1 00:00:00,160 --> 00:00:02,840 Speaker 1: Hey everyone, it's Robert and Joe here. Today we've got 2 00:00:02,880 --> 00:00:04,760 Speaker 1: something a little bit different to share with you. It 3 00:00:04,840 --> 00:00:08,039 Speaker 1: is a new season of the Smart Talks with IBM 4 00:00:08,160 --> 00:00:09,160 Speaker 1: podcast series. 5 00:00:09,640 --> 00:00:11,639 Speaker 2: Today we are witnessed to one of those rare moments 6 00:00:11,680 --> 00:00:14,400 Speaker 2: in history, the rise of an innovative technology with the 7 00:00:14,400 --> 00:00:18,680 Speaker 2: potential to radically transform business and society forever. The technology, 8 00:00:18,760 --> 00:00:22,200 Speaker 2: of course, is artificial intelligence, and it's the central focus 9 00:00:22,239 --> 00:00:24,840 Speaker 2: for this new season of Smart Talks with IBM. 10 00:00:25,320 --> 00:00:28,400 Speaker 1: Join hosts from your favorite Pushkin podcasts as they talk 11 00:00:28,480 --> 00:00:31,680 Speaker 1: with industry experts and leaders to explore how businesses can 12 00:00:31,720 --> 00:00:35,400 Speaker 1: integrate AI into their workflows and help drive real change 13 00:00:35,400 --> 00:00:38,240 Speaker 1: in this new era of AI. And of course, host 14 00:00:38,280 --> 00:00:40,479 Speaker 1: Malcolm Gladwell will be there to guide you through the 15 00:00:40,479 --> 00:00:42,640 Speaker 1: season and throw in his two cents as well. 16 00:00:43,120 --> 00:00:46,200 Speaker 2: Look out for new episodes of Smart Talks with IBM 17 00:00:46,400 --> 00:00:49,559 Speaker 2: every other week on the iHeartRadio app, Apple Podcasts, or 18 00:00:49,600 --> 00:00:53,360 Speaker 2: wherever you get your podcasts. And learn more at IBM 19 00:00:53,479 --> 00:00:57,720 Speaker 2: dot com, slash smart Talks. 20 00:01:02,280 --> 00:01:03,160 Speaker 3: Pushkin. 21 00:01:06,800 --> 00:01:14,200 Speaker 4: Welcome, Welcome, Welcome to Smart Talks with IBM. 22 00:01:14,240 --> 00:01:17,880 Speaker 3: Hello, Hello, Welcome to Smart Talks with IBM, a podcast 23 00:01:18,040 --> 00:01:23,920 Speaker 3: from Pushkin Industries. iHeartRadio and IBM. I'm Malcolm Gladwell. This season, 24 00:01:24,160 --> 00:01:27,440 Speaker 3: we're diving back into the world of artificial intelligence, but 25 00:01:27,520 --> 00:01:32,840 Speaker 3: with a focus on the powerful concept of open its possibilities, implications, 26 00:01:33,200 --> 00:01:36,960 Speaker 3: and misconceptions. We'll look at openness from a variety of 27 00:01:37,000 --> 00:01:40,880 Speaker 3: angles and explore how the concept is already reshaping industries, 28 00:01:41,319 --> 00:01:44,920 Speaker 3: ways of doing business, and our very notion of what's possible. 29 00:01:45,360 --> 00:01:47,880 Speaker 3: And for the first episode of this season, we're bringing 30 00:01:47,920 --> 00:01:52,080 Speaker 3: you a special conversation. I recently sat down with Rob Thomas. 31 00:01:52,560 --> 00:01:55,560 Speaker 3: Rob is the senior vice president of Software and chief 32 00:01:55,600 --> 00:01:59,160 Speaker 3: Commercial Officer of IBM. I spoke to him in front 33 00:01:59,160 --> 00:02:01,800 Speaker 3: of a live audience as part of New York Tech Week. 34 00:02:02,440 --> 00:02:06,520 Speaker 3: We discussed how businesses can harness the immense productivity benefits 35 00:02:06,560 --> 00:02:11,000 Speaker 3: of AI while implementing it in a responsible and ethical manner. 36 00:02:11,520 --> 00:02:15,240 Speaker 3: We also broke down a fascinating concept that Rob believes 37 00:02:15,280 --> 00:02:20,600 Speaker 3: about AI, known as the productivity paradox. Okay, let's get 38 00:02:20,639 --> 00:02:31,120 Speaker 3: to the conversation. How are we doing good? Rob? This 39 00:02:31,200 --> 00:02:35,280 Speaker 3: is our our second time. We did one of these 40 00:02:35,560 --> 00:02:37,720 Speaker 3: in the middle of the pandemic, but now it's all 41 00:02:37,760 --> 00:02:39,440 Speaker 3: such a blur now that us can figure out when 42 00:02:39,440 --> 00:02:39,680 Speaker 3: it was. 43 00:02:39,800 --> 00:02:42,160 Speaker 4: I know it's hard to those are like a blurry years. 44 00:02:42,200 --> 00:02:43,600 Speaker 5: You don't know what happened, right. 45 00:02:43,760 --> 00:02:46,680 Speaker 3: But well, it's good to see you, to meet you again. 46 00:02:47,680 --> 00:02:50,240 Speaker 3: I wanted to start by going back. You've been at 47 00:02:50,240 --> 00:02:52,560 Speaker 3: IBM twenty years? Is that right? 48 00:02:52,960 --> 00:02:54,880 Speaker 5: Twenty five in July, believe it or not. 49 00:02:55,040 --> 00:02:56,760 Speaker 3: So you were a kid when you joined. 50 00:02:56,840 --> 00:02:57,360 Speaker 5: I was four. 51 00:03:00,200 --> 00:03:04,639 Speaker 3: W I want to contrast present day Rob and twenty 52 00:03:04,680 --> 00:03:09,560 Speaker 3: five years ago Rob. When you arrive at IBM, what 53 00:03:09,639 --> 00:03:11,520 Speaker 3: do you think your job is going to be? It, 54 00:03:11,600 --> 00:03:13,400 Speaker 3: your career is going? Where do you think the kind 55 00:03:13,440 --> 00:03:15,760 Speaker 3: of problems you're going to be addressing are? 56 00:03:17,560 --> 00:03:19,799 Speaker 4: Well, it's kind of surreal because I I joined IBM 57 00:03:19,880 --> 00:03:25,600 Speaker 4: Consulting and I'm coming out of school, and you quickly realize, wait, 58 00:03:25,639 --> 00:03:28,079 Speaker 4: the job of a consultant is to tell other companies 59 00:03:28,120 --> 00:03:30,720 Speaker 4: what to do. And I was like, I literally know nothing, 60 00:03:32,320 --> 00:03:34,160 Speaker 4: and so you're immediately trying to figure out, so how 61 00:03:34,160 --> 00:03:35,760 Speaker 4: am I going to be relevant given that I know 62 00:03:35,800 --> 00:03:39,160 Speaker 4: absolutely nothing to advise other companies on what they should 63 00:03:39,200 --> 00:03:42,000 Speaker 4: be doing. And I remember it well, like we were 64 00:03:42,040 --> 00:03:45,120 Speaker 4: sitting in a room. When you're a consultant, you're waiting 65 00:03:45,160 --> 00:03:46,520 Speaker 4: for somebody else to find work for you. 66 00:03:47,600 --> 00:03:48,240 Speaker 5: A bunch of us. 67 00:03:48,160 --> 00:03:50,920 Speaker 4: Sitting in a room and somebody walks in and says, 68 00:03:52,160 --> 00:03:53,560 Speaker 4: we need somebody that knows visio. 69 00:03:53,760 --> 00:03:56,520 Speaker 5: Does anybody know visio? I'd never heard of visio. 70 00:03:56,560 --> 00:03:59,040 Speaker 4: I don't know if anybody in the room has. So 71 00:03:59,120 --> 00:04:02,800 Speaker 4: everybody's likes sit around looking at their shoes. So finally 72 00:04:02,800 --> 00:04:05,640 Speaker 4: I was like, I know it. So I raised my hand. 73 00:04:05,640 --> 00:04:08,000 Speaker 4: They're like, great, we got a project for you next week. 74 00:04:08,880 --> 00:04:10,920 Speaker 4: So I was like, all right, I have like three 75 00:04:11,000 --> 00:04:14,920 Speaker 4: days to figure out what visio is, and I hope 76 00:04:14,920 --> 00:04:17,000 Speaker 4: I can actually figure out how to use it now. Luckily, 77 00:04:17,520 --> 00:04:20,479 Speaker 4: it wasn't like a programming language. I mean, it's pretty 78 00:04:20,520 --> 00:04:24,279 Speaker 4: much a drag and drop capability. And so I literally 79 00:04:24,600 --> 00:04:25,800 Speaker 4: left the office. 80 00:04:25,760 --> 00:04:29,080 Speaker 5: Went to a bookstore, bought the first three books. 81 00:04:28,760 --> 00:04:31,080 Speaker 4: On Visio I could find, spent the whole week in 82 00:04:31,120 --> 00:04:34,720 Speaker 4: reading the books, and showed up and got their work 83 00:04:34,760 --> 00:04:37,080 Speaker 4: on the project. And so it was a bit of 84 00:04:37,080 --> 00:04:40,760 Speaker 4: a risky move, but I think that's kind of you 85 00:04:42,200 --> 00:04:44,400 Speaker 4: doing this well. But if you don't take risk, you'll 86 00:04:44,480 --> 00:04:48,600 Speaker 4: never you'll never achieve. And so does some extent. Everybody's 87 00:04:48,640 --> 00:04:51,320 Speaker 4: making everything up all the time. It's like, can you 88 00:04:51,440 --> 00:04:54,880 Speaker 4: learn faster than somebody else. Is what the difference is 89 00:04:55,320 --> 00:04:58,440 Speaker 4: in almost every part of life. And so it was 90 00:04:58,440 --> 00:05:01,040 Speaker 4: not planned, but it was an accident, but it kind 91 00:05:01,080 --> 00:05:02,840 Speaker 4: of forced me to figure out that you're going to 92 00:05:02,920 --> 00:05:03,719 Speaker 4: have to figure things out. 93 00:05:04,080 --> 00:05:06,560 Speaker 3: You know, we're here to talk about AI, and I'm 94 00:05:06,600 --> 00:05:12,960 Speaker 3: curious about the evolution of your understanding or IBM's understanding 95 00:05:13,000 --> 00:05:15,440 Speaker 3: of my AI. At what point in the last twenty 96 00:05:15,440 --> 00:05:18,000 Speaker 3: five years do you begin to think, oh, this is 97 00:05:18,040 --> 00:05:19,760 Speaker 3: really going to be at the core of what we 98 00:05:20,120 --> 00:05:22,120 Speaker 3: think about and work on at this company. 