WEBVTT 1 00:00:00.080 --> 00:00:06.240 Mall with a Heart, kc AA Radio Lomalinda, where no listener is ever 2 00:00:06.480 --> 00:00:13.119 left behind. NBC News Radio, I'm Julie Ryan. The core of tropical 3 00:00:13.160 --> 00:00:18.480 Storm Hillary is nearing southern California after making landfall earlier today in Mexico. The 4 00:00:18.559 --> 00:00:22.719 National Hurricane Center says winds are now at sixty miles per hour as it barrels 5 00:00:22.719 --> 00:00:27.679 towards southern California about one hundred and fifteen miles south southeast of San Diego. 6 00:00:27.920 --> 00:00:32.880 Hillary could bring catastrophic and life threatening floods through Monday in the region. Parts 7 00:00:32.880 --> 00:00:37.359 of Nevada and Arizona will also be impacted by Hillary. It's the first tropical 8 00:00:37.399 --> 00:00:42.640 storm to hit California in more than eighty years. Southern California officials are warning 9 00:00:42.679 --> 00:00:47.719 to sustained power outages as Tropical Storm Hillary makes landfall. San Diego Gas and 10 00:00:47.799 --> 00:00:52.759 Electric chief exect Caroline Wynn says high winds could prevent bucket crews from getting to 11 00:00:52.799 --> 00:00:56.359 power lines, so I really want to make sure that our customers are prepared 12 00:00:56.439 --> 00:01:00.320 for prolonged outages, She says. Even though field crews have been strategically positioned, 13 00:01:00.439 --> 00:01:04.319 Hillary will dictate what they'll be able to accomplish the head of the Federal 14 00:01:04.319 --> 00:01:10.760 Emergency Management Agency says there are numerous possible reasons why so many are still unaccounted 15 00:01:10.799 --> 00:01:15.280 for from the Maui wildfire. Dan Criswell said on CNN State of the Union 16 00:01:15.400 --> 00:01:19.480 that other agencies such as the FBI, Department of Defense, and Department of 17 00:01:19.519 --> 00:01:22.959 Health and Human Services have been brought in to help account for the missing, 18 00:01:23.280 --> 00:01:27.000 to help identify and also reunify people that maybe in shelters, may have gone 19 00:01:27.040 --> 00:01:32.000 somewhere else, making sure that we can work together to identify who is still 20 00:01:32.040 --> 00:01:34.879 on accounted for where people are. She and President Biden are scheduled to go 21 00:01:34.920 --> 00:01:40.599 to Hawaii on Monday. Former Vice President Mike Pence says he was never made 22 00:01:40.640 --> 00:01:46.200 aware of any broad based declassification of documents by former President Trump. Pence's comments 23 00:01:46.239 --> 00:01:49.680 on ABC's This Week were consistent with what former White House Chief of Staff Mark 24 00:01:49.760 --> 00:01:56.040 Meadows reportedly told investigators for Special counsel Jack Smith. The former Vice president added 25 00:01:56.079 --> 00:02:00.000 that it's possible Trump did declassify the documents found at Marrow Lag without his or 26 00:02:00.079 --> 00:02:06.159 Meadows knowledge. You're listening to the Latest on NBC News Radio. This segment 27 00:02:06.200 --> 00:02:10.159 sponsored by the All new Rams Market, Rams Express Car Wash and the RAMS 28 00:02:10.280 --> 00:02:15.159 Ultimate one Stop Shop at twenty ninety mentone Boulevard on the corner of Crafton Hills 29 00:02:15.159 --> 00:02:20.800 Avenue, just across from Jacento Farms. RAMS is now open all new, 30 00:02:21.000 --> 00:02:24.840 well stocked, bringing fresh and sharp convenience Demento with their clean, safe chemicals, 31 00:02:24.879 --> 00:02:29.840 They'll leave your car all shiny and protected. When you buy gas inside 32 00:02:29.879 --> 00:02:32.240 the one stop shop, you'll get savings on your car wash. Make life 33 00:02:32.280 --> 00:02:37.599 easy and sowed it all with Rams. You can save with the Dinosaur and 34 00:02:37.680 --> 00:02:43.000 your top tier Sinclaire Fuel. Rams has great car wash membership dealing and you 35 00:02:43.000 --> 00:02:46.439 can save on gas when you wash your car. Satisfy your appetite when you 36 00:02:46.520 --> 00:02:52.840 drop in the one stop shop convenience store. Smell the chicken delectable mouthwatering, 37 00:02:53.199 --> 00:02:57.919 crispy, crunchy fried chicken that is. Rams is now open at twenty ninety 38 00:02:57.919 --> 00:03:02.479 mentone Boulevard and Mentone. Thank you to RAMS for their community spirit and sponsoring 39 00:03:02.680 --> 00:03:14.360 this radio station. Rams and Menton is waiting for you. Now Here's a 40 00:03:14.400 --> 00:03:21.800 new concept digital network advertising for businesses. Display your ad inside their building. 41 00:03:22.319 --> 00:03:24.439 If a picture's worth of a thousand words, your company is going to thrive 42 00:03:24.520 --> 00:03:30.719 with digital network advertising. Choose your marketing sites or jump on the DNA system 43 00:03:30.759 --> 00:03:37.680 and advertise with all participants. Your business ad or logo is rotated multiple times 44 00:03:37.680 --> 00:03:45.319 an hour inside local businesses where people will discover your company. Digital network advertising 45 00:03:45.800 --> 00:03:51.719 DNA a novel way to be seen and remembered. Digital network advertising with networks 46 00:03:51.719 --> 00:03:55.680 in Redlands and KAIPA call in the nine on nine area two two nine two 47 00:03:55.759 --> 00:04:01.479 nine three for introductory pricing. That's nine nine two two two nine two nine 48 00:04:01.599 --> 00:04:10.400 three for digital network advertising. One last time Digital network advertising nine O nine 49 00:04:10.520 --> 00:04:15.000 two two two nine two nine three. Tahebow Tea Club's original pure power Darco 50 00:04:15.079 --> 00:04:19.120 supertea helps build red corpusals in the blood, which carry oxygen to organs and 51 00:04:19.199 --> 00:04:24.720 cells. Our organs and cells need oxygen to regenerate themselves. The immune system 52 00:04:24.720 --> 00:04:28.399 needs oxygen to develop, and cancer dyes in oxygen. So the t is 53 00:04:28.439 --> 00:04:30.720 great for healthy people because it helps build the immune system, and it can 54 00:04:30.759 --> 00:04:35.360 truly be miraculous for someone fighting a potentially life threatening disease due to an infection, 55 00:04:35.480 --> 00:04:40.759 diabetes, or cancer. The tea is also organic and naturally caffeine free. 56 00:04:40.800 --> 00:04:44.480 A one pound package of tea is forty nine ninety five, which includes 57 00:04:44.519 --> 00:04:47.639 shipping two order. Please visit Tehebow Tea Club dot com. Tihebow is spelled 58 00:04:47.680 --> 00:04:53.519 tea like tom, a, h ee b like boyo. They continue with 59 00:04:53.560 --> 00:04:57.759 the word t and then the word club. The complete website is Tohebow Tea 60 00:04:57.800 --> 00:05:01.079 Club dot com or call us at eight y eight six one zero eight zero 61 00:05:01.240 --> 00:05:06.199 eight eight Monday through Saturday, nine am to five pm California time. That's 62 00:05:06.360 --> 00:05:13.639 eight one eight six one zero eight zero eight eight tohibow Tclub dot com. 63 00:05:13.720 --> 00:05:15.279 Do you love to dine out and travel, then join me the number one 64 00:05:15.319 --> 00:05:18.079 food critic, Allen Borgan, host of the Let's Sign Out TV show on 65 00:05:18.120 --> 00:05:21.759 PBS every Saturday at four pm. For the new Let's Sign Out Radio Show 66 00:05:21.839 --> 00:05:26.920 on KCAA ten fifty am, along with my cost Isabelle Bussy, we will 67 00:05:26.920 --> 00:05:30.279 be reviewing local restaurants, interviews with culinary professionals, review new cooking products, 68 00:05:30.360 --> 00:05:34.319 cookbooks, plus so much more, including giving away free gift certificates. That's 69 00:05:34.360 --> 00:05:39.720 the new Let's Sign Out Radio Show right here on CACAA ten fifty am every 70 00:05:39.759 --> 00:06:20.560 Saturday at four pm. Happy Eating, Okay, c A Economy as a 71 00:06:20.680 --> 00:06:27.079 rod. The world is teeming with innovation as new business models reinvent every industry. 72 00:06:27.120 --> 00:06:31.600 Inside Analysis is your source of information and insight about how to make the 73 00:06:31.680 --> 00:06:36.279 most of this exciting new era. Learn more at inside analysis dot com, 74 00:06:36.800 --> 00:06:48.319 Analysis dot com. And now here's your host, Eric Kavanaugh. So, 75 00:06:49.040 --> 00:06:55.079 yeah, soks, welcome to the future. Indeed, as my hero William 76 00:06:55.120 --> 00:06:59.439 Gibson says, the future is here already, It's just not evenly distributed, 77 00:07:00.000 --> 00:07:02.199 at least not yet. And that's the purpose of our show is to distribute 78 00:07:02.199 --> 00:07:05.480 the future, at least tell you about what's coming down the pike. And 79 00:07:05.519 --> 00:07:10.319 I gotta say, folks, I have never seen the future come faster than 80 00:07:10.360 --> 00:07:13.759 it has been coming in the last few months. So here on our show, 81 00:07:13.800 --> 00:07:16.079 of course, we talk all about the information economy, new jobs, 82 00:07:16.120 --> 00:07:20.199 new roles, new technologies, new business models, the ways of doing things, 83 00:07:20.519 --> 00:07:26.079 and wow, we kind of hit the inflection point, I guess, 84 00:07:26.120 --> 00:07:30.199 not five six months ago. And really you can peg it to chat GPT 85 00:07:30.759 --> 00:07:33.160 as a large language model. It's not new. Large language model has been 86 00:07:33.160 --> 00:07:36.040 around for a while, for a couple of years at least, but this 87 00:07:36.160 --> 00:07:43.959 particular version came out mature enough and big enough and wild enough to really shock 88 00:07:44.040 --> 00:07:47.079 the business world. I can tell you that people are kind of freaked out. 89 00:07:47.120 --> 00:07:50.279 Even the folks at Google are nervous. And of course they have their 90 00:07:50.279 --> 00:07:55.160 own large language model called Bard, which is different than chat GPT in a 91 00:07:55.199 --> 00:07:59.519 number of different ways. But the bottom line is that this form of artificial 92 00:07:59.519 --> 00:08:03.959 intelligence AI is absolutely shaking the foundation of how we do business. It's not 93 00:08:05.040 --> 00:08:09.839 perfect, and it really is important to understand the limitations of these things and 94 00:08:09.199 --> 00:08:13.079 understand the use cases when you use them, when not to use them. 95 00:08:13.120 --> 00:08:16.639 And there is all this talk of so called hallucinations. Well what is a 96 00:08:16.680 --> 00:08:20.399 hallucination? Well, for those who don't know yet, Chat, GPT, 97 00:08:20.720 --> 00:08:26.079 BAR, these large language models, they are basically predictive engines for words, 98 00:08:26.360 --> 00:08:31.480 and they will deliver words based upon the prompts you give them. They've been 99 00:08:31.519 --> 00:08:35.879 trained on millions and even billions of points and tons and tons mountains of data, 100 00:08:37.519 --> 00:08:41.440 and so they can reflect back in pros all kinds of different concepts. 