1 00:00:02,080 --> 00:00:06,120 Speaker 1: To understand why the cosmetics supergiant Lorel Group is teaming 2 00:00:06,200 --> 00:00:09,200 Speaker 1: up with IBM, you must first take a closer look 3 00:00:09,280 --> 00:00:13,160 Speaker 1: at its products. Take lipstick, for example, It's one of 4 00:00:13,160 --> 00:00:16,360 Speaker 1: those things that seems straightforward, a waxy cylinder that you 5 00:00:16,440 --> 00:00:19,440 Speaker 1: rub on your lips to turn them a different color. Easy, right, 6 00:00:20,000 --> 00:00:23,759 Speaker 1: Well maybe not, as my colleague Lucy Sullivan found out 7 00:00:23,960 --> 00:00:27,040 Speaker 1: when I sent her an assignment to Lorel's North America 8 00:00:27,120 --> 00:00:28,840 Speaker 1: Research and Innovation Center. 9 00:00:29,920 --> 00:00:35,000 Speaker 2: All right, I am reporting live from the Looreal visitor's 10 00:00:35,159 --> 00:00:39,040 Speaker 2: parking lot. Malcolm told me that he would be sending 11 00:00:39,040 --> 00:00:44,919 Speaker 2: me to Paris, France for this Looreal excursion, but instead 12 00:00:45,000 --> 00:00:48,360 Speaker 2: I am in Clark, New Jersey. Pass a lot of 13 00:00:48,400 --> 00:00:51,760 Speaker 2: strip malls on the way here. But to be fair 14 00:00:51,800 --> 00:00:55,600 Speaker 2: to Clark, New Jersey and Lorel, this is a beautiful compound. 15 00:00:55,600 --> 00:00:58,800 Speaker 2: It kind of looks like a spa. 16 00:00:59,000 --> 00:01:02,560 Speaker 1: Lucy went into the center and was blown away. The 17 00:01:02,600 --> 00:01:07,840 Speaker 1: facility houses about six hundred scientists and experts across skincare, makeup, fragrance, 18 00:01:07,920 --> 00:01:12,000 Speaker 1: hair care, innovative packaging, and tech. It is one of 19 00:01:12,040 --> 00:01:15,319 Speaker 1: the largest formulation lab spaces in the industry. It's the 20 00:01:15,360 --> 00:01:19,800 Speaker 1: size of six basketball courts. The reason Loreel's facility is 21 00:01:19,840 --> 00:01:22,800 Speaker 1: so big and has so many people is that everything 22 00:01:22,880 --> 00:01:25,720 Speaker 1: Loriel does to bring a product to market happens here, 23 00:01:26,319 --> 00:01:30,920 Speaker 1: from molecule discovery and product development to consumer testing. The 24 00:01:30,959 --> 00:01:35,040 Speaker 1: center even has its own mini factory. My conception of lipstick, 25 00:01:35,360 --> 00:01:39,160 Speaker 1: that it's just a waxy stick was plain wrong. Lipstick 26 00:01:39,400 --> 00:01:42,520 Speaker 1: is a high performance product born from years of research, 27 00:01:42,880 --> 00:01:48,760 Speaker 1: consumer insights, and precision science. Lipstick isn't simple. It's incredibly complex, 28 00:01:49,360 --> 00:01:51,840 Speaker 1: and one of the main reasons it's so complex is 29 00:01:51,880 --> 00:01:54,720 Speaker 1: just a nature of fashion trends. The kind of lipstick 30 00:01:54,760 --> 00:01:57,240 Speaker 1: consumers want is constantly changing. 31 00:01:57,760 --> 00:01:59,880 Speaker 3: A lot of our consumer insights with Floreal is like, 32 00:02:00,280 --> 00:02:02,160 Speaker 3: where are consumers going in the future. 33 00:02:02,480 --> 00:02:06,440 Speaker 1: This is Nadine Gomez. She's vice president for Loreel's research 34 00:02:06,480 --> 00:02:08,240 Speaker 1: and innovation development team. 35 00:02:08,639 --> 00:02:11,040 Speaker 3: Our chemists are working on five six years down the line. 36 00:02:11,240 --> 00:02:14,839 Speaker 3: We predicted that consumers wanted more of a softer look 37 00:02:14,840 --> 00:02:15,760 Speaker 3: on their lips as well. 38 00:02:15,880 --> 00:02:17,400 Speaker 2: So how do you predict something like that. 39 00:02:17,960 --> 00:02:22,320 Speaker 3: We see slow signals from fashion houses and social media 40 00:02:22,360 --> 00:02:24,239 Speaker 3: and things like that. We kind of see that trend 41 00:02:24,320 --> 00:02:27,440 Speaker 3: evolving a little bit, and then we know at five 42 00:02:27,480 --> 00:02:28,799 Speaker 3: six years it's going to become. 43 00:02:28,560 --> 00:02:32,480 Speaker 1: Big Lucy talked with her about the origins of one 44 00:02:32,520 --> 00:02:36,200 Speaker 1: of their products, Mabelne matt Inc Liquid lipstick. 45 00:02:36,760 --> 00:02:39,080 Speaker 3: Our competitors have two steps. The first step is a 46 00:02:39,120 --> 00:02:41,720 Speaker 3: base go it's super opaque. You get the color and 47 00:02:41,760 --> 00:02:45,320 Speaker 3: you get the maddy, but it's very, very drying ellips. 48 00:02:45,360 --> 00:02:47,360 Speaker 3: You cannot wear that, honestly more than ten minutes. It 49 00:02:47,360 --> 00:02:50,120 Speaker 3: feels like your lips are like aching at one point. 