99 00:05:24,040 --> 00:05:28,200 Speaker 4: The computer scientist John McCarthy, he was he's the person 100 00:05:28,240 --> 00:05:31,880 Speaker 4: that's credited with coining the phrase artificial intelligence. It's like 101 00:05:31,920 --> 00:05:37,120 Speaker 4: in the fifties, and he made an interesting comedy said 102 00:05:37,279 --> 00:05:39,600 Speaker 4: he said, once it works, it's no longer called AI, 103 00:05:41,839 --> 00:05:45,040 Speaker 4: and that then became it's called like the AI effect, 104 00:05:45,120 --> 00:05:48,880 Speaker 4: which is it seems very difficult, very mysterious, but once 105 00:05:48,920 --> 00:05:53,080 Speaker 4: it becomes commonplace, it's just no longer what it is. 106 00:05:53,440 --> 00:05:55,880 Speaker 4: And so if you put that frame on it, I 107 00:05:55,880 --> 00:05:58,240 Speaker 4: think we've always been doing AI at some level. And 108 00:05:58,320 --> 00:06:00,479 Speaker 4: I even think back to when I joined I in 109 00:06:00,600 --> 00:06:05,120 Speaker 4: ninety nine. At that point there was work on rules 110 00:06:05,160 --> 00:06:11,160 Speaker 4: based engines, analytics, all of this was happening, So it 111 00:06:11,200 --> 00:06:14,280 Speaker 4: all depends on how you really define that term. 112 00:06:14,520 --> 00:06:16,640 Speaker 5: You could argue that elements of. 113 00:06:16,600 --> 00:06:21,719 Speaker 4: Statistics, probability, it's not exactly AI, but it certainly feeds 114 00:06:21,760 --> 00:06:22,640 Speaker 4: into it. 115 00:06:22,680 --> 00:06:23,840 Speaker 5: And so I feel like. 116 00:06:23,839 --> 00:06:26,960 Speaker 4: We've been working on this topic of how do we 117 00:06:27,040 --> 00:06:33,040 Speaker 4: deliver better insights better automation since IBM was formed. If 118 00:06:33,040 --> 00:06:35,359 Speaker 4: you read about what Thomas Watson Junior did, that was 119 00:06:35,440 --> 00:06:40,320 Speaker 4: all about automating tasks that AI will probably certainly not 120 00:06:40,360 --> 00:06:44,320 Speaker 4: by today's definition, but it's in the same zip code. 121 00:06:44,520 --> 00:06:46,680 Speaker 3: So from your perspective, it feels a lot more like 122 00:06:46,720 --> 00:06:48,160 Speaker 3: an evolution than a revolution. 123 00:06:48,279 --> 00:06:51,440 Speaker 4: Is that a fair statement, yes, which I think most 124 00:06:51,520 --> 00:06:55,680 Speaker 4: great things in technology tend to happen that way. Many 125 00:06:55,760 --> 00:06:58,760 Speaker 4: of the revolutions, if you will, tend to fizzle out. 126 00:07:00,080 --> 00:07:02,360 Speaker 3: Even given that is there, I guess what I'm asking is, 127 00:07:02,560 --> 00:07:05,720 Speaker 3: I'm curious about whether there was a moment in that 128 00:07:05,760 --> 00:07:09,720 Speaker 3: evolution when you had to readjust your expectations about what 129 00:07:11,160 --> 00:07:13,160 Speaker 3: AI was going to be capable of. I mean, was there, 130 00:07:13,600 --> 00:07:18,000 Speaker 3: you know, was there a particular innovation or a particular 131 00:07:18,640 --> 00:07:21,560 Speaker 3: problem that was solved that made you think, oh, this 132 00:07:21,720 --> 00:07:24,360 Speaker 3: is different than what I thought. 133 00:07:26,120 --> 00:07:29,800 Speaker 4: I would say the moments that caught our attention certainly 134 00:07:31,120 --> 00:07:35,160 Speaker 4: casper Off winning the chess tournament Nobody or Deep Blue 135 00:07:35,160 --> 00:07:38,080 Speaker 4: beating casper Off. I should say, nobody really thought that 136 00:07:38,160 --> 00:07:43,760 Speaker 4: was possible before that, and then it was Watson winning Jeopardy. 137 00:07:44,160 --> 00:07:46,840 Speaker 4: These were moments that said, maybe there's more here than 138 00:07:46,840 --> 00:07:51,080 Speaker 4: we even thought was possible. And so I do think 139 00:07:51,080 --> 00:07:57,440 Speaker 4: there's points in time where we realized maybe way more could. 140 00:07:57,320 --> 00:07:59,400 Speaker 5: Be done than we had even imagined. 141 00:08:00,120 --> 00:08:04,360 Speaker 4: But I do think it's consistent progress every month and 142 00:08:04,400 --> 00:08:10,400 Speaker 4: every year versus some seminal moment. Now, certainly large language 143 00:08:10,440 --> 00:08:13,040 Speaker 4: models as of recent have caught everybody's attention because it 144 00:08:13,080 --> 00:08:17,600 Speaker 4: has a direct consumer application, but I would almost think 145 00:08:17,640 --> 00:08:21,640 Speaker 4: of that as what Netscape was for the for the 146 00:08:21,640 --> 00:08:25,520 Speaker 4: web browser. Yeah, it brought the Internet to everybody, but 147 00:08:26,160 --> 00:08:28,400 Speaker 4: that didn't become the Internet per se. 148 00:08:29,200 --> 00:08:29,520 Speaker 5: Yeah. 149 00:08:29,840 --> 00:08:32,440 Speaker 3: I have a cousin who worked for IBM for forty 150 00:08:32,480 --> 00:08:35,280 Speaker 3: one years. I saw him this weekend. He's in Toronto. 151 00:08:35,880 --> 00:08:38,760 Speaker 3: By the way, I said, do you work for Rob Thomas? 152 00:08:39,160 --> 00:08:44,400 Speaker 3: He went like this, he goes, He said, I'm five 153 00:08:44,480 --> 00:08:48,200 Speaker 3: layers down. But so I always whenever I see my 154 00:08:48,200 --> 00:08:50,280 Speaker 3: cousin I ask him, can you tell me again what 155 00:08:50,320 --> 00:08:52,600 Speaker 3: you do? Because it's always changing, right, I guess this 156 00:08:52,720 --> 00:08:55,880 Speaker 3: is a function of working at IBM. So eventually he 157 00:08:56,000 --> 00:08:58,160 Speaker 3: just gives up and says, you know, we're just solving problems. 158 00:08:58,240 --> 00:09:01,040 Speaker 3: So what we're doing, which I sort of as a 159 00:09:01,120 --> 00:09:04,760 Speaker 3: kind of frame, And I was curious, what's the coolest 160 00:09:04,800 --> 00:09:07,800 Speaker 3: problem you ever worked on? Not biggest, not most important, 161 00:09:08,960 --> 00:09:11,800 Speaker 3: but the coolest, the one that's like that sort of 162 00:09:11,800 --> 00:09:13,560 Speaker 3: makes you smile when you think back on it. 163 00:09:13,640 --> 00:09:18,400 Speaker 4: Probably when I was in microelectronics, because it was a 164 00:09:18,400 --> 00:09:21,480 Speaker 4: world I had no exposure to. I hadn't studied computer science, 165 00:09:22,559 --> 00:09:28,679 Speaker 4: and we were building a lot of high performance semiconductor technology, 166 00:09:28,800 --> 00:09:31,600 Speaker 4: so just chips that do a really great job of 167 00:09:32,080 --> 00:09:36,080 Speaker 4: processing something or other. And we figured out that there 168 00:09:36,160 --> 00:09:39,960 Speaker 4: was a market in consumer gaming that was starting to happen, 169 00:09:40,720 --> 00:09:43,440 Speaker 4: and we got to the point where we became the 170 00:09:43,520 --> 00:09:51,200 Speaker 4: chip inside the Nintendo, We the Microsoft Xbox Sony PlayStation, 171 00:09:51,400 --> 00:09:54,440 Speaker 4: so we basically had the entire gaming market running on 172 00:09:54,520 --> 00:09:55,400 Speaker 4: ib AND chips. 173 00:09:56,080 --> 00:10:00,640 Speaker 3: And to use every parent basically is pointing at you 174 00:10:00,679 --> 00:10:01,959 Speaker 3: and saying you're the call. 175 00:10:03,120 --> 00:10:06,760 Speaker 5: Probably well they would have found it from anybody. 176 00:10:06,559 --> 00:10:10,800 Speaker 4: But it was the first time I could explain my 177 00:10:10,960 --> 00:10:12,920 Speaker 4: job to my kids, who were quite young at that time, 178 00:10:13,080 --> 00:10:15,720 Speaker 4: like what I did, Like it was more tangible for 179 00:10:15,760 --> 00:10:18,200 Speaker 4: them than saying we solve problems or douce you know, 180 00:10:18,240 --> 00:10:22,560 Speaker 4: build solutions like it became very tangible for them, and 181 00:10:22,920 --> 00:10:25,640 Speaker 4: I think that's, you know, a rewarding part of the 182 00:10:25,720 --> 00:10:28,520 Speaker 4: job is when you can help your family actually understand 183 00:10:28,559 --> 00:10:30,080 Speaker 4: what you do. Most people can't do that. It's probably 184 00:10:30,120 --> 00:10:31,880 Speaker 4: easier for you. They can they can see the books, 185 00:10:33,320 --> 00:10:36,160 Speaker 4: but for for some of us in the business, the 186 00:10:36,160 --> 00:10:38,360 Speaker 4: business world, it's not always as obvious. So that was 187 00:10:38,400 --> 00:10:41,200 Speaker 4: like one example where the dots really connected. 