101 00:08:41.840 --> 00:08:46.320 You can use chat, GPT or BAR to write a business plan, or 102 00:08:46.360 --> 00:08:50.679 to write marketing emails, or to write articles or blogs or short stories or 103 00:08:50.720 --> 00:08:54.360 all kinds of stuff. But it is generative in nature, which means it 104 00:08:54.399 --> 00:08:58.120 creates new stuff. And basically the way it works is it fuses together vectors 105 00:08:58.159 --> 00:09:03.639 of information based upon its model, its training, and also your prompts what 106 00:09:03.759 --> 00:09:07.639 you give it. Well. Very quickly in this new world, we started 107 00:09:07.679 --> 00:09:13.320 seeing people talk about embeddings and trying to figure out ways to get these engines 108 00:09:13.320 --> 00:09:20.320 to be more accurate in specific domains of content, so in banking, or 109 00:09:20.320 --> 00:09:26.320 in healthcare, or in retail or marketing or whatever. Embeddings refer to actual, 110 00:09:26.519 --> 00:09:30.600 real corporate data that you either embed into the model itself, which is 111 00:09:30.679 --> 00:09:35.360 kind of complicated, or most people are talking about embedding them in a vector 112 00:09:35.440 --> 00:09:39.320 database, which is then adjacent to the large language model. And what happens 113 00:09:39.399 --> 00:09:43.080 is when you give your prompts, the engine will reach into both areas, 114 00:09:43.080 --> 00:09:48.559 will reach into your vector database to see your trusted data, and then it'll 115 00:09:48.600 --> 00:09:52.559 reach over into its corpus of language and understanding to kind of fuse together some 116 00:09:52.600 --> 00:09:58.279 content that is reflective then of your corporate data. This is called training the 117 00:09:58.399 --> 00:10:01.960 models, and you do have to train these things for them to be very 118 00:10:01.960 --> 00:10:03.679 good at what you want them to be good at. But they are very 119 00:10:03.720 --> 00:10:07.960 good right out of the box and all kinds of stuff. But what's fun 120 00:10:07.000 --> 00:10:11.519 from my perspective is that when we start talking about pointing these algorithms at your 121 00:10:11.519 --> 00:10:16.039 corporate data will, you have to be careful about that because a lot of 122 00:10:16.039 --> 00:10:20.080 companies really don't know everything that they have in terms of data, and historically 123 00:10:20.120 --> 00:10:24.679 it's been very difficult to figure that out because what would you do. Run 124 00:10:24.759 --> 00:10:26.360 some query and you know, just run a report. Okay, here's our 125 00:10:26.440 --> 00:10:31.799 data. It's a very difficult thing to break down, to parse and to 126 00:10:31.879 --> 00:10:37.039 demonstrate, and frankly, it's just hard to crawl across all those objects, 127 00:10:37.080 --> 00:10:39.919 all those documents in your share point instance, or wherever it else you have 128 00:10:39.960 --> 00:10:46.240 information stored. So it's important to remember that for basically every organization of any 129 00:10:46.279 --> 00:10:50.279 significant size or age, you're going to have a ton of data in the 130 00:10:50.320 --> 00:10:56.159 form of spreadsheets, in the form of PDFs and word documents and all kinds 131 00:10:56.159 --> 00:10:58.919 of different things. So you're gonna have all this information that can be very 132 00:11:00.080 --> 00:11:03.039 useful, but there may be sensitive innovation, then there may be wrong information 133 00:11:03.120 --> 00:11:07.600 too. So this training is really important stuff. And we're going to talk 134 00:11:07.600 --> 00:11:11.840 to a company today that can help you figure out if you're at the starting 135 00:11:11.840 --> 00:11:13.799 line now this is not all that they do. But when I was talking 136 00:11:13.840 --> 00:11:18.799 to them about their technology and their services, I thought to myself, Hey, 137 00:11:18.840 --> 00:11:22.919 you guys have a really powerful story to tell in this whole large language 138 00:11:22.919 --> 00:11:26.960 model narrative is unfolding right now. And so to that and end, we've 139 00:11:26.000 --> 00:11:31.039 got Lewis wyn Jones on the line from Think Data Works, and of course 140 00:11:31.080 --> 00:11:33.360 we have our buddy Eve Mulkers into Wings. He'll be joining us in a 141 00:11:33.360 --> 00:11:35.639 second. But Lewis, tell us a bit about yourself and Think Data Works 142 00:11:35.840 --> 00:11:41.480 and how you are able to help companies determine if they're ready for large language 143 00:11:41.480 --> 00:11:43.480 models and then kind of go down that road. Thanks so much, Eric, 144 00:11:43.519 --> 00:11:46.720 and and thanks for the sort of primmer. I think that you hit 145 00:11:46.759 --> 00:11:50.320 the nail on the head. There's so much excitement right now, and a 146 00:11:50.320 --> 00:11:54.120 lot of that has been because of what GPT has done in the past few 147 00:11:54.159 --> 00:12:00.200 months. My company is, as you sort of suggested, is more in 148 00:12:00.200 --> 00:12:03.840 the sort of data management space, and that's definitely a less sexy place than 149 00:12:05.480 --> 00:12:07.759 what you're sort of seeing when you when you plug a prompt into chat GBT 150 00:12:07.840 --> 00:12:11.080 and see what comes out. But the reality is, if you scratchually hard 151 00:12:11.080 --> 00:12:16.240 at the surface, you start to realize that you do need to have well 152 00:12:16.279 --> 00:12:20.720 structured data in order to get well structured responses coming out the other end. 153 00:12:20.759 --> 00:12:24.240 I think that what we see with chat GBT is the generalization of large language 154 00:12:24.240 --> 00:12:28.240 model and generative AI tech, which is why it's so exciting, right The 155 00:12:28.279 --> 00:12:31.639 fact that there's something out there that I can go, when I can plug 156 00:12:31.639 --> 00:12:35.600 a question into and get a pretty good response back, that's really exciting. 157 00:12:35.679 --> 00:12:37.759 But the reason it's exciting for me is because I haven't been working in AI 158 00:12:37.919 --> 00:12:41.559 for the past, you know, ten years, trying to build these models, 159 00:12:41.879 --> 00:12:45.120 and the fact that my grandmother can go and do this is really sort 160 00:12:45.159 --> 00:12:48.679 of the exciting part of this. I think that if you want to start 161 00:12:48.679 --> 00:12:54.720 to have that tech embedded into the internal systems of most companies, there's a 162 00:12:54.720 --> 00:12:56.279 lot of work that needs to be done to get you to that place. 163 00:12:56.639 --> 00:13:01.600 And the things that get you to that place our management of your data assets. 164 00:13:01.639 --> 00:13:03.360 I mean, you kind of mentioned it. Most people have data all 165 00:13:03.360 --> 00:13:07.480 over the place. They don't They've got a better inventory of their office furniture 166 00:13:07.480 --> 00:13:11.200 than they do of their data sets. And there's a lot of things that 167 00:13:11.200 --> 00:13:18.639 are required to make good on the promise of AI and actually start to generate 168 00:13:18.720 --> 00:13:22.279 value at the back end of your business. And so that's really where where 169 00:13:22.320 --> 00:13:26.519 our company sits on the VP of product here, and we've been building a 170 00:13:26.840 --> 00:13:31.960 data platform for the past roughly ten years. We were born out of the 171 00:13:31.960 --> 00:13:37.639 open data space actually, which you know, for those people who are listening 172 00:13:37.639 --> 00:13:41.080 who don't remember what that was all about, but it was governments that were 173 00:13:41.120 --> 00:13:43.879 sort of giving data into the public domain and saying this could be valuable. 174 00:13:45.120 --> 00:13:46.720 The problem with it was that it was all over the place, it was 175 00:13:46.759 --> 00:13:52.120 a mess, there was no standardization, and you know is file formats from 176 00:13:52.159 --> 00:13:54.279 the nineteen seventies. So what we did is we would sort of sit on 177 00:13:54.320 --> 00:13:58.080 top of all this data, standardize and make it accessible to people. And 178 00:13:58.120 --> 00:14:03.159 that's when we really realized that the thing that was happening with open data globally 179 00:14:03.519 --> 00:14:07.200 was also happening within the enterprise itself. So the data is all over the 180 00:14:07.200 --> 00:14:09.320 place, it's a mess, it's you know, you don't know how to 181 00:14:09.360 --> 00:14:11.840 get the value out of it. And so the tool that we built to 182 00:14:11.960 --> 00:14:18.360 sort of standardize all this government data, we sort of turned towards business data 183 00:14:18.480 --> 00:14:22.440 and all of the tool and functional functionality that go along with that, which 184 00:14:22.480 --> 00:14:26.480 really does support it's sort of that data refinery in order to get the AI 185 00:14:26.559 --> 00:14:30.559 analysis coming out the other end of it. So it's I would say that 186 00:14:30.559 --> 00:14:35.200 it's a prerequisite to start to get the benefit of large language models or any 187 00:14:35.200 --> 00:14:39.240 other kind of a modeling that you require. Yeah, and I think your 188 00:14:39.279 --> 00:14:43.519 background and kind of where you all came from is perfectly suited for this challenge 189 00:14:43.600 --> 00:14:50.080 because you were out there trying to reconcile, classifying, organize this vast amount 190 00:14:50.120 --> 00:14:54.799 of data, unstructured data, all kind of different formats, very unwieldy. 191 00:14:54.960 --> 00:14:56.919 So that was your you know, kind of how you cut your teeth on 192 00:14:58.039 --> 00:15:03.159 that environment. And now you're able to do the same thing and help organizations 193 00:15:03.360 --> 00:15:07.960 again understand what they have, which for those who haven't tried, I promise, 194 00:15:07.120 --> 00:15:13.279 is a very difficult and challenging task to undertake, but it's very important 195 00:15:13.320 --> 00:15:15.840 for lots of reasons. I mean, I kind of think that this is 196 00:15:15.919 --> 00:15:18.799 the sort of come to Jesus moment, if you will, for enterprises and 197 00:15:18.840 --> 00:15:20.840 all the data they've been sweeping out of the rug for the last thirty years. 198 00:15:20.919 --> 00:15:24.919 It's like we've known that we had to do something about it. But 199 00:15:24.000 --> 00:15:28.559 now you really really do have to do something about it, and you guys 200 00:15:28.559 --> 00:15:31.960 can help with that process, right, Yeah, absolutely. I mean I 201 00:15:31.960 --> 00:15:39.039 think that it really depends when you're talking to because in certain boardrooms people are 202 00:15:39.080 --> 00:15:41.120 saying, well, we want streamlined analytics, we want to make data driven 203 00:15:41.120 --> 00:15:45.879 decisions, we want all this up, and they don't understand that in order 204 00:15:45.879 --> 00:15:48.480 to get there, you need to have your house in order. And it's 205 00:15:48.480 --> 00:15:52.759 a step that you can't actually skip that sort of data operation step, the 206 00:15:52.799 --> 00:15:58.120 metadata management step, the good indexing, the good tagging, and the secure 207 00:15:58.159 --> 00:16:00.799 distribution throughout the enterprise. It's not a nice to have, it is a 208 00:16:00.840 --> 00:16:06.039 prerequisite. And what we're seeing across the enterprise is that people who have skipped 209 00:16:06.039 --> 00:16:08.960 that step and maybe set up very good data and analytics divisions, who are 210 00:16:10.000 --> 00:16:14.320 trying to unleash the power of AI on the enterprise, what they're doing is 211 00:16:14.320 --> 00:16:18.519 they're generally sort of building one or two products a year. This is not 212 00:16:18.799 --> 00:16:23.399 fully automated, and this is not actually a system that is going to scale. 