WEBVTT 1 00:00:00.120 --> 00:00:03.960 Eight eight seven. Tony even stays awake all night, twenty four hours a 2 00:00:04.000 --> 00:00:07.639 day, seven days a week, so you can sleep better and rest easy. 3 00:00:08.320 --> 00:00:12.240 South Pacific sleep left, start feeling better and get in a great night 4 00:00:12.320 --> 00:00:21.199 of sleep. Today. You're on board caseyaas Inland Express caseyaa home, Linda 5 00:00:21.440 --> 00:00:32.719 en fifty am the station that needs notice here behind the information economy has a 6 00:00:32.840 --> 00:00:38.240 ride. The world is teeming with innovation as new business models reinvent every industry 7 00:00:38.280 --> 00:00:43.600 industry. Inside Analysis is your source of information and insight about how to make 8 00:00:43.640 --> 00:00:48.439 the most of this exciting new era. Learn more at inside analysis dot com, 9 00:00:48.520 --> 00:00:53.039 Inside analysis dot com. And now here's your host, Eric Kavanaugh. 10 00:00:57.679 --> 00:01:00.039 All right, ladies and gentlemen, Hello and welcome back once again the only 11 00:01:00.159 --> 00:01:04.680 coast to coast show all about the information economy. It's called Inside Analysis, 12 00:01:04.760 --> 00:01:08.000 or is truly Eric Kavanaugh Here with one of my best friends in the business, 13 00:01:08.079 --> 00:01:12.079 a guy who's been around we're a couple of years now and has made 14 00:01:12.079 --> 00:01:18.040 his way into all sorts of different data scenarios, data products, data concepts, 15 00:01:18.120 --> 00:01:22.599 data models, data designs. Mister Kent Graziano, I asked him before 16 00:01:22.599 --> 00:01:26.680 we started, does that mean? Does that mean thank you very much in 17 00:01:26.719 --> 00:01:33.040 Italian because of course gratimile, but it means grace or beloved or deer, 18 00:01:33.560 --> 00:01:38.599 so kent deer gratiano. How was going out there, buddy? Good? 19 00:01:38.640 --> 00:01:41.760 How are you doing rich? I'm doing good. So yeah, folks, 20 00:01:41.799 --> 00:01:46.200 I had this idea to do a series called profiles in data, because it's 21 00:01:46.200 --> 00:01:48.200 all about the people when you get right down to it. I mean, 22 00:01:48.239 --> 00:01:51.319 of course the data is important, and the tools and the technologies and the 23 00:01:51.359 --> 00:01:53.920 methods and how to do all that stuff. But at the end of the 24 00:01:53.000 --> 00:01:57.000 day, even in this era of AI, it's still about the people. 25 00:01:57.239 --> 00:02:00.359 And it's always going to be about the people. I mean, the jobs 26 00:02:00.359 --> 00:02:02.680 are going to change. I think we're going to go through a very disruptive 27 00:02:02.719 --> 00:02:07.159 period. It's already happening from what I've seen. It's it's already taking place. 28 00:02:07.719 --> 00:02:09.360 There are lots of reasons for it, but the one thing you can 29 00:02:09.360 --> 00:02:14.199 count on is the wisdom of sages, and so Kan't. I'm going to 30 00:02:14.240 --> 00:02:16.599 throw you in that category as a data sage. You've been around a long 31 00:02:16.599 --> 00:02:22.319 time, You've seen the ups and downs, and we're in a very different 32 00:02:22.360 --> 00:02:23.680 world right now. You know, I'm going to throw kind of a curb. 33 00:02:23.719 --> 00:02:29.080 I'll let you to start. I feel like these large language models represent 34 00:02:29.199 --> 00:02:35.240 a very serious inflection point in the history of business, quite frankly, but 35 00:02:35.560 --> 00:02:39.680 primarily of the data world. And i'll explain why. So in the very 36 00:02:39.719 --> 00:02:44.520 earliest days of DM radio, I remember wrapping my head around ETL, all 37 00:02:44.560 --> 00:02:47.280 this ETL moving around and around and around and around, all these batch windows 38 00:02:47.280 --> 00:02:51.280 being hit, and I thought to myself, this is crazy, like that. 39 00:02:51.680 --> 00:02:54.360 You're probably moving data that doesn't get used. You're probably moving the same 40 00:02:54.439 --> 00:02:59.960 data multiple times. You're probably sometimes overwriting good data with bad I mean, 41 00:03:00.000 --> 00:03:01.759 and there just must be so many things happening here on the data front. 42 00:03:02.000 --> 00:03:07.159 But business just demands things. The business demands data, It demands execution, 43 00:03:07.240 --> 00:03:09.599 it demands performance. Business people say we got to get our data in they're 44 00:03:09.639 --> 00:03:13.080 okay, fine, let's move it in there. Oh we didn't get the 45 00:03:13.120 --> 00:03:16.000 schema right, and then things changed, right like with snowflake, and you 46 00:03:16.039 --> 00:03:20.800 know a lot about Snowflake. Maybe we'll start there. One thing that I 47 00:03:20.840 --> 00:03:25.680 thought they really solved and really figured out was that schemas change because the world 48 00:03:25.800 --> 00:03:30.120 changes. So you want to be able to change the schema of your warehouse. 49 00:03:30.479 --> 00:03:34.240 Well, in the pre Snowflake days, that was a huge problem. 50 00:03:34.280 --> 00:03:36.879 I mean, you know, you'd be asking for lots of trouble, a 51 00:03:36.879 --> 00:03:39.800 lot of money, a lot of very unhappy engineers, like lots of bad 52 00:03:39.840 --> 00:03:44.240 things would happen, and Snowflake's like, now, let's make it easy to 53 00:03:44.319 --> 00:03:46.520 pull this thing down, change the schema in the spin it back up again. 54 00:03:47.280 --> 00:03:51.000 It was it your take as well. That was a very big change 55 00:03:51.039 --> 00:03:53.840 in the in the industry. Yeah, I mean the I mean Snowflake was 56 00:03:54.960 --> 00:03:59.400 that you had to use the standard phrase a game changer, certainly in the 57 00:03:59.479 --> 00:04:06.199 data warehouse world because of a number of the features that been voluntary architected into 58 00:04:06.400 --> 00:04:11.759 the platform right and you know what you're talking about there, which made a 59 00:04:11.840 --> 00:04:15.240 huge difference and still obviously does to a lot of people was their zero copy 60 00:04:15.240 --> 00:04:20.240 clone. The ability to take an existing schema and with one command clone the 61 00:04:20.399 --> 00:04:27.079 entire thing and basically make a complete replica of it with no additional storage costs. 62 00:04:27.120 --> 00:04:30.160 That was the thing, because that was something I couldn't do in the 63 00:04:30.240 --> 00:04:34.519 data warehousing world if I wanted to have a dev instance that had the same 64 00:04:34.560 --> 00:04:40.959 amount of data as my PROD instance. I never ever in my career was 65 00:04:41.000 --> 00:04:45.399 able to get that because the cost was too much and the time, you 66 00:04:45.399 --> 00:04:49.959 know, once you started getting into terabytes of data to replicate that on storage 67 00:04:50.000 --> 00:04:58.000 with the systems at the time, that could take days, weeks sometimes depending 68 00:04:58.000 --> 00:05:02.079 on the system, and you really couldn't get which meant you couldn't actually innovate 69 00:05:02.879 --> 00:05:09.279 with confidence that you weren't going to break something right right. You couldn't test 70 00:05:09.319 --> 00:05:11.519 it. It's like, oh, the query works great. Yeah, well 71 00:05:11.560 --> 00:05:15.639 you're got one tenth of the data in your QA instance as you do in 72 00:05:15.680 --> 00:05:17.399 production. And then it goes to production and the users are screaming, going, 73 00:05:18.160 --> 00:05:20.600 I can't live with this, this is this is too slow. And 74 00:05:20.639 --> 00:05:26.720 now you're doing, you know again, additional iterations and refactoring to figure out 75 00:05:26.720 --> 00:05:28.920 how can I tune it to make it better, how can I make it 76 00:05:28.959 --> 00:05:31.160 faster? And you know, the users are already mad at you because you 77 00:05:31.199 --> 00:05:33.879 gave them, you told them you had what they wanted, and they're like, 78 00:05:33.920 --> 00:05:38.959 well, this is unusable, it's too slow. And the zero copy 79 00:05:39.000 --> 00:05:42.560 cloning thing that was like, wow, this this is awesome. This means 80 00:05:42.600 --> 00:05:46.480 I can actually not shoot myself in the foot before I move something into production 81 00:05:46.879 --> 00:05:49.959 and have to do these cycles again, we could do real user acceptance testing 82 00:05:50.839 --> 00:05:57.079 without this exorbitant cost or you know, waiting weeks for the infrastructure team to 83 00:05:57.079 --> 00:06:01.439 put more disc online and for the DVAS to you know, restore something from 84 00:06:01.480 --> 00:06:05.079 backup, assuming the backup tape was actually good, right, you know that 85 00:06:05.160 --> 00:06:08.959 they could do all of that, So that that that really did change a 86 00:06:09.000 --> 00:06:14.160 lot for a lot of companies out there. Then, and of course the 87 00:06:14.199 --> 00:06:18.040 whole separation to compute from storage that allowed you to scale the storage if you 88 00:06:18.120 --> 00:06:23.399 needed without scaling the compute, and scale the compute without scaling the storage just 89 00:06:23.439 --> 00:06:27.720 made it so flexible. It just changed. It changed so many things. 90 00:06:28.480 --> 00:06:31.000 Then you throw in there the variant data type that allowed you to load Jason. 91 00:06:31.199 --> 00:06:35.000 So you talk about schema changes. You know, we actually got real 92 00:06:35.120 --> 00:06:41.839 schema on read load in a Jason document into a table in relational database with 93 00:06:41.879 --> 00:06:46.360 the table access it was SQL. And you know, as your data pipeline 94 00:06:46.399 --> 00:06:49.800 is running and you got new data coming in, if there's a little change 95 00:06:49.800 --> 00:06:53.360 in the scheme, it didn't break anything. I mean, I'm sure you 96 00:06:53.439 --> 00:06:56.920 remember the days that you were talking about ETL. It's like the source system 97 00:06:57.079 --> 00:07:01.680 changed. They dropped one column right, and didn't tell anybody because they're thinking 98 00:07:01.720 --> 00:07:05.480 about their operational system, they're not talking about the data warehouse. And then 99 00:07:05.519 --> 00:07:10.879 your informatica job blows up in the middle of the night and now everybody's screaming 100 00:07:10.920 --> 00:07:13.800 the next morning because the data hasn't been refreshed. You told us that'd be 101 00:07:13.920 --> 00:07:16.519 refreshed overnight and it's not been refreshed. What happened, Eh, you changed 102 00:07:16.519 --> 00:07:19.399 the source. They changed the source system. Didn't tell us. Now we 103 00:07:19.439 --> 00:07:25.079 got to go back and re engineer the ETL and change the tables and the 104 00:07:25.160 --> 00:07:29.279 database, and you know, so much of that went away in the last 105 00:07:29.639 --> 00:07:34.160 Well see, Snowflake was founded in twenty twelve, so last ten years. 