188 00:10:42,160 --> 00:10:45,360 Speaker 3: There were a couple's a couple of stuck about a 189 00:10:45,400 --> 00:10:48,120 Speaker 3: little bit of this into context of of AI love 190 00:10:48,200 --> 00:10:51,360 Speaker 3: because I love the frame of problem solving as a 191 00:10:51,400 --> 00:10:54,280 Speaker 3: way of understanding what the function of the technology is. 192 00:10:54,600 --> 00:10:57,320 Speaker 3: So I know that you guys did something, did some 193 00:10:57,400 --> 00:11:01,280 Speaker 3: work with I never know how to announced it is 194 00:11:01,320 --> 00:11:05,280 Speaker 3: it Sevia, Sevia, Sevia with the football club Sevia in Spain. 195 00:11:05,559 --> 00:11:08,160 Speaker 3: Tell me about Tell me a little bit about that. 196 00:11:08,240 --> 00:11:11,040 Speaker 3: What problem were they trying to solve and why did 197 00:11:11,120 --> 00:11:11,640 Speaker 3: they call you? 198 00:11:11,720 --> 00:11:19,720 Speaker 4: In every sports franchise is trying to get an advantage, right, 199 00:11:19,760 --> 00:11:25,760 Speaker 4: Let's just be that clear. Everybody's how can I use data, analytics, insights, 200 00:11:25,840 --> 00:11:28,960 Speaker 4: anything that will make us one percent better on the 201 00:11:29,000 --> 00:11:34,760 Speaker 4: field at some point in the future. And Sevie reached 202 00:11:34,760 --> 00:11:36,959 Speaker 4: out to us because they had. 203 00:11:36,800 --> 00:11:37,559 Speaker 5: Seen some of that. 204 00:11:37,640 --> 00:11:39,840 Speaker 4: We've done some work with the Toronto Raptors in the 205 00:11:39,880 --> 00:11:44,559 Speaker 4: past and others, and their thought was maybe there's something 206 00:11:44,559 --> 00:11:48,000 Speaker 4: we could do. They'd heard all about generative AI, that 207 00:11:48,080 --> 00:11:49,840 Speaker 4: heard about large language models. 208 00:11:50,320 --> 00:11:52,800 Speaker 5: And the problem, back to your point on. 209 00:11:53,120 --> 00:11:55,880 Speaker 4: Solving problems, was we want to do a way better 210 00:11:55,960 --> 00:12:01,560 Speaker 4: job of assessing talent, because really the lifeblood of a 211 00:12:01,559 --> 00:12:05,720 Speaker 4: sports franchise is can you continue to cultivate talent? Can 212 00:12:05,800 --> 00:12:09,000 Speaker 4: you find talent that others don't find? Can you see 213 00:12:09,040 --> 00:12:11,599 Speaker 4: something in somebody that they don't see in themselves or 214 00:12:11,640 --> 00:12:12,240 Speaker 4: maybe no other. 215 00:12:12,160 --> 00:12:13,280 Speaker 5: Team season them. 216 00:12:13,640 --> 00:12:17,560 Speaker 4: And we ended up building some of them called Scout Advisor, 217 00:12:18,440 --> 00:12:22,840 Speaker 4: which is built on Watson X, which basically just ingests 218 00:12:24,000 --> 00:12:27,600 Speaker 4: tons and tons of data, and we like to think 219 00:12:27,600 --> 00:12:29,800 Speaker 4: of it as finding, you know, the needle in the 220 00:12:29,840 --> 00:12:34,280 Speaker 4: haystack of you know, here's three players that aren't being considered. 221 00:12:34,600 --> 00:12:38,920 Speaker 4: They're not on the top teams today, and I think 222 00:12:39,040 --> 00:12:41,360 Speaker 4: working with them together, we found some pretty good insights 223 00:12:41,360 --> 00:12:43,600 Speaker 4: that's helped them out how What was intriguing to. 224 00:12:43,559 --> 00:12:47,400 Speaker 3: Me was we're not just talking about quantitative data. We're 225 00:12:47,400 --> 00:12:51,719 Speaker 3: also talking about qualitative data. But that's the puzzle part 226 00:12:51,760 --> 00:12:54,239 Speaker 3: of the thing that fastens me. How does what incorporate 227 00:12:54,360 --> 00:12:58,120 Speaker 3: qualitative analysis into that sort of so you just feeding 228 00:12:58,160 --> 00:13:00,440 Speaker 3: in scouting reports and things like that. 229 00:13:02,320 --> 00:13:03,959 Speaker 4: I got to realize think about how much I can 230 00:13:04,000 --> 00:13:10,040 Speaker 4: actually disclose it. But if you think about so, quantitative 231 00:13:10,160 --> 00:13:12,280 Speaker 4: is relatively easy. 232 00:13:12,240 --> 00:13:13,480 Speaker 5: Yeah, every team collects that. 233 00:13:14,920 --> 00:13:17,920 Speaker 4: You know, what's the forty yard dashable think they use 234 00:13:17,960 --> 00:13:22,679 Speaker 4: that term, certainly not in Spain. That's all quantitative. Qualitative 235 00:13:22,760 --> 00:13:26,839 Speaker 4: is what's happening off the field. It could be diet, 236 00:13:27,160 --> 00:13:30,960 Speaker 4: it could be habits, it could be behavior. You can 237 00:13:31,040 --> 00:13:33,720 Speaker 4: imagine a range of things that would all feed into 238 00:13:34,640 --> 00:13:38,439 Speaker 4: an athlete's performance and so relationships. 239 00:13:39,640 --> 00:13:41,160 Speaker 5: There's many different aspects, and. 240 00:13:41,120 --> 00:13:44,480 Speaker 4: So it's trying to figure out the right blend of 241 00:13:44,600 --> 00:13:47,720 Speaker 4: quantitative and qualitative that gives you a unique insight. 242 00:13:48,360 --> 00:13:50,720 Speaker 3: How transparent is that kind of system? I mean, is 243 00:13:50,720 --> 00:13:55,080 Speaker 3: it telling you it's saying pick this guy, not this guy, 244 00:13:55,120 --> 00:13:56,920 Speaker 3: But is it telling you why it prefers this guy 245 00:13:56,960 --> 00:13:57,520 Speaker 3: to this guy. 246 00:13:57,520 --> 00:14:00,800 Speaker 4: Is that I think for anything the realm of AI, 247 00:14:00,960 --> 00:14:04,480 Speaker 4: you have to answer the why question. Otherwise you've fallen 248 00:14:04,480 --> 00:14:08,840 Speaker 4: into the trap of the you know, the proverbial black box. 249 00:14:09,120 --> 00:14:11,880 Speaker 4: And then wait, I made this decision, I'd never understood 250 00:14:11,880 --> 00:14:14,320 Speaker 4: why it didn't work out, So you always have to 251 00:14:14,320 --> 00:14:15,959 Speaker 4: answer why without a doubt? 252 00:14:16,880 --> 00:14:18,160 Speaker 3: And how is why answered? 253 00:14:20,720 --> 00:14:24,720 Speaker 4: Sources of data, the reasoning that went into it, and 254 00:14:24,840 --> 00:14:28,080 Speaker 4: so it's basically just tracing back the chain of how 255 00:14:28,120 --> 00:14:31,000 Speaker 4: you got to the answer. And in the case of 256 00:14:31,200 --> 00:14:33,680 Speaker 4: what we do in Watson X is we have IBM models. 257 00:14:34,120 --> 00:14:36,760 Speaker 4: We also use some other open source models, So it 258 00:14:36,760 --> 00:14:39,600 Speaker 4: would be which model was used, what was the data 259 00:14:39,640 --> 00:14:41,840 Speaker 4: set that was fed into that model, How is it 260 00:14:41,880 --> 00:14:42,640 Speaker 4: making decisions? 261 00:14:42,640 --> 00:14:45,920 Speaker 5: How is it performing? Is it robust? 262 00:14:46,080 --> 00:14:48,320 Speaker 4: Meaning is it reliable in terms of if you feed 263 00:14:48,360 --> 00:14:50,120 Speaker 4: it two of the same data set, do you get 264 00:14:50,160 --> 00:14:53,080 Speaker 4: the same answer? These are all the you know, the 265 00:14:53,120 --> 00:14:55,080 Speaker 4: technical aspects of understanding the why. 266 00:14:56,160 --> 00:15:00,320 Speaker 3: How quickly do you expect all professional sports franchises to 267 00:15:00,360 --> 00:15:02,480 Speaker 3: adopt some kind of are they already there? If I 268 00:15:02,520 --> 00:15:06,160 Speaker 3: went out and pulled the general managers of the one 269 00:15:06,240 --> 00:15:09,120 Speaker 3: hundred most valuable sports franchises in the world. How many 270 00:15:09,160 --> 00:15:11,760 Speaker 3: of them would be using some kind of AI system 271 00:15:11,800 --> 00:15:13,000 Speaker 3: to assist in their efforts. 272 00:15:14,920 --> 00:15:18,640 Speaker 4: One hundred and twenty percent would meaning that everybody's doing it, 273 00:15:18,680 --> 00:15:20,560 Speaker 4: and some think they're doing way more than they probably 274 00:15:20,560 --> 00:15:24,160 Speaker 4: actually are. So everybody's doing it. I think what's weird 275 00:15:24,200 --> 00:15:29,720 Speaker 4: about sports is everybody's so convinced that what they're doing 276 00:15:29,840 --> 00:15:34,160 Speaker 4: is unique that they generally speaking, don't want to work 277 00:15:34,200 --> 00:15:36,600 Speaker 4: with a third party to do it because they're afraid 278 00:15:36,720 --> 00:15:39,360 Speaker 4: that that would expose them. But in reality, I think 279 00:15:39,400 --> 00:15:42,240 Speaker 4: most are doing eighty to ninety percent of the same things. 280 00:15:43,880 --> 00:15:46,040 Speaker 4: So but without a doubt, everybody's doing it. 281 00:15:46,440 --> 00:15:51,080 Speaker 3: Yeah. Yeah. The other I say that I loved was 282 00:15:51,120 --> 00:15:54,240 Speaker 3: there was one but a shipping line tricon on the 283 00:15:54,280 --> 00:15:57,560 Speaker 3: Mississippi River. Tell me a little bit about that project. 284 00:15:57,640 --> 00:15:58,880 Speaker 3: What problem were they trying to solve? 