213 00:16:23.440 --> 00:16:26.919 Well, it's a system that under the voot if you scratch it, 214 00:16:26.000 --> 00:16:30.240 like, if you get underneath it, it's a lot of ad hoc work, 215 00:16:30.399 --> 00:16:33.120 and what everyone wants to do to really maximize the benefit of this stuff 216 00:16:33.320 --> 00:16:37.960 is to automate those processes. And the way to automate them is to have 217 00:16:37.360 --> 00:16:42.519 your data environment set up in a way that will facilitate sort of the pulling 218 00:16:42.559 --> 00:16:47.120 in of data, the pushing back, the sharing, the sort of collaboration 219 00:16:47.559 --> 00:16:49.519 and the management of that data. Otherwise, what you're going to be doing 220 00:16:49.679 --> 00:16:55.200 is basically manually curating data sets and pushing them into Tableau. And that is 221 00:16:55.279 --> 00:17:00.639 not AI as far as I'm considering. But that's here laughing, I'll bring 222 00:17:00.639 --> 00:17:03.840 you into comment on that. I mean, I think lewis to sit the 223 00:17:03.919 --> 00:17:07.640 nail on the head there. You really want to be strategic about this process, 224 00:17:07.680 --> 00:17:11.839 and it could be extremely valuable going through the inventory to understand where you've 225 00:17:11.880 --> 00:17:17.599 come up from. What's happening, right, go aheady, Yeah, what 226 00:17:17.799 --> 00:17:21.839 I see. I think back of my years of doing software development, and 227 00:17:22.160 --> 00:17:26.559 this is what we see now in data management. But during it, back 228 00:17:26.599 --> 00:17:29.920 in the days, you wrote a piece of codes as such, and then 229 00:17:30.160 --> 00:17:33.839 some libraries came along you started using those libraries to be more proficient and more 230 00:17:33.880 --> 00:17:38.960 productive in creating your applications and your software and this is now happening with all 231 00:17:40.000 --> 00:17:42.720 the libraries. For example, if you take pythons that are available to access 232 00:17:42.759 --> 00:17:48.039 APIs, to access the large language models. But still, like you say, 233 00:17:48.319 --> 00:17:52.440 Lewis, you take these things, you take one product, you build 234 00:17:52.480 --> 00:17:56.319 one product. But people don't think about the architecture to be capable of really 235 00:17:56.359 --> 00:18:02.279 scaling the products. And that's something where I see that is sometimes a misconnect 236 00:18:02.279 --> 00:18:07.960 between more technical people, especially in the IT department, compared to business people. 237 00:18:07.960 --> 00:18:11.440 Why they see the magic shining object and they think, well, it's 238 00:18:11.480 --> 00:18:15.160 like you say, it's upload band. We've got our dashboard. Yeah, 239 00:18:15.240 --> 00:18:18.759 next time something changes, how faster we're going to Are you going to change 240 00:18:18.759 --> 00:18:22.319 the dashboard? How much effort will it take? You? Do you understand 241 00:18:22.359 --> 00:18:26.359 where your data comes from? And that is what when we talk about architecture, 242 00:18:26.400 --> 00:18:30.279 that you need to put these really solid foundations in place to be able 243 00:18:30.359 --> 00:18:34.720 to manage it. And that's very oftentimes for god and so I'm quite happy 244 00:18:34.799 --> 00:18:40.119 that I've seen it through my career as well. These are the things we've 245 00:18:40.160 --> 00:18:44.000 been building over the time, and now all of a sudden it gets attention, 246 00:18:44.079 --> 00:18:47.880 it gets a name. Like for example, if you talk about observability, 247 00:18:48.079 --> 00:18:49.599 Well, we need to know where the data comes from. Yes, 248 00:18:49.720 --> 00:18:56.240 we've been interrogating the method that are over time into our systems to understand how 249 00:18:56.319 --> 00:19:00.480 it looks like to Atto made it to deploy it in a continue look at 250 00:19:00.599 --> 00:19:07.680 software development as well, where now we have continuous integration and continuous development where 251 00:19:07.000 --> 00:19:11.720 this was not possible twenty thirty years ago, and I see that happening now 252 00:19:11.759 --> 00:19:17.720 with especially with the large language models and all the AI tools that we are 253 00:19:17.759 --> 00:19:22.680 definitely looking at that. So once we've get this really into a continuous development 254 00:19:22.720 --> 00:19:27.400 mode where off the hook, I think it's going to be even more amazing 255 00:19:27.480 --> 00:19:33.920 and the speed is going to increase even more than we over the lost few 256 00:19:33.960 --> 00:19:37.039 months with the large language models. Yeah, and you know, I'll bring 257 00:19:37.119 --> 00:19:41.319 Lewis back into comment on something in particular. You mentioned architecture, and Lewis 258 00:19:41.359 --> 00:19:45.960 that's really important, right. What Eaves is talking about here is building an 259 00:19:47.000 --> 00:19:51.599 information foundation that will allow you to then build apps more quickly, as opposed 260 00:19:51.640 --> 00:19:56.559 to just building one off bespoke apps which may solve a business problem for now, 261 00:19:56.039 --> 00:20:02.720 but will not scale very well and won't be optimum in terms of allowing 262 00:20:02.759 --> 00:20:06.000 you to do other apps that use the same data. For example, we're 263 00:20:06.000 --> 00:20:08.559 talking about data products here, right, and so if you build the foundation, 264 00:20:08.680 --> 00:20:12.599 well then the individual products can be spun up pretty quickly and efficiently, 265 00:20:12.960 --> 00:20:18.240 whereas if you don't, if it's all the spoke and spaghetti wire is going 266 00:20:18.279 --> 00:20:21.279 in between them. Again, it may solve a particular problem, but it's 267 00:20:21.279 --> 00:20:25.039 not going to last for too long and it will amount to technical debt before 268 00:20:25.079 --> 00:20:27.000 long. Right, yeah, And I mean the thing is is this isn't 269 00:20:27.079 --> 00:20:30.839 theory. This is we've seen this happen before. You know, ten years 270 00:20:30.839 --> 00:20:33.039 ago, fifteen years ago, people said, oh, if you dump all 271 00:20:33.039 --> 00:20:36.559 your data into a data lake, magic stuff is going to come out of 272 00:20:36.599 --> 00:20:38.279 it. Well what happened there. We got a bunch of really really messy 273 00:20:38.359 --> 00:20:42.720 data environments and people still struggling with this problem. And then said people said, 274 00:20:42.720 --> 00:20:45.519 okay, well the cloud will be the solution to all this problems. 275 00:20:45.799 --> 00:20:49.079 Well, okay, people started to centralize to the cloud, and now they're 276 00:20:49.079 --> 00:20:53.359 going back to having a more sort of on prem and hybrid environment. People 277 00:20:53.359 --> 00:20:57.200 are looking for the magic bullet, the thing that's going to sit on top 278 00:20:57.279 --> 00:21:03.000 of an unstructured environment and pulled insights. That's not what's going to happen, 279 00:21:03.039 --> 00:21:07.440 and AI can never get there because it needs to be trained in a well 280 00:21:07.599 --> 00:21:12.680 architected environment, in a good sort of well balanced place where data flows evenly. 281 00:21:14.400 --> 00:21:18.119 And that I'm not saying that the previous technologies that have come out from 282 00:21:18.119 --> 00:21:21.200 clouds the data lakes have been mistakes. They have and they've been extremely useful 283 00:21:21.240 --> 00:21:23.440 and they've taken us sort of forward. But I think now what we're looking 284 00:21:23.440 --> 00:21:27.720 at is you don't want to have to centralize everything into one place for a 285 00:21:27.759 --> 00:21:33.599 bunch of reasons, but you also can't have everything existing in random regions everywhere. 286 00:21:33.640 --> 00:21:37.519 So really, what we're seeing, and this is just from an architecture 287 00:21:37.519 --> 00:21:41.559 perspective, is a new sort of reliance and focus on data virtualization, so 288 00:21:41.599 --> 00:21:48.640 you can actually connect to data wherever it is and you are centralizing access without 289 00:21:48.640 --> 00:21:52.960 actually centralizing assets themselves. And that's the sort of thing that you can do 290 00:21:52.079 --> 00:22:00.279 to really start to create this central observation platform for the data, and that's 291 00:22:00.279 --> 00:22:03.599 going to be the key to unlock management and distribution within the enterprise. And 292 00:22:03.599 --> 00:22:07.480 then of course, yes, all the downstream benefits of connecting it to your 293 00:22:07.519 --> 00:22:11.039 analytical tools and if you need to your models that you're going to be building 294 00:22:11.039 --> 00:22:14.720 out the back end, but without it, you're just sort of trying to 295 00:22:14.720 --> 00:22:18.279 build on pretty murky, swampy ground. Yeah no, that's a really really 296 00:22:18.319 --> 00:22:25.279 good point where you're democratizing access through a governed zone, if you will, 297 00:22:25.519 --> 00:22:29.680 Because we've talked about this for years now on this show and other shows too, 298 00:22:30.039 --> 00:22:33.920 that something like data governance cannot happen if you have eighteen to two hundred 299 00:22:34.079 --> 00:22:38.599 entry points where you're only managing five to seven of them, well then you 300 00:22:38.640 --> 00:22:44.839 don't have governance. Now. Historically, getting to that point was extremely difficult 301 00:22:44.880 --> 00:22:48.960 because the technology it really wasn't up to snuff. If you try to go 302 00:22:48.079 --> 00:22:52.759 through that one choke point, things just wouldn't work fast enough. But I 303 00:22:52.799 --> 00:22:56.799 think we have seen the capability to expand out the scale out dynamically and be 304 00:22:56.839 --> 00:23:00.000 able to handle those things real quick. I think Lewis, we are at 305 00:23:00.000 --> 00:23:06.039 the beginning of another era because we now can pull that off efficiently and effectively, 306 00:23:06.119 --> 00:23:07.799 Whereas you know, five ten years ago, it wasn't really that possible. 307 00:23:07.839 --> 00:23:11.359 What do you think, Yeah, No, I think you're right. 