50 00:02:50,360 --> 00:02:51,960 Speaker 3: So we had to develop a top cot and you'll 51 00:02:51,960 --> 00:02:53,720 Speaker 3: see many of our competitors did the same thing. It's 52 00:02:53,720 --> 00:02:55,760 Speaker 3: like a bomb. You put it on top, it's super comfortable, 53 00:02:56,400 --> 00:02:59,680 Speaker 3: but we also noticed that consumers kind of get tired 54 00:03:00,120 --> 00:03:02,440 Speaker 3: reapplying a bomb. So we're like, what can we do 55 00:03:02,560 --> 00:03:04,680 Speaker 3: to create this two step into one step? 56 00:03:05,440 --> 00:03:08,280 Speaker 1: So Loriel had a challenge, how do you make a 57 00:03:08,360 --> 00:03:12,960 Speaker 1: comfortable liquid matt lipstick that doesn't require consumers to reapply 58 00:03:13,080 --> 00:03:16,560 Speaker 1: a top layer of bomb. Solving this type of problem 59 00:03:16,760 --> 00:03:19,600 Speaker 1: takes a lot of resources and a lot of expertise, 60 00:03:19,960 --> 00:03:24,160 Speaker 1: and crucially, it takes time. Remember, Nadine said that working 61 00:03:24,200 --> 00:03:27,160 Speaker 1: on a breakthrough product such as matt Inc can take 62 00:03:27,400 --> 00:03:31,240 Speaker 1: years before it comes out. But can this process be 63 00:03:31,360 --> 00:03:36,240 Speaker 1: accelerated taken further? Be even more sustainable. That's what IBM 64 00:03:36,280 --> 00:03:40,920 Speaker 1: and Loreel are hoping to find out. My name is 65 00:03:40,920 --> 00:03:44,000 Speaker 1: Malcolm Gladwell. You're listening to the latest episode of Smart 66 00:03:44,040 --> 00:03:47,000 Speaker 1: Talks with IBM, where we offer our listeners a glimpse 67 00:03:47,040 --> 00:03:50,640 Speaker 1: behind the curtain of the world of technology. In our 68 00:03:50,720 --> 00:03:54,240 Speaker 1: last episode, we talked about how an AI assistant created 69 00:03:54,240 --> 00:03:58,360 Speaker 1: with IBM Watson X helps future teachers practice responsive teaching 70 00:03:58,760 --> 00:04:04,200 Speaker 1: by simulating class interactions with students. In this episode, we 71 00:04:04,360 --> 00:04:07,760 Speaker 1: take you on an even more unexpected journey into the 72 00:04:07,760 --> 00:04:13,320 Speaker 1: world of cosmetics, hair care, skincare, fragrance, makeup, and how 73 00:04:13,360 --> 00:04:17,359 Speaker 1: a custom AI model could help Loriel's researchers shape the 74 00:04:17,400 --> 00:04:20,159 Speaker 1: future of what we put on our faces every morning. 75 00:04:28,040 --> 00:04:30,159 Speaker 1: I want to say on lipstick a moment longer to 76 00:04:30,240 --> 00:04:34,080 Speaker 1: help illustrate what goes into Loriel's product development, and let's 77 00:04:34,080 --> 00:04:38,320 Speaker 1: focus on matt inc lipstick. Loriel wanted to create something 78 00:04:38,520 --> 00:04:41,599 Speaker 1: that was comfortable and could be applied in one step. 79 00:04:42,240 --> 00:04:44,279 Speaker 4: So to go from two step to one step, we 80 00:04:44,320 --> 00:04:46,760 Speaker 4: had to look cross functionally and try to figure out 81 00:04:46,760 --> 00:04:49,160 Speaker 4: what can we bring into the product to make it 82 00:04:49,200 --> 00:04:52,400 Speaker 4: more comfortable, and luckily we have many different types of 83 00:04:52,400 --> 00:04:53,320 Speaker 4: products at Lorel. 84 00:04:53,720 --> 00:04:56,960 Speaker 1: That's Alex Good, a senior chemist who leads the lip 85 00:04:56,960 --> 00:05:00,159 Speaker 1: products team in North America. She says the trick to 86 00:05:00,240 --> 00:05:04,520 Speaker 1: making matt incwork was finding an elastomer, a substance they 87 00:05:04,560 --> 00:05:06,560 Speaker 1: were already using in foundation. 88 00:05:07,480 --> 00:05:10,360 Speaker 4: We have this elastomer that can give you like more 89 00:05:10,360 --> 00:05:13,040 Speaker 4: comfortable and make it feel like there's like something on 90 00:05:13,040 --> 00:05:14,000 Speaker 4: your lips, like a cushion. 91 00:05:14,800 --> 00:05:18,320 Speaker 1: She handed Lucy two jars. The first jar contained the 92 00:05:18,360 --> 00:05:20,839 Speaker 1: former version of the product that was used in super 93 00:05:20,880 --> 00:05:24,280 Speaker 1: State twenty four. By the way, this is exactly why 94 00:05:24,320 --> 00:05:26,800 Speaker 1: I sent Lucy to the lab in my place the. 95 00:05:26,839 --> 00:05:30,000 Speaker 4: Samples, and I actually have something for you to try here, 96 00:05:30,520 --> 00:05:34,599 Speaker 4: so you can try this is what was the initial product. 97 00:05:34,920 --> 00:05:38,480 Speaker 2: Okay, so this is like it sort of looks like okay, 98 00:05:38,480 --> 00:05:41,360 Speaker 2: it is. It looks like vacline that has like more 99 00:05:41,400 --> 00:05:44,239 Speaker 2: of a color. It's kind of a beige, looks like 100 00:05:44,279 --> 00:05:44,799 Speaker 2: some skin. 101 00:05:44,960 --> 00:05:45,359 Speaker 5: Okay. 102 00:05:45,600 --> 00:05:47,920 Speaker 2: So this was from the two steps this would go 103 00:05:47,960 --> 00:05:55,080 Speaker 2: on after Oh okay, right islet Okay, So. 104 00:05:55,080 --> 00:05:58,280 Speaker 4: It feels like very wet. As you can see, it's 105 00:05:58,320 --> 00:06:01,279 Speaker 4: kind of it's gonna absorb into your skin and leaves 106 00:06:01,800 --> 00:06:03,520 Speaker 4: and then you're gonna feel the dryness. 107 00:06:03,120 --> 00:06:05,560 Speaker 2: Of the product. Okay, so we're gonna move. 108 00:06:05,560 --> 00:06:07,640 Speaker 4: From the clay product that you have on your hand 109 00:06:07,720 --> 00:06:09,640 Speaker 4: now to the elastimmer or. 110 00:06:09,680 --> 00:06:10,520 Speaker 6: You try half hour. 111 00:06:11,279 --> 00:06:14,760 Speaker 1: This jar held the elastomer that Laurel had spent years 112 00:06:14,800 --> 00:06:15,800 Speaker 1: developing in the lab. 113 00:06:16,800 --> 00:06:20,960 Speaker 2: This one is it clear, looks like aqua for a 114 00:06:21,080 --> 00:06:23,040 Speaker 2: much bligger and you. 115 00:06:23,040 --> 00:06:26,480 Speaker 4: Can pay a physical layer that you're putting on your aid. 116 00:06:26,680 --> 00:06:30,920 Speaker 2: Yeah, so that's much thicker. It kind of like clumps together. Yeah, 117 00:06:31,880 --> 00:06:35,640 Speaker 2: it's more of a cloudy. It's less shimmery though that's intended. 