285 00:16:00,960 --> 00:16:04,680 Speaker 4: Think about the problem that I would say everybody noticed 286 00:16:04,720 --> 00:16:08,280 Speaker 4: if you go back to twenty twenty was things are 287 00:16:08,320 --> 00:16:10,920 Speaker 4: getting hold held up in ports. It was actually an 288 00:16:10,960 --> 00:16:13,200 Speaker 4: article in the paper this morning kind of tracing the 289 00:16:13,240 --> 00:16:17,000 Speaker 4: history of what happened twenty twenty twenty one and why 290 00:16:17,120 --> 00:16:19,640 Speaker 4: ships were basically sitting at seas for months at a 291 00:16:19,680 --> 00:16:23,320 Speaker 4: time and at that stage, we just we had a 292 00:16:23,360 --> 00:16:28,840 Speaker 4: massive throughput issue. But moving even beyond the pandemic, you 293 00:16:28,880 --> 00:16:33,800 Speaker 4: can see it now with ships getting through like Panama Canal, 294 00:16:33,880 --> 00:16:36,640 Speaker 4: there's like a narrow window where you can get through, 295 00:16:37,080 --> 00:16:40,800 Speaker 4: and if you don't have your paperwork done, you don't 296 00:16:40,800 --> 00:16:42,880 Speaker 4: have the right approvals, you're not going through and it 297 00:16:42,920 --> 00:16:44,320 Speaker 4: may cost you a day or two and that's a 298 00:16:44,360 --> 00:16:48,200 Speaker 4: lot of money. In the shipping industry and the tricon example, 299 00:16:48,840 --> 00:16:52,440 Speaker 4: it's really just about when you're pulling into a port, 300 00:16:53,560 --> 00:16:56,840 Speaker 4: if you have the right paperwork done, you can get 301 00:16:56,920 --> 00:17:01,880 Speaker 4: goods off the ship very quickly. Ship a lot of food, 302 00:17:02,520 --> 00:17:06,240 Speaker 4: which by definition, since it's not packaged food, it's fresh food. 303 00:17:06,640 --> 00:17:10,359 Speaker 4: There is an expiration period and so if it takes 304 00:17:10,359 --> 00:17:15,680 Speaker 4: them an extra two hours, certainly multiple hours or a day, 305 00:17:16,119 --> 00:17:17,879 Speaker 4: they have a massive problem because then you're going to 306 00:17:17,920 --> 00:17:20,400 Speaker 4: deal with spoilage and so it's going to set you back. 307 00:17:21,040 --> 00:17:24,400 Speaker 4: And what we've worked with them on is using an 308 00:17:24,440 --> 00:17:28,440 Speaker 4: assistant that we've built in Watson X called Orchestrate, which 309 00:17:28,480 --> 00:17:33,840 Speaker 4: basically is just AI doing digital labor, so we can 310 00:17:34,000 --> 00:17:39,960 Speaker 4: replicate nearly any repetitive task and do that with software. 311 00:17:39,760 --> 00:17:40,600 Speaker 5: Instead of humans. 312 00:17:41,520 --> 00:17:45,000 Speaker 4: So as you may imagine shipping industry still has a 313 00:17:45,040 --> 00:17:48,000 Speaker 4: lot of paperwork that goes on and so being able 314 00:17:48,040 --> 00:17:51,040 Speaker 4: to take forms that normally would be multiple hours of 315 00:17:51,119 --> 00:17:53,160 Speaker 4: filling it out. Oh this isn't right, send it back. 316 00:17:53,680 --> 00:17:57,320 Speaker 4: We've basically built that as a digital skill inside of 317 00:17:57,640 --> 00:18:02,080 Speaker 4: WATSONEX orchestrate, and so now it's done in minutes. 318 00:18:03,119 --> 00:18:06,040 Speaker 3: They did they realize that they could have that kind 319 00:18:06,080 --> 00:18:08,240 Speaker 3: of efficiency by teaming up with you? Or is that 320 00:18:08,280 --> 00:18:12,320 Speaker 3: something you came to them and said, guys, we can 321 00:18:12,359 --> 00:18:14,120 Speaker 3: do this way better than you think. What's the. 322 00:18:15,960 --> 00:18:19,479 Speaker 4: I'd say it's always, it's always both sides coming together 323 00:18:19,640 --> 00:18:22,560 Speaker 4: at a moment that for some reason makes sense because 324 00:18:23,760 --> 00:18:25,560 Speaker 4: you could say, why didn't this happen like five years ago, 325 00:18:25,640 --> 00:18:29,920 Speaker 4: like this seems so obvious. Well, technology wasn't quite ready then, 326 00:18:30,440 --> 00:18:32,600 Speaker 4: I would say, But they knew they had a need 327 00:18:33,080 --> 00:18:36,639 Speaker 4: because I forget what the precise number is, but you know, 328 00:18:36,880 --> 00:18:40,680 Speaker 4: reduction of spoilage has massive impact on their bottom line, 329 00:18:42,680 --> 00:18:44,160 Speaker 4: and so they knew they had a need. 330 00:18:45,200 --> 00:18:48,120 Speaker 5: We thought we could solve it and the two together. 331 00:18:48,960 --> 00:18:50,760 Speaker 3: Who did you guys go to them? 332 00:18:50,800 --> 00:18:50,960 Speaker 5: Though? 333 00:18:51,840 --> 00:18:52,560 Speaker 3: Did they come to you? 334 00:18:52,840 --> 00:18:56,199 Speaker 4: I recall that this one was an inbound meaning they 335 00:18:56,200 --> 00:18:59,520 Speaker 4: had reached out to IBM and we'd like to solve 336 00:18:59,520 --> 00:19:01,359 Speaker 4: this problem. I think it went into one of our 337 00:19:01,400 --> 00:19:04,440 Speaker 4: digital centers if I recall it a literary phone. 338 00:19:04,200 --> 00:19:09,359 Speaker 3: Call, but the other the reverse is more interesting to 339 00:19:09,440 --> 00:19:11,640 Speaker 3: me because there seems to be a very, very large 340 00:19:11,720 --> 00:19:14,520 Speaker 3: universe of people who have problems that could be solved 341 00:19:14,680 --> 00:19:19,480 Speaker 3: this way and they don't realize it. What's your Is 342 00:19:19,520 --> 00:19:22,600 Speaker 3: there a shining example of this of someone you just 343 00:19:22,680 --> 00:19:25,240 Speaker 3: can't you just think could benefit so much and isn't 344 00:19:25,240 --> 00:19:26,160 Speaker 3: benefiting right now? 345 00:19:28,920 --> 00:19:30,359 Speaker 5: Maybe I'll answer it slightly differently. 346 00:19:30,520 --> 00:19:35,320 Speaker 4: I'm I'm surprised by how many people can benefit that 347 00:19:35,400 --> 00:19:38,359 Speaker 4: you wouldn't even logically think of. First, let me give 348 00:19:38,359 --> 00:19:46,200 Speaker 4: you an example. There's a franchiser of hair salons, sport 349 00:19:46,240 --> 00:19:49,520 Speaker 4: Clips is the name. My sons used to go there 350 00:19:49,520 --> 00:19:51,280 Speaker 4: for haircuts because they have like TVs and you can 351 00:19:51,320 --> 00:19:54,080 Speaker 4: watch sports, so they loved that they got entertained while 352 00:19:54,080 --> 00:19:57,080 Speaker 4: they would get their haircut. I think the last place 353 00:19:57,080 --> 00:19:59,600 Speaker 4: that you would think is using AI today would be 354 00:20:00,160 --> 00:20:01,880 Speaker 4: a franchiser of hair salons. 355 00:20:03,840 --> 00:20:05,560 Speaker 5: But just follow it through. 356 00:20:06,640 --> 00:20:09,639 Speaker 4: The biggest part of how they run their business is 357 00:20:09,680 --> 00:20:12,840 Speaker 4: can I get people to cut hair? And this is 358 00:20:12,840 --> 00:20:15,320 Speaker 4: the high turnover industry because there's a lot of different 359 00:20:15,320 --> 00:20:16,840 Speaker 4: places you can work if you want to cut hair. 360 00:20:17,240 --> 00:20:19,359 Speaker 4: People actually get injured cutting hair because you're on your 361 00:20:19,400 --> 00:20:22,879 Speaker 4: feet all day, that type of thing. And they're using 362 00:20:23,119 --> 00:20:28,240 Speaker 4: same technology orchestrate as part of their recruiting process. How 363 00:20:28,240 --> 00:20:32,440 Speaker 4: can they automate a lot of people submitting resumes, who 364 00:20:32,480 --> 00:20:35,800 Speaker 4: they speak to, how they qualify them for the position. 365 00:20:36,560 --> 00:20:39,119 Speaker 4: And so the reason I give that example is the 366 00:20:39,560 --> 00:20:45,000 Speaker 4: opportunity for AI, which is unlike other technologies, is truly unlimited. 367 00:20:46,600 --> 00:20:50,200 Speaker 4: It will touch every single business. It's not the realm 368 00:20:50,240 --> 00:20:53,040 Speaker 4: of the fortune five hundred or the fortune one thousand. 369 00:20:53,840 --> 00:20:54,679 Speaker 5: This is the. 370 00:20:54,880 --> 00:20:57,720 Speaker 4: Fortune any size. And I think that may be one 371 00:20:57,760 --> 00:20:59,639 Speaker 4: thing that people underestimate about. 372 00:21:00,680 --> 00:21:03,399 Speaker 3: Yeah, what about I mean I was thinking about education 373 00:21:03,760 --> 00:21:07,159 Speaker 3: as a kind of I mean, education is a perennial 374 00:21:09,800 --> 00:21:12,320 Speaker 3: whipping boy for you guys that are living in the 375 00:21:12,400 --> 00:21:18,160 Speaker 3: nineteenth century, right. I'm just curious about if a superintendent 376 00:21:18,160 --> 00:21:20,159 Speaker 3: of a public school system or the president of the 377 00:21:20,240 --> 00:21:23,359 Speaker 3: university sat down and had lunch with you and said, 378 00:21:25,000 --> 00:21:27,520 Speaker 3: do the university first. My cost are out of control, 379 00:21:28,200 --> 00:21:34,000 Speaker 3: my enrollment is down, my students hate me, and my 380 00:21:34,080 --> 00:21:40,719 Speaker 3: board is revolting help. How would you think about helping 381 00:21:40,720 --> 00:21:41,680 Speaker 3: someone in that situation. 