106 00:07:34.199 --> 00:07:39.040 I started with them in twenty fifteen, so it's been less than a decade, 107 00:07:39.519 --> 00:07:43.319 and you know, they solved those really big problems that we had in 108 00:07:43.360 --> 00:07:47.079 data warehousing for multiple decades. Well, I mean you kind of hinted that 109 00:07:46.959 --> 00:07:51.639 the big downside of the old way, which is basically an opportunity costs around 110 00:07:51.720 --> 00:07:59.120 innovation because it was fragile, because it was difficult to work with, because 111 00:07:59.240 --> 00:08:03.240 changes were very hard, people didn't try to make changes. I mean, 112 00:08:03.600 --> 00:08:05.439 it's like a kid who keeps getting reprimanded for trying new things. They just 113 00:08:05.480 --> 00:08:09.040 stopped like, fine, I won't do anything right. So the psychology of 114 00:08:09.639 --> 00:08:16.000 data use really took a whole new turn, in large part thanks to Snowflake. 115 00:08:16.040 --> 00:08:18.319 Now, of course we have data bricks, we have a whole bunch 116 00:08:18.319 --> 00:08:20.920 of other options of ways to do things. But the point is that now 117 00:08:22.000 --> 00:08:26.000 there is more confidence to move forward to try changes to do things because you're 118 00:08:26.040 --> 00:08:31.160 not so worried about bringing the whole thing to a crashing halt. Right right, 119 00:08:31.360 --> 00:08:37.639 You're not creating instant technical debt by doing something right, and so you 120 00:08:37.679 --> 00:08:41.639 can take a more agile approach. You can be more responsive to the business 121 00:08:43.159 --> 00:08:46.200 and the business requirements when they want to change, with like yeah, without 122 00:08:46.240 --> 00:08:50.679 having to worry that, oh yeah, we're gonna break it. And then 123 00:08:50.720 --> 00:08:54.080 the tools have evolved so much too, between the data catalogs and the data 124 00:08:54.120 --> 00:08:58.679 engineering tools. You know, the thing that we always needed was really good 125 00:08:58.720 --> 00:09:03.320 ima analysis reports. But to get the impact analysis reports, you needed something 126 00:09:03.360 --> 00:09:09.039 tracking the metadata around data lineage. You had to know where did that data 127 00:09:09.080 --> 00:09:11.440 come from and where did it go? We pulled a data element in from 128 00:09:11.480 --> 00:09:16.080 one source to a system, how many places did it go And if we 129 00:09:16.200 --> 00:09:18.519 changed that, how many things is it going to impact? Well, you've 130 00:09:18.519 --> 00:09:22.279 got to have the lineage to do the impact analysis. And now we've got 131 00:09:22.399 --> 00:09:26.519 a lot more tools in er what we call our modern data stack that allow 132 00:09:26.600 --> 00:09:30.159 us to do that so much easier, so much quicker. You're no longer 133 00:09:30.200 --> 00:09:35.919 relying on a couple of data engineers to know the code inside and out. 134 00:09:35.960 --> 00:09:37.279 And go, oh, yeah, that's right. I remember over in that 135 00:09:37.919 --> 00:09:41.759 routine we did this, and we did this based on the user requirements. 136 00:09:41.799 --> 00:09:45.600 But then they had us do this in a different routine and yeah, if 137 00:09:45.600 --> 00:09:48.720 we touch that, we're going to break three different things. You don't have 138 00:09:48.759 --> 00:09:52.440 to do that anymore. You can find it through the metadata. Yeah. 139 00:09:52.519 --> 00:09:56.279 Well, and you hinted it something else tonight. Throughout the largest language models 140 00:09:56.320 --> 00:10:01.559 to start us off here, and you think about the impact of these engines 141 00:10:01.639 --> 00:10:05.000 now, I think there were numerous other shoes left to drop in this equation, 142 00:10:05.720 --> 00:10:09.120 just one way of putting it, basically, because they are very powerful. 143 00:10:09.279 --> 00:10:13.200 But you know, Sam Altman had that very curious comment he made a 144 00:10:13.279 --> 00:10:15.799 couple of months ago, say, oh, the era of large language models 145 00:10:15.879 --> 00:10:20.639 is over. Everyone was like, what, what are you talking about I 146 00:10:20.720 --> 00:10:22.840 think, No, I don't know. This is just my personal theory. 147 00:10:22.879 --> 00:10:28.360 But we've now seen that they've had some challenges people, And this is research 148 00:10:28.399 --> 00:10:30.799 that I've done and people have talked to have said, there are things that 149 00:10:30.840 --> 00:10:33.120 it did very well three months ago, four months ago that now has a 150 00:10:33.159 --> 00:10:37.240 hard time doing. A good buddy of mine, who actually almost worked with 151 00:10:37.320 --> 00:10:41.600 open Ai, said that he looked at their models three years ago or so 152 00:10:41.720 --> 00:10:46.279 and said, you're gonna have all kinds of problems because you're not respecting entropy. 153 00:10:46.679 --> 00:10:48.399 He said, you're gonna have short term memory problems, long term memory 154 00:10:48.399 --> 00:10:52.240 problems. Basically, he blew it up at them, and they just said, 155 00:10:52.240 --> 00:10:54.320 all right, we'll get out of here. We're going to do it 156 00:10:54.320 --> 00:10:56.399 our way. We'll deal with that later. Yeah, we'll do exactly, 157 00:10:56.440 --> 00:11:00.639 We'll figure that out later. And of course they rolled out and huge. 158 00:11:01.039 --> 00:11:03.799 I mean, you know, the friends I have at Google, I can 159 00:11:03.840 --> 00:11:07.279 tell they are stressed out. The company is very stressed out right now, 160 00:11:07.279 --> 00:11:09.120 which is kind of hard to believe because it's Google, Like, I mean, 161 00:11:09.159 --> 00:11:13.159 goodness, you're one of the biggest companies in the world. You change 162 00:11:13.240 --> 00:11:16.960 the world with the stuff that you did, and of course what happened. 163 00:11:16.159 --> 00:11:20.720 Microsoft goes throws a whole bunch of money at open AI, then the opening 164 00:11:20.759 --> 00:11:26.080 eye board curiously kicks him out, and then such Inno Della hires him, 165 00:11:26.159 --> 00:11:30.399 and like, wow, that was one of the biggest tennis matches I've ever 166 00:11:30.399 --> 00:11:35.480 seen. Like what is going back and forth? Point being? And where 167 00:11:35.480 --> 00:11:39.960 I'm kind of going with all this is, even though there are some challenges, 168 00:11:39.000 --> 00:11:43.600 they're so powerful. You can load code and ask the machine to tell 169 00:11:43.639 --> 00:11:48.440 you what the code means. You can load a whole bunch of documents and 170 00:11:48.480 --> 00:11:50.320 ask it to summarize. So I think those are the two better use cases. 171 00:11:50.759 --> 00:11:54.840 But kind of where I'm going is I see these llms, the engines 172 00:11:54.879 --> 00:11:58.159 of them at least, and maybe it's SLM small language models to become more 173 00:11:58.200 --> 00:12:05.679 prominent, really serving as a very powerful component in a workflow. So they're 174 00:12:05.720 --> 00:12:09.279 good at ETL. For example, I scraped a whole bunch of data off 175 00:12:09.360 --> 00:12:13.720 the CEES website because I wanted to get a clean list of companies from that 176 00:12:13.879 --> 00:12:16.960 site. So I scraped a whole I did a just command a command C 177 00:12:16.600 --> 00:12:20.360 command V into a Google doc and then I pasted that in the chat into 178 00:12:22.159 --> 00:12:24.559 Gemini was barred back then, and I said, please give me a clean 179 00:12:24.639 --> 00:12:28.200 list of just the company names and all this text, and there's all this 180 00:12:28.279 --> 00:12:31.919 other text inside. It was like, okay, that just banged it out. 181 00:12:31.960 --> 00:12:37.080 I'm like, all right, that's pretty impressive. You know that it 182 00:12:37.120 --> 00:12:41.120 can take an instruction like that ascertain I only want company names, know what 183 00:12:41.159 --> 00:12:46.759 those are in the context, and then bang out a whole list, And 184 00:12:46.799 --> 00:12:48.679 then I asked that to give me a Twitter handle for each one. It 185 00:12:48.720 --> 00:12:52.639 did that too, but half of them were bad. So it does make 186 00:12:52.679 --> 00:12:56.360 stuff up. But nonetheless, I mean, think about if you'd seen this 187 00:12:56.440 --> 00:12:58.559 blob of text, you know what I'm talking about, just all kinds of 188 00:12:58.600 --> 00:13:01.840 other words in between places and tabs and commas, and is also kind of 189 00:13:01.840 --> 00:13:07.600 crap. I had no problem just stripping all that stuff out. That's a 190 00:13:07.720 --> 00:13:11.320 very powerful function, which would have taken a good ETL developer. I don't 191 00:13:11.320 --> 00:13:13.720 even know how long to sort through and to figure out it would be hard 192 00:13:13.720 --> 00:13:18.200 to do. Not for the llms. So what kind of impact do you 193 00:13:18.200 --> 00:13:22.559 think they're going to have on the data management business? Writ large Well, 194 00:13:22.080 --> 00:13:28.879 I think that what we're seeing is the next step in data democratization right by 195 00:13:30.240 --> 00:13:33.799 wee can now with even what you described, you basically you took the data 196 00:13:33.879 --> 00:13:37.679 engineer out of the loop, right, You were able to do something very 197 00:13:37.759 --> 00:13:41.799 quickly on your own. And I think that's part of the promise and the 198 00:13:41.840 --> 00:13:48.320 power of these things is to allow business analysts now to do things with data 199 00:13:48.720 --> 00:13:52.480 without having to decoders. Right, they don't even necessarily need to know SQL. 200 00:13:52.720 --> 00:13:56.480 In one of the recent demos I saw, I think it's called Cortex 201 00:13:58.559 --> 00:14:01.360 when the new sale featured, right, is like you can you can ask 202 00:14:01.399 --> 00:14:05.320 a question say I want I want the uh, you know, the average 203 00:14:05.360 --> 00:14:13.320 sales over the last three years for my top five regions, and it writes 204 00:14:13.320 --> 00:14:18.159 the sequel for you, right, And you know that's starting need to come 205 00:14:18.200 --> 00:14:22.200 out. And I've seen a couple of demos like that recently from the data 206 00:14:22.200 --> 00:14:26.320 analytics perspective that now again you don't you don't have to know SQL. Uh, 207 00:14:26.639 --> 00:14:31.639 you don't even necessarily have to understand the schema of the database in order 208 00:14:31.720 --> 00:14:37.080 to get some basic analytics done using an ll m U, using a you 209 00:14:37.120 --> 00:14:41.840 know it's prompt. And my son is in school now and he he mentioned 210 00:14:41.840 --> 00:14:48.240 something about one of the careers that they've talked about in the career education stuff 211 00:14:48.240 --> 00:14:52.279 that he's going through at college was a thing called prompt engineering, right that 212 00:14:52.720 --> 00:14:56.360 you could you could take classes in prompt engineering. It's like, well, 213 00:14:56.360 --> 00:15:00.919 what does that mean? That's okay? How to ask the question? So 214 00:15:01.080 --> 00:15:05.000 I think we used to call that critical thinking, right learning how to ask 215 00:15:05.039 --> 00:15:09.799 the question, But now you're going to learn how to ask the question of 216 00:15:09.960 --> 00:15:15.559 an AI basically right, and how you ask the question will have an effect 217 00:15:15.639 --> 00:15:20.320 on the outcome because again, the AI the LM can't read your mind and 218 00:15:20.360 --> 00:15:26.039 it can only interpret so much. So being able to ask very clear, 219 00:15:26.200 --> 00:15:33.840 concise questions using language that the AI recognizes is going to be the skill, 220 00:15:33.240 --> 00:15:37.120 right, That's going to be the skill rather than knowing how to write, 221 00:15:37.559 --> 00:15:43.279 you know, select some count star group by right in order to get it 222 00:15:43.000 --> 00:15:48.600 in order to get the answer, which is so huge because and this is 223 00:15:48.679 --> 00:15:50.919 leading up to maybe segment two. I'm going to give you one of my 224 00:15:50.960 --> 00:15:56.320 big ideas and sear to think about it. But because now you've democratized access 225 00:15:56.399 --> 00:16:00.480 to analytics, is what you've done. Basically, you had to know some 226 00:16:00.519 --> 00:16:03.320 sqel or have a good idea at least how to use the technologies to be 227 00:16:03.360 --> 00:16:07.960 able to build the queries. Now you can just ask it a question. 