308 00:23:11.839 --> 00:23:17.079 Governance matured faster than the technology, and so the principles of governance were saying, 309 00:23:17.119 --> 00:23:18.920 Okay, you need to do this, this and this, and so 310 00:23:18.039 --> 00:23:23.559 people were writing extremely robust data privacy programs, but they couldn't actually implement them 311 00:23:23.559 --> 00:23:26.640 with the tech that existed. Now that tech does exist, and I think 312 00:23:26.680 --> 00:23:33.359 that the kind of egalitarian approach to distribution within the enterprise and good governance are 313 00:23:33.400 --> 00:23:37.480 possible at the same time, So you can get increased access and increase governance 314 00:23:37.880 --> 00:23:41.000 at the same time, which before it was I think one or the other. 315 00:23:41.640 --> 00:23:44.519 Yeah, that's a really good point too, because you did have to 316 00:23:44.559 --> 00:23:47.839 make a choice before. Now you don't have to make that choice. There 317 00:23:47.880 --> 00:23:51.480 are other choices you'll need to make, of course, but you can pull 318 00:23:51.559 --> 00:23:55.599 this off. And if governance is important to you, and it's important to 319 00:23:55.640 --> 00:24:00.279 most organizations, not just for security but also for quality, for quality product 320 00:24:00.279 --> 00:24:03.759 out the other side, you know you want to have some governance, so 321 00:24:03.799 --> 00:24:06.640 you know what's going into this, you know how to manage it. And 322 00:24:06.759 --> 00:24:07.880 we're getting there, folks, we're getting there. Well, don't touch that, 323 00:24:07.920 --> 00:24:11.200 do I'll be right back. We're talking through a couple of experts on 324 00:24:11.279 --> 00:24:18.160 inside. An alfis right back to me. What if you could own a 325 00:24:18.240 --> 00:24:22.160 piece of the future, What if you could build your next castle not on 326 00:24:22.279 --> 00:24:27.039 sand, but on the bedrock of a modern blockchain ecosystem. The first Internet 327 00:24:27.079 --> 00:24:33.160 gold rush made millionaires, The second wave is minting billionaires, but the third 328 00:24:33.160 --> 00:24:37.920 wave is just gathering now and anyone can get in on the action. Hop 329 00:24:37.960 --> 00:24:41.200 online to crowdpointtech dot com to learn how you can secure a foothold in the 330 00:24:41.240 --> 00:24:47.119 blockchain revolution. Whatever your passion, wherever you want to go in life, 331 00:24:47.400 --> 00:24:52.759 there's an opportunity awaiting you right now. Go to crowdpointtech dot com to learn 332 00:24:52.799 --> 00:24:57.119 how the blockchain will fuel the next generation of innovation in this globally connected world. 333 00:24:57.640 --> 00:25:03.839 That's crowd pointtech dot Your trusted agent in an untrusted world. What's the 334 00:25:03.920 --> 00:25:08.160 longest running radio show in the world focused on data? DM Radio? Want 335 00:25:08.160 --> 00:25:12.519 to be a guest sometime? Send an email to info at DM radio dot 336 00:25:12.559 --> 00:25:18.680 biz. That's Info at DM radio dot biz. Can your IRA stand up 337 00:25:18.720 --> 00:25:22.519 to the next financial crisis that our top economists are saying is at our doorsteps. 338 00:25:22.519 --> 00:25:27.160 By allocating a percentage of your IRA into physical gold and silver with a 339 00:25:27.319 --> 00:25:33.480 tax free rollover, you can diversify in safeguard your holdings from turbulent markets and 340 00:25:33.559 --> 00:25:38.079 economic downturns by putting your IRA back on the gold standard. Find out how 341 00:25:38.079 --> 00:25:44.039 to safeguard your assets with a tax free rollover with a Genesis Gold IRA, 342 00:25:44.480 --> 00:25:48.119 the only IRA that can hold physical precious metals. Call now for your free 343 00:25:48.119 --> 00:25:53.519 gold and silver report. Protect your IRA today with one simple phone call and 344 00:25:53.640 --> 00:25:59.599 learn how to qualify for up to ten thousand dollars in free silver called Genesis 345 00:25:59.640 --> 00:26:04.720 Gold. Empowering faith driven stewardship. Eight hundred six four four eight six one 346 00:26:04.839 --> 00:26:11.160 one eight hundred six four four eight six one one. Eight hundred six four 347 00:26:11.319 --> 00:26:18.480 four eight six one one. That's eight hundred six four four eighty six eleven. 348 00:26:19.920 --> 00:26:23.960 When a player's sudden cardiac event brought a national football game to a halt, 349 00:26:25.119 --> 00:26:29.359 it's shown a spotlight on the importance of CPR readiness. Now, with 350 00:26:29.440 --> 00:26:33.519 youth sports in full swing, the American Heart Association is rallying parents and coaches 351 00:26:33.559 --> 00:26:37.880 to be ready in an emergency. To be ready, learn hands only CPR. 352 00:26:38.160 --> 00:26:41.960 It's a skill anyone can learn in minutes. Just visit Heart dot org 353 00:26:42.000 --> 00:26:48.359 slash hands only CPR. Hands only CPR is nationally supported by an Elephant's Health 354 00:26:48.400 --> 00:26:52.759 Foundation. Each year, three hundred and fifty thousand Americans died from a cardiac 355 00:26:52.880 --> 00:26:57.359 arrest. When seconds matter most, CPR can be the difference in whether a 356 00:26:57.400 --> 00:27:03.920 friend or family member survives. That's why the American Heart Association is challenging every 357 00:27:03.000 --> 00:27:07.759 household to elect at least one person to learn CPR. If you have ninety 358 00:27:07.759 --> 00:27:12.359 seconds, you can be your family CPR hero. Just watch the American Heart 359 00:27:12.359 --> 00:27:18.880 Association's hands only CPR video at hart dot org and become a hero. Do 360 00:27:18.920 --> 00:27:22.319 you need to get your hands on some extra money right now? Maybe twenty 361 00:27:22.359 --> 00:27:26.039 five thousand or more. If you're a homeowner, Now is a perfect time 362 00:27:26.079 --> 00:27:30.319 to get cash out while homes in many neighborhoods like yours have gone up in 363 00:27:30.440 --> 00:27:33.880 value. You can use the money for anything it's yours. You can buy 364 00:27:33.880 --> 00:27:38.119 an investment property, pay off higher interest at or make home improvements. If 365 00:27:38.160 --> 00:27:42.319 you need twenty five thousand, fifty thousand or more, Now is the time 366 00:27:42.519 --> 00:27:47.119 home values are up and so is your equity. We offer you a way 367 00:27:47.160 --> 00:27:49.680 to use it. No need to use your savings. Call New American Funding 368 00:27:49.720 --> 00:27:53.200 now and see how much cash out you can get. Call eight hundred seven 369 00:27:53.279 --> 00:27:59.920 one oh three seven three nine eight hundred seven one OZHO three seven three nine 370 00:28:00.079 --> 00:28:03.799 eight hundred seven one h three seven three nine. That's eight hundred seven one 371 00:28:03.920 --> 00:28:08.720 h thirty seven thirty nine n MLS sixty six h six wxs dot org. 372 00:28:08.799 --> 00:28:12.440 This is not an offer or commitment to end subject to borrow or improperty qualifications. 373 00:28:12.519 --> 00:28:18.759 Not all borrowers will qualify. Terms and conditions apply equal housing opportunity. 374 00:28:19.839 --> 00:28:30.400 Welcome back to Inside Analysis. Here's your host Eric Tavanaugh show. All right, 375 00:28:32.079 --> 00:28:34.400 welcome back once again to Inside Analysis. Take us from the future. 376 00:28:34.400 --> 00:28:38.720 Indeed, your host Eric Kavanaugh here with Eve Mulkers and Lewis Wynn Jones. 377 00:28:38.799 --> 00:28:42.319 Who are You're talking about the power of AI? Are you ready for it 378 00:28:42.359 --> 00:28:47.799 in your organization? I've been joking that my router may not be ready for 379 00:28:47.839 --> 00:28:49.799 the power of AI because it keeps crapping out on me. That's not very 380 00:28:49.839 --> 00:28:53.440 good. I have to call Armstrong Cable after this show and to read them 381 00:28:53.480 --> 00:28:57.480 the right act. In the meantime, let's get back to our guests, 382 00:28:57.480 --> 00:29:03.200 and we were just talking about how the data management side really is so crucial, 383 00:29:03.240 --> 00:29:07.279 and I'll throw this over to you first. Lewis, you know data 384 00:29:07.319 --> 00:29:11.559 management has always been important data quality, understanding your process as your workflows, 385 00:29:11.720 --> 00:29:15.440 all that stuff, what you have where it came from lineage, etc. 386 00:29:15.160 --> 00:29:19.039 And it's even more important now in the age of AI because these engines are 387 00:29:19.119 --> 00:29:25.200 very powerful and they will take whatever you give them and create something from that. 388 00:29:25.640 --> 00:29:27.440 And if what you're giving them is bad data, you're going to get 389 00:29:27.519 --> 00:29:32.839 very bad results. Right, go ahead, Yeah, absolutely, And I 390 00:29:32.839 --> 00:29:37.160 mean I think that talking about the results. The thing that's kind of happening 391 00:29:37.240 --> 00:29:41.160 right now is that the people who are in charge of data management historically have 392 00:29:41.319 --> 00:29:45.759 been kind of the IT side of the organization, whereas the people who are 393 00:29:45.759 --> 00:29:48.279 in charge of the results, you know, the data products and those things 394 00:29:48.319 --> 00:29:52.279 were more business. And what's happening over the past ten years, fifteen years, 395 00:29:52.319 --> 00:29:53.880 it's those two sides of the business have grown further and further apart. 396 00:29:55.799 --> 00:29:59.880 What we need to do, both like just sort of culturally but also through 397 00:30:00.160 --> 00:30:03.480 infrastructure, is to bring those two sides of the business back together again. 398 00:30:03.559 --> 00:30:07.960 Right, You need to have data management intimately tied to data productization and building 399 00:30:07.960 --> 00:30:12.079 AI and things like that, because alternatively, what you have if you don't 400 00:30:12.160 --> 00:30:15.920 do that is the business side of the organization. Talking about AI and all 401 00:30:15.920 --> 00:30:21.759 of its potential to transform business and the IT side of the organization stuck in 402 00:30:22.200 --> 00:30:23.960 sort of data governance land, and you need to start bringing those two things 403 00:30:25.000 --> 00:30:30.119 together to sort of create a cohesive umbrella of data and analytics together, which 404 00:30:30.200 --> 00:30:36.359 are governed by single tools that are applicable to business users and IT users in 405 00:30:36.400 --> 00:30:38.880 equal measure. And that's I think that that's something that patchypt has actually done 406 00:30:38.880 --> 00:30:41.960 really really well. Is made this something that can be used by anyone. 