382 00:21:43,280 --> 00:21:45,240 Speaker 5: I spend some time with universities. 383 00:21:45,280 --> 00:21:48,720 Speaker 4: I like to go back and visit Alma Maters, where 384 00:21:48,720 --> 00:21:50,720 Speaker 4: I went to school, and so. 385 00:21:50,800 --> 00:21:51,680 Speaker 5: I do that every year. 386 00:21:52,560 --> 00:21:55,320 Speaker 4: The challenge I have hall of Seeming university is there 387 00:21:55,359 --> 00:21:58,240 Speaker 4: has to be a will. Yeah, and I'm not sure 388 00:21:58,240 --> 00:22:03,840 Speaker 4: the incentives are quite right today because bringing in new technology, 389 00:22:03,920 --> 00:22:05,720 Speaker 4: let's say we want to go after we can help 390 00:22:05,840 --> 00:22:10,480 Speaker 4: you figure out student recruiting or how you automate more 391 00:22:10,480 --> 00:22:15,160 Speaker 4: of your education, everybody suddenly feels threatened that university. 392 00:22:15,720 --> 00:22:16,800 Speaker 5: Hold on, that's my job. 393 00:22:17,359 --> 00:22:19,720 Speaker 4: I'm the one that decides that, or I'm the one 394 00:22:19,760 --> 00:22:22,800 Speaker 4: that wants to dictate the course. So there has to 395 00:22:22,800 --> 00:22:26,679 Speaker 4: be a will. So I think it's very possible, and 396 00:22:27,560 --> 00:22:30,000 Speaker 4: I do think over the next decade you will see 397 00:22:30,040 --> 00:22:32,400 Speaker 4: some universities that jump all over this and they will 398 00:22:32,440 --> 00:22:34,800 Speaker 4: move ahead, and you see others that do not. 399 00:22:35,680 --> 00:22:37,600 Speaker 5: Because it's very possible. 400 00:22:39,200 --> 00:22:42,159 Speaker 3: Where how does when you say there has to be 401 00:22:42,200 --> 00:22:44,760 Speaker 3: a will? Is that the kind of is that a 402 00:22:44,880 --> 00:22:46,800 Speaker 3: kind of thing that that people that I beb to 403 00:22:46,840 --> 00:22:51,320 Speaker 3: think about? Like when in this conversation you hypothetical conversation, 404 00:22:51,359 --> 00:22:53,880 Speaker 3: you might have with the university president, would you give 405 00:22:53,920 --> 00:22:59,360 Speaker 3: advice on on where the will comes from. 406 00:22:59,440 --> 00:23:01,760 Speaker 4: I don't do that as much in a university context. 407 00:23:01,760 --> 00:23:05,560 Speaker 4: I do that every day in a business context, because 408 00:23:06,320 --> 00:23:08,360 Speaker 4: if you can find the right person in a business 409 00:23:08,359 --> 00:23:12,480 Speaker 4: that wants to focus on growth or the bottom line 410 00:23:13,000 --> 00:23:15,600 Speaker 4: or how do you create more productivity. Yes, it's going 411 00:23:15,600 --> 00:23:19,719 Speaker 4: to create a lot of organizational resistance potentially, but you 412 00:23:19,720 --> 00:23:21,960 Speaker 4: can find somebody that will figure out how to push 413 00:23:21,960 --> 00:23:27,040 Speaker 4: that through. I think for universities, I think that's also possible. 414 00:23:27,320 --> 00:23:30,000 Speaker 4: I'm not sure there's there's there's a return on investment 415 00:23:30,040 --> 00:23:30,720 Speaker 4: for us to do that. 416 00:23:31,040 --> 00:23:38,720 Speaker 3: Yeah, yeah, yeah, God, let's let's find some terms. AI years. 417 00:23:39,119 --> 00:23:41,400 Speaker 3: I told you'd like to use. What does that mean? 418 00:23:43,160 --> 00:23:45,879 Speaker 4: We just started using this term literally in the last 419 00:23:45,920 --> 00:23:47,640 Speaker 4: three months, and. 420 00:23:49,320 --> 00:23:51,560 Speaker 5: It was it was what we observed. 421 00:23:51,200 --> 00:23:55,800 Speaker 4: Internally, which is most technology you build, you say, all right, 422 00:23:55,840 --> 00:23:58,480 Speaker 4: what's going to happen in year one, year two, year three, 423 00:23:58,840 --> 00:24:02,400 Speaker 4: and it's, you know, largely by a calendar. 424 00:24:02,880 --> 00:24:04,640 Speaker 5: AI years are the idea that what. 425 00:24:04,760 --> 00:24:07,240 Speaker 4: Used to be a year is now like a week, 426 00:24:08,840 --> 00:24:11,879 Speaker 4: and that is how fast the technology is moving. Do 427 00:24:11,920 --> 00:24:14,280 Speaker 4: you give you an example. We had one client we're 428 00:24:14,320 --> 00:24:18,679 Speaker 4: working with. They're using one of our granite models, and 429 00:24:18,720 --> 00:24:21,280 Speaker 4: the results they were getting we're not very good. Accuracy 430 00:24:21,359 --> 00:24:24,119 Speaker 4: was not there, their performance was not there. So I 431 00:24:24,160 --> 00:24:25,639 Speaker 4: was like scratching my head. I was like, what is 432 00:24:25,640 --> 00:24:30,760 Speaker 4: going on? They were Financial services, the bank, So I'm 433 00:24:30,760 --> 00:24:32,640 Speaker 4: scratching my head, like what is going on? Everybody else 434 00:24:32,720 --> 00:24:36,240 Speaker 4: is getting this and like these results are horrible. And 435 00:24:36,800 --> 00:24:39,120 Speaker 4: I said to the team, which version of the model 436 00:24:39,119 --> 00:24:43,400 Speaker 4: are you using? This was in February, Like, we're using 437 00:24:43,440 --> 00:24:46,560 Speaker 4: the one from October. I was like, all right, now 438 00:24:46,560 --> 00:24:50,280 Speaker 4: we know precisely the problem because the model from October 439 00:24:50,320 --> 00:24:53,200 Speaker 4: is effectively useless now since we're here in February. 440 00:24:53,280 --> 00:24:57,400 Speaker 3: Serious, actually useless, completely useless. 441 00:24:57,560 --> 00:25:00,520 Speaker 4: Yeah, that is how fast this has changed. And so 442 00:25:01,000 --> 00:25:05,200 Speaker 4: the minute, same use case, same data, you give them 443 00:25:05,200 --> 00:25:09,760 Speaker 4: the model from late January instead of October, the results 444 00:25:09,800 --> 00:25:10,520 Speaker 4: are off the charts. 445 00:25:11,080 --> 00:25:14,680 Speaker 3: Yeah. Wait, so what exactly happened between October and January? 446 00:25:14,880 --> 00:25:15,960 Speaker 5: The model got way better? 447 00:25:16,520 --> 00:25:18,239 Speaker 3: Could dig into that, Like, what do you mean by 448 00:25:18,280 --> 00:25:18,679 Speaker 3: the way. 449 00:25:18,560 --> 00:25:19,760 Speaker 5: We are constant? 450 00:25:19,800 --> 00:25:24,400 Speaker 4: We have built large compute infrastructure where we're doing model training, 451 00:25:25,040 --> 00:25:27,439 Speaker 4: and to be clear, model training is the realm of 452 00:25:27,600 --> 00:25:31,800 Speaker 4: probably in the world. My guess is five to ten companies. 453 00:25:32,880 --> 00:25:33,240 Speaker 5: And so. 454 00:25:34,760 --> 00:25:37,920 Speaker 4: You build a model, you're constantly training it, You're doing 455 00:25:37,960 --> 00:25:41,480 Speaker 4: fine tuning, you're doing more training, you're adding data every day, 456 00:25:41,520 --> 00:25:45,760 Speaker 4: every hour it gets better. And so how does it 457 00:25:45,800 --> 00:25:48,320 Speaker 4: do that. You're feeding it more data, you're feeding it 458 00:25:48,400 --> 00:25:53,359 Speaker 4: more live examples. We're using things like synthetic data at 459 00:25:53,359 --> 00:25:55,560 Speaker 4: this point, which is we're basically creating data to do 460 00:25:55,600 --> 00:25:58,840 Speaker 4: the training as well. All of this feeds into how 461 00:25:58,920 --> 00:26:00,160 Speaker 4: useful the model is. 462 00:26:00,760 --> 00:26:02,720 Speaker 5: And so using the. 463 00:26:02,680 --> 00:26:06,040 Speaker 4: October model, those were the results in October, just a fact, 464 00:26:06,040 --> 00:26:09,320 Speaker 4: that's how good it was then. But back to the 465 00:26:09,359 --> 00:26:12,840 Speaker 4: concept of AI years, two weeks is a long time. 466 00:26:13,960 --> 00:26:16,960 Speaker 3: Does that are we in a steep part of the 467 00:26:17,000 --> 00:26:19,800 Speaker 3: model learning carve or do you expect this to continue 468 00:26:19,840 --> 00:26:21,560 Speaker 3: along this at this pace? 469 00:26:23,160 --> 00:26:27,399 Speaker 5: I think that is the big question and don't have 470 00:26:27,440 --> 00:26:28,120 Speaker 5: an answer yet. 471 00:26:28,560 --> 00:26:30,520 Speaker 4: By definition, at some point you would think it would 472 00:26:30,520 --> 00:26:33,040 Speaker 4: have to slow down a bit, but it's not obvious 473 00:26:33,080 --> 00:26:35,440 Speaker 4: that that is on the horizon. 474 00:26:35,040 --> 00:26:38,600 Speaker 3: Still speeding up. Yes, how fast can it get? 475 00:26:41,040 --> 00:26:44,199 Speaker 4: We've debated can you actually have better results in the 476 00:26:44,240 --> 00:26:49,000 Speaker 4: afternoon than you did in the morning. Really it's nuts, Yeah, 477 00:26:49,000 --> 00:26:51,760 Speaker 4: I know, but that's why we came up with this 478 00:26:51,840 --> 00:26:53,840 Speaker 4: term because I think you also have to think of 479 00:26:53,920 --> 00:26:55,600 Speaker 4: like concepts that. 480 00:26:57,720 --> 00:26:58,719 Speaker 5: Gets people's attention. 