118 00:06:35,880 --> 00:06:40,760 Speaker 4: Yes, so this is a like a powder, this dispersed 119 00:06:40,880 --> 00:06:45,000 Speaker 4: in dimetic code, and it creates like a comfort on 120 00:06:45,000 --> 00:06:47,080 Speaker 4: your lips, a field like there's something there for a 121 00:06:47,120 --> 00:06:50,760 Speaker 4: barrier to keep the film form on. And that's like 122 00:06:50,800 --> 00:06:53,680 Speaker 4: the key ingredient that came from Foundation that we transferred 123 00:06:53,680 --> 00:06:56,719 Speaker 4: into lipstick to give us this innovative product ahead of 124 00:06:56,760 --> 00:07:00,360 Speaker 4: the market. Yeah, this is what gives it comfort. So 125 00:07:00,440 --> 00:07:03,200 Speaker 4: the difference between super State twenty four and matt Inc 126 00:07:03,320 --> 00:07:05,800 Speaker 4: is really the comfort. They both last a long time, 127 00:07:06,160 --> 00:07:08,800 Speaker 4: but this matt Inc you don't have to apply the 128 00:07:08,839 --> 00:07:11,320 Speaker 4: bomb over and over again, So you can apply matt 129 00:07:11,320 --> 00:07:12,440 Speaker 4: ink once for the day and. 130 00:07:12,400 --> 00:07:16,200 Speaker 1: You're good, right, Alex Good is under selling it here 131 00:07:16,840 --> 00:07:19,440 Speaker 1: once for the day, and you're good. That's a liquid 132 00:07:19,480 --> 00:07:24,000 Speaker 1: lipstick revolution. Literally millions of Loreal consumers around the world 133 00:07:24,240 --> 00:07:28,080 Speaker 1: have worn matt ink. It's a blockbuster. It's also a 134 00:07:28,120 --> 00:07:33,200 Speaker 1: marvel of science. The world's first liquid lipstick was developed 135 00:07:33,240 --> 00:07:35,520 Speaker 1: in the nineteen thirties, and it was actually just a 136 00:07:35,600 --> 00:07:39,680 Speaker 1: stain for your lips, barely counts as lipstick. Then came 137 00:07:39,760 --> 00:07:42,400 Speaker 1: another wave of liquid lipstick when they were able to 138 00:07:42,440 --> 00:07:45,600 Speaker 1: make it matt That was a two step version. It 139 00:07:45,600 --> 00:07:48,800 Speaker 1: felt heavy on your lips. You had to keep reapplying 140 00:07:48,800 --> 00:07:53,600 Speaker 1: the top coat. It was inconvenient. Lorel tackled that challenge 141 00:07:53,640 --> 00:07:56,679 Speaker 1: in the lab, with chemists like Alex and Nadine leaving 142 00:07:56,720 --> 00:08:01,840 Speaker 1: the charge their breakthrough matt Inc. But creating matt Inc 143 00:08:01,920 --> 00:08:05,040 Speaker 1: took a long time, trial and error, the hard work 144 00:08:05,080 --> 00:08:10,000 Speaker 1: of scientific experimentation. As Nadine told Lucy, the lipstick team 145 00:08:10,240 --> 00:08:13,160 Speaker 1: had to put the new product to extensive tests. 146 00:08:13,600 --> 00:08:16,520 Speaker 3: We do a very robustability system here. You know, we 147 00:08:16,560 --> 00:08:19,960 Speaker 3: have color odor appearance. We monitor this in extreme conditions. 148 00:08:20,320 --> 00:08:22,920 Speaker 3: We simulate a forty five degrees celsius and that can 149 00:08:22,960 --> 00:08:25,080 Speaker 3: be something like a three year shelf life. I'm saying, 150 00:08:25,680 --> 00:08:28,600 Speaker 3: we simulate your real life product. Like if you leave 151 00:08:28,640 --> 00:08:31,240 Speaker 3: your lip gloss in the car in Arizona's one hundred 152 00:08:31,240 --> 00:08:33,040 Speaker 3: and twelve degrees for three days, is it still going 153 00:08:33,080 --> 00:08:36,000 Speaker 3: to perform? Is it gonna smell? Is it gonna look granted? 154 00:08:36,120 --> 00:08:37,199 Speaker 3: Is it gonna change colors? 155 00:08:37,400 --> 00:08:38,160 Speaker 7: We do all that. 156 00:08:38,800 --> 00:08:42,680 Speaker 1: See what I mean. Lipstick is complex. Most people would 157 00:08:42,679 --> 00:08:45,600 Speaker 1: never consider it a piece of technology, but one lip 158 00:08:45,600 --> 00:08:48,319 Speaker 1: product has millions of data. 159 00:08:48,040 --> 00:08:50,439 Speaker 3: Points, so much science behind. And you can see here 160 00:08:50,480 --> 00:08:52,280 Speaker 3: how many scientists we have. You know, some of them 161 00:08:52,280 --> 00:08:56,120 Speaker 3: have PhDs, some of them have masters degrees chemistry, biology, psychology. 162 00:08:56,160 --> 00:08:59,839 Speaker 1: Also, when I first heard about this collaboration between LORI 163 00:09:00,120 --> 00:09:03,679 Speaker 1: L and IBM, I was surprised. I thought, these are 164 00:09:03,720 --> 00:09:07,080 Speaker 1: two very different companies. What do they really have in common? 165 00:09:07,880 --> 00:09:10,720 Speaker 1: Meet you guys, Yeah. To find out, I went to 166 00:09:10,760 --> 00:09:13,840 Speaker 1: the IBM Research Center outside New York City, which I 167 00:09:13,880 --> 00:09:15,600 Speaker 1: have to say is one of the coolest buildings I've 168 00:09:15,640 --> 00:09:19,920 Speaker 1: ever been in, A semi circular modernist masterpiece with a 169 00:09:19,920 --> 00:09:22,679 Speaker 1: long curving wall of windows, looks like something out of 170 00:09:22,679 --> 00:09:25,560 Speaker 1: a Stanley Kubrick movie. I was there to talk with 171 00:09:25,679 --> 00:09:29,880 Speaker 1: two experts from research and innovation at Lorel, Methu Cassier 172 00:09:30,320 --> 00:09:35,079 Speaker 1: and Gabriel Bertoli metthew is VP for Digital and Transformation. 