407 00:30:44.559 --> 00:30:45.839 Yeah, and you know, let me just kind of chime in here and 408 00:30:45.880 --> 00:30:51.559 say that, you know, there are lots of different values that you can 409 00:30:51.599 --> 00:30:56.480 get from data virtualization, one of which is just knowing which data is most 410 00:30:56.559 --> 00:31:00.039 important. I'll throw it over to Lewis first, but if you have a 411 00:31:00.039 --> 00:31:06.160 good data virtualization solution, you're going to be seeing in very clear fashion what 412 00:31:06.279 --> 00:31:08.720 data is being used, what data is important. That can help you that 413 00:31:08.920 --> 00:31:12.400 architect the rest of your data foundation. Right, Oh, absolutely, I 414 00:31:12.440 --> 00:31:17.799 mean a data virtualization is one of those things that I think is close to 415 00:31:17.839 --> 00:31:21.119 that silver bullet that people have been looking for, because what it allows you 416 00:31:21.160 --> 00:31:25.920 to do is instead of getting tied up into this etl land where you're constantly 417 00:31:25.920 --> 00:31:29.720 having to move things from here and there you're just projecting it from where it 418 00:31:29.759 --> 00:31:36.440 lives into a centralized place and creating that sort of possibility for someone who maybe 419 00:31:36.440 --> 00:31:40.319 doesn't have access to your sequel server to go see the data deps in it 420 00:31:40.400 --> 00:31:42.960 without it moving, without it breaking in of your governance rules. And that's 421 00:31:44.000 --> 00:31:48.440 why virtualization is a tool within a sort of larger data management framework that's really 422 00:31:48.480 --> 00:31:52.960 really powerful to just provide that democratic access piece. Now, of course you 423 00:31:53.039 --> 00:31:57.000 want some layers of control on top of that about how you actually distribute the 424 00:31:57.039 --> 00:32:02.400 data, but just having the ability to project or create a lens through to 425 00:32:02.440 --> 00:32:07.079 the data from where it resides into another location that is more user accessible is 426 00:32:07.160 --> 00:32:10.279 incredibly powerful technology. And then the next piece of that, of course, 427 00:32:10.559 --> 00:32:14.920 is once you've got access to that that piece of data, that virtualized data, 428 00:32:15.000 --> 00:32:17.559 being able to then connect it to your damn stream analytics from that same 429 00:32:17.640 --> 00:32:22.680 centralized place. Right, Because if you can't finish off that storyline, then 430 00:32:22.720 --> 00:32:24.519 you've just sort of checked the book out from the library, but you can't 431 00:32:24.559 --> 00:32:28.400 open it, right, it's not useful for you. So you need to 432 00:32:28.440 --> 00:32:31.640 have that end to end integrated set up. And Eve earlier was talking about 433 00:32:31.680 --> 00:32:35.759 the architecture needs to be right there. And the problem is if you try 434 00:32:35.759 --> 00:32:38.799 to do this with seven different discrete pieces of functionality, they all need to 435 00:32:38.799 --> 00:32:42.440 play nice with each other, right, And what our approach has always been 436 00:32:42.839 --> 00:32:45.279 is that you need to have to be able to do all of these things 437 00:32:45.599 --> 00:32:49.200 from a single system that can virtualize the data, make it discoverable, and 438 00:32:49.240 --> 00:32:52.480 then push it down where it needs to go. If you try to break 439 00:32:52.480 --> 00:32:53.960 that apart too far, it's going to get You're gonna get in the weeds 440 00:32:54.000 --> 00:32:59.799 again. Yeah. Well, and there is this question of scale. I 441 00:33:00.039 --> 00:33:02.359 think eve that's what's really shaking a lot of businesses right now, is that 442 00:33:02.400 --> 00:33:07.759 the scale of data that's available that needs to be processed to get some signal 443 00:33:08.319 --> 00:33:13.319 for AI is so much larger than it was five years ago, ten years 444 00:33:13.319 --> 00:33:15.160 ago, and all systems are just going to break. They're just not going 445 00:33:15.200 --> 00:33:20.720 to be able to suffice. And so we're in this sort of transformative period 446 00:33:20.759 --> 00:33:24.279 of time now where organizations have to be very strategic and figure out what is 447 00:33:24.319 --> 00:33:29.599 my stepping stone to the next generation of technologies. And I think that's where 448 00:33:29.640 --> 00:33:36.559 data virtualization is extremely compelling, because again, you use it as this pathway 449 00:33:36.599 --> 00:33:39.720 of bridge to get to the next way of doing things that still is connected 450 00:33:39.759 --> 00:33:44.519 to your existing environment, because you know, rip replace is always very very 451 00:33:44.519 --> 00:33:49.319 difficult, so difficult that it almost never happens. We've joked before about sun 452 00:33:49.400 --> 00:33:53.160 setting applications and other fantasies, right or other fairy tales. So it's difficult 453 00:33:53.160 --> 00:34:00.319 to do. So that data virtualization is a really powerful mechanism to solve these 454 00:34:00.319 --> 00:34:02.359 problems and open the door to the next generation of how to do things. 455 00:34:02.359 --> 00:34:07.680 What do you think even yeah, we've been discussing about that, and what 456 00:34:07.720 --> 00:34:12.280 I see and in a virtualization layer, and which is not so understandable for 457 00:34:12.400 --> 00:34:16.000 the most of the people, is it's it's simplified or it's abstraction of the 458 00:34:16.079 --> 00:34:22.199 integration behind the scenes as such. If you're back in the days they've been 459 00:34:22.239 --> 00:34:25.559 building building point point systems, so you had to integrate each system with each 460 00:34:25.840 --> 00:34:30.039 and every other system within the organization. That takes a lot of time. 461 00:34:30.280 --> 00:34:35.159 If something changes on one end, you need to change all the interacting systems. 462 00:34:35.440 --> 00:34:37.760 With virtualization, this is kind of glue what you put on top of 463 00:34:37.760 --> 00:34:42.519 that. If you think about object oriented programming, you had your clauses and 464 00:34:42.639 --> 00:34:45.239 and what's kind of abstraction layer and what you had through which you were talking 465 00:34:45.280 --> 00:34:50.079 to the other systems compared to an API. In such a way, this 466 00:34:50.159 --> 00:34:55.039 is where integration solves well, where virtualization solves a part of the integration problem 467 00:34:55.119 --> 00:34:59.719 as such, but at the other end as well, allows you to s 468 00:35:00.199 --> 00:35:04.760 your systems because it's composable. You can change your database behind the scenes or 469 00:35:04.800 --> 00:35:08.000 depending on which type of query you want to run, is it more analytical 470 00:35:08.280 --> 00:35:12.840 or more operational. Wise, you can talk to do different systems, but 471 00:35:12.920 --> 00:35:16.719 it looks like one and the same system if you're a user using the virtualization 472 00:35:16.800 --> 00:35:22.519 layer, and that helps into going much faster in building applications and systems on 473 00:35:22.599 --> 00:35:27.320 top of a virtualization layer. So for the future as well, you're kind 474 00:35:27.360 --> 00:35:31.559 of safe because you develop everything on the simplification layer, and if you need 475 00:35:31.559 --> 00:35:37.280 a more performance system, you simply let's call it still simply plug it in 476 00:35:37.440 --> 00:35:43.199 the other system that is more perform more scalable as such. If we now 477 00:35:43.280 --> 00:35:45.840 see at the cloud, if you need the cloud, you can switch it 478 00:35:45.840 --> 00:35:49.639 off to the cloud and make it more scalable and move back to one prem 479 00:35:49.760 --> 00:35:55.039 So that will simplify your IT landscape in such a way. But my feeling 480 00:35:55.079 --> 00:36:00.320 is that we're still pretty far off on the virtualization level. A lot of 481 00:36:00.480 --> 00:36:07.679 organization are still very reluctant towards virtualization because they have back experiences in the day 482 00:36:07.679 --> 00:36:12.559 where it was not performing it was pretty expensive. But if you see now 483 00:36:12.599 --> 00:36:19.079 the virtualization landscape together with some semantic layers, this is the time to really 484 00:36:19.119 --> 00:36:23.599 look into that and consider that if you're building modern data applications. You know, 485 00:36:23.639 --> 00:36:29.320 and a lot of terms get thrown around like data lake house and data 486 00:36:29.400 --> 00:36:31.639 lake and data warehouse, and of course a very popular term has been data 487 00:36:31.760 --> 00:36:36.920 fabric over the last number of years. And when I think about the functionality 488 00:36:37.000 --> 00:36:39.199 that the data fabric should have, well, what does it boilt down to 489 00:36:39.440 --> 00:36:45.960 bost down to access governance? It boils down to automation or automation makes sense, 490 00:36:46.000 --> 00:36:50.440 and I think that's the ultimate data fabric that has the automation built into 491 00:36:50.480 --> 00:36:53.960 it to know when things are happening and to go and grab data sort before 492 00:36:54.199 --> 00:36:57.960 it's needed or before it's going to be needed, that kind of thing. 493 00:36:58.400 --> 00:37:01.480 It seems to me that virtualization, if it gets robuffed enough, can fulfill 494 00:37:01.599 --> 00:37:04.920 that role of a data fabric. But I don't know, what do you 495 00:37:04.960 --> 00:37:07.960 think lewis am my splitting hairs or conflating things. No, I mean I 496 00:37:07.960 --> 00:37:12.199 think that that's I think that that's a fair point. I think we're still 497 00:37:12.360 --> 00:37:15.800 going to see where we land with the whole data fabric, data lakehouse. 498 00:37:15.920 --> 00:37:21.719 The new nomenclature that I put my money right now on data fabric being being 499 00:37:21.760 --> 00:37:27.119 the one that sort of emerges triumphant because I think it's very realistic in terms 500 00:37:27.119 --> 00:37:34.400 of what is possible. Access, discovery, governance, observability are all made 501 00:37:34.519 --> 00:37:39.960 possible by having data virtualization and a data catalog blended together. Be throwing some 502 00:37:40.039 --> 00:37:44.960 data ops and some health monitoring there, and you've got a pretty slick d 503 00:37:45.079 --> 00:37:49.719 you know platform. And then the final key I think is that actually, 504 00:37:49.760 --> 00:37:52.880 at the end of the day, being able to access that underlying data becomes 505 00:37:52.880 --> 00:37:54.480 really really critical. And this is you've mentioned, you know, some people 506 00:37:54.480 --> 00:37:58.760 have been bitten before with sort of virtualization tools and things like that. I 507 00:37:58.800 --> 00:38:05.440 think that that was based on a sort of system where just cataloging the data 508 00:38:05.519 --> 00:38:07.239 was enough, but at the end of the day, you still need to 509 00:38:07.280 --> 00:38:10.079 be able to provide access to that data, need to push it down to 510 00:38:10.239 --> 00:38:14.320 where it needs to go in order to provide benefit to the organization. And 511 00:38:14.360 --> 00:38:16.679 these are the types of things that we've learned over the past decade that now 512 00:38:16.719 --> 00:38:21.920 we're getting right. And yeah, I think that fabric is really the solution 513 00:38:21.960 --> 00:38:23.559 to that. I think the next step is sort of also getting a solution 514 00:38:23.599 --> 00:38:29.480 that enables you to distribute your data and more holistically throughout the enterprise and in 515 00:38:29.519 --> 00:38:35.320 a secure way to really enable that democratization among business and technical units. Yeah, 516 00:38:35.360 --> 00:38:37.320 that's a really good point. Even maybe I'll throw it over to you 517 00:38:37.360 --> 00:38:42.760 for commentary. Again, these are fairly loose terms, but data fabric, 518 00:38:42.800 --> 00:38:46.320 I think does boil down to the access layer that you want your apps to 519 00:38:46.320 --> 00:38:50.000 connect to. Because the end of the day, that's what we're talking about. 