481 00:26:58,920 --> 00:27:02,080 Speaker 3: So you basically earning into a bakery, you're like the 482 00:27:02,119 --> 00:27:04,359 Speaker 3: bread from yesterday. You know you can have it for 483 00:27:04,440 --> 00:27:08,080 Speaker 3: twenty five cents. But I mean you do proferential pricing. 484 00:27:08,119 --> 00:27:12,720 Speaker 3: You could say we'll judge you X for yesterday's model, 485 00:27:13,160 --> 00:27:14,600 Speaker 3: two X for today's model. 486 00:27:16,200 --> 00:27:20,119 Speaker 4: I think that's dangerous as a merchandising strategy, but I 487 00:27:20,119 --> 00:27:20,720 Speaker 4: guess your point. 488 00:27:21,119 --> 00:27:24,240 Speaker 3: Yeah, but that's crazy. And this, by the way, so 489 00:27:24,280 --> 00:27:26,840 Speaker 3: this model is the same true for almost You're talking 490 00:27:26,880 --> 00:27:30,240 Speaker 3: specifically about a model that was created to help some 491 00:27:30,320 --> 00:27:34,600 Speaker 3: aspect of a financial services So is that kind of 492 00:27:34,720 --> 00:27:37,760 Speaker 3: model accelerating faster and running faster than other models for 493 00:27:37,840 --> 00:27:39,240 Speaker 3: other kinds of problems? 494 00:27:39,600 --> 00:27:44,800 Speaker 4: So this domain was code. Yeah, so by definition, if 495 00:27:44,800 --> 00:27:48,359 Speaker 4: you're feel feeding in more data some more code, you 496 00:27:48,359 --> 00:27:49,440 Speaker 4: get those kind of results. 497 00:27:50,400 --> 00:27:51,720 Speaker 5: It does depend on the model type. 498 00:27:52,000 --> 00:27:54,439 Speaker 4: Yeah, there's a lot of code in the world, and 499 00:27:54,480 --> 00:27:57,280 Speaker 4: so we can find that we can create it. Like 500 00:27:57,320 --> 00:28:01,800 Speaker 4: I said, there's other aspect x where there's probably less 501 00:28:01,960 --> 00:28:05,000 Speaker 4: inputs available, which means you probably won't get the same 502 00:28:05,080 --> 00:28:08,200 Speaker 4: level of iteration. But for code that's certainly the cycle 503 00:28:08,240 --> 00:28:09,000 Speaker 4: times that we're seeing. 504 00:28:09,040 --> 00:28:11,639 Speaker 3: Yeah, and how do you know that Let's stick with 505 00:28:11,680 --> 00:28:14,320 Speaker 3: this one example of this model you have. How do 506 00:28:14,359 --> 00:28:18,640 Speaker 3: you know that your model is better than big company 507 00:28:18,680 --> 00:28:21,680 Speaker 3: B down the street? The client asks you, why would 508 00:28:21,680 --> 00:28:24,680 Speaker 3: I go with IBM as opposed to some the s 509 00:28:24,880 --> 00:28:27,000 Speaker 3: firm in the valley that says, as they have a 510 00:28:27,000 --> 00:28:31,240 Speaker 3: model on this, what's your how do you frame your advantage? 511 00:28:32,560 --> 00:28:35,719 Speaker 4: Well, we benchmark all of this, and I think the 512 00:28:35,720 --> 00:28:41,000 Speaker 4: most important is metric is price performance, not price, not performance, 513 00:28:41,000 --> 00:28:45,360 Speaker 4: but the combination of the two. And we're super competitive there. Well, 514 00:28:45,920 --> 00:28:48,360 Speaker 4: for what we just released, with what we've done in 515 00:28:48,440 --> 00:28:50,920 Speaker 4: open source, we know that nobody's close to us right 516 00:28:50,960 --> 00:28:53,960 Speaker 4: now on code now. To be clear, that will probably change, yeah, 517 00:28:54,120 --> 00:28:57,120 Speaker 4: because it's like leapfrog. People will jump ahead, then we 518 00:28:57,280 --> 00:29:02,640 Speaker 4: jump back ahead. But we're very confident that with everything 519 00:29:02,680 --> 00:29:04,520 Speaker 4: we've done in the last few months, we've taken a 520 00:29:04,600 --> 00:29:05,640 Speaker 4: huge leap forward here. 521 00:29:05,840 --> 00:29:08,880 Speaker 3: Yeah, it's I mean, this goes back to the point 522 00:29:08,880 --> 00:29:11,640 Speaker 3: I was making in the beginning, so about the difference 523 00:29:11,680 --> 00:29:16,360 Speaker 3: between your twenty something self in ninety nine and yourself today. 524 00:29:16,680 --> 00:29:21,120 Speaker 3: But this time compression has to be a crazy adjustment. 525 00:29:21,560 --> 00:29:24,640 Speaker 3: So the concept of what you're working on and how 526 00:29:24,680 --> 00:29:28,000 Speaker 3: you make decisions internally and things has to undergo this 527 00:29:28,080 --> 00:29:31,800 Speaker 3: kind of revolution if you're switching from I mean back 528 00:29:31,800 --> 00:29:35,240 Speaker 3: in the day, a model might be useful for how 529 00:29:35,280 --> 00:29:36,640 Speaker 3: long years. 530 00:29:36,440 --> 00:29:40,040 Speaker 4: Years I think about you know, statistical models that set 531 00:29:40,080 --> 00:29:44,360 Speaker 4: inside things like SPSS, which is a product that a 532 00:29:44,360 --> 00:29:45,200 Speaker 4: lot of students. 533 00:29:45,000 --> 00:29:45,640 Speaker 5: Use around the world. 534 00:29:45,680 --> 00:29:47,640 Speaker 4: I mean, those have been the same models for twenty 535 00:29:47,720 --> 00:29:49,959 Speaker 4: years and they're still very good at what they do. 536 00:29:50,760 --> 00:29:54,640 Speaker 4: And so yes, it's a completely it's a completely different 537 00:29:55,520 --> 00:29:58,160 Speaker 4: moment for how fast this is moving. And I think 538 00:29:58,160 --> 00:30:02,120 Speaker 4: it just raises the bar for everybody, whether you're a 539 00:30:02,120 --> 00:30:06,360 Speaker 4: technology provider like us, or you're a bank or an 540 00:30:06,400 --> 00:30:09,960 Speaker 4: insurance company or a shipping company, to say, how do 541 00:30:10,040 --> 00:30:14,360 Speaker 4: you really change your culture to be way more aggressive 542 00:30:15,560 --> 00:30:16,680 Speaker 4: than you normally would be. 543 00:30:18,720 --> 00:30:21,320 Speaker 3: Does this means it's a weird question, but does this 544 00:30:21,360 --> 00:30:25,360 Speaker 3: mean a different set of kind of personality or character 545 00:30:25,400 --> 00:30:28,840 Speaker 3: traits are necessary for a decision maker in tech now 546 00:30:28,880 --> 00:30:30,520 Speaker 3: than twenty five years ago. 547 00:30:33,640 --> 00:30:37,400 Speaker 4: There's a book I saw recently, it's called The Geek Way, 548 00:30:37,720 --> 00:30:42,520 Speaker 4: which talked about how technology companies have started to operate 549 00:30:42,560 --> 00:30:46,840 Speaker 4: in different ways maybe than many you know, traditional companies, 550 00:30:48,280 --> 00:30:55,040 Speaker 4: and more about being data driven, more about delegation. Are 551 00:30:55,080 --> 00:30:58,640 Speaker 4: you willing to have the smartest person in the room 552 00:30:58,640 --> 00:31:00,560 Speaker 4: make decisions opposed to the high paid. 553 00:31:00,360 --> 00:31:01,280 Speaker 5: Person in the room. 554 00:31:01,760 --> 00:31:04,560 Speaker 4: I think these are all different aspects that every company 555 00:31:04,600 --> 00:31:05,280 Speaker 4: is going to face. 556 00:31:05,720 --> 00:31:10,520 Speaker 3: Yeah, yeah, next term, talk about open. When you use 557 00:31:10,560 --> 00:31:11,680 Speaker 3: that word open, what do you mean. 558 00:31:14,200 --> 00:31:16,960 Speaker 4: I think there's really only one definition of open, which 559 00:31:17,000 --> 00:31:22,040 Speaker 4: is for technology, is open source? An open source means 560 00:31:22,680 --> 00:31:28,560 Speaker 4: the code is freely available. Anybody can see it, access it, 561 00:31:29,480 --> 00:31:30,320 Speaker 4: contribute to it. 562 00:31:30,600 --> 00:31:33,520 Speaker 3: And what is Tell me about why that's an important principle. 