173 00:09:35,600 --> 00:09:40,600 Speaker 1: Gabriel is the Chief Digital Transformation Officer for Formulation. These 174 00:09:40,600 --> 00:09:43,680 Speaker 1: are the people whose jobs are to oversee big changes 175 00:09:43,720 --> 00:09:46,840 Speaker 1: within the company. And Methu told me to try on 176 00:09:46,880 --> 00:09:47,560 Speaker 1: some lipstick. 177 00:09:48,000 --> 00:09:49,840 Speaker 7: I'm gonna make you try this one. 178 00:09:50,240 --> 00:09:54,360 Speaker 1: Okay, this is super stay viniting final inc. 179 00:09:54,559 --> 00:09:55,600 Speaker 7: Yeah, so that's a glosse. 180 00:09:55,720 --> 00:09:57,040 Speaker 1: Never in my life put on lipsey. You have no 181 00:09:57,080 --> 00:09:57,640 Speaker 1: idea what I'm doing. 182 00:09:57,679 --> 00:09:59,440 Speaker 7: You don't have to put it. You can try it virtu. 183 00:10:00,240 --> 00:10:03,640 Speaker 1: Oh this may not be news to people who buy makeup, 184 00:10:03,880 --> 00:10:06,320 Speaker 1: but it was news to me. You can try on 185 00:10:06,400 --> 00:10:11,600 Speaker 1: Loreal products virtually. They call it augmented beauty. Oh my goodness. 186 00:10:12,280 --> 00:10:14,600 Speaker 1: That is the strangest thing I've ever said. I look 187 00:10:14,679 --> 00:10:16,480 Speaker 1: quite fetching, that's the way. 188 00:10:16,480 --> 00:10:16,920 Speaker 2: It amazing. 189 00:10:18,160 --> 00:10:21,120 Speaker 7: And I can just hit you can choose your color absolutely. 190 00:10:21,840 --> 00:10:24,199 Speaker 1: So I'm on a little app it's looking at me 191 00:10:24,520 --> 00:10:26,839 Speaker 1: and it's just showing me exactly how I would look 192 00:10:26,840 --> 00:10:29,360 Speaker 1: with different shades of lipstick. So the odd idea of 193 00:10:29,400 --> 00:10:31,319 Speaker 1: going into a store and trying on each one, you 194 00:10:31,400 --> 00:10:32,960 Speaker 1: cannot do that from home, if you're not even at 195 00:10:33,000 --> 00:10:33,360 Speaker 1: the store. 196 00:10:33,840 --> 00:10:36,840 Speaker 7: Yeah, absolutely, that's all purpose. If you want to manage 197 00:10:36,880 --> 00:10:38,880 Speaker 7: a trend, I would go for something more like pitch. 198 00:10:40,040 --> 00:10:41,240 Speaker 1: You think I'm a peach person. 199 00:10:42,240 --> 00:10:42,520 Speaker 4: I don't know. 200 00:10:43,000 --> 00:10:45,560 Speaker 1: That looks I have to say that looks kind of natural. 201 00:10:45,600 --> 00:10:48,720 Speaker 1: It just is enhanced. It's given me a boys share 202 00:10:48,760 --> 00:10:53,880 Speaker 1: I would not otherwise have. This is why Loreel says 203 00:10:53,920 --> 00:10:58,480 Speaker 1: it creates beauty products and beauty experiences. Loriel is a 204 00:10:58,480 --> 00:11:02,920 Speaker 1: beauty tech company. Over the last decade, Laurel has seized 205 00:11:02,920 --> 00:11:07,319 Speaker 1: the power of AI and more recently, generative AI technology 206 00:11:07,360 --> 00:11:11,600 Speaker 1: has become a driving force alongside science and creativity. And 207 00:11:11,640 --> 00:11:15,520 Speaker 1: while some of this digital technology is relatively new, Matthew 208 00:11:15,760 --> 00:11:19,480 Speaker 1: helped me see that IBM and Lorel have always had 209 00:11:19,480 --> 00:11:20,240 Speaker 1: a lot in common. 210 00:11:20,800 --> 00:11:23,920 Speaker 7: I saw the original creator of Loyal, Jentulier, was a 211 00:11:23,960 --> 00:11:26,880 Speaker 7: chemist in nineteen or nine, so one hundred and sixteen 212 00:11:26,920 --> 00:11:30,760 Speaker 7: years ago, and he created this new air color type 213 00:11:31,000 --> 00:11:34,080 Speaker 7: for the market in France, and then little by little, 214 00:11:34,400 --> 00:11:36,679 Speaker 7: it has been always a very scientific company. So if 215 00:11:36,679 --> 00:11:39,120 Speaker 7: you look a little bit at key facts, we invented 216 00:11:39,160 --> 00:11:43,040 Speaker 7: sun filters in the nineteen thirties. There was a very 217 00:11:43,120 --> 00:11:45,760 Speaker 7: very big milestone where we also invented not only product, 218 00:11:45,800 --> 00:11:49,400 Speaker 7: but a reconstructed skin. So if you look at nineteen 219 00:11:49,800 --> 00:11:53,440 Speaker 7: seventeen nine, we've been the created this reconstructed kin that 220 00:11:53,559 --> 00:11:55,839 Speaker 7: helped us to go out of animal testing very fast 221 00:11:55,840 --> 00:11:57,920 Speaker 7: and by the way, before the law even asked it 222 00:11:57,920 --> 00:12:01,880 Speaker 7: to cosmetic companies. And then more recently, because it's a 223 00:12:01,960 --> 00:12:05,440 Speaker 7: story of innovation, we launch on new molecules like one 224 00:12:05,440 --> 00:12:07,800 Speaker 7: that you can find in laroche pose Milabi three, which 225 00:12:07,840 --> 00:12:11,280 Speaker 7: is really helping people to find against some you know, 226 00:12:11,880 --> 00:12:13,920 Speaker 7: spots that they could have on their skin. It's all 227 00:12:13,960 --> 00:12:16,200 Speaker 7: about like big mountation, how to regulate it. 228 00:12:16,880 --> 00:12:20,440 Speaker 1: Loreel and IBM were both started in the early twentieth century, 229 00:12:20,800 --> 00:12:24,400 Speaker 1: Lorel in nineteen oh nine and IBM in nineteen eleven. 230 00:12:24,880 --> 00:12:28,760 Speaker 1: Both companies have long standing histories of innovation, of using 231 00:12:28,800 --> 00:12:31,960 Speaker 1: trial and error to improve everything they do. The two 232 00:12:31,960 --> 00:12:34,720 Speaker 1: companies have been doing that in parallel for more than 233 00:12:34,800 --> 00:12:38,959 Speaker 1: a century until recently. When does it start, When do 234 00:12:39,120 --> 00:12:40,720 Speaker 1: Lorel and IBM start working together? 