:16:24,040 --> 00:16:26,360 Speaker 1: a shot with you and who you can like go 267 00:16:26,400 --> 00:16:28,560 Speaker 1: back and forth on how terrible she was. Like. You 268 00:16:28,600 --> 00:16:31,240 Speaker 1: want someone who really understands what you're looking for. And 269 00:16:31,320 --> 00:16:33,520 Speaker 1: he says that these things are going to start to 270 00:16:33,680 --> 00:16:36,040 Speaker 1: command a premium in the labor market, and there's gonna 271 00:16:36,040 --> 00:16:39,280 Speaker 1: be way fewer of them, but they will still exist. 272 00:16:39,320 --> 00:16:42,600 Speaker 1: They'll be a luxury sort of. But also in that 273 00:16:42,680 --> 00:16:46,160 Speaker 1: new world, we're gonna totally have to rethink how people 274 00:16:46,320 --> 00:16:50,440 Speaker 1: get money, how people live. Um. There's this idea floating 275 00:16:50,480 --> 00:16:55,200 Speaker 1: around about basic income, and it's this this theory that 276 00:16:55,280 --> 00:16:58,560 Speaker 1: you know, as robots basically take over the way that 277 00:16:58,640 --> 00:17:01,280 Speaker 1: we earn a livelihood, and you're gonna have to be 278 00:17:01,360 --> 00:17:05,000 Speaker 1: able to guarantee people some sort of fixed amount of 279 00:17:05,040 --> 00:17:07,840 Speaker 1: income um that they can spend, whether that comes from 280 00:17:07,840 --> 00:17:10,560 Speaker 1: the government, whether that comes from taxing the people who 281 00:17:10,600 --> 00:17:13,440 Speaker 1: create robots that are gonna take all our jobs. We're 282 00:17:13,440 --> 00:17:16,880 Speaker 1: gonna have to change sort of the distribution of capital 283 00:17:16,920 --> 00:17:19,679 Speaker 1: because right now, you earn money based on you know, 284 00:17:19,720 --> 00:17:23,360 Speaker 1: the scarcity of your labor um, and and once robots 285 00:17:23,359 --> 00:17:26,439 Speaker 1: are taking that over, the money is going to be 286 00:17:26,480 --> 00:17:29,720 Speaker 1: held by the people who control the robots, who invented them. 287 00:17:29,760 --> 00:17:32,080 Speaker 1: So we're gonna have to rethink that. Going back to 288 00:17:32,119 --> 00:17:36,080 Speaker 1: your bartending example, like if you're really really amazing bartender 289 00:17:36,200 --> 00:17:41,360 Speaker 1: with these incredible, interpersonable skills and you're really warm and 290 00:17:41,440 --> 00:17:45,000 Speaker 1: you make these really cool creative cocktails, then yeah, your 291 00:17:45,080 --> 00:17:48,600 Speaker 1: future is golden. You're gonna be fine, and if anything, 292 00:17:48,640 --> 00:17:50,800 Speaker 1: you're probably going to be making even more money in 293 00:17:50,840 --> 00:17:53,320 Speaker 1: the future. But if you're kind of like the middle 294 00:17:53,440 --> 00:17:56,199 Speaker 1: of the road bartender, like you know, you're you're kind 295 00:17:56,240 --> 00:17:59,560 Speaker 1: of friendly, but not that friendly, and your cocktails aren't 296 00:17:59,560 --> 00:18:02,240 Speaker 1: that created of it's pretty much, you know, the same 297 00:18:02,280 --> 00:18:07,160 Speaker 1: stuff as what's on other menus around the city, then uh, 298 00:18:07,280 --> 00:18:10,040 Speaker 1: your future isn't that bright. That probably means that your 299 00:18:10,119 --> 00:18:16,679 Speaker 1: job is um much more vulnerable to robots and software. Right, 300 00:18:16,840 --> 00:18:19,959 Speaker 1: creativity is going to be just a huge game changer. 