228 00:16:07.200 --> 00:16:11.519 Now again, it does make things up, so you have to be careful. 229 00:16:11.519 --> 00:16:12.600 You have to vet things. But we've always had to vet things. 230 00:16:12.639 --> 00:16:15.480 I mean, there are times when the data engineering was done wrong. So 231 00:16:15.799 --> 00:16:19.120 it's not like humans didn't make mistakes, and now the AI is creating some 232 00:16:19.440 --> 00:16:22.639 new function called a mistake. No, like, it makes stuff up, 233 00:16:22.639 --> 00:16:26.759 and it has some problems, but the fact that you can sit there and 234 00:16:26.799 --> 00:16:32.200 just have such an interactive discourse with the data, to me is incredibly game 235 00:16:32.320 --> 00:16:37.320 changing. And I think it's going to put even more excitement in the workplace 236 00:16:37.399 --> 00:16:41.759 to use the data because a lot of times people don't use the technologies because 237 00:16:41.799 --> 00:16:45.960 they're hard to use, because they're slow, because they don't trust the data 238 00:16:47.000 --> 00:16:49.080 for example, and we do have the trust issue. But this stuff is 239 00:16:49.120 --> 00:16:52.360 not hard to use. I mean, it can be hard to get really 240 00:16:52.399 --> 00:16:56.320 good at your prompt engineering, but even that, just to ask a different 241 00:16:56.360 --> 00:17:00.080 question, it doesn't get tired, it doesn't get annoyed at you. I 242 00:17:00.120 --> 00:17:03.000 will admit I tried Groc a little bit. I thought Grock was kind of 243 00:17:03.039 --> 00:17:06.519 annoying. That's the one that's in Twitter these days. It's supposed to be 244 00:17:06.640 --> 00:17:10.599 sarcastic and goofy, And I'm like, is there a big market for an 245 00:17:10.680 --> 00:17:14.319 engine that gives me sarcastic answers? I just don't think so. I think, 246 00:17:14.680 --> 00:17:17.480 you know, maybe Elon smoked a couple of dubies before that one came 247 00:17:17.519 --> 00:17:19.279 out, which is fine. I mean, the guy's done amazing things. 248 00:17:19.440 --> 00:17:25.000 He's an engineer himself, and you look at the accomplishments this guy's had. 249 00:17:25.000 --> 00:17:27.519 It's just absolutely shocking. But you do have competition out there. And I 250 00:17:27.519 --> 00:17:33.720 guess what I'm driving at is by democratizing access to systems of record with these 251 00:17:34.039 --> 00:17:38.599 llms, we are fundamentally changing the game about how people interact with data, 252 00:17:38.799 --> 00:17:42.359 how they're able to consume data learn from data. I'm writing an abstract right 253 00:17:42.359 --> 00:17:45.160 now, in fact, for the show I'll do with you on Thursday, 254 00:17:47.240 --> 00:17:52.319 all about this concept of data literacy. And I think that just playing around 255 00:17:52.400 --> 00:17:56.880 with these tools connected to your data sources, you are fostering better data literacy 256 00:17:56.960 --> 00:18:00.759 because you're learning things about the data and you don't have to take a class 257 00:18:00.839 --> 00:18:04.079 necessarily, just have to dedicate the time to hop online, play around, 258 00:18:04.319 --> 00:18:07.720 click a few buttons, type into some questions and see the different answers, 259 00:18:07.880 --> 00:18:11.720 and you have to work with it to see how it operates and to kind 260 00:18:11.720 --> 00:18:15.240 of know sort of the contours of what it does. But folks, don't 261 00:18:15.279 --> 00:18:18.079 touch out that. I'll be right back in a moment. We're talking to 262 00:18:18.920 --> 00:18:30.039 a hero of data. Ken Graziano will be right back. Welcome back to 263 00:18:30.200 --> 00:18:37.559 Inside Analysis. Here's your host and Eric Tavanaugh. All right, folks, 264 00:18:37.559 --> 00:18:41.920 back here on Inside Analysis with Kent Graziano, who has been a data avenger 265 00:18:41.000 --> 00:18:44.799 for years. Quite frankly, you've been out there on the front line. 266 00:18:44.960 --> 00:18:48.559 I remember I met you at the Data Vault conference a number of years ago. 267 00:18:48.640 --> 00:18:52.000 That was fun. That was before the big COVID, And you know, 268 00:18:52.200 --> 00:18:53.680 COVID gave us a lot of time to think about stuff, for sure, 269 00:18:53.720 --> 00:18:56.720 and to sort of reevaluate what was going on. I'm sure there's a 270 00:18:56.720 --> 00:19:00.559 bit of a COVID surge in terms of innovation because people are time to think 271 00:19:00.559 --> 00:19:03.240 about stuff and do cool new things. And with that in mind, I'm 272 00:19:03.279 --> 00:19:07.839 going to throw my big idea over to the sage graz Siattle. I want 273 00:19:07.839 --> 00:19:11.839 to get your thoughts on this, so I keep thinking about how executives use 274 00:19:11.920 --> 00:19:17.079 data and how we've used data historically, and if there are these long processes 275 00:19:17.160 --> 00:19:21.599 of building reports and doing all these different things, I think the llm's kind 276 00:19:21.599 --> 00:19:23.680 of turn all that on its head. And if you do it right, 277 00:19:25.200 --> 00:19:26.720 I think what's going to happen is companies are going to get their own idea, 278 00:19:26.759 --> 00:19:30.440 to get their own private model. Maybe it's a small language model for 279 00:19:30.480 --> 00:19:34.759 a particular industry like legal or healthcare or retail, because they have their own 280 00:19:34.839 --> 00:19:40.599 lexicon, and so it's good to kind of weed out the semantic issues that 281 00:19:40.680 --> 00:19:42.200 you'll find with a large language model. That's kind of the point of the 282 00:19:42.240 --> 00:19:45.400 small language models. So you get that, and then you train, you 283 00:19:45.599 --> 00:19:51.000 get your vector database. You basically take your curated, trusted data and you 284 00:19:51.039 --> 00:19:53.880 start feeding it into the vector database as you're embeddings. These are your anchors 285 00:19:53.880 --> 00:19:59.000 of truth. And what I think is going to happen is this is sort 286 00:19:59.000 --> 00:20:02.720 of the Valhalla, if you will, is that you get enough of your 287 00:20:03.039 --> 00:20:07.480 corporate data in your embeddings. You've got your anchors of truth now, and 288 00:20:07.519 --> 00:20:11.519 then you set up whether it's COFKA topics or some kind of stream of changes 289 00:20:11.640 --> 00:20:18.000 coming through like CDC, basically into the vector database and what's going to happen 290 00:20:18.119 --> 00:20:22.960 is the executives are just going to sit there and have this amazing intern with 291 00:20:22.039 --> 00:20:26.599 tremendous amounts of knowledge at their fingertips. They'll be able to ask any number 292 00:20:26.599 --> 00:20:30.079 of questions, you know, how how are we doing this month? What 293 00:20:30.119 --> 00:20:33.680 can I do to change? Who's really excelling in my sales team? Like 294 00:20:33.759 --> 00:20:37.160 all these kind of questions that you can ask and just get information pouring into 295 00:20:37.200 --> 00:20:41.119 them. And it's going to be very useful because it's connected to your data 296 00:20:41.119 --> 00:20:44.799 warehouse, to your CRM system, to your sales sporce, to your clickstream 297 00:20:44.839 --> 00:20:48.680 analysis, whatever it is that you've got, that senior executive is tapped into 298 00:20:48.720 --> 00:20:52.799 it and can now ask any question he or she wants. How viable is 299 00:20:52.839 --> 00:20:56.559 that? Do you think? And is that the future? Yeah? I 300 00:20:56.640 --> 00:21:00.839 think you've hit on something there, because really this is, you know, 301 00:21:00.880 --> 00:21:07.240 the extension evolution. Maybe it's the the fruition of all the ideas that we 302 00:21:07.319 --> 00:21:15.519 had around data warehousing and decision support systems and advanced analytics and business intelligence in 303 00:21:15.599 --> 00:21:21.920 general. Right that to have that at the fingertips of an executive, that's 304 00:21:22.160 --> 00:21:23.599 that's where the value comes in, Right, is you've got to be able 305 00:21:23.640 --> 00:21:29.759 to use the data to make effective business decisions. And I think that what 306 00:21:29.839 --> 00:21:38.200 you're describing could be the differentiator that allows an organization to get their competitive advantage. 307 00:21:38.519 --> 00:21:45.240 And specifically when they're looking at their internal data right and training a model 308 00:21:45.319 --> 00:21:48.759 on their data, if they do that right, that's the thing that's going 309 00:21:48.799 --> 00:21:52.720 to give them the competitive advantage. Because so many things have been commoditized, 310 00:21:52.799 --> 00:21:56.279 right, and we've got lots of data out there that's being shared via data 311 00:21:56.279 --> 00:21:57.720 marketplaces, and sure, you're probably going to want to pull some of that 312 00:21:57.839 --> 00:22:03.359 in. But it's the combination of that external data with your internal data, 313 00:22:03.440 --> 00:22:07.079 your proprietary data, that's the thing that's got to give you the competitive advantage. 314 00:22:07.279 --> 00:22:12.839 And if you've got like you said this uh basically online on demand uh 315 00:22:14.000 --> 00:22:22.599 executive intern data intern that's putting all this together automatically right through through a small 316 00:22:22.680 --> 00:22:26.039 language model that the executive got a question they just you know, like like 317 00:22:26.079 --> 00:22:29.759 Alexa, Well they'll probably give it. They'll have to give it their own 318 00:22:29.799 --> 00:22:34.759 name and say, hey, uh, I heard from my sales guy that 319 00:22:34.759 --> 00:22:40.960 we're having a problem in the western region. Give me give me a summary 320 00:22:41.000 --> 00:22:45.000 of what happened in sales in the last three days in the West Region and 321 00:22:45.359 --> 00:22:49.279 boom, it's like and eventually it should be able to give us. You 322 00:22:49.279 --> 00:22:52.359 know, well, what's the implication of that, Oh, the trend. 