235 00:12:41,559 --> 00:12:44,440 Speaker 8: So we started in twenty twenty three at the end 236 00:12:44,480 --> 00:12:46,600 Speaker 8: of the year. But you know, really the discussion is 237 00:12:46,640 --> 00:12:51,240 Speaker 8: really recent, absolutely, absolutely, it's really recent in reality, you know, 238 00:12:51,320 --> 00:12:54,640 Speaker 8: I would say the first really interaction happened at the 239 00:12:54,640 --> 00:12:56,040 Speaker 8: beginning of twenty twenty four. 240 00:12:56,840 --> 00:13:00,240 Speaker 1: This is Gabriel Bertoli who I spoke to alongside too. 241 00:13:01,000 --> 00:13:04,199 Speaker 8: What really played a key role here is we wanted 242 00:13:04,200 --> 00:13:08,400 Speaker 8: to bring from a logic perspective to R and D together, 243 00:13:09,640 --> 00:13:13,080 Speaker 8: which normally you know companies like us, you just go 244 00:13:13,200 --> 00:13:16,040 Speaker 8: to a provider. You know it's a customer and the 245 00:13:16,080 --> 00:13:18,640 Speaker 8: supplier and your work they delivered to you. Here the 246 00:13:18,679 --> 00:13:20,080 Speaker 8: concept was totally different. 247 00:13:20,920 --> 00:13:24,360 Speaker 1: Mid two said that the collaboration began with simple conversations. 248 00:13:24,720 --> 00:13:27,560 Speaker 7: So if you look at the way IBM entered into 249 00:13:28,160 --> 00:13:32,400 Speaker 7: loyal Labs, it's started by interviewing people what would help 250 00:13:32,440 --> 00:13:34,680 Speaker 7: you to do your job? What is your business need? 251 00:13:35,200 --> 00:13:37,680 Speaker 7: So it was, by the way, two months ago, a 252 00:13:37,880 --> 00:13:41,680 Speaker 7: long series of interviews and from all the people around 253 00:13:41,720 --> 00:13:44,880 Speaker 7: the world we have in research in Brazil, in India, 254 00:13:45,000 --> 00:13:49,720 Speaker 7: in China, Japan, US, France of course, so we really 255 00:13:49,720 --> 00:13:51,320 Speaker 7: want to make sure that at the end of the day, 256 00:13:51,679 --> 00:13:53,760 Speaker 7: this new model, this new tool that we will give 257 00:13:53,800 --> 00:13:56,520 Speaker 7: to people is really people centrick in the way that 258 00:13:56,600 --> 00:13:58,040 Speaker 7: it selves their daily need. 259 00:13:58,679 --> 00:14:03,480 Speaker 1: More. The point has leveraged technology for decades and accumulated 260 00:14:03,760 --> 00:14:09,080 Speaker 1: amounted of scientific knowledge, everything from consumer aspirations and market trends, 261 00:14:09,400 --> 00:14:12,800 Speaker 1: to the results of all the experiments conducted during product development, 262 00:14:13,400 --> 00:14:17,080 Speaker 1: to which formulations melt in a hot car. It's hard 263 00:14:17,080 --> 00:14:20,880 Speaker 1: to get your head around. Looreal isn't just a cosmetics company. 264 00:14:21,360 --> 00:14:23,640 Speaker 1: It's a beauty data powerhouse. 265 00:14:24,560 --> 00:14:29,400 Speaker 8: If we have sixteen thousand terabat of data coming from 266 00:14:29,760 --> 00:14:37,280 Speaker 8: consumer insights, coming from market research coming from sales, well 267 00:14:37,440 --> 00:14:42,280 Speaker 8: with the new technology, maybe by aligning those two and 268 00:14:42,400 --> 00:14:45,400 Speaker 8: using best in class technology you can solve that problem. 269 00:14:45,480 --> 00:14:48,200 Speaker 1: So you say you have sixteen terabytes of data, put 270 00:14:48,200 --> 00:14:50,160 Speaker 1: that in perspective. How much data is that? 271 00:14:50,720 --> 00:14:51,040 Speaker 2: Give me? 272 00:14:52,360 --> 00:14:56,320 Speaker 8: This is one hundred year of Floreal data based on 273 00:14:56,520 --> 00:15:01,400 Speaker 8: the last forty years of data the systems. So this 274 00:15:01,520 --> 00:15:03,400 Speaker 8: is really I mean, we're talking about one hundred years 275 00:15:03,400 --> 00:15:06,440 Speaker 8: of data that only Loreal have. Let's take the example 276 00:15:06,520 --> 00:15:08,880 Speaker 8: of the ellipsis. I mean, you know, if ellipsex can 277 00:15:08,920 --> 00:15:13,080 Speaker 8: be between twenty and thirty year ow material, each raw 278 00:15:13,120 --> 00:15:16,760 Speaker 8: material will have I would say ten or fifteen way 279 00:15:16,960 --> 00:15:19,840 Speaker 8: of doing things. 280 00:15:21,320 --> 00:15:23,800 Speaker 1: Gabrielle is talking about how things used to be done. 281 00:15:24,120 --> 00:15:28,040 Speaker 1: Researchers at Lorel needed roughly twenty five ingredients for a 282 00:15:28,040 --> 00:15:31,160 Speaker 1: new lipstick formulation, but they have to choose from a 283 00:15:31,160 --> 00:15:35,480 Speaker 1: pool of hundreds, if not thousands, of raw materials, And 284 00:15:35,560 --> 00:15:37,800 Speaker 1: even after they settle on the ones they want, they 285 00:15:37,800 --> 00:15:40,200 Speaker 1: have to figure out how much of each ingredient they 286 00:15:40,240 --> 00:15:44,600 Speaker 1: need and in what form, what molecular weight, what combination. 