563 00:31:36,760 --> 00:31:39,880 Speaker 4: When you take a topic like AI, I think it 564 00:31:39,920 --> 00:31:43,720 Speaker 4: would be really bad for the world if this was 565 00:31:43,760 --> 00:31:48,320 Speaker 4: in the hands of one or two companies, or three 566 00:31:48,400 --> 00:31:51,680 Speaker 4: or four, doesn't matter the number some small number. Think 567 00:31:51,680 --> 00:31:56,280 Speaker 4: about like in history sometimes early nineteen hundreds, the Interstate 568 00:31:56,360 --> 00:32:00,120 Speaker 4: Commerce Commission was created, and the whole idea was to 569 00:32:00,160 --> 00:32:05,640 Speaker 4: protect farmers from railroads. Meaning they wanted to allow free trade, 570 00:32:06,040 --> 00:32:08,440 Speaker 4: but they knew that, well, there's only so many railroad tracks, 571 00:32:08,440 --> 00:32:11,800 Speaker 4: so we need to protect farmers from the shipping costs 572 00:32:11,800 --> 00:32:15,720 Speaker 4: that railroads could impose. So good idea, but over time 573 00:32:16,040 --> 00:32:19,800 Speaker 4: that got completely overtaken by the railroad lobby, and then 574 00:32:19,800 --> 00:32:23,680 Speaker 4: they use that to basically just increase prices, and it 575 00:32:23,720 --> 00:32:27,800 Speaker 4: made the lives of farmers way more difficult. I think 576 00:32:27,800 --> 00:32:31,240 Speaker 4: you could play the same analogy through with AI. If 577 00:32:31,280 --> 00:32:35,520 Speaker 4: you allow a handful of companies to have the technology, 578 00:32:35,680 --> 00:32:38,240 Speaker 4: you regulate around the principles of with those one or 579 00:32:38,240 --> 00:32:39,040 Speaker 4: two companies, then. 580 00:32:38,920 --> 00:32:41,680 Speaker 5: You've trapped the entire world. Think that would be very bad. 581 00:32:43,040 --> 00:32:46,360 Speaker 5: So the danger of that happening for sure. 582 00:32:46,480 --> 00:32:49,640 Speaker 4: I mean there's companies in Watson in Washington every week 583 00:32:50,120 --> 00:32:54,640 Speaker 4: trying to achieve that outcome. And so the opposite of 584 00:32:54,680 --> 00:32:56,360 Speaker 4: that is to say it's going to be an open 585 00:32:56,360 --> 00:33:01,480 Speaker 4: source because nobody can dispute opens because it's right there, 586 00:33:01,520 --> 00:33:05,640 Speaker 4: everybody can see it. And so I'm a strong believer 587 00:33:05,640 --> 00:33:07,120 Speaker 4: that open source will win for AI. 588 00:33:07,160 --> 00:33:07,880 Speaker 5: It has to win. 589 00:33:08,640 --> 00:33:13,720 Speaker 4: It's not just important for business, but it's important for humans. 590 00:33:14,480 --> 00:33:17,600 Speaker 3: On the I'm curious about on the list of things 591 00:33:17,640 --> 00:33:21,400 Speaker 3: you worry about, actually, let me before I ask, let 592 00:33:21,440 --> 00:33:23,880 Speaker 3: me ask this question very generally. What is the list 593 00:33:23,920 --> 00:33:26,760 Speaker 3: of things you worry about? What's your top five business 594 00:33:26,760 --> 00:33:27,920 Speaker 3: related worries right now? 595 00:33:29,400 --> 00:33:31,720 Speaker 5: Tops from those are the first question. We could be 596 00:33:31,760 --> 00:33:33,400 Speaker 5: here for hours for me to answer. 597 00:33:34,760 --> 00:33:36,680 Speaker 3: I did say business related. We could leave you know, 598 00:33:37,960 --> 00:33:40,240 Speaker 3: your kids haircuts got it out of. 599 00:33:40,120 --> 00:33:45,080 Speaker 4: The Number one is always it's the thing that's probably 600 00:33:45,120 --> 00:33:50,240 Speaker 4: always been true, which is just people. Do we have 601 00:33:50,280 --> 00:33:52,440 Speaker 4: the right skills? Are we doing a good job of 602 00:33:52,480 --> 00:33:55,920 Speaker 4: training our people? Are our people doing a good job 603 00:33:55,960 --> 00:33:59,840 Speaker 4: of working with clients? Like that's number one? Number two 604 00:33:59,880 --> 00:34:06,360 Speaker 4: is innovation? Are we pushing the envelope enough? Are we 605 00:34:06,440 --> 00:34:11,200 Speaker 4: staying ahead? Number three is which kind of feeds into 606 00:34:11,239 --> 00:34:13,200 Speaker 4: the innovation one is risk taking? 607 00:34:13,200 --> 00:34:16,680 Speaker 5: Are we taking enough risk? Without risk, there is no growth? 608 00:34:17,600 --> 00:34:22,080 Speaker 4: And I think the trap that every larger company inevitably 609 00:34:22,120 --> 00:34:27,440 Speaker 4: falls into is conservatism. Things are good enough, and so 610 00:34:27,600 --> 00:34:30,759 Speaker 4: it's are we pushing the envelope? Are we taking enough 611 00:34:30,880 --> 00:34:33,640 Speaker 4: risk to really have an impact? I'd say those are 612 00:34:33,640 --> 00:34:35,400 Speaker 4: probably the top three that I spend. 613 00:34:35,680 --> 00:34:39,239 Speaker 3: Last turn to define productivity paradox something. I know you've 614 00:34:39,400 --> 00:34:40,960 Speaker 3: thought a lot about what does that mean? 615 00:34:42,400 --> 00:34:45,040 Speaker 4: So I started thinking hard about this because all I 616 00:34:45,160 --> 00:34:50,920 Speaker 4: saw and read every day was fear about AI, and 617 00:34:52,000 --> 00:34:55,640 Speaker 4: I studied economics, and so I kind of went back 618 00:34:55,640 --> 00:34:59,480 Speaker 4: to like basic economics, and there's been like a macro 619 00:34:59,600 --> 00:35:03,600 Speaker 4: invest formula. I guess I would say it's been around 620 00:35:03,600 --> 00:35:11,680 Speaker 4: forever that says growth comes from productivity growth plus population 621 00:35:11,800 --> 00:35:18,520 Speaker 4: growth plus debt growth. So if those three things are working, you'll. 622 00:35:18,360 --> 00:35:19,320 Speaker 5: Get GDP growth. 623 00:35:20,600 --> 00:35:22,239 Speaker 4: And so then you think about that and you say, well, 624 00:35:23,120 --> 00:35:26,800 Speaker 4: debt growth, we're probably not going back to zero percent 625 00:35:26,880 --> 00:35:29,400 Speaker 4: interest rates, so to some extent there's going to be 626 00:35:29,400 --> 00:35:30,160 Speaker 4: a ceiling on that. 627 00:35:31,400 --> 00:35:32,080 Speaker 5: And then you. 628 00:35:32,000 --> 00:35:36,960 Speaker 4: Look at population growth. There are shockingly few countries or 629 00:35:37,040 --> 00:35:39,440 Speaker 4: places in the world that will see population growth over 630 00:35:39,480 --> 00:35:42,280 Speaker 4: the next thirty to fifty years. In fact, most places 631 00:35:42,320 --> 00:35:46,560 Speaker 4: are not even at replacement rates. And so I'm like, 632 00:35:46,560 --> 00:35:48,560 Speaker 4: all right, so population growth is not going to be there. 633 00:35:49,880 --> 00:35:51,840 Speaker 4: So that would mean if you just take it to 634 00:35:52,480 --> 00:35:59,920 Speaker 4: the extreme, the only chance of continued GDP growth is productivity. 635 00:36:00,960 --> 00:36:07,040 Speaker 4: And the best way to solve productivity is AI That's 636 00:36:07,040 --> 00:36:09,759 Speaker 4: why I say it's a paradox. On one hand, everybody's 637 00:36:09,800 --> 00:36:13,520 Speaker 4: scared after death it's going to take over the world, 638 00:36:14,040 --> 00:36:18,240 Speaker 4: take all of our jobs, ruin us. But in reality 639 00:36:18,280 --> 00:36:20,200 Speaker 4: maybe it's the other way, which is it's the only 640 00:36:20,239 --> 00:36:21,320 Speaker 4: thing that can save us. 641 00:36:21,600 --> 00:36:23,880 Speaker 5: Yeah, and if you believe. 642 00:36:23,640 --> 00:36:26,600 Speaker 4: That economic equation, which I think has proven quite true 643 00:36:26,640 --> 00:36:29,080 Speaker 4: over hundreds of years, I do think it's probably the 644 00:36:29,120 --> 00:36:30,080 Speaker 4: only thing that can save us. 645 00:36:31,560 --> 00:36:34,680 Speaker 3: Actually looked at the numbers yesterday for totally random reason 646 00:36:35,000 --> 00:36:38,160 Speaker 3: on population growth in Europe and receive this is a 647 00:36:38,160 --> 00:36:40,799 Speaker 3: special bonus question. We'll see how smart you are. Which 648 00:36:40,840 --> 00:36:45,160 Speaker 3: country in Europe? Condeently, Europe has the highest population growth? 649 00:36:46,880 --> 00:36:52,920 Speaker 4: It's small continental Europe, probably one of the Nordics, I would. 650 00:36:52,680 --> 00:36:58,640 Speaker 3: Guess, close. Luxembourg. Okay, something that's going on in Luxembourg. 651 00:37:00,600 --> 00:37:02,680 Speaker 3: I feel like, well, all of us need to investigate. 652 00:37:03,000 --> 00:37:05,000 Speaker 3: They're at one point four nine, which in the day, 653 00:37:05,040 --> 00:37:08,280 Speaker 3: by the way, would be a relatively that's the best 654 00:37:08,320 --> 00:37:11,200 Speaker 3: performing country. I mean in the day, you'd countries had 655 00:37:11,280 --> 00:37:14,560 Speaker 3: routinely had two points something, you know, percent growth in 656 00:37:15,000 --> 00:37:18,480 Speaker 3: a given year. Last question, you're writing a book. Now 657 00:37:18,800 --> 00:37:21,680 Speaker 3: we were talking chatting about it backstage, and now I 658 00:37:21,719 --> 00:37:25,880 Speaker 3: appreciate the paradox of this book, which is in a 659 00:37:26,120 --> 00:37:28,239 Speaker 3: universe with a model, is better in the afternoon than 660 00:37:28,280 --> 00:37:30,080 Speaker 3: it is in the morning. How do you write a 661 00:37:30,080 --> 00:37:33,680 Speaker 3: book that's like printed on paper? I expected to reuseful. 662 00:37:37,400 --> 00:37:41,320 Speaker 4: This is the challenge. And I am an incredible author 663 00:37:41,360 --> 00:37:44,399 Speaker 4: of useless books. I mean, most of what I've spent 664 00:37:44,520 --> 00:37:47,760 Speaker 4: time on in the last decade of stuff that's completely useless, 665 00:37:47,880 --> 00:37:52,160 Speaker 4: like a year after it's written. And so when we 666 00:37:52,160 --> 00:37:53,560 Speaker 4: were talking about it, I was like, I would like 667 00:37:53,560 --> 00:37:56,960 Speaker 4: to do something around AI that's timeless. 