323 00:22:52.400 --> 00:22:56.319 The trend is if the sales, if what's happening in sales in the West 324 00:22:56.319 --> 00:23:02.200 Region continues for X number of more weeks, you're going to lose this much 325 00:23:02.200 --> 00:23:07.440 profit, right, right, These are the kinds of questions that we built 326 00:23:07.480 --> 00:23:10.720 data warehouses to answer. Yes, it's kind of where I'm going with this. 327 00:23:10.920 --> 00:23:14.640 Absolutely, I don't I don't think that data warehouse is going to go 328 00:23:14.680 --> 00:23:18.599 away, but it's to me, it has to get simpler, I think. 329 00:23:18.839 --> 00:23:22.279 And you know, so you look at the marketplace. What happened. 330 00:23:22.680 --> 00:23:26.119 Data Bricks went out and bought Mosaic mL for like one point four billion or 331 00:23:26.119 --> 00:23:29.599 something like that when this was just taking off, and Terra Data and some 332 00:23:29.640 --> 00:23:32.799 of the others were still kind of laughing at the jen Ai stuff, which 333 00:23:32.839 --> 00:23:37.200 I thought was a bit foolish frankly, But those guys were laughing at the 334 00:23:37.200 --> 00:23:41.000 cloud originally too, if you remember, Yeah, well that's wrong, that's 335 00:23:41.000 --> 00:23:45.759 pretty funny. It didn't work either, right, Well, I don't know, 336 00:23:45.839 --> 00:23:48.920 man, I just I keep thinking about the disruptive power of these engines. 337 00:23:48.960 --> 00:23:52.880 And again, yes, they get some stuff wrong. So do the 338 00:23:52.920 --> 00:23:56.039 old systems. The old systems got stuff wrong because their data was wrong or 339 00:23:56.079 --> 00:23:59.440 the model was wrong, and it takes even longer to find it and fix 340 00:23:59.480 --> 00:24:03.359 it in the old systems. Well, that's a different other thing, right, 341 00:24:03.400 --> 00:24:06.880 that's the other thing. So it's like, here's what fascinates me is 342 00:24:07.400 --> 00:24:11.839 trying to figure out how do these models figure out what to choose when they 343 00:24:11.920 --> 00:24:15.359 give you their answer. Now, kindy that got bought by click I took 344 00:24:15.359 --> 00:24:21.160 a demo of their engine. It's pretty interesting because it connects to an LLM 345 00:24:21.559 --> 00:24:26.920 and basically you use it for like frequently asked questions for your manual for some 346 00:24:26.039 --> 00:24:30.319 new product that you bought, like the new iPhone or something, and you 347 00:24:30.440 --> 00:24:33.880 load the whole thing in there. It will automatically give you frequently asked questions 348 00:24:33.920 --> 00:24:37.799 that it suggests and the answers, and you curate that and then when someone 349 00:24:37.920 --> 00:24:40.759 uses it, you'll get the answer you asked. But they'll give you a 350 00:24:40.839 --> 00:24:44.720 drop down that'll show you where it got those bits from, So it'll say 351 00:24:44.799 --> 00:24:47.799 the first sentence came from page three, paragraph two. The other one came 352 00:24:47.839 --> 00:24:51.599 from here. I'm like, now that is compelling because now you're getting no 353 00:24:51.599 --> 00:24:56.039 annotation, right, that's the annotation that we were taught in our language. 354 00:24:56.079 --> 00:25:00.599 Our language classes, right in our English classes are learning how to do research 355 00:25:00.640 --> 00:25:03.480 papers. It's like, don't put anything in there that you can't prove where 356 00:25:03.519 --> 00:25:07.240 it came from, right, And it's and whether there's endnotes or footnotes, 357 00:25:07.559 --> 00:25:11.359 all those annotations are there, and that's yeah, that's part of what we 358 00:25:11.400 --> 00:25:17.200 need. That's I guess the part of the QA on these things is can 359 00:25:17.240 --> 00:25:19.599 we can we see where this data came from rather than it even though it 360 00:25:19.640 --> 00:25:25.440 might be a black box that generated the result, it's still traceable to the 361 00:25:25.480 --> 00:25:29.319 source and say, these are the references, this is where these numbers came 362 00:25:29.319 --> 00:25:32.759 from, This is where this concept came from, this is this is how 363 00:25:32.799 --> 00:25:37.759 we're coming up with this particular recommendation. And I think that's critical. That 364 00:25:37.759 --> 00:25:42.480 that is critical, you know, to differentiate between a hallucination and a good 365 00:25:42.519 --> 00:25:47.079 answer that you actually want to run your business off of. Right. Well, 366 00:25:47.119 --> 00:25:48.759 and that's the thing, right, is that to run your business off 367 00:25:48.799 --> 00:25:52.319 these things, you're going to have to have some certitude about what they're telling 368 00:25:52.359 --> 00:25:56.200 you. So that is a concern, But I just I feel like there's 369 00:25:56.240 --> 00:26:00.720 going to be downward pressure on pricing data warehousing. But at the same time, 370 00:26:02.000 --> 00:26:04.480 you've got so much automation now, it's so much easier to do things 371 00:26:04.519 --> 00:26:07.200 than it was in the old days. I mean, I guess that one 372 00:26:07.240 --> 00:26:11.240 of the biggest hurdles to overcome here is just old fashioned mindsets. What do 373 00:26:11.319 --> 00:26:15.920 you think? Absolutely? Oh yeah, I mean that's the Even when I 374 00:26:15.960 --> 00:26:22.839 started off with Snowflake, the most frequent question or might actually comment I got 375 00:26:22.880 --> 00:26:25.920 when I was out, you know, being the evangelist and talking to people 376 00:26:25.960 --> 00:26:29.400 about the separation compute from storage and zero copy cloning and all the things that 377 00:26:29.440 --> 00:26:32.039 we talked about. And the first thing is like, oh, that'd be 378 00:26:32.160 --> 00:26:37.039 that'd be awesome if it was true. And then they say, okay, 379 00:26:37.039 --> 00:26:38.720 well, well how do we how do we index our queries? It's like 380 00:26:38.720 --> 00:26:41.960 you don't, Well, then this thing can't work. It can't work. 381 00:26:42.200 --> 00:26:45.839 There's no way it can work because you don't have Indexes's like, no, 382 00:26:47.000 --> 00:26:52.240 this is a completely different architecture. And it was getting that that mindset change. 383 00:26:52.519 --> 00:26:56.799 Same thing when we went from Waterfall to agile, right, trying to 384 00:26:56.839 --> 00:27:00.119 think about how do we do a project in iterations, versus spending six months 385 00:27:00.119 --> 00:27:04.000 writing up the requirements document, getting everybody signed off, and then spending two 386 00:27:04.079 --> 00:27:07.480 years building it, or spending a year and a half doing an enterprise data 387 00:27:07.519 --> 00:27:11.200 model and then having to hand it to a DBA to convert it to a 388 00:27:11.240 --> 00:27:15.440 schema in a database, and three years later you've got a data warehouse that 389 00:27:15.440 --> 00:27:21.880 has nothing in it because nobody cares anymore. Right, we had to change 390 00:27:21.880 --> 00:27:26.079 the way people were thinking. I think what we're talking about here, it's 391 00:27:26.119 --> 00:27:29.920 the same thing. It's the you know, we've talked over the last couple 392 00:27:29.920 --> 00:27:33.240 of years about companies about data literacy and data culture, and it really is 393 00:27:33.279 --> 00:27:41.319 the organizational culture and expectations that you've got to find a way to shift those 394 00:27:41.839 --> 00:27:47.200 right and get out of the well the classic and this is this is every 395 00:27:47.200 --> 00:27:49.839 industry. It's not just it, it is every industry. Well, we've 396 00:27:49.880 --> 00:27:53.599 never done it that way before, or the reverse, Well, we've always 397 00:27:53.640 --> 00:27:57.920 done it this way, and I'm comfortable doing it this way, right, 398 00:27:59.000 --> 00:28:02.920 so I want to just keep going this way. I actually, early on 399 00:28:03.160 --> 00:28:11.119 Snowflake, we worked with a large retailer in England and they after their evaluation, 400 00:28:11.279 --> 00:28:17.559 decided to switch from one of the existing big data warehouse companies to Snowflake, 401 00:28:18.440 --> 00:28:22.640 and several DBAs literally resigned and went and found other jobs. And this 402 00:28:22.720 --> 00:28:26.119 is in Europe, where you know, people don't change jobs hardly at all. 403 00:28:26.559 --> 00:28:30.960 Because they didn't want to spend the last five years of their career learning 404 00:28:32.000 --> 00:28:34.480 Snowflake. They wanted to stay with what they were comfortable with, which was 405 00:28:34.880 --> 00:28:41.000 the older system. And they found another company in London that needed a DBA 406 00:28:41.240 --> 00:28:44.599 that had their expertise, and they figured they were just they were going to 407 00:28:44.640 --> 00:28:48.200 go over there and sail off into the sunset, that that was going to 408 00:28:48.200 --> 00:28:53.559 be less stressful for them than to stay where they were and learn this really 409 00:28:53.599 --> 00:29:00.079 exciting new technology that they literally left the organization rather than change the way they 410 00:29:00.079 --> 00:29:06.000 were doing things and thinking about things differently. Yeah, that's not uncommon. 411 00:29:06.079 --> 00:29:08.559 I was. I was floored. I was like, you do they did? 412 00:29:08.599 --> 00:29:15.640 What? Okay, I'm done, I'm out, No, thank you. 413 00:29:15.640 --> 00:29:18.559 That's uh, that's wild, I mean, and that that's I feel 414 00:29:18.559 --> 00:29:22.799 like, that's where we are right now because of these lllms, I mean, 415 00:29:22.960 --> 00:29:26.680 so much is being shaken at the moment, and people stuff to do 416 00:29:26.680 --> 00:29:30.480 their jobs. These have to get stuff done. I mean, I've played 417 00:29:30.519 --> 00:29:34.480 around with these things and I'm really amazed at what they're able to capture. 418 00:29:34.519 --> 00:29:37.519 Even though they do hallucinate, even though they do make things up. It's 419 00:29:37.559 --> 00:29:41.880 really impressive what they can find. And you just have to sit there and 420 00:29:41.920 --> 00:29:45.440 wonder, Wow, all of this stuff is embedded in these models, and 421 00:29:45.519 --> 00:29:48.400 it's just deep in there. It's gonna take us years to find out what's 422 00:29:48.440 --> 00:29:52.119 in there because there's so much stuff in there already, right, I mean, 423 00:29:52.200 --> 00:29:56.960 like, how do you even how do you begin? Yeah, and 424 00:29:56.480 --> 00:30:00.480 like everything is, you got to start one foot in front of the other. 