287 00:15:45,360 --> 00:15:48,280 Speaker 1: It's not just a math problem. It's a problem that 288 00:15:48,400 --> 00:15:55,920 Speaker 1: requires balancing multiple perspectives safety, performance, quality, compliance standards, sustainability, 289 00:15:56,000 --> 00:15:59,720 Speaker 1: and more. It can take years. But what if you 290 00:15:59,760 --> 00:16:03,560 Speaker 1: could simulate hundreds of cars parked in a sweltering heat. 291 00:16:03,920 --> 00:16:06,480 Speaker 1: What if you could do all those trials and errors 292 00:16:06,800 --> 00:16:10,840 Speaker 1: virtually over and over and over again. What if instead 293 00:16:10,840 --> 00:16:14,640 Speaker 1: of mixing materials together by hand, you could ask AI 294 00:16:14,760 --> 00:16:18,680 Speaker 1: to predict what combinations might work best and then try 295 00:16:18,760 --> 00:16:19,720 Speaker 1: those out first. 296 00:16:20,280 --> 00:16:24,440 Speaker 8: This is ten on the power of twenty five. This 297 00:16:24,520 --> 00:16:29,000 Speaker 8: is one hundred billion of years for a human to 298 00:16:29,120 --> 00:16:33,200 Speaker 8: do a change in the formula or the possibility they have. 299 00:16:34,000 --> 00:16:39,040 Speaker 8: You can only do this by using technology, power of 300 00:16:39,120 --> 00:16:40,640 Speaker 8: technology and data that you have. 301 00:16:41,280 --> 00:16:44,560 Speaker 1: This, Matthew says, is where IBM can come in to 302 00:16:44,640 --> 00:16:49,320 Speaker 1: help take things further. Using artificial intelligence, IBM can help 303 00:16:49,360 --> 00:16:53,080 Speaker 1: Loriel create a custom AI model that helps to crunch 304 00:16:53,120 --> 00:16:57,040 Speaker 1: those numbers to be a companion to the researchers to 305 00:16:57,080 --> 00:16:58,160 Speaker 1: give them superpowers. 306 00:16:58,400 --> 00:17:00,440 Speaker 7: We don't want to replace the intuition of sent this. 307 00:17:00,520 --> 00:17:03,480 Speaker 7: We just want to make sure that this intuition is 308 00:17:03,520 --> 00:17:07,840 Speaker 7: really augmented by some calculation poor that, as Gabrielle said, 309 00:17:07,880 --> 00:17:10,400 Speaker 7: then does all ten and the poor of twenty five 310 00:17:10,600 --> 00:17:13,840 Speaker 7: solution and se probably try this one, this one, this one, 311 00:17:14,080 --> 00:17:16,760 Speaker 7: it looks like a better solution, and then ultimately that's 312 00:17:16,800 --> 00:17:19,200 Speaker 7: really the decision of the chemist to make it happen. 313 00:17:21,880 --> 00:17:24,960 Speaker 1: Well. To make a predictive AI model that can give 314 00:17:25,080 --> 00:17:29,040 Speaker 1: Loreal researchers those superpowers, you'd need that mountain of data, 315 00:17:29,600 --> 00:17:33,800 Speaker 1: years worth of laboratory testing and all Loreal's data digitized 316 00:17:33,960 --> 00:17:37,960 Speaker 1: and AI ready. You need to train artificial intelligence on 317 00:17:38,160 --> 00:17:41,000 Speaker 1: everything the company has already done in order for it 318 00:17:41,119 --> 00:17:42,639 Speaker 1: to predict what it could do. 319 00:17:43,160 --> 00:17:47,440 Speaker 6: Loriial has one hundred years a course of data, fifty 320 00:17:47,560 --> 00:17:49,400 Speaker 6: years of digitized EGGDA. 321 00:17:49,960 --> 00:17:53,600 Speaker 1: This is Miriam Ashuri, Senior director of Product Management for 322 00:17:53,720 --> 00:17:57,159 Speaker 1: IBM Watson X. Lorel has the data, and part of 323 00:17:57,200 --> 00:18:00,120 Speaker 1: IBM's job is to help put that data to work, 324 00:18:00,520 --> 00:18:04,800 Speaker 1: which involves ensuring data quality. Mariam talked about the concept 325 00:18:05,040 --> 00:18:06,480 Speaker 1: of AI ready data. 326 00:18:07,160 --> 00:18:10,000 Speaker 6: The sole purpose of this data engineering pipeline is to 327 00:18:10,119 --> 00:18:14,600 Speaker 6: clean the data and we call them AI ready data 328 00:18:14,800 --> 00:18:18,600 Speaker 6: makes them ready to be consumed by AI. So basically 329 00:18:18,640 --> 00:18:22,399 Speaker 6: looking into biases and the data to fix the distribution, 330 00:18:22,600 --> 00:18:25,960 Speaker 6: looking into guard brains that we are putting into place 331 00:18:26,000 --> 00:18:28,560 Speaker 6: in terms of removing personal information. 332 00:18:29,440 --> 00:18:32,480 Speaker 1: Marriam that explained that a custom model like the one 333 00:18:32,560 --> 00:18:35,760 Speaker 1: IBM is creating with Loril can be more efficient and 334 00:18:35,920 --> 00:18:39,200 Speaker 1: targeted than the larger general purpose AI models. 335 00:18:39,400 --> 00:18:42,760 Speaker 6: You've heard about large language models. The reason that they 336 00:18:42,800 --> 00:18:46,359 Speaker 6: call them large language model is they are exposed into 337 00:18:47,760 --> 00:18:51,800 Speaker 6: really large amount of data. So the larger the model, 338 00:18:51,840 --> 00:18:54,840 Speaker 6: the more take of all the models are, but also 339 00:18:55,280 --> 00:18:59,440 Speaker 6: the larger computed requires that translates and increase carbon footprint 340 00:18:59,520 --> 00:19:04,440 Speaker 6: and entergy consumption, that translatestan increase latency that's your response 341 00:19:04,480 --> 00:19:08,520 Speaker 6: time that translatestand increased costs. So you started seeing that 342 00:19:09,240 --> 00:19:14,320 Speaker 6: enterprises started grabbing a much smaller model customize it on 343 00:19:14,400 --> 00:19:18,680 Speaker 6: their proprietary data that's the data, their DOMAINO specific data, 344 00:19:18,800 --> 00:19:22,600 Speaker 6: or the data about their users to create something differentiated 345 00:19:23,200 --> 00:19:26,359 Speaker 6: that is applicable to a real world use case but 346 00:19:26,560 --> 00:19:30,159 Speaker 6: also delivers the performance that they needed for a fraction 347 00:19:30,280 --> 00:19:33,040 Speaker 6: of the costs, and that's why there's been a lot 348 00:19:33,080 --> 00:19:37,439 Speaker 6: of push around using custom models versus very large general 349 00:19:37,480 --> 00:19:38,560 Speaker 6: purpose models. 