520 00:38:50.000 --> 00:38:52.679 We're talking about applications. They're going to connect to the data that they 521 00:38:52.679 --> 00:38:58.159 need to get something done. So instead of the old fashioned process of just 522 00:38:58.239 --> 00:39:02.679 persisting data in mongo eb or mic sequel or whatever, that's a very linear, 523 00:39:02.920 --> 00:39:07.480 sort of limited approach, and instead you want this richer environment that any 524 00:39:07.519 --> 00:39:12.639 number of apps can tap into and quickly get what they want and get more 525 00:39:12.719 --> 00:39:15.760 than they would have gotten from just a single connection to a single database, 526 00:39:15.880 --> 00:39:20.719 righting, Yeah, the most important part is here that the centralized or the 527 00:39:20.800 --> 00:39:25.280 central available governance what you can tie to your data products in such a way 528 00:39:25.400 --> 00:39:30.239 and that you either consume it to an applicational operational application an API, or 529 00:39:30.440 --> 00:39:35.079 you just do an export. You always have that same type of approach on 530 00:39:35.440 --> 00:39:38.800 who can access which data. Before it was only within the organization, and 531 00:39:38.880 --> 00:39:44.880 we see more and more organizations exchanging it with external organizations. So there is 532 00:39:44.880 --> 00:39:49.920 the important part that you have an easy way of onboarding your external bodies and 533 00:39:50.280 --> 00:39:53.880 decide who can see what type of data in an easy way that you can 534 00:39:54.000 --> 00:39:58.320 govern them that data, where is it going to, who is using it 535 00:39:58.360 --> 00:40:01.320 for what? So you stay comply and know where your data is going there, 536 00:40:01.639 --> 00:40:06.119 especially in marketing, you want to know who are your consumers and who 537 00:40:06.239 --> 00:40:12.280 is using your data. Especially in Europe, we're pretty strict on on understanding 538 00:40:12.440 --> 00:40:16.519 where that data is being used for privacy reasons and I think in that direction, 539 00:40:16.599 --> 00:40:24.400 data fabric is definitely a very nice architecture to support governance in a pretty 540 00:40:25.159 --> 00:40:29.679 easy way. Easy in such a way, it's not out of the blue, 541 00:40:29.760 --> 00:40:32.559 it's not snapping with your fingers, but the tools are there and learned 542 00:40:32.599 --> 00:40:37.679 a lot from the past where you said in the beginning it was purely cataloging 543 00:40:37.719 --> 00:40:40.400 your data, understanding where it is and who maybe has access to that. 544 00:40:40.920 --> 00:40:46.840 But access was mostly a very ideal related thing in place, very technical, 545 00:40:47.079 --> 00:40:52.199 pretty complex, so getting your user account to have access to systems that took 546 00:40:52.280 --> 00:40:58.199 months in some some organizations. So this is definitely not something if you want 547 00:40:58.239 --> 00:41:01.679 to put a product into the walked in the next coming month. Now, 548 00:41:01.760 --> 00:41:05.320 these are all really good points, and you know, before the break comes 549 00:41:05.400 --> 00:41:08.280 up here, I do want to dive a bit deeper into the discovery side 550 00:41:08.280 --> 00:41:13.239 of the equation because we know we're talking again about how to get ready for 551 00:41:13.400 --> 00:41:15.599 leveraging a large language model, and that means you got to do some discovery 552 00:41:15.599 --> 00:41:21.760 and understand what's out there whereas there personally identifiable information, what is the nature 553 00:41:21.760 --> 00:41:23.800 of this business information that we have, and create some sort of a topology 554 00:41:23.880 --> 00:41:28.559 or a map to better understand, and then you create your short list of 555 00:41:28.599 --> 00:41:30.599 things to work on. I scaring this data or do we really need that 556 00:41:30.679 --> 00:41:35.719 data? But it really does start with that discovery process, right lewis what 557 00:41:35.719 --> 00:41:38.000 do you think? Yeah, I think you're right. Discovery this sort of 558 00:41:38.039 --> 00:41:43.719 the pillar here, and again it's not the thing that's going to turn heads 559 00:41:43.719 --> 00:41:46.760 in the boardroom being able to discover things. But at the end of the 560 00:41:46.880 --> 00:41:52.599 day, any sort of catalog or management solution is going to need to index 561 00:41:52.960 --> 00:41:55.079 everything to do with your data, and it's going to need to automate that 562 00:41:55.079 --> 00:42:01.719 that indexing process and then allow you to create sort of combinations of terms. 563 00:42:01.920 --> 00:42:05.400 And those terms might be based on what's in the data, but it might 564 00:42:05.440 --> 00:42:08.039 also be based on where does the data live, who's the owner of the 565 00:42:08.119 --> 00:42:12.639 data and my team, you know, for some of our users, it's 566 00:42:12.679 --> 00:42:15.519 like what's the data cost and when's the contract term coming up? These are 567 00:42:15.559 --> 00:42:17.519 the things that I actually need to know. And that's metadata. That's not 568 00:42:17.599 --> 00:42:22.239 just about the you know, the structural build of the data or its descriptive 569 00:42:22.280 --> 00:42:28.559 elements. That's also sort of the administrative side of metadata, all of which 570 00:42:28.599 --> 00:42:30.840 needs to be captured by the solution, indexed by the search. And then 571 00:42:30.880 --> 00:42:36.199 it can't just be a blinking cursor you know Google type search that where you 572 00:42:36.199 --> 00:42:39.440 you plug in, you know, personally identify a whole information and automatically everything 573 00:42:39.440 --> 00:42:44.079 comes back. You want to be able to create sort of rich combinations of 574 00:42:44.079 --> 00:42:46.880 things to get to what you're looking for, so that you can create these 575 00:42:47.000 --> 00:42:52.599 very highly curated lists, and then once you've done that, you need to 576 00:42:52.639 --> 00:42:54.719 sort of have recourse to move data around if you need to. We talked 577 00:42:54.760 --> 00:42:59.599 a little bit about the expense that's that's driving a lot of this stuff, 578 00:42:59.639 --> 00:43:01.559 and you're able to do hot and cold storage from this platform, move it 579 00:43:01.599 --> 00:43:07.199 from a really performed warehouse to sort more open source one based on how much 580 00:43:07.199 --> 00:43:09.559 it's being used, and things like that. That all starts with discovery. 581 00:43:09.599 --> 00:43:14.440 All those benefits start from being able to just open the door and see what's 582 00:43:14.480 --> 00:43:16.920 going on in your environment. Yeah. No, that's exactly right. And 583 00:43:17.000 --> 00:43:20.480 you know, you talked about the importance of metadata, and I love the 584 00:43:20.480 --> 00:43:23.320 comment you made about contracts and dating. What its is costing me and what 585 00:43:23.519 --> 00:43:28.079 value are we getting? You know, when you bring together the power of 586 00:43:28.079 --> 00:43:31.480 observability with the power of discovery and analysis and of course the human brain. 587 00:43:31.599 --> 00:43:35.519 Let's not forget us humans. I don't care what it says. We're gonna 588 00:43:35.519 --> 00:43:37.360 be around for a while now and we'll be very important in this space. 589 00:43:37.679 --> 00:43:40.639 When you bring all that together, it starts to get very interesting. In 590 00:43:40.719 --> 00:43:49.280 things like ROI and TCO, which historically have been almost impossible to really quantify. 591 00:43:49.360 --> 00:43:52.360 I mean, you can tell the story and you can get people excited 592 00:43:52.360 --> 00:43:55.239 about stuff. That's one way to keep investment in your projects. But we're 593 00:43:55.280 --> 00:44:00.559 getting so much better now at being able to determine the value of this particular 594 00:44:00.599 --> 00:44:07.519 solution against key metrics. For example, customer service improved by thirty percent because 595 00:44:07.519 --> 00:44:09.760 the tickets came down. Stuff like that, we're getting into space where it's 596 00:44:09.880 --> 00:44:13.639 very measurable and that it's very good news for this business. Will be right 597 00:44:13.639 --> 00:44:19.199 back. You're listening to inside Analysis. Do you own an annuity, either 598 00:44:19.280 --> 00:44:22.159 fixed rate, indexed or variable? Are you paying high fees and getting low 599 00:44:22.199 --> 00:44:27.239 returns? If so, Annuity General would like you to have this free book 600 00:44:27.280 --> 00:44:30.280 to learn the pitfalls and mistakes of buying an annuity. The Annuity Dues and 601 00:44:30.360 --> 00:44:35.519 Don'ts for Baby Boomers contains the little known truths about annuities, like how to 602 00:44:35.559 --> 00:44:38.840 help reduce your fees and increase retirement income. And it's free, that's right 603 00:44:39.320 --> 00:44:44.119 free. As a bonus, we'll also throw in a free annuity rate report 604 00:44:44.199 --> 00:44:49.440 just for calling. We researched over one thousand annuities and summarized rates and benefits 605 00:44:49.480 --> 00:44:53.280 from financially strong insurers. You get annuity dues and don'ts for baby boomers and 606 00:44:53.360 --> 00:44:58.880 the Annuity of Rate Report, both absolutely free for calling Annuity General Today. 607 00:44:59.360 --> 00:45:04.880 Hurry flies are limited. Call now eight hundred two four five one six nine 608 00:45:05.039 --> 00:45:09.360 seven, eight hundred two four five one six nine seven, eight hundred two 609 00:45:09.519 --> 00:45:15.840 four five one six ninety seven. That's eight hundred two four five sixteen ninety 610 00:45:15.880 --> 00:45:20.800 seven. Do you own a timeshare? Well, face the facts. You 611 00:45:20.880 --> 00:45:24.039 made a mistake, You made a bad purchase. A timeshare is not an 612 00:45:24.079 --> 00:45:29.599 investment. It's a money pit that continues forever. If you use your time 613 00:45:29.639 --> 00:45:34.159 share, that's great. But if you don't and you want illegally get out 614 00:45:34.199 --> 00:45:37.960 of your contract, call my friends right now at the Timeshare Exit Hotline. 