425 00:30:00.319 --> 00:30:03.920 But you've got to be willing to try. And I think that's that's 426 00:30:03.960 --> 00:30:07.720 the big that's the big lesson in all of this, and certainly, you 427 00:30:07.720 --> 00:30:11.400 know, even the last ten years with the you know, going from a 428 00:30:11.559 --> 00:30:18.599 dupe to cloud data warehouses like Snowflake and then data bricks, is you've got 429 00:30:18.640 --> 00:30:23.160 to be willing to take a look at these new technologies and think about it 430 00:30:23.200 --> 00:30:27.160 critically. Is like how can I take advantage of this technology rather than go, 431 00:30:27.200 --> 00:30:30.440 oh my god, I'm going to lose my job, right, They're 432 00:30:30.480 --> 00:30:32.599 not going to need me anymore. It's like, well, yeah, they're 433 00:30:32.599 --> 00:30:34.119 not going to need you anymore if you don't learn how to do new things, 434 00:30:34.559 --> 00:30:37.359 if you don't learn how to use these new technologies, you know, 435 00:30:37.359 --> 00:30:42.400 figure out how how to apply yourself to the new technology with the knowledge that 436 00:30:42.440 --> 00:30:47.319 you already have, you know, whether it's you know, in the data 437 00:30:47.319 --> 00:30:51.559 world, we're talking about data management. You understand things like schemas and SQL 438 00:30:51.599 --> 00:30:55.079 and all that. Well, how can you apply that in the ll M 439 00:30:55.200 --> 00:31:00.240 world and the SLM world? You know, how can you be of value 440 00:31:00.640 --> 00:31:08.400 to your organization helping them make these transitions to the newer technologies for the advantage 441 00:31:08.400 --> 00:31:12.599 of the organization. Yeah, that's very interesting. I mean you mentioned earlier 442 00:31:12.599 --> 00:31:18.400 you schema on read, and now with these llms, it's almost it's almost 443 00:31:18.440 --> 00:31:21.680 like you can do that to a certain extent, right and schema, but 444 00:31:22.000 --> 00:31:25.279 not in the way that any of those sequel people ever dreamed of, right, 445 00:31:25.440 --> 00:31:30.640 right, So that we now have an a an AI that can read 446 00:31:30.720 --> 00:31:34.759 the schema right right, that that somebody doesn't have to look at the er 447 00:31:34.960 --> 00:31:41.559 diagram and know how to write the joints. Now, the downside, you're 448 00:31:41.640 --> 00:31:45.480 I guess the caveat and all of this and you mentioned it kind of in 449 00:31:45.839 --> 00:31:48.759 one of your earlier statements about having your data warehouses and all that is. 450 00:31:48.839 --> 00:32:00.440 This presumes you've got a good, a well structured data architecture, that there's 451 00:32:00.599 --> 00:32:04.559 metadata at least that tells you how the data is related and what the data 452 00:32:04.720 --> 00:32:07.480 means. The semantics of the data you talked about, you know, the 453 00:32:07.920 --> 00:32:13.319 you know, having a lexicon of say a law firm. You know you 454 00:32:13.440 --> 00:32:17.000 got that terminology there, well, you know the underlying data. There's got 455 00:32:17.039 --> 00:32:23.640 to be a relationship there between that legal legal ease and the data in order 456 00:32:23.720 --> 00:32:28.759 to be able to ask a question and create a prompt that will get you 457 00:32:28.839 --> 00:32:32.440 the answer you want. And so it presumes it it's your data structure that 458 00:32:32.559 --> 00:32:37.359 is documented and that it's high quality, right, because the the ll M 459 00:32:37.599 --> 00:32:43.200 is not going to discriminate between a bad row of data and good row of 460 00:32:43.279 --> 00:32:46.480 data unless you've somehow programmed it to say, you know, ignore all the 461 00:32:46.599 --> 00:32:52.000 roles where data is null. Right, So you're you've put a data quality 462 00:32:52.079 --> 00:32:54.559 check into the prompt itself. Right. I know I don't want to do 463 00:32:54.599 --> 00:33:00.440 analysis on things where the sales date is blank, right, because obviously I've 464 00:33:00.440 --> 00:33:05.319 got a problem. But that you would have to know enough about data quality 465 00:33:06.000 --> 00:33:09.160 and know enough about the data to ask a question that way. So otherwise 466 00:33:09.200 --> 00:33:14.880 you got it somewhere. Somebody has to still do you know the hard part 467 00:33:14.960 --> 00:33:20.839 that we've always done there with with building that data platform and ingesting the data 468 00:33:21.079 --> 00:33:27.799 and curating the data to get you use that word curating curated data that we're 469 00:33:27.839 --> 00:33:30.079 going to use to feed all these things, right, And I think that's 470 00:33:30.119 --> 00:33:36.119 going to be that the next big push is figuring out these pipelines to feed 471 00:33:36.160 --> 00:33:38.920 the l ms. Do you have to get your your sort of foundation in 472 00:33:39.000 --> 00:33:43.119 place, but then it's going to be updating these things and then monitoring over 473 00:33:43.160 --> 00:33:45.160 time. But folks, don't touch out. That'll be right back. You're 474 00:33:45.160 --> 00:33:55.519 listening to Inside Analysis. Welcome back to Inside Analysis. Here's your host, 475 00:33:57.039 --> 00:34:01.640 Eric tabnaugh show. All right, folks, back here on Inside Analysis talking 476 00:34:01.640 --> 00:34:05.319 to the one, the only. Kent Gratziano is going to be at Data 477 00:34:05.440 --> 00:34:08.840 Universe April tenth and eleventh in New York City the Javit Center. Don't miss 478 00:34:08.840 --> 00:34:12.880 it, folks, It's going to be fun. Yours truly will be there 479 00:34:13.239 --> 00:34:15.159 and I'll throw one of my other curve balls at you. Can't just because 480 00:34:15.320 --> 00:34:17.760 you're such a good guy and I know you can hit good curve balls. 481 00:34:19.320 --> 00:34:22.440 What I'm going to talk about one of my talks is the death of journalism, 482 00:34:22.559 --> 00:34:25.679 as I call it, And what I'm really referring to is the fact 483 00:34:25.679 --> 00:34:32.079 that there's no media company who can stand up to these large language models and 484 00:34:32.119 --> 00:34:38.519 the power of this AI in terms of personalization, in terms of covering everything 485 00:34:38.559 --> 00:34:43.400 that could be interesting to someone. And I think what's going to happen here 486 00:34:43.440 --> 00:34:45.719 is that the smart media companies are going to figure out how to do what 487 00:34:45.760 --> 00:34:50.079 we mentioned. They're going to get their own model. They're going to train 488 00:34:50.400 --> 00:34:54.760 their model on their voice by loading all the past articles as embeddings. And 489 00:34:54.800 --> 00:35:00.440 then what's going to happen is journalists are going to be more like curat and 490 00:35:00.559 --> 00:35:05.079 editors, and you're going to be spinning out stories from facts. You have 491 00:35:05.119 --> 00:35:07.920 to have your fact sets. This is really really important fact sets. And 492 00:35:08.079 --> 00:35:13.880 that's systems of record. So I think like sales tax systems, for example, 493 00:35:13.920 --> 00:35:16.400 I think a municipality has to collect sales tax. They're collecting all this 494 00:35:16.519 --> 00:35:21.760 data from all the different stores and you determine what products are sold. Most 495 00:35:21.840 --> 00:35:24.280 have barcodes so you can be able to do some analysis. Wouldn't that be 496 00:35:24.360 --> 00:35:29.519 a fantastic service to subscribe to as a business person. Let's say I run 497 00:35:29.559 --> 00:35:32.119 a small like an easy Mart or something like that, to be able to 498 00:35:32.159 --> 00:35:36.440 see who's buying what. And the example I give is squish mellows. Like 499 00:35:36.440 --> 00:35:37.920 one day you're just looking at your report, like, what are these squish 500 00:35:37.920 --> 00:35:42.119 metal things that everyone's making tons of money on? Look into this, Oh, 501 00:35:42.199 --> 00:35:45.719 there's some new toy for kids that kids freaking love. So now kids 502 00:35:45.760 --> 00:35:51.480 have to have every squish mellow under the sun babies, it is. It's 503 00:35:51.480 --> 00:35:53.800 a new It's exactly what it is. It's the new beanie baby with lots 504 00:35:53.840 --> 00:35:58.320 of different sizes. My point is that it's not going to be so much 505 00:35:58.400 --> 00:36:02.480 individual reporters going out spending all day writing something, as it's going to be 506 00:36:02.639 --> 00:36:09.159 dynamically generated bits of information from systems of record, spoken in text that is, 507 00:36:09.280 --> 00:36:13.800 from a large language model that can be altered. Because now, like, 508 00:36:14.280 --> 00:36:16.320 it's like having a reporter that you can ask questions of all day long. 509 00:36:16.559 --> 00:36:19.119 No, wait a minute, tell me more about this. Tell me 510 00:36:19.119 --> 00:36:22.239 more about that. No reporter can sit there and answer questions for everyone all 511 00:36:22.320 --> 00:36:24.199 day long you just can't do it. So I think there's going to be 512 00:36:24.239 --> 00:36:30.039 a whole movement of data engineering around media, of connecting to these large language 513 00:36:30.039 --> 00:36:34.880 models. And you still have reporters and journalists who will sort of review this 514 00:36:34.960 --> 00:36:38.159 stuff and write some of their own original content for their voice. But to 515 00:36:38.239 --> 00:36:43.000 me, there's no stopping this train. It's like a high speed train coming 516 00:36:43.000 --> 00:36:45.719 at all of us. What do you think? Well, my one question 517 00:36:45.800 --> 00:36:51.760 there is like, but where how do we collect the facts to go into 518 00:36:51.880 --> 00:36:55.960 generating these stories? I like the idea of being able to ask the questions. 519 00:36:57.000 --> 00:37:00.000 And you know, right now we've got all kinds of clickbait headline, 520 00:37:00.159 --> 00:37:01.920 so it's like, you want to ask the question. You read the article 521 00:37:01.920 --> 00:37:06.039 and it doesn't quite tell you. It's like, well, right, where 522 00:37:06.480 --> 00:37:14.079 what was the vaccination rate in New York State between March twenty twenty one in 523 00:37:14.119 --> 00:37:17.880 April twenty twenty two, broken down by county? And you might want to 524 00:37:17.880 --> 00:37:22.320 get that detail because you live in a particular county in New York. You 525 00:37:22.360 --> 00:37:24.920 want to say, well, what really happened? The headline says vaccination rates 526 00:37:25.000 --> 00:37:30.760 dropped fifty percent over what it was pre COVID or something like that. And 527 00:37:30.920 --> 00:37:34.360 you want to get the details, but the article doesn't quite cover it, 528 00:37:34.599 --> 00:37:37.159 but you want to ask those questions. But again back to what's the source 529 00:37:37.199 --> 00:37:44.320 of the data though, where's that information going to come from? If you 530 00:37:44.360 --> 00:37:46.480 don't have you still have to have I think the reporters in the field. 