350 00:19:39,160 --> 00:19:42,560 Speaker 1: So how is a custom model created? Miriam says, you 351 00:19:42,640 --> 00:19:45,720 Speaker 1: start with the base model. Imagine you're buying a car. 352 00:19:46,200 --> 00:19:48,359 Speaker 1: You could get a minivan, or a sedan or a 353 00:19:48,359 --> 00:19:51,240 Speaker 1: sports car, and then you get to customize it. You 354 00:19:51,240 --> 00:19:54,520 Speaker 1: could add a sunroof, leather seats, or a rearview camera. 355 00:19:55,040 --> 00:19:57,080 Speaker 1: Turns out you could do the same thing with your 356 00:19:57,119 --> 00:19:59,960 Speaker 1: AI model. You pick a base and then you cut 357 00:20:00,000 --> 00:20:03,159 Speaker 1: customize it. You tune it on the data unique to 358 00:20:03,200 --> 00:20:04,080 Speaker 1: your organization. 359 00:20:04,520 --> 00:20:08,360 Speaker 6: We do believe that one model doesn't fit all use cases. 360 00:20:09,040 --> 00:20:13,000 Speaker 6: You want to truly have access to any model anywhere, 361 00:20:13,080 --> 00:20:17,480 Speaker 6: and by any model anywhere, I really mean any model anywhere, 362 00:20:17,560 --> 00:20:22,439 Speaker 6: open source, proprietary, low call out your machine wherever the 363 00:20:22,480 --> 00:20:25,919 Speaker 6: model is. You want to host it yourself, because then 364 00:20:26,119 --> 00:20:28,960 Speaker 6: you would be able to take advantage of the best 365 00:20:29,000 --> 00:20:31,840 Speaker 6: of the technology at any point and pick the right 366 00:20:31,920 --> 00:20:33,399 Speaker 6: model for the target use case. 367 00:20:33,760 --> 00:20:37,159 Speaker 1: So a custom model tuned on Lorel's data would be 368 00:20:37,240 --> 00:20:41,399 Speaker 1: more targeted and efficient than a general purpose model. It 369 00:20:41,400 --> 00:20:46,600 Speaker 1: would understand the researchers world and provide transparency into its workings. 370 00:20:46,960 --> 00:20:49,800 Speaker 1: That's part of the magic. And what could a custom 371 00:20:49,920 --> 00:20:53,880 Speaker 1: AI foundation model do for a company like lorel. 372 00:20:54,400 --> 00:20:59,840 Speaker 9: Accord with more contain the complexity of the formulation. 373 00:21:00,520 --> 00:21:04,600 Speaker 1: That's Gillomme lais Moline, an IBM distinguished engineer and one 374 00:21:04,640 --> 00:21:06,800 Speaker 1: of the people working on the AI model. 375 00:21:07,040 --> 00:21:12,040 Speaker 9: And to hype ower the formulator to go not only 376 00:21:12,359 --> 00:21:17,200 Speaker 9: past but also I would say, be able to include 377 00:21:17,320 --> 00:21:21,520 Speaker 9: more complexity or so in the formulation, more personalization, more 378 00:21:21,680 --> 00:21:26,639 Speaker 9: certain ability, better selected ingredient. So it's really a tool 379 00:21:26,720 --> 00:21:30,560 Speaker 9: to help them and to also help them to unniche 380 00:21:30,720 --> 00:21:31,560 Speaker 9: the creativity. 381 00:21:34,640 --> 00:21:38,360 Speaker 1: Kayomi is saying that with its custom AI model, Loreel 382 00:21:38,400 --> 00:21:42,000 Speaker 1: can improve every step of its product development pipeline, make 383 00:21:42,080 --> 00:21:45,760 Speaker 1: the process faster and more sustainable. But he's also saying 384 00:21:46,000 --> 00:21:48,880 Speaker 1: that the model could help Loriel create something that's never 385 00:21:48,920 --> 00:21:56,120 Speaker 1: been done before. What could that product be? So I'm 386 00:21:56,119 --> 00:21:57,800 Speaker 1: mourning you with that. All my questions are going to 387 00:21:57,800 --> 00:21:58,760 Speaker 1: be really dumb. 388 00:21:59,560 --> 00:22:01,240 Speaker 5: Okay, no, please by all me. 389 00:22:03,000 --> 00:22:05,159 Speaker 1: To find out what people at Loril are dreaming of. 390 00:22:05,640 --> 00:22:09,520 Speaker 1: I spoke with Trisha Iyagari, global general manager at Loriel's 391 00:22:09,520 --> 00:22:13,120 Speaker 1: Mabeline brand and they asked her about her own dreams 392 00:22:13,240 --> 00:22:16,280 Speaker 1: and how technology and science could help bring those dreams 393 00:22:16,600 --> 00:22:19,520 Speaker 1: into the world. Do you have a secret wish list 394 00:22:19,800 --> 00:22:23,000 Speaker 1: of things you think that this partnership could produce, Like, 395 00:22:23,119 --> 00:22:25,200 Speaker 1: is there a product out there that's been technically too 396 00:22:25,200 --> 00:22:28,920 Speaker 1: difficult that you think could be a worthy target? 397 00:22:29,240 --> 00:22:31,280 Speaker 5: There is one that I think could be really amazing. 398 00:22:31,480 --> 00:22:31,800 Speaker 1: What's that? 399 00:22:32,600 --> 00:22:35,479 Speaker 5: So? Shine products in general are harder to create, and 400 00:22:35,560 --> 00:22:42,159 Speaker 5: we're unable to create a shiny, long wearing eyeshadow. So 401 00:22:42,320 --> 00:22:44,920 Speaker 5: basically like a shadow that could stay on your eyelids, 402 00:22:44,960 --> 00:22:47,320 Speaker 5: that won't settle into creases, that won't move all over 403 00:22:47,359 --> 00:22:50,399 Speaker 5: your face, that has a glossy effect. It's like the 404 00:22:50,400 --> 00:22:51,000 Speaker 5: holy grail. 