668 00:37:56,200 --> 00:37:59,919 Speaker 5: Yeah, that would be useful ten or twenty years. 669 00:38:00,160 --> 00:38:04,400 Speaker 4: No, But then to your point, so, how is that 670 00:38:04,440 --> 00:38:08,759 Speaker 4: even remotely possible if the model is better in the 671 00:38:08,760 --> 00:38:11,600 Speaker 4: afternoon and in the morning. So that's the challenge in 672 00:38:11,640 --> 00:38:14,240 Speaker 4: front of us. But the book is around AI value creation, 673 00:38:14,880 --> 00:38:18,160 Speaker 4: so kind of links to this productivity paradox, and how 674 00:38:18,200 --> 00:38:25,040 Speaker 4: do you actually get sustained value out of AI, out 675 00:38:25,080 --> 00:38:29,600 Speaker 4: of automation, out of data science. And so the biggest 676 00:38:29,640 --> 00:38:31,440 Speaker 4: challenge in front of us is can we make this 677 00:38:31,680 --> 00:38:34,719 Speaker 4: relevant that's the day that it's published. 678 00:38:34,800 --> 00:38:36,040 Speaker 3: How are you setting out to do that? 679 00:38:38,160 --> 00:38:41,160 Speaker 4: I think you have to to some extent level it 680 00:38:41,239 --> 00:38:43,880 Speaker 4: up to bigger concepts, which is kind of why I 681 00:38:43,920 --> 00:38:49,520 Speaker 4: go to things like macroeconomics, population geography as opposed to 682 00:38:49,600 --> 00:38:53,000 Speaker 4: going into the weeds of the technology itself. If you 683 00:38:53,080 --> 00:38:55,759 Speaker 4: write about this is how you get better performance out 684 00:38:55,760 --> 00:38:59,640 Speaker 4: of a model, we can agree that will be completely 685 00:38:59,680 --> 00:39:02,600 Speaker 4: useful two years from now, maybe even two months from now, 686 00:39:03,120 --> 00:39:07,280 Speaker 4: and so it will be less in the technical detail 687 00:39:08,000 --> 00:39:11,520 Speaker 4: and more of what is sustained value creation for AI, 688 00:39:12,200 --> 00:39:14,839 Speaker 4: which if you think on what is hopefully a ten 689 00:39:14,920 --> 00:39:18,520 Speaker 4: or twenty year period, it's probably we're kind of substituting 690 00:39:18,640 --> 00:39:21,520 Speaker 4: AI for technology. Now I've realized, because I think this 691 00:39:21,560 --> 00:39:25,080 Speaker 4: has always been true for technology. It's just now AI 692 00:39:25,160 --> 00:39:28,400 Speaker 4: is the thing that everybody wants to talk about. But 693 00:39:28,440 --> 00:39:30,560 Speaker 4: let's see if we can do it. Time will tell. 694 00:39:31,440 --> 00:39:34,120 Speaker 3: Did you get any inkling that the pace that this 695 00:39:34,239 --> 00:39:37,919 Speaker 3: AI year's phenomenon was gonna that things with the pace 696 00:39:37,920 --> 00:39:40,480 Speaker 3: of change was going to accelerate so much because you 697 00:39:40,560 --> 00:39:43,239 Speaker 3: had More's law, right, you had a model in the 698 00:39:43,280 --> 00:39:48,600 Speaker 3: technology world for this kind of exponential increase in so 699 00:39:48,640 --> 00:39:53,080 Speaker 3: were you were you thinking about that kind of accelerate 700 00:39:53,320 --> 00:39:55,440 Speaker 3: similar kind of acceleration in. 701 00:39:55,440 --> 00:40:00,319 Speaker 4: The I think anybody had said they expect did what 702 00:40:00,320 --> 00:40:05,800 Speaker 4: we're seeing today is probably exaggerating. I think it's way 703 00:40:05,840 --> 00:40:11,560 Speaker 4: faster than anybody expected. Yeah, but technologies, back to your 704 00:40:11,600 --> 00:40:15,520 Speaker 4: point at More's Law has always accelerated through the years, 705 00:40:15,520 --> 00:40:16,040 Speaker 4: So I. 706 00:40:16,040 --> 00:40:19,320 Speaker 5: Wouldn't say it's a shock, but it is surprising. 707 00:40:19,920 --> 00:40:25,440 Speaker 3: Yeah, You've had a kind of extraordinary privileged position to 708 00:40:25,680 --> 00:40:28,359 Speaker 3: watch and participate in this revolution, right, I mean, how 709 00:40:28,360 --> 00:40:33,960 Speaker 3: many other people have been in that have ridden this wave. 710 00:40:33,800 --> 00:40:34,160 Speaker 5: Like you have. 711 00:40:35,520 --> 00:40:38,680 Speaker 4: I do wonder is this really that much different or 712 00:40:38,719 --> 00:40:40,479 Speaker 4: does it feel different just because we're here? 713 00:40:41,520 --> 00:40:43,520 Speaker 5: I mean, I do think on one level. Yes. 714 00:40:44,400 --> 00:40:46,680 Speaker 4: So in the time I've been an IBM, internet happened, 715 00:40:48,239 --> 00:40:54,440 Speaker 4: Mobile happened, social network happened, blockchain happened. 716 00:40:55,000 --> 00:40:56,400 Speaker 5: AI. So a lot has happened. 717 00:40:56,719 --> 00:40:58,080 Speaker 4: But then you go back and say, well, but if 718 00:40:58,120 --> 00:41:03,759 Speaker 4: I'd been here between nineteen seventy and ninety five, there 719 00:41:03,760 --> 00:41:06,640 Speaker 4: were a lot of things that are pretty fundamental. Then too, say, 720 00:41:06,640 --> 00:41:09,840 Speaker 4: I wondered, almost do we do we always exaggerate the 721 00:41:09,880 --> 00:41:10,759 Speaker 4: timeframe that we're in. 722 00:41:13,160 --> 00:41:13,640 Speaker 5: I don't know. 723 00:41:14,080 --> 00:41:17,120 Speaker 4: Yeah, but it's a good idea though. 724 00:41:19,040 --> 00:41:22,040 Speaker 3: I think the ending with the phrase I don't know 725 00:41:22,800 --> 00:41:26,279 Speaker 3: it's a good idea though. That's the great way to 726 00:41:26,320 --> 00:41:32,560 Speaker 3: wrap this up. Thank you so much, Thank you, Malcolm. 727 00:41:32,760 --> 00:41:35,960 Speaker 3: In a field that is evolving as quickly as artificial intelligence, 728 00:41:36,320 --> 00:41:39,160 Speaker 3: it was inspiring to see how adaptable Rob has been 729 00:41:39,200 --> 00:41:43,000 Speaker 3: over his career. The takeaways from my conversation with Rob 730 00:41:43,280 --> 00:41:47,160 Speaker 3: had been echoing in my head ever since. He emphasized 731 00:41:47,280 --> 00:41:50,960 Speaker 3: how open source models allow AI technology to be developed 732 00:41:50,960 --> 00:41:56,399 Speaker 3: by many players. Openness also allows for transparency. Rob told 733 00:41:56,440 --> 00:42:01,360 Speaker 3: me about AI use cases like IBM's collaborate with Sevilla's 734 00:42:01,400 --> 00:42:05,080 Speaker 3: football club. That example really brought home for me how 735 00:42:05,160 --> 00:42:10,360 Speaker 3: AI technology will touch every industry. Despite the potential benefits 736 00:42:10,360 --> 00:42:15,320 Speaker 3: of AI, challenges exist in its widespread adoption. Rob discussed 737 00:42:15,520 --> 00:42:20,520 Speaker 3: how resistance to change, concerns about job security and organizational 738 00:42:20,600 --> 00:42:26,480 Speaker 3: inertia can slow down implementation of AI solutions. The paradox, though, 739 00:42:26,520 --> 00:42:29,040 Speaker 3: according to Rob, is that rather than being afraid of 740 00:42:29,080 --> 00:42:32,720 Speaker 3: a world with AI, people should actually be more afraid 741 00:42:32,719 --> 00:42:36,480 Speaker 3: of a world without it. AI, he believes, has the 742 00:42:36,520 --> 00:42:39,400 Speaker 3: potential to make the world a better place in a 743 00:42:39,440 --> 00:42:43,879 Speaker 3: way that no other technology can. Rob painted an optimistic 744 00:42:44,000 --> 00:42:47,520 Speaker 3: version of the future, one in which AI technology will 745 00:42:47,560 --> 00:42:51,839 Speaker 3: continue to improve at an exponential rate. This will free 746 00:42:51,920 --> 00:42:56,000 Speaker 3: up workers to dedicate their energy to more creative tasks. 747 00:42:56,600 --> 00:43:01,120 Speaker 3: I for one am on board Smart Talks with IBM 748 00:43:01,239 --> 00:43:05,360 Speaker 3: is produced by Matt Romano, Joey Fishground and Jacob Goldstein. 749 00:43:05,760 --> 00:43:09,560 Speaker 3: We're edited by Lydia gene Kott. Our engineers are Sarah 750 00:43:09,560 --> 00:43:14,720 Speaker 3: Bruguer and Ben Tolliday. Theme song by Gramscow. Special thanks 751 00:43:14,719 --> 00:43:17,239 Speaker 3: to the eight Bar and IBM teams, as well as 752 00:43:17,239 --> 00:43:20,799 Speaker 3: the Pushkin marketing team. Smart Talks with IBM is a 753 00:43:20,840 --> 00:43:25,600 Speaker 3: production of Pushkin Industries and Ruby Studio at iHeartMedia. To 754 00:43:25,640 --> 00:43:31,280 Speaker 3: find more Pushkin podcasts, listen on the iHeartRadio app, Apple Podcasts, 755 00:43:31,400 --> 00:43:36,200 Speaker 3: or wherever you listen to podcasts. I'm Malcolm Gladwell. This 756 00:43:36,280 --> 00:43:39,880 Speaker 3: is a paid advertisement from IBM. The conversations on this 757 00:43:40,000 --> 00:43:52,640 Speaker 3: podcast don't necessarily represent IBM's positions, strategies, or opinions.