531 00:37:46.599 --> 00:37:51.239 Somebody has to go out and interview people somehow, and it might be via 532 00:37:51.320 --> 00:37:53.880 zoom like this, right that you find, Hey, here's a person who 533 00:37:53.920 --> 00:38:00.360 has information about something that just happened, and so we get them on a 534 00:38:00.400 --> 00:38:06.119 zoom call and we talked to them, and then the zoom AI transcribes that 535 00:38:07.239 --> 00:38:10.159 transcribes that conversation, so nobody's actually having to type it all up, right, 536 00:38:12.360 --> 00:38:15.119 And then that could be a source, I guess for the LM. 537 00:38:15.199 --> 00:38:19.639 But somewhere there's still got to be that interaction with the external world somehow, 538 00:38:19.800 --> 00:38:22.360 right, Yeah, And I think that's going to be these systems of records. 539 00:38:22.360 --> 00:38:28.920 So taxes is one, because every county collects taxes. There are other 540 00:38:29.079 --> 00:38:34.440 sources for government, like the Federal Register for example, wherever actions are taken. 541 00:38:34.519 --> 00:38:37.880 So I think like your ERP system, basically that's tracking the movement of 542 00:38:38.000 --> 00:38:42.079 goods and the sales and all this kind of stuff, salesforce, That's all 543 00:38:42.079 --> 00:38:45.199 I was mentioning. You want to tap into all these different systems and get 544 00:38:45.239 --> 00:38:49.119 a feed from them to be able to ask questions of your business and so 545 00:38:49.280 --> 00:38:52.000 of the government. You could do the same thing. And I just think 546 00:38:52.039 --> 00:38:57.679 about how much there is to be gleaned from these from these environments of ours. 547 00:38:57.719 --> 00:39:00.079 And you think about even policy documents. Right. A buddy of mine, 548 00:39:00.119 --> 00:39:02.920 Jim Harris, was saying he was talking to a friend who works for 549 00:39:04.199 --> 00:39:06.880 I think a Canadian government organization, and he was like, Oh, I 550 00:39:06.920 --> 00:39:07.920 need to fire this person. How can I do that? I don't know 551 00:39:08.360 --> 00:39:12.239 what the paperwork is going to look like. And he said, just asking 552 00:39:12.400 --> 00:39:15.000 LLM. And it's like, okay, he asked, and it's like okay. 553 00:39:15.039 --> 00:39:16.920 Step one you have to file a letter of grievance and say hey, 554 00:39:16.960 --> 00:39:20.679 your job is not performing well at et cetera. Step two is to wait 555 00:39:20.719 --> 00:39:23.119 two weeks. Step three is do this and all these things like, because 556 00:39:23.360 --> 00:39:28.039 if you've fed all this documentation into a large language model, it can give 557 00:39:28.039 --> 00:39:31.039 you a good summary. And so I mean, wow, think about, 558 00:39:31.199 --> 00:39:35.880 like in the court system, how you have all these different motions you can 559 00:39:35.920 --> 00:39:38.679 file and you have to found the motion on something based on this, based 560 00:39:38.719 --> 00:39:43.800 on that case law for example. I mean there are these stories of an 561 00:39:43.800 --> 00:39:47.639 attorney who just used whatever the LM gave him and got in trouble. I 562 00:39:47.639 --> 00:39:52.880 almost think that's law. Yeah, it actually referenced, it didn't exist, 563 00:39:53.599 --> 00:39:58.119 right, and and that's a problem. Right. But a small language model 564 00:39:58.119 --> 00:40:00.960 that's trained on all the case laws, it's going to be a very different 565 00:40:00.000 --> 00:40:02.920 story. And now you can check things. And I mean you can even 566 00:40:02.960 --> 00:40:07.599 work that into the workflow to say, as your RAG model, check and 567 00:40:07.719 --> 00:40:13.000 make sure that these are actual cases at the So again that gets back to 568 00:40:13.039 --> 00:40:17.440 the workflow stuff. Well that I mean we've paid lawyers hundreds even thousands of 569 00:40:17.480 --> 00:40:22.159 dollars an hour sometimes to know the nuances of these codes and these policies. 570 00:40:22.519 --> 00:40:25.119 Well that you know, you don't have to do that anymore. To me, 571 00:40:25.239 --> 00:40:30.320 that's just a huge change. It's I think about just you your manual, 572 00:40:30.440 --> 00:40:32.440 like your you know, your code of conducts or something. For these 573 00:40:32.440 --> 00:40:36.800 big companies to have like big thing or even the loss that they're passing these 574 00:40:36.880 --> 00:40:39.280 days a thousand pages. Oh we can't. There was a famous politician who 575 00:40:39.280 --> 00:40:42.000 said, oh, what's the point of reading it if you don't have two 576 00:40:42.039 --> 00:40:45.679 days and two lawyers, Well now you can. Now you can just feed 577 00:40:45.719 --> 00:40:47.920 that sucker until on them and ask it all kinds of questions. What are 578 00:40:47.960 --> 00:40:52.119 the strangest things in here, what's the most expensive, what's the least expensive? 579 00:40:52.559 --> 00:40:57.480 Point being that is a massive gains changer for day to day workflow. 580 00:40:57.599 --> 00:41:00.960 And it's not just text generation, it's not just image generation. I mean, 581 00:41:00.000 --> 00:41:04.840 those are two categories of use case. But to me, the analytical 582 00:41:04.960 --> 00:41:08.400 side is the most interesting because it does have all these different ways of looking 583 00:41:08.400 --> 00:41:13.800 at things because of all the parameters. Right, Yeah, no, I 584 00:41:13.840 --> 00:41:17.480 think I think you're You're not wrong on that. I'm still thinking about the 585 00:41:17.840 --> 00:41:23.159 journalism experience example, though, as to how that's going to work, Like 586 00:41:23.320 --> 00:41:27.400 you can see the other end of it. But I still say you have 587 00:41:27.480 --> 00:41:31.519 a sourcing issue. You've still got to have the humans involved somehow to get 588 00:41:31.559 --> 00:41:37.079 the basic information for it. Though. I guess, you know, these 589 00:41:37.199 --> 00:41:42.920 days you could, I guess you could scrape everything off of Twitter because there's 590 00:41:43.639 --> 00:41:45.599 uh, you know, we talked about citizen data scientists in the past. 591 00:41:45.639 --> 00:41:51.440 Here with democratizing data, we've got citizen journalists out there. Unfortunately, you 592 00:41:51.440 --> 00:41:55.960 don't know necessarily how accurate the reporting is, but I guess if you have 593 00:41:57.559 --> 00:42:01.280 something happening in the world and you've got you know, twenty or thirty Twitter 594 00:42:01.320 --> 00:42:07.360 feeds that are reporting it, that are that are not just re reposting what 595 00:42:07.519 --> 00:42:13.039 somebody else posted, but are actually taking pictures themselves. And now we've got 596 00:42:13.079 --> 00:42:17.199 ais that are now able to start doing things like analyze right, pictures right, 597 00:42:17.280 --> 00:42:22.760 and say what's in that picture? Say these these five these five posts 598 00:42:23.159 --> 00:42:28.719 all have pictures supposedly of this event, and it should be able to overlay 599 00:42:28.719 --> 00:42:31.639 them and go, yeah, those pictures, they're unique pictures, but they 600 00:42:31.679 --> 00:42:34.800 are of the same area, and you know, you get the kind of 601 00:42:34.800 --> 00:42:37.519 the landscaping overlay of this line here that yep, that's that same building in 602 00:42:37.559 --> 00:42:42.960 the background, and be able to verify all of that, right, Well, 603 00:42:43.079 --> 00:42:45.840 that is a very very interesting point, and so kind of to that 604 00:42:46.000 --> 00:42:50.719 end, I thought of this a while ago. And cameras, all these 605 00:42:50.719 --> 00:42:53.079 digital cameras, they have their own metadata, right they have. Some of 606 00:42:53.119 --> 00:42:55.880 them have GPS, so they know where you are, they know what time 607 00:42:55.880 --> 00:43:00.760 of day you took the picture. They and they have the medidated. I'm 608 00:43:00.760 --> 00:43:04.199 sure that's encoded in the file, like we talked about parquet files. Right, 609 00:43:04.199 --> 00:43:07.159 it's got the medidata baked in there so you can vet and see uha, 610 00:43:07.599 --> 00:43:12.000 this one was doctor that one wasn't. But to your point, you 611 00:43:12.039 --> 00:43:15.599 could aggregate on demand when there's a big event, like when there's a riot 612 00:43:15.639 --> 00:43:20.679 someplace, or when you know there's a smash and grab or something, to 613 00:43:20.760 --> 00:43:24.800 be able to dynamically pick up little bits and pieces of that. I could 614 00:43:24.840 --> 00:43:30.599 see an AI engine writing a story there was a row at Bob's bar last 615 00:43:30.719 --> 00:43:36.239 nights and you know, twenty seven people were involved, Like you could get 616 00:43:36.239 --> 00:43:39.559 some juicy details and Bob Jones was taken to jail because you get his mugshot 617 00:43:39.639 --> 00:43:43.280 or something. I mean, it is possible, But to your point, 618 00:43:43.280 --> 00:43:46.119 you have to have some source. So Twitter becomes a source, social media 619 00:43:46.119 --> 00:43:52.199 becomes a source. Official records like police reports and things of that nature becomes 620 00:43:52.199 --> 00:43:57.719 sources, and those those things have to be accessible. That's a's the ither 621 00:43:57.800 --> 00:44:01.760 thing that the whole data sharing thing, whether it's a government source, linconomics 622 00:44:01.840 --> 00:44:06.000 data you were talking about, the tax data you're talking about, it has 623 00:44:06.039 --> 00:44:12.599 to be accessible somewhere so that so that the LLM can get to the data, 624 00:44:12.719 --> 00:44:15.960 or you could build a data pipeline to pull it in. That's right, 625 00:44:15.039 --> 00:44:17.360 No, that's exactly correct. And I think there's gonna be a lot 626 00:44:17.400 --> 00:44:21.880 of pressure on those systems because when this stuff starts to take off. I 627 00:44:21.880 --> 00:44:27.159 mean, apparently that's why Elon instituted that rate limiting stuff for a period of 628 00:44:27.159 --> 00:44:30.199 time, because they realized that someone was sucking all their data out and using 629 00:44:30.280 --> 00:44:34.760 it to train other models because here it was a free source and you could 630 00:44:34.840 --> 00:44:37.119 use it. So he was trying to stop that. It's just crazy the 631 00:44:37.199 --> 00:44:38.599 things we're dealing with these days. But don't touch out. The podcast bonus 632 00:44:38.599 --> 00:44:45.199 segment is up next. We'll be right back. All right, folks, 633 00:44:45.239 --> 00:44:50.000 time for the podcast bonus segment Here on a fantastic inside analysis with our profiles 634 00:44:50.039 --> 00:44:53.800 in data Sage Kent Gardziano. We've been talking all about the Data Universe conference. 