405 00:22:51,119 --> 00:22:53,960 Speaker 1: That's the holy grail. Yeah, yeah, you may have seen 406 00:22:53,960 --> 00:22:58,720 Speaker 1: that look in fashion shows, but that look isn't real, 407 00:22:59,320 --> 00:23:00,639 Speaker 1: not for people be and lucy. 408 00:23:00,680 --> 00:23:03,280 Speaker 5: Anyway, if you're walking down a runway, you see a 409 00:23:03,280 --> 00:23:05,399 Speaker 5: lot of makeup artists doing techniques where they put some 410 00:23:05,560 --> 00:23:08,879 Speaker 5: shadow on, they layer vasoline over it, and like slather 411 00:23:09,000 --> 00:23:12,119 Speaker 5: vassaline on somebody's eyes to create this very like glossy look. 412 00:23:12,400 --> 00:23:14,480 Speaker 5: But you know, within five minutes after they walk down 413 00:23:14,480 --> 00:23:16,280 Speaker 5: the runway, I'm sure it's all over their face or 414 00:23:16,320 --> 00:23:23,240 Speaker 5: being washed off. So the look is kind of more 415 00:23:23,280 --> 00:23:25,720 Speaker 5: of like a fashion look that we've been unable to create, 416 00:23:25,760 --> 00:23:28,720 Speaker 5: and real, real consumers can't wear it because it would 417 00:23:28,720 --> 00:23:29,600 Speaker 5: get it everywhere. 418 00:23:30,040 --> 00:23:32,320 Speaker 1: Trisha had another thing on her wish list too. 419 00:23:32,680 --> 00:23:36,360 Speaker 5: The other that we would really like is semi permanence makeup. 420 00:23:37,280 --> 00:23:42,600 Speaker 5: So we've talked a lot about really really comfortable thin 421 00:23:42,800 --> 00:23:45,320 Speaker 5: film makeup that you could wear all over your face 422 00:23:45,440 --> 00:23:47,360 Speaker 5: and that you can sleep in and then it will 423 00:23:47,440 --> 00:23:50,760 Speaker 5: last a couple of days basically, so whether it be 424 00:23:50,840 --> 00:23:53,520 Speaker 5: on your face, on your lashes, on your brows. So 425 00:23:53,600 --> 00:23:56,479 Speaker 5: anything that's like more of a semi permanent meaning lasting 426 00:23:56,560 --> 00:23:58,680 Speaker 5: for three days or more, would be amazing. 427 00:23:59,160 --> 00:24:01,399 Speaker 1: Yeah. Yeah, And you say those two things, they have been 428 00:24:01,440 --> 00:24:04,160 Speaker 1: the whole How long have they been on the wish 429 00:24:04,160 --> 00:24:05,720 Speaker 1: list of lorel Oh. 430 00:24:05,560 --> 00:24:08,120 Speaker 5: My gosh. I have been trying to develop this shiny 431 00:24:08,119 --> 00:24:12,040 Speaker 5: eyeshadow since I started. What year did I start? Like 432 00:24:12,359 --> 00:24:15,600 Speaker 5: twenty ten? And I'm sure many people had asked before me, 433 00:24:15,680 --> 00:24:19,680 Speaker 5: and we tried so many iterations of it and nobody's 434 00:24:19,680 --> 00:24:23,960 Speaker 5: been able to achieve it. 435 00:24:23,960 --> 00:24:27,280 Speaker 1: It's clear that Loreel's experts like Tricia have a lot 436 00:24:27,280 --> 00:24:33,040 Speaker 1: of ideas. I once said what I called a magic 437 00:24:33,080 --> 00:24:36,280 Speaker 1: Wand project, where I called up scientists and technologists in 438 00:24:36,320 --> 00:24:39,359 Speaker 1: as many different fields as possible and asked them what 439 00:24:39,520 --> 00:24:42,080 Speaker 1: they could create if they could just wave a magic 440 00:24:42,119 --> 00:24:46,280 Speaker 1: wand and make it real, and everyone had something they'd 441 00:24:46,280 --> 00:24:49,760 Speaker 1: want to create everyone, that's not the issue. The issue 442 00:24:49,840 --> 00:24:52,840 Speaker 1: is that there are a million different impediments to make 443 00:24:52,880 --> 00:24:56,040 Speaker 1: the ideas on the wish list reel. Lack of resources, 444 00:24:56,200 --> 00:24:59,159 Speaker 1: lack of time, some crucial bit of know how is lacking. 445 00:24:59,440 --> 00:25:01,960 Speaker 1: There's a ga between what we want and what we 446 00:25:02,000 --> 00:25:05,199 Speaker 1: can actually have, and one of the simplest ways to 447 00:25:05,200 --> 00:25:07,560 Speaker 1: think of the promise of AI is that it can 448 00:25:07,680 --> 00:25:11,480 Speaker 1: narrow that gap, not close it, of course, but do 449 00:25:11,640 --> 00:25:15,119 Speaker 1: enough that people with dreams realize there are more things 450 00:25:15,160 --> 00:25:34,639 Speaker 1: within their grasp than they could ever have imagined. Smart 451 00:25:34,640 --> 00:25:38,280 Speaker 1: Talks with IBM is produced by Matt Romano, Amy Gaines, McQuaid, 452 00:25:38,760 --> 00:25:43,360 Speaker 1: Lucy Sullivan, and Jake Harper. Were edited by Lacy Roberts. 453 00:25:43,640 --> 00:25:47,919 Speaker 1: Engineering by Nina Bird Lawrence, mastering by Sarah Brugaier, Music 454 00:25:47,960 --> 00:25:52,320 Speaker 1: by Gramoscope. Special thanks to Tatiana Lieberman and Cassidy Meyer. 455 00:25:52,880 --> 00:25:55,760 Speaker 1: Smart Talks with IBM is a production of Pushkin Industries 456 00:25:56,000 --> 00:26:00,440 Speaker 1: and Ruby's studio at iHeartMedia. To find more Pushkin podcast 457 00:26:00,880 --> 00:26:04,600 Speaker 1: listen on the iHeartRadio app, Apple Podcasts, or wherever you 458 00:26:04,640 --> 00:26:09,000 Speaker 1: get your podcasts. I'm Malcolm Laba. This is a paid 459 00:26:09,040 --> 00:26:13,880 Speaker 1: advertisement from IBM. The conversations on this podcast don't necessarily 460 00:26:13,920 --> 00:26:28,040 Speaker 1: represent IBM's positions, strategies, or opinions.