615 00:45:38.039 --> 00:45:43.360 They're an experienced team of lawyers who help good people like you get out of 616 00:45:43.360 --> 00:45:46.880 a time share contract that they just don't want. Don't throw away your money 617 00:45:46.920 --> 00:45:51.639 on maintenance fees, use it for things you really want. We can help 618 00:45:51.679 --> 00:45:55.639 you end your time share contract and stop the money drain immediately. If you 619 00:45:55.639 --> 00:46:00.840 are ready to move on with your time share, call our team right now. 620 00:46:00.079 --> 00:46:05.239 Cancel your time share now with a free call eight hundred two nine oh 621 00:46:05.519 --> 00:46:09.719 six seven O five eight hundred two nine oh six seven O five. Eight 622 00:46:09.840 --> 00:46:15.519 hundred two nine zero six seven O five. That's eight hundred two nine oh 623 00:46:15.840 --> 00:46:21.159 sixty seven O five. Do you own a time share, Well, face 624 00:46:21.239 --> 00:46:23.719 the facts. You made a mistake, You made a bad purchase. A 625 00:46:23.840 --> 00:46:29.880 timeshare is not an investment. It's a money pit that continues forever. If 626 00:46:29.880 --> 00:46:32.639 you use your time share, that's great. But if you don't and you 627 00:46:32.719 --> 00:46:37.320 want illegally get out of your contract, call my friends right now at the 628 00:46:37.440 --> 00:46:43.199 time Share Exit Hotline. They're an experienced team of lawyers who help good people 629 00:46:43.239 --> 00:46:46.320 like you get out of a time share contract that they just don't want. 630 00:46:46.599 --> 00:46:51.440 Don't throw away your money on maintenance fees. Use it for things you really 631 00:46:51.440 --> 00:46:54.679 want. We can help you end your time share contract and stop the money 632 00:46:54.760 --> 00:46:59.800 drain immediately. If you are ready to move on with your time share, 633 00:47:00.159 --> 00:47:02.920 call our team right now. Yes, so your time share now with a 634 00:47:04.039 --> 00:47:08.239 free call eight hundred two eight nine O four one three eight hundred two eight 635 00:47:08.360 --> 00:47:14.519 nine O four one three eight hundred two eight nine O four one three. 636 00:47:14.800 --> 00:47:19.519 That's eight hundred two eight nine zero four thirteen. If you served in the 637 00:47:19.559 --> 00:47:22.599 Marine Corps. By now you know about the contaminated water problem at Camp Lejeune. 638 00:47:22.800 --> 00:47:27.400 If you were stationed or worked at Camp Lejeune from nineteen fifty three to 639 00:47:27.519 --> 00:47:30.159 nineteen eighty seven, you probably have a lot of questions. We have some 640 00:47:30.280 --> 00:47:35.679 answers. You could be entitled to compensation. Billions of dollars are being allocated 641 00:47:35.719 --> 00:47:39.280 to pay for damages to anyone stationed at Camp Lejeune during that time. Unfortunately, 642 00:47:39.440 --> 00:47:44.559 it appears that officials may have known the contaminated water problem existed and did 643 00:47:44.639 --> 00:47:47.760 little to protect their men. The Semper five code was not honored. If 644 00:47:47.800 --> 00:47:52.920 you or someone in your family has developed a serious illness, including various forms 645 00:47:52.920 --> 00:47:57.440 of called this Camp Lejeune Legal Support Line right now. You can't turn back 646 00:47:57.440 --> 00:48:00.960 the clock and change what happened, but you can cert they call right now 647 00:48:00.039 --> 00:48:04.760 and learn your rights as a marine. Here's the number. Call eight hundred 648 00:48:04.920 --> 00:48:10.480 two five four three two one eight eight hundred two five four three two one 649 00:48:10.599 --> 00:48:19.800 eight. That's eight hundred two five four thirty two eighteen paid for express. 650 00:48:19.840 --> 00:48:30.000 Welcome back to Inside Analysis. Here's your host Eric kabinat right. Folks back 651 00:48:30.000 --> 00:48:35.079 here on Inside Analysis talking to Lewis Winnomes from Think Data Works and Eve Mulkers 652 00:48:35.159 --> 00:48:38.000 from seven W Data and the Brink. There we're chatting about the power of 653 00:48:38.039 --> 00:48:44.000 AI and the importance of data discovery, and I keep thinking to myself that 654 00:48:44.079 --> 00:48:46.760 these foundational models, which is not just the large language models, but these 655 00:48:46.800 --> 00:48:52.960 foundational models are going to fundamentally change a lot of different things. And if 656 00:48:52.000 --> 00:48:57.400 you think about how systems talk to each other on the average day, just 657 00:48:57.519 --> 00:49:00.800 transacting the work that needs to get done over the network, for example, 658 00:49:00.320 --> 00:49:05.880 there are so many efficiencies that we can gain from collaboration, from being able 659 00:49:05.920 --> 00:49:08.280 to see what other people are doing. I'm talking to a company right now, 660 00:49:08.320 --> 00:49:12.159 in fact, writing a paper in my head for a company called dev 661 00:49:12.239 --> 00:49:19.159 rev which has built this really interesting object model and they're really fusing development with 662 00:49:19.280 --> 00:49:22.280 revenue management and isn't that what everyone wants? And through their system, you 663 00:49:22.320 --> 00:49:28.119 can see across roles and see exactly who's working on what at any given time. 664 00:49:28.679 --> 00:49:31.079 And because they thought through that metadata side of the equation ahead of time. 665 00:49:31.400 --> 00:49:35.440 I'll fill this over to lewis what you're able to see is not just 666 00:49:35.519 --> 00:49:38.920 what all your individual developers are working on, but what kinds of projects are 667 00:49:38.960 --> 00:49:44.199 they are they maintenance projects, are they net new functionality projects? And then 668 00:49:44.239 --> 00:49:45.480 where is the value you'll be able to track? Oh, well, this 669 00:49:45.519 --> 00:49:51.039 project that we launched two months ago is already generating some value. So you 670 00:49:51.079 --> 00:49:53.800 can actually start seeing ROI And to me, that's going to be the key 671 00:49:53.840 --> 00:50:00.199 is seeing across organizations who's working on not just up, down, but sideways. 672 00:50:00.599 --> 00:50:02.960 And once you can see up, down, left and right and really 673 00:50:04.039 --> 00:50:08.840 understand how your teams are coming together, that's a pretty compelling storyline. And 674 00:50:08.880 --> 00:50:14.039 that comes from discovery, but it comes from data management. It comes from 675 00:50:14.079 --> 00:50:17.519 setting up your models efficiently such that people can see around and understand what's going 676 00:50:17.519 --> 00:50:20.760 on. But Lewis, I'll throw it over to you. I think that's 677 00:50:20.800 --> 00:50:25.239 going to be a real push forward in terms of productivity gains. Just the 678 00:50:25.280 --> 00:50:30.719 observability across workloads, across teams. What do you think? Yeah, I 679 00:50:30.719 --> 00:50:35.039 mean you mentioned the critical word there, which is observability, and I think 680 00:50:35.079 --> 00:50:39.320 that a pound of observability is going to be a dash of observe ability is 681 00:50:39.360 --> 00:50:43.960 going to be worth a pound of cure or whatever it is. And the 682 00:50:44.360 --> 00:50:47.159 point here being that teams have always been building interesting things, right, You've 683 00:50:47.159 --> 00:50:51.599 got data teams. You've got data scientists who are doing great stuff, but 684 00:50:51.639 --> 00:50:53.079 they might be doing it in a vacuum. And I think that when I 685 00:50:53.119 --> 00:50:58.320 speak to our clients and the people who use our platform, there's often some 686 00:50:58.559 --> 00:51:01.639 person who might be in governance, might be in charge of ROI, who's 687 00:51:01.679 --> 00:51:05.440 having to knock on a lot of doors and email a lot of people saying 688 00:51:05.559 --> 00:51:07.880 hey, can you give me an update on what you're working on? What 689 00:51:07.960 --> 00:51:10.159 data sets? Are you saying? What's going on here? That they can 690 00:51:10.239 --> 00:51:14.599 still do their thing. I think the model shouldn't be too that you need 691 00:51:14.639 --> 00:51:19.360 to completely upend the good work that your teams are doing and force them all 692 00:51:19.400 --> 00:51:22.400 into a central place. The model should be that if you can have something 693 00:51:22.480 --> 00:51:25.400 that can sit on top of all the work that all of your different teams 694 00:51:25.440 --> 00:51:30.400 are doing, lurp up that information and then make it more generally accessible to 695 00:51:30.400 --> 00:51:35.760 people without upending the work that's already being done. Then you're letting someone from 696 00:51:35.800 --> 00:51:39.679 your team who's in charge of establishing things like ROI or making good connections right 697 00:51:39.760 --> 00:51:43.239 saying, oh, this team's working on this data set. This team's working 698 00:51:43.239 --> 00:51:46.079 on this data set. If I sort of play a bit of matchmaker between 699 00:51:46.079 --> 00:51:50.199 them, maybe they can both run a bit faster. That's the real net 700 00:51:50.199 --> 00:51:53.199 benefit here. It's not in improving individual teams, although they will get that 701 00:51:53.239 --> 00:51:58.800 benefit as well. It's about sort of across the enterprise having that sort of 702 00:51:58.840 --> 00:52:05.079 democratized view and observation platform and what you get the benefit that you get from 703 00:52:05.280 --> 00:52:07.960 from that. Yeah, knowing how teams can work together. I mean, 704 00:52:08.159 --> 00:52:12.599 I'll throw this over to Eve. AI can help with these things. I 705 00:52:12.639 --> 00:52:15.800 think AI is going to be very compelling for just suggesting things to people, 706 00:52:15.920 --> 00:52:20.400 pointing out stuff that maybe you hadn't thought of, because you'll never think of 707 00:52:20.480 --> 00:52:23.280 everything. There was a great quote in a movie with William Hurt, like 708 00:52:23.360 --> 00:52:27.159 thirty years ago. I'll ever forget this. One of the characters says something 709 00:52:27.199 --> 00:52:29.840 like who they were criminals? And then one guy said, hey, man, 710 00:52:29.880 --> 00:52:31.639 you told me there's a hundred ways to screw up any crime. If 711 00:52:31.679 --> 00:52:35.400 you could think of fifty of them, you're a genius, right, So 712 00:52:35.840 --> 00:52:38.840 point being, you're always going to miss something, and that's where AI comes 713 00:52:38.880 --> 00:52:43.039 into help. But to me, again, the power of the human mind 714 00:52:43.159 --> 00:52:47.639 just thinking through things and processing and figuring out what Lewis just referred to, 715 00:52:47.840 --> 00:52:51.960 Hey, these teams are working on something pretty darn similar. If I can 716 00:52:52.039 --> 00:52:57.320 just get them to connect and start sharing information sharing that at data sharing plans, 717 00:52:57.480 --> 00:53:00.199 roadmaps. I mean, just knowing what people are working on is really 718 00:53:00.280 --> 00:53:04.639 compelling and just and team size, right, if you have a hundred developers, 719 00:53:04.880 --> 00:53:07.159 that doesn't mean you're accomplishing one hundred times as much as one developer, 720 00:53:07.440 --> 00:53:12.039 or even fifty times as much as two. It really depends on how well 721 00:53:12.079 --> 00:53:15.360 those teams work