635 00:44:53.800 --> 00:44:59.280 Coming up is the April tenth and eleventh dates just a few weeks away 636 00:44:59.280 --> 00:45:02.960 now. Will do a fireside chat with yours truly about privacy, which of 637 00:45:04.000 --> 00:45:08.960 course conjures up images of governance and security and data management and ethics and all 638 00:45:09.000 --> 00:45:13.800 that fun stuff. And there's also the Data Vault conference coming up in the 639 00:45:13.800 --> 00:45:15.920 beginning of May, Is that right, Kent? Yeah, yeah, w 640 00:45:16.199 --> 00:45:22.679 W DVC. It's gonna be the tenth anniversary that flies when you're having fun, 641 00:45:22.800 --> 00:45:28.559 Oh yeah, when you're not in Stove. Vermont's beautiful, beautiful location. 642 00:45:29.360 --> 00:45:31.840 It is a little harder to get to than New York City, but 643 00:45:32.039 --> 00:45:36.679 in al for but it's it's worth it. It's worth it to get there. 644 00:45:36.760 --> 00:45:39.639 Well, it's funny speaking of getting there. When we drove, when 645 00:45:39.679 --> 00:45:43.320 you and I met there, which is like I guess the year before COVID 646 00:45:43.360 --> 00:45:47.760 twenty nineteen maybe nineteen, and if we drove and you know, higher forces 647 00:45:47.800 --> 00:45:51.199 were watching out four of us because I'm trying to look at the snap like 648 00:45:51.239 --> 00:45:53.679 to understand what this is, like the straight thing we get and we realize 649 00:45:54.079 --> 00:45:59.440 it's a fairy, Like we better get that fairy because and we were there 650 00:45:59.519 --> 00:46:04.559 just in time to get the last one and across the southern end of Lake 651 00:46:04.639 --> 00:46:07.480 Champlain. Yeah, I guess so, because we would have had to drive 652 00:46:07.519 --> 00:46:08.920 all the way rount it would have been a really you'll be glad to know 653 00:46:09.000 --> 00:46:13.280 there's a bridge there. Now it's there, Okay, there there is a 654 00:46:13.320 --> 00:46:19.320 bridge. It's a it's a Whitey drive through the countryside there in in the 655 00:46:19.480 --> 00:46:22.920 far east border of New York State, the very southern end of Lake Champlain. 656 00:46:23.199 --> 00:46:25.719 But there's actually a bridge that goes over now, so you don't you 657 00:46:25.760 --> 00:46:30.599 don't have to wait for a ferry. There's actually a real bridge that got 658 00:46:30.599 --> 00:46:34.000 put in there. I must have been during COVID, because I you know, 659 00:46:34.039 --> 00:46:36.599 when I went after COVID and drove they say, oh, hey, 660 00:46:36.639 --> 00:46:39.360 this is great. And I used I used Google Google Maps, and I 661 00:46:39.400 --> 00:46:44.280 was just following the map going it looks like there's a bridge on the map. 662 00:46:44.400 --> 00:46:46.400 I say, okay, I'll go that way, sure enough, and 663 00:46:46.440 --> 00:46:51.360 it goes into southern Vermont. Well. And it's funny because, you know, 664 00:46:51.840 --> 00:46:53.960 in terms of following bots and just doing what the bot tells you. 665 00:46:54.039 --> 00:46:58.480 I heard someone give a real good example and say you're already doing it when 666 00:46:58.480 --> 00:47:01.280 you use your map speature on your phone, just trusting what it tells you 667 00:47:01.320 --> 00:47:04.400 to do, and like, oh, here we go. And I think 668 00:47:04.480 --> 00:47:08.960 that is going to be a very interesting other change in our industries, these 669 00:47:08.960 --> 00:47:14.320 little AI agents doing little things, just scanning around looking for stuff. I 670 00:47:14.320 --> 00:47:16.679 mean, you and I are to talk about privacy, and of course that 671 00:47:16.719 --> 00:47:22.039 again brings governance into the picture and ethics what you what you should do ethically. 672 00:47:22.280 --> 00:47:25.679 I think that there are all these conversations happening now because we need them 673 00:47:25.679 --> 00:47:30.119 now, because guess what, this is a big deal. And you know, 674 00:47:30.280 --> 00:47:35.280 I'll tell you one of my other big visions here is these companies like 675 00:47:35.320 --> 00:47:38.880 Amazon and Google and Facebook and LinkedIn and others have done very well by using 676 00:47:38.920 --> 00:47:43.280 our data to train their algorithms, which of course they then use, they 677 00:47:43.280 --> 00:47:46.079 say, to give us better stuff. I've never yet bought the argument I'm 678 00:47:46.079 --> 00:47:50.320 going to get better advertising, like you know, I don't think I've seen 679 00:47:50.440 --> 00:47:54.599 better advertising exactly. It's like, you know, really, is that that's 680 00:47:54.639 --> 00:47:58.000 the that's the caret the end of the stick here, as I give up 681 00:47:58.000 --> 00:48:00.639 all my data and then you I know that you're gonna give me good things 682 00:48:00.679 --> 00:48:05.960 that people are trying to sell me, and I haven't seen that. But 683 00:48:06.280 --> 00:48:08.760 I will say, there is some cool stuff on you know, some cool 684 00:48:08.840 --> 00:48:12.920 gadgets these days. So I've come across some of those on Facebook and some 685 00:48:12.960 --> 00:48:15.360 other places, so there is some benefit to it. But I want to 686 00:48:15.360 --> 00:48:19.719 be able to own our algorithm, like everyone talks about being able to own 687 00:48:19.760 --> 00:48:22.400 your data and get some money like secondary income. I mean, I think 688 00:48:22.480 --> 00:48:28.079 that the numbers aren't quite there to really make that interesting, Like how much 689 00:48:28.159 --> 00:48:30.880 data will it be? Is that? Like how much Spotify pays small bands? 690 00:48:30.920 --> 00:48:34.239 I hear people joke, Yeah, I got my check for a dollar 691 00:48:34.320 --> 00:48:37.239 forty nine in the mail, and how exciting is that? There is this 692 00:48:37.360 --> 00:48:42.119 scale problem, But I do feel like still all the power is centralized with 693 00:48:42.159 --> 00:48:45.760 these big organizations, and I'd love to find some way it's kind of pull 694 00:48:45.840 --> 00:48:50.239 that out and get some more power back in the individual's hands or even in 695 00:48:50.239 --> 00:48:53.519 aggregate you know, Like I wrote about what I referred to as a consumer 696 00:48:53.559 --> 00:49:00.760 facing data leg of information about transactions. So you think credit card companies sell 697 00:49:00.800 --> 00:49:05.559 their exhaust data to investment banks so they can optimize what they buy and the 698 00:49:05.599 --> 00:49:08.800 stock market and other reasons like that. I think when they sell our data 699 00:49:09.280 --> 00:49:14.719 to some third party like that, they should also provide an anonymized version for 700 00:49:14.840 --> 00:49:17.920 this consumer data leg where you and I and anyone else could kind of log 701 00:49:17.960 --> 00:49:22.079 in and just see interesting data about what's happening in the world. How many 702 00:49:22.079 --> 00:49:25.480 apples are being sold, how many cars are being sold. That's interesting information 703 00:49:25.559 --> 00:49:29.239 for business people, and this kind of gets back to the journalism thing, 704 00:49:29.800 --> 00:49:34.119 because, especially for business people, what I want more than anything is information. 705 00:49:34.280 --> 00:49:37.840 I don't really want attitude or spin or narrative or all that stuff. 706 00:49:37.840 --> 00:49:40.760 What I want is facts about what are people paying for this, what are 707 00:49:40.760 --> 00:49:44.760 people paying for that? How long does it take this company to do that 708 00:49:44.840 --> 00:49:46.559 job, how long does they take that company to do this job. That's 709 00:49:46.719 --> 00:49:52.440 really useful information, and it's just facts. It's just basic facts that come 710 00:49:52.440 --> 00:49:55.760 from transactional systems. So that's my big vision for the future. But what 711 00:49:55.800 --> 00:50:00.320 do you think about that? Oh? Yeah, I'm with you because you 712 00:50:00.320 --> 00:50:05.519 know, Dan lind said, who invented data vault? He he used to 713 00:50:05.519 --> 00:50:10.480 say, you know, the data vault is a historical repository of the facts 714 00:50:10.960 --> 00:50:15.079 from your source system. It's not a single source of truth. It's a 715 00:50:15.119 --> 00:50:19.679 single source of facts because you know, truth becomes malleable to a certain extent 716 00:50:19.719 --> 00:50:24.199 with your interpretation, right. You know, But to to create the information, 717 00:50:24.320 --> 00:50:28.519 for to be information and to be informative, it has to be based 718 00:50:28.559 --> 00:50:30.159 on facts. So you've got to start, You've got to set that foundation 719 00:50:30.840 --> 00:50:35.880 right. And so having a you know, a we don't have to have 720 00:50:35.960 --> 00:50:39.360 a single source of facts like we used to. That was our goal with 721 00:50:39.440 --> 00:50:44.599 data warehousing was the single source of truth. You know the technologies. Well, 722 00:50:44.800 --> 00:50:49.079 not everything has to necessarily be in the data warehouse, provided you've got 723 00:50:49.119 --> 00:50:52.800 the data pipelines and you know the right business use cases for it, but 724 00:50:52.880 --> 00:50:58.199 you do have to have a source of fact in order to do this, 725 00:50:58.199 --> 00:51:02.079 because you don't want you know, guesswork going on, you know, unless, 726 00:51:02.079 --> 00:51:07.639 of course you're dealing with customer sentiment, which is different. But then 727 00:51:07.840 --> 00:51:13.880 even that, getting customer sentiment off of say, off of x or Facebook 728 00:51:13.960 --> 00:51:19.920 or LinkedIn by people posting about your product, those are facts. This customer 729 00:51:20.039 --> 00:51:25.159 said this, right right, And from that fact and a lot of other 730 00:51:25.199 --> 00:51:31.719 similar facts you can start to evolve then an understanding of customer sentiment about your 731 00:51:31.760 --> 00:51:37.119 product, right right. That's so cool. I love that as single source 732 00:51:37.159 --> 00:51:39.920 of facts. It makes a lot of sense because, and I'll throw one 733 00:51:40.000 --> 00:51:44.920 last teaser out here, I came across as stat as I was preparing for 734 00:51:44.960 --> 00:51:49.079 my talk on the death of journalism, and the biggest challenges are bias. 735 00:51:49.239 --> 00:51:52.559 Quite frankly, there's bias. There's also something in might like there's misinformation, 736 00:51:52.840 --> 00:51:59.119 there's disinformation, and there's missed information, and to me, that's the biggest 737 00:51:59.119 --> 00:52:01.679 one. Go ahead, you get I'll give you another one, poisoned information. 738 00:52:02.519 --> 00:52:07.800 Poisoned Yeah, yeah, Dan alerted me to that one. There's been 739 00:52:07.800 --> 00:52:12.599 a couple of articles about that, about poisoned data. So think of it 740 00:52:12.639 --> 00:52:16.800 as reverse hacking. Instead of somebody hacking into your system and stealing your data, 741 00:52:17.119 --> 00:52:22.559 they hack into your system and inject bad data so that your algorithms come 742 00:52:22.559 --> 00:52:27.199 out with the wrong answer. Wow, that's crazy. Isn't that crazy? 