301 00:18:20,600 --> 00:18:23,159 Speaker 1: It already is now, but that's going to be really 302 00:18:23,160 --> 00:18:26,320 Speaker 1: how you can leverage yourself, is what these experts said. Right, 303 00:18:26,359 --> 00:18:28,160 Speaker 1: So I can't help but feel like it would lead 304 00:18:28,200 --> 00:18:33,000 Speaker 1: to more inequality in terms of who wins and who loses. Yeah, 305 00:18:33,119 --> 00:18:34,919 Speaker 1: you know, it reminds me of some of the issues 306 00:18:34,960 --> 00:18:38,879 Speaker 1: we talked about in our inaugural episode with our friend 307 00:18:38,920 --> 00:18:43,280 Speaker 1: Barry Bosworth of Brookings. He sketched for us a history 308 00:18:43,280 --> 00:18:49,399 Speaker 1: of automation going back to the spinning wheel, steam engine, electricity, 309 00:18:49,800 --> 00:18:56,000 Speaker 1: personal computers, where do you think Amy and machines like 310 00:18:56,119 --> 00:19:01,359 Speaker 1: her fit into that historical sweepe. There's this weird dichotomy 311 00:19:01,400 --> 00:19:03,879 Speaker 1: where we have people who are both afraid that robots 312 00:19:03,880 --> 00:19:06,480 Speaker 1: are going to take all our jobs and also afraid 313 00:19:06,560 --> 00:19:11,000 Speaker 1: that we're not innovating enough. Uh, that productivity is just 314 00:19:11,080 --> 00:19:16,320 Speaker 1: gonna be sluggish forever. I don't I don't know which 315 00:19:16,359 --> 00:19:20,080 Speaker 1: one is right. I don't think it's probably either extreme. 316 00:19:20,160 --> 00:19:23,560 Speaker 1: I think it's going to be somewhere in the middle. Um, 317 00:19:23,760 --> 00:19:26,920 Speaker 1: the so called Internet of things is really changing how 318 00:19:26,960 --> 00:19:31,119 Speaker 1: we look at productivity and how we use technology to 319 00:19:31,280 --> 00:19:34,639 Speaker 1: get data and to drive decision making and to to 320 00:19:34,800 --> 00:19:39,399 Speaker 1: make our processes more efficient and smarter. Um. It just 321 00:19:39,680 --> 00:19:41,600 Speaker 1: it may take some time for these things to show 322 00:19:41,680 --> 00:19:44,120 Speaker 1: up in the data. I think that's what people are 323 00:19:44,160 --> 00:19:45,879 Speaker 1: trying to figure out. But Dan, I think I have 324 00:19:45,920 --> 00:19:49,440 Speaker 1: a more important question, and that is are you a robot? 325 00:19:50,280 --> 00:20:02,159 Speaker 1: And Victoria we should probably wrap up, Tori. Tori, you 326 00:20:02,160 --> 00:20:05,399 Speaker 1: know we started this episode with a question of whether 327 00:20:05,480 --> 00:20:08,880 Speaker 1: a robot will be able to take our jobs? Should 328 00:20:08,920 --> 00:20:11,880 Speaker 1: we ask Sirie? Oh? Yeah, we should definitely ask Seria. Okay, 329 00:20:11,960 --> 00:20:16,240 Speaker 1: let's see what she says. Is a robot going to 330 00:20:16,359 --> 00:20:22,960 Speaker 1: take all of our jobs. Interesting question, Victoria, all right, 331 00:20:23,080 --> 00:20:30,479 Speaker 1: that was a super late answer, Siri. Yes, yes, old 332 00:20:30,520 --> 00:20:35,879 Speaker 1: technology for the win. Sirie needs some work. Thanks again 333 00:20:35,960 --> 00:20:38,920 Speaker 1: for listening to Bloomberg Benchmark. We will be back next 334 00:20:38,960 --> 00:20:44,120 Speaker 1: week and you can find us on Bloomberg dot com, iTunes, Podcasts, SoundCloud, Stitcher, 335 00:20:44,600 --> 00:20:47,959 Speaker 1: all the places um as well as on the Bloomberg terminal. 336 00:20:48,400 --> 00:20:51,520 Speaker 1: And if you're on any of those platforms, please take 337 00:20:51,520 --> 00:20:53,640 Speaker 1: a moment to rate and review the show so other 338 00:20:53,720 --> 00:20:56,040 Speaker 1: listeners can find us and let us know what you 339 00:20:56,119 --> 00:20:58,760 Speaker 1: thought of this show. You can reach us and follow 340 00:20:58,880 --> 00:21:02,280 Speaker 1: us on Twitter at Daniel mass d C, Tori Stillwell, 341 00:21:02,440 --> 00:21:06,320 Speaker 1: and Kita seven. We'll see you next week. And I 342 00:21:06,359 --> 00:21:07,440 Speaker 1: am not a robot.