together, what the plan is, how that plan fits in 722 00:53:15.400 --> 00:53:19.679 with what other people are doing. I mean, this is the strategy side 723 00:53:20.079 --> 00:53:23.280 of development, and I think all of that is enabled by these large language 724 00:53:23.280 --> 00:53:28.480 models, by these foundational models and the AI, and of course the data 725 00:53:28.519 --> 00:53:30.920 management and the virtualization. But Lewis'll turn that first over to you and then 726 00:53:30.960 --> 00:53:34.840 over to Eat to comment on. Go ahead. Yeah, I mean, 727 00:53:34.880 --> 00:53:37.760 I think that's that's the key is if we want to make good on anything 728 00:53:37.800 --> 00:53:43.079 to do with AI, you can't have a bunch of developers in closed rooms 729 00:53:43.079 --> 00:53:45.719 trying to figure this stuff out on their own. It needs to be an 730 00:53:45.719 --> 00:53:49.360 open door policy and there needs to be governance attached to it. Right, 731 00:53:49.400 --> 00:53:51.760 We're not in the wild West anymore, and you can't just say we're going 732 00:53:51.800 --> 00:53:54.559 to throw a large language model over our entire database, because there's PII in 733 00:53:54.599 --> 00:53:57.440 there. There's stuff in there that you don't want it to be trained on. 734 00:53:57.519 --> 00:54:00.159 So you need to have governance, which for a lot of people has 735 00:54:00.199 --> 00:54:01.679 been a big hairy problem because they don't know how to do it, and 736 00:54:01.679 --> 00:54:05.400 so they've just said, well, our version of governance is no one gets 737 00:54:05.440 --> 00:54:07.599 access to anything. We've solved that problem. Now, that's not the way 738 00:54:07.599 --> 00:54:12.519 we're going to move forward, right, More people need access to more data 739 00:54:13.239 --> 00:54:16.440 and in a secure way. So that observation platform and letting it be mission 740 00:54:16.519 --> 00:54:21.679 control for data teams. Right, So the person who's tasked with revenue gets 741 00:54:21.679 --> 00:54:23.639 to do revenue based stuff, and the person who's tasked with doing data stuff 742 00:54:23.960 --> 00:54:25.880 still gets to do their data stuff, and they get to do it in 743 00:54:25.920 --> 00:54:30.719 a quicker, faster, more integrated way. Yeah, that makes a lot 744 00:54:30.760 --> 00:54:34.199 of sense. Eging to comment on there. Yeah, a quicker and foster 745 00:54:34.360 --> 00:54:37.119 way I think to insights, and that's an important thing that you were discussing 746 00:54:37.199 --> 00:54:43.159 just before. It's an easy way of looking at the one hundred tasks or 747 00:54:43.199 --> 00:54:46.320 even one thousand tasks. We see it in medical care as well, where 748 00:54:46.360 --> 00:54:51.480 the scale really helps you build those insights and give the insights at a hyper 749 00:54:51.559 --> 00:54:54.360 speed. And I think that's that's the power of what we're seeing and sometimes 750 00:54:54.360 --> 00:55:00.840 missing these interactions and communication lines within organizations. If you don't talk to the 751 00:55:00.840 --> 00:55:04.519 person, you don't get these insights. It's by accident if you're around the 752 00:55:04.559 --> 00:55:07.239 water cooler that you have these type of discussions and say, hey, this 753 00:55:07.840 --> 00:55:09.639 is interesting what you're working on. Let's have a look and see if we 754 00:55:09.679 --> 00:55:15.559 can do something together. And with the large language models, you get these 755 00:55:15.599 --> 00:55:19.639 these advices in such a way on the other hand as well, where it 756 00:55:20.519 --> 00:55:23.599 gives you a pretty complete list and you say throw in another of five points 757 00:55:23.920 --> 00:55:28.280 and you get the out of five points and after a while you understand, 758 00:55:28.480 --> 00:55:31.039 okay, we really exhausted the model. This is all we can get from 759 00:55:31.079 --> 00:55:36.039 it. But that's much harder from a human mind to get all of these 760 00:55:36.079 --> 00:55:38.559 aspects listed in such a way. So I think that the scale and the 761 00:55:38.599 --> 00:55:44.239 speed that's the important thing. Where we see that it can augment on how 762 00:55:44.280 --> 00:55:47.320 we work. Like every was saying, thinking through the critical thinking, which 763 00:55:47.360 --> 00:55:51.199 is not yet in the models. I think we will see that in the 764 00:55:51.199 --> 00:55:54.639 future that the models will evolve and get more critical thinking as well, but 765 00:55:54.760 --> 00:55:59.599 this is still a very strong capability of what we have as a human being. 766 00:56:00.239 --> 00:56:01.920 Yeah, and you know, I'll get back to the theme of the 767 00:56:01.960 --> 00:56:07.480 show here, which is using these technologies to discover what data you have, 768 00:56:07.559 --> 00:56:10.320 to classify it, to organize it, to understand it, and then maybe 769 00:56:10.400 --> 00:56:15.199 you're ready to leverage some of these large language models. But I am pretty 770 00:56:15.199 --> 00:56:19.559 excited, Lewis, to think about what you could learn about your business by 771 00:56:19.679 --> 00:56:23.960 connecting understand that you can possibly remediate so you can publish something in a vector 772 00:56:24.039 --> 00:56:27.719 data base and go, oh, we shouldn't have done that. Let's pull 773 00:56:27.760 --> 00:56:31.079 that out now to not have that exposed. But I have to think that 774 00:56:31.519 --> 00:56:37.519 these language models are a very profound mirror that you can use to reflect back 775 00:56:37.639 --> 00:56:42.599 what your business has been doing, because you will see stuff in there and 776 00:56:42.639 --> 00:56:44.840 it's almost like you want, if you're a big enough organization, you want 777 00:56:44.840 --> 00:56:47.280 to have a skunkworks team or you know someone who's just in there poking around 778 00:56:47.760 --> 00:56:52.599 seeing what's going on, to be able to develop your plan to govern and 779 00:56:52.719 --> 00:56:55.000 access and virtualize better. What do you think, Lewis, I think that's 780 00:56:55.440 --> 00:57:00.679 such a great point I would if I was in charge of a large enterprise. 781 00:57:00.840 --> 00:57:04.880 The way I'd be looking at LMS right now is let's use this internally 782 00:57:04.960 --> 00:57:07.760 to understand a few things internally before we even think about the data products with 783 00:57:07.800 --> 00:57:12.320 it. I think the problem that I have with generative AI right now is 784 00:57:12.360 --> 00:57:15.920 the generative portal part. I think I'd like to use it for classification, 785 00:57:16.519 --> 00:57:21.239 train it on very good models and things like that. But immediately dropping it 786 00:57:21.280 --> 00:57:24.000 in and thinking that some production ready tool is going to come at the other 787 00:57:24.079 --> 00:57:28.599 end of it, I think it is dangerous thinking. And I don't think 788 00:57:28.599 --> 00:57:32.559 that this limits the use for the enterprise of starting to integrate it, to 789 00:57:32.559 --> 00:57:37.000 bring in and do some automatic sort of classifications of things. And chat pots 790 00:57:37.000 --> 00:57:42.159 are a great use for it. But let's not get ourselves like this is 791 00:57:42.159 --> 00:57:46.039 still pretty nascent technology, and things like large scale statistical analysis have been around 792 00:57:46.039 --> 00:57:50.400 for a long time and people haven't necessarily made good on those either. Why 793 00:57:50.880 --> 00:57:53.719 Because the management problem is still sort of very very at the core of this 794 00:57:53.840 --> 00:57:58.599 issue. So I think chat ept the biggest thing that it's done is it's 795 00:57:58.599 --> 00:58:02.880 shown that business focus people need to be able to integrate with data in meaningful 796 00:58:02.920 --> 00:58:07.199 ways. And they did a really good job of making that possible for us. 797 00:58:07.440 --> 00:58:08.880 But we still got a lot of work to do in order to make 798 00:58:08.880 --> 00:58:12.800 sure that everyone can benefit from an equally Yeah. Well, there's a lot 799 00:58:12.800 --> 00:58:19.599 of education and guiding basically guidance provided to organizations to understand how to use it. 800 00:58:19.599 --> 00:58:22.760 It's like any tool. You want to be playing around it to understand 801 00:58:22.800 --> 00:58:25.239 the contours basically, to understand what it does. We do have a show 802 00:58:25.280 --> 00:58:29.159 coming up on Thursday. As a matter of fact, eviill be delivering a 803 00:58:29.239 --> 00:58:32.760 keynote on master of the prompt. I'll give you some advice on how to 804 00:58:34.280 --> 00:58:38.760 leverage that prompt because it is very important how you phrase those prompts and which 805 00:58:38.920 --> 00:58:45.159 specific words you use and how you deliver the information to the large language model. 806 00:58:45.199 --> 00:58:47.480 And it's just taking time. It's going to take time to see what 807 00:58:47.599 --> 00:58:52.239 comes back at you and learn how to fine tune these things. And remember, 808 00:58:52.280 --> 00:58:54.679 you can regenerate. They all have this capacity to regenerate. So it's 809 00:58:54.679 --> 00:58:58.360 like I didn't like that one, Let's do it again at a couple more 810 00:58:58.440 --> 00:59:00.760 keywords. To do it again, it comes out, it'll get you eighty 811 00:59:00.760 --> 00:59:04.239 percent of the way there. Then you have to get yourself the last twenty 812 00:59:04.280 --> 00:59:09.079 percent. But the folks you're listening to inside and out, you've eating lots 813 00:59:09.079 --> 00:59:14.239 of great food and lots of great food at restaurants. Cowboy Burgers in Fontana 814 00:59:14.599 --> 00:59:19.039 and now on Arlington and Riverside will fast become one of your favorites with their 815 00:59:19.079 --> 00:59:23.960 delicious, mouth watering burgers and breakfast burritos. Cowboy Burgers and Barbecue also serves 816 00:59:23.960 --> 00:59:29.679 fantastic smoke barbecue, baby back ribs, dry tip chicken, pulled pork sandwiches, 817 00:59:29.760 --> 00:59:31.800 as well as lunch and dinner plates. Everything is made from scratch, 818 00:59:31.920 --> 00:59:37.159 including their delicious side dishes like cole slap, potato salad, barbecue beans, 819 00:59:37.199 --> 00:59:40.159 and much much more. Check out their rich, decadent chocolate brownies. Hi. 820 00:59:40.280 --> 00:59:44.880 I'm food critic Allenborgan, and you can dine in, take food out, 821 00:59:45.000 --> 00:59:49.320 or have them cater your neck special event. I highly recommend Cowboy Burgers 822 00:59:49.360 --> 00:59:53.960 and Barbecue at their new location at five five seven three Arlington Avenue in Riverside. 823 00:59:54.239 --> 00:59:58.760 Just look them up on the Internet. That's Cowboy Burgers and Barbecue, 824 00:59:58.960 --> 01:00:04.639 happy eating and perfect for the holidays. Cowboy Burgers and Barbecue is also available 825 01:00:04.800 --> 01:00:08.559 for catering that's Cowboy Burgers and barbecue in Fontana and now in River