743 00:52:27.880 --> 00:52:30.719 That is crazy. Well, it's been so much fun talking to you. 744 00:52:30.800 --> 00:52:32.920 Can't You're such a rock star in the industry. I've always looked to you 745 00:52:32.920 --> 00:52:37.159 for advice on what's happening. I look forward to seeing you in New York. 746 00:52:37.199 --> 00:52:39.079 And it's a couple of weeks. Folks up online of Data Universe and 747 00:52:39.280 --> 00:52:42.760 we'll see it a couple of weeks. You've been listening to Inside Analysis. 748 00:52:42.960 --> 00:52:46.280 KCIA Radio has openings for one hour talk shows. If you want to host 749 00:52:46.320 --> 00:52:51.480 a radio show, now is the time. Make CACIA your flag ship station. 750 00:52:51.760 --> 00:52:54.679 Our rates are affordable and our services are second to none. We broadcast 751 00:52:54.679 --> 00:52:59.880 to a population of five million people plus. We stream and podcast on all 752 00:53:00.079 --> 00:53:04.719 major online audio and video systems. If you've been thinking about broadcasting a weekly 753 00:53:04.840 --> 00:53:09.039 radio program on real radio plus the Internet, contact our CEO at two eight 754 00:53:09.119 --> 00:53:15.119 one five nine nine ninety eight hundred two eight one five nine nine ninety eight 755 00:53:15.239 --> 00:53:17.679 hundred. You could skype your show from your home to our Redlands, California 756 00:53:17.679 --> 00:53:22.119 studio, where our live producers and engineers are ready to work with you personally. 757 00:53:22.320 --> 00:53:28.039 A radio program on KCAA is the perfect work from home avocation in these 758 00:53:28.039 --> 00:53:32.039 stressful times. Just type KCAA Radio dot com into your browser to learn more 759 00:53:32.079 --> 00:53:36.599 about hosting a show on the best station in the nation, or call our 760 00:53:36.679 --> 00:53:43.360 CEO for details to eight one five nine nine ninety eight hundred. I'm doctor 761 00:53:43.400 --> 00:53:50.639 Anthony Lyiserwitz, and this is Climate Connections. Plastic is used in many products, 762 00:53:50.760 --> 00:53:54.760 from containers and bags to electronics and vehicles, and it's a significant source 763 00:53:54.880 --> 00:53:59.960 of climate warming pollution. Plastic's production. At the moment, it's come through 764 00:54:00.480 --> 00:54:06.199 to four point five percent of global greenhouse gas emissions. Livia Carbernaut of the 765 00:54:06.239 --> 00:54:09.639 Technical University in Munich, Germany, explains that almost all plastics are made from 766 00:54:09.679 --> 00:54:15.320 oil, natural gas or coal. Extracting and transporting those fuels emits carbon pollution. 767 00:54:16.000 --> 00:54:20.719 Then more fossil fuels are burned to supply the heat and electricity used to 768 00:54:20.760 --> 00:54:25.159 refine those raw materials and manufacture the plastic products, so the entire process is 769 00:54:25.320 --> 00:54:30.559 very carbon intensive. Carbonnaout's research shows that in the past few decades, carbon 770 00:54:30.599 --> 00:54:36.719 pollution from making plastic has doubled because production has grown and shifted to parts of 771 00:54:36.760 --> 00:54:39.239 the world that burn a lot of coal for energy. So she says the 772 00:54:39.280 --> 00:54:45.000 industry needs to change. Making plastics from algae or plants can reduce the need 773 00:54:45.079 --> 00:54:50.199 for fossil fuels as raw ingredients, and switching to renewable energy can reduce carbon 774 00:54:50.239 --> 00:54:54.639 pollution from production facilities. Also, for sure, we should avoid plastics whenever 775 00:54:54.840 --> 00:55:00.280 possible, like the single use plastics to help limit how much is made in 776 00:55:00.320 --> 00:55:05.719 the first place. Climate Connections is produced by the Yell Center for Environmental Communication. 777 00:55:06.159 --> 00:55:12.440 To learn more about climate change, visit climatec connections dot org. And 778 00:55:12.519 --> 00:55:17.039 now the Voices of KCAA was an exciting announcement. Want to hear NBC News 779 00:55:17.199 --> 00:55:21.440 or KCAA anywhere you go, Well, now there's an app for that. 780 00:55:21.719 --> 00:55:25.679 KCAA is celebrating twenty five years in our silver Anniversary with a brand new app. 781 00:55:25.840 --> 00:55:30.679 The new KCAA app is now available on your smart device, cell phone, 782 00:55:30.920 --> 00:55:35.920 in your car, or any place. Just search KCAA on Google Play 783 00:55:36.199 --> 00:55:38.599 or in the Apple Store one touch and you can listen on your car radio, 784 00:55:38.679 --> 00:55:44.599 Bluetooth device, Android Auto or Applecar Play. Catch the KCAA buzz in 785 00:55:44.639 --> 00:55:47.800 your earbuds or on the streets. Celebrating twenty five years of talk news and 786 00:55:47.880 --> 00:55:53.360 excellence with our new KCAA app, Just do it and download it. KCAA 787 00:55:53.440 --> 00:56:00.440 celebrating twenty five years. I'm Listening reminds you that talk saves lives and nine 788 00:56:00.440 --> 00:56:04.239 eight eight makes it even easier to reach out and talk nine to one to 789 00:56:04.239 --> 00:56:07.360 one for emergency services nine eight eight for mental health needs nine eight eight connects 790 00:56:07.360 --> 00:56:12.239 you with trained counselors and over two hundred crisis centers nationwide. Find out more 791 00:56:12.360 --> 00:56:15.920 at I'm Listening dot org. AM radio provides always on new sports, talk, 792 00:56:16.039 --> 00:56:21.239 traffic, and weather reports. It also delivers vital emergency information when your 793 00:56:21.280 --> 00:56:24.639 community needs it most. A new bill in Congress would ensure AM radio stays 794 00:56:24.679 --> 00:56:29.880 in your car because when sale and internet services are down, this free emergency 795 00:56:29.920 --> 00:56:34.320 service is critical. Text AM to five two eight eighty six and tell Congress 796 00:56:34.360 --> 00:56:37.239 to support the AM radio. For every Vehicle act message in data RATESMA AMPLA. 797 00:56:37.360 --> 00:56:39.199 You may receive up to four messages a month, and you may text 798 00:56:39.199 --> 00:56:44.320 stop to stop this message. Furnished by the National Association of Broadcasters. T 799 00:56:44.480 --> 00:56:49.159 Hebot Club's original purepowd to rco Super tea comes from the only tree in the 800 00:56:49.239 --> 00:56:52.719 world that fungus does not grow on. As a result, it naturally has 801 00:56:52.800 --> 00:56:58.559 anti fungal, anti infection, anti viral, antibacterial, anti inflammation, and 802 00:56:58.800 --> 00:57:01.760 anti parasite property. So the tea is great for healthy people because it helps 803 00:57:01.760 --> 00:57:06.800 build the immune system, and it can be truly miraculous for someone fighting a 804 00:57:06.840 --> 00:57:10.320 potentially life threatening disease due to an infection, diabetes, or cancer. The 805 00:57:10.360 --> 00:57:15.599 tea is also organic and naturally caffeine free. A one pound package of tea 806 00:57:15.679 --> 00:57:17.760 is forty nine to ninety five, which includes shipping. To order, please 807 00:57:17.840 --> 00:57:22.880 visit to Hebota Club dot com. T hebo is spelled tea like tom a, 808 00:57:22.480 --> 00:57:27.719 h ee b like boy o. They continue with the word t and 809 00:57:27.760 --> 00:57:30.920 then the word club. The complete website is to Hebot Club dot com or 810 00:57:30.960 --> 00:57:36.519 call us at eight one eight six one zero eight zero eight eight Monday through 811 00:57:36.559 --> 00:57:40.159 Saturday nine am to five pm California time. That's eight one eight six one 812 00:57:40.239 --> 00:57:45.599 zero eight zero eight eight to Hebot club dot com. Are you graduating high 813 00:57:45.639 --> 00:57:51.199 school soon and wondering what to do next? College is one option, but 814 00:57:51.400 --> 00:57:58.119 why not consider the high paying jobs made possible by union power Labor Union Teamsters 815 00:57:58.199 --> 00:58:01.400 Local nineteen thirty two is open to training center to get you into the high 816 00:58:01.480 --> 00:58:07.400 school to high paying job pipeline. You'll learn all the skills needed to excel 817 00:58:07.440 --> 00:58:15.039 in opportunities across industries. Visit nineteen thirty two Trainingcenter dot org to enroll today. 818 00:58:15.639 --> 00:58:23.159 That's nineteen thirty two Trainingcenter dot org. Look around your office? Is 819 00:58:23.159 --> 00:58:27.760 it time to change things up? Start a new home office or reorganize your 820 00:58:27.760 --> 00:58:31.639 professional office space. Visit Office Furniture Outlet and Corona and you'll feel great. 821 00:58:31.760 --> 00:58:36.880 With a huge inventory of both new and pre owned office furniture you can buy, 822 00:58:37.119 --> 00:58:39.559 sell, or even trade to get the job done. Office Furniture Outlet 823 00:58:39.599 --> 00:58:44.960 and Corona will get you looking and feeling good, and that simply means success 824 00:58:44.960 --> 00:58:49.880 and great business results. From executive office collections to home office options, you 825 00:58:49.960 --> 00:58:53.960 can find exactly what you need at an affordable price office furniture outlet. They 826 00:58:54.000 --> 00:58:59.599 have desk selections that range from modern and contemporary to traditional and elegant. With 827 00:58:59.679 --> 00:59:02.320 the law large selection of sizes, finishes, and styles, you can design 828 00:59:02.360 --> 00:59:06.760 an office just the way you need it, an office you can be proud 829 00:59:06.800 --> 00:59:08.880 of. Pre office Finisher outlet and Corona is just south of the ninety one 830 00:59:08.880 --> 00:59:15.280 Freeway in McKinley at two eighty four DuPont Street, or visit ofousa dot com. 831 00:59:15.320 --> 00:59:22.599 That's ofousa dot Com for the office furniture outlet in Corona today. For 832 00:59:22.679 --> 00:59:29.079 several years, KCAA has been marketing the Youngevity brand of nutritional and personal care 833 00:59:29.119 --> 00:59:34.320 products. Our experience with Youngevity has been one hundred percent positive, so we 834 00:59:34.400 --> 00:59:38.159 are pleased to recommend them to you. Regarding nutritional supplements, we recommend pollen 835 00:59:38.199 --> 00:59:44.559 Burst in the berry flavor and tangy Tangerine two point zero in the tablet form. 836 00:59:44.760 --> 00:59:49.119 For regularity issues, we recommend three day cleanse, and for personal care 837 00:59:49.400 --> 00:59:54.360 we recommend morning hydration cream. You can shop online for Youngevity at www dot 838 00:59:54.440 --> 00:59:59.960 KCAA team dot com, or you can order by phone by calling eight hundre 839 01:00:00.039 --> 01:00:02.360 Ndred ninet eight two three, one, nine seven, and tell customer support 840 01:00:02.400 --> 01:00:07.320 that you are part of the KCAA team. Youngevity is an American company based 841 01:00:07.360 --> 01:00:08.119 in San Diego.