WEBVTT 1 00:00:00.160 --> 00:00:05.480 I'm Chris Karagio, NBC News Radio kc AA Radio, Loma Linda, where 2 00:00:05.519 --> 00:00:10.960 no listener is ever left behind. The information economy has a ride. The 3 00:00:11.039 --> 00:00:16.920 world is teeming with innovation as new business models reinvent every industry industry. Inside 4 00:00:16.960 --> 00:00:21.960 Analysis is your source of information and insight about how to make the most of 5 00:00:22.000 --> 00:00:27.440 this exciting new era. Learn more at inside analysis dot com, Insideanalysis dot 6 00:00:27.480 --> 00:00:35.880 com. And now here's your host, through Eric Kavanaugh. All Right, 7 00:00:36.000 --> 00:00:39.799 ladies and gentlemen, Hello and welcome back once again to another episode of Inside 8 00:00:39.840 --> 00:00:43.520 Analysis, the only coast to coast radio show in the US of A that's 9 00:00:43.520 --> 00:00:48.240 all about the information economy. Yours truly, Eric Kavanaugh here, and what 10 00:00:48.439 --> 00:00:53.039 a treat. I'm actually dialing in from way across the pond. I'm over 11 00:00:53.079 --> 00:00:58.600 in Vietnam, big thanks to FPT Software. Their headquarters is right here in 12 00:00:58.679 --> 00:01:03.000 Hanoi, Vietnam. So I'm here for their big annual event what they call 13 00:01:03.079 --> 00:01:06.760 their own cees, if you will, and they're doing some amazing things. 14 00:01:06.760 --> 00:01:11.840 They're going to announce a big partnership later on tomorrow with Landing Ai, which 15 00:01:11.879 --> 00:01:15.159 is a huge AI company. So just like lots of other firms, like 16 00:01:15.280 --> 00:01:21.159 Data Bricks and Amazon and Microsoft. These folks are going all in on AI. 17 00:01:21.400 --> 00:01:25.159 Google of course as well, is doing the same Oracle as talking about 18 00:01:25.159 --> 00:01:29.040 it, IBM is talking about it. Everybody is going all in on AI, 19 00:01:29.239 --> 00:01:32.760 and y shouldn't they. And for today's show, we're going to take 20 00:01:32.760 --> 00:01:38.280 a little bit of a twist. I'm going to be interviewed myself. That's 21 00:01:38.359 --> 00:01:42.640 right, yours truly will be the subject of an interview with DJ Murphy of 22 00:01:42.760 --> 00:01:48.239 Ourx Global. They are the folks who recently took over the Big Data London 23 00:01:48.319 --> 00:01:53.280 conference in of course the UK, and they're also planning right now Data Universe, 24 00:01:53.599 --> 00:01:57.680 which is going to be a nice big event on data and AI and 25 00:01:57.719 --> 00:02:01.719 analytics and so forth in New York City next April eleventh and twelfth. I'm 26 00:02:01.799 --> 00:02:06.879 very excited about that because we missed the Strata conference and these folks have come 27 00:02:06.920 --> 00:02:10.439 along to fill that void. And my good buddy and longtime friend and colleague, 28 00:02:10.599 --> 00:02:14.240 Mike Ferguson is the one who got me connected with this group. He 29 00:02:14.360 --> 00:02:16.639 was the conference chair over at Big Data London, and I can tell you 30 00:02:16.719 --> 00:02:23.120 it was a huge success. They had something like twenty thousand registrants and fifteen 31 00:02:23.199 --> 00:02:27.319 thousand people show up and they were very clever. They gave the tickets away 32 00:02:27.759 --> 00:02:30.879 to the delegates, so you don't have to pay to go into Big Data 33 00:02:30.960 --> 00:02:34.479 London. Anyone can go in and from what I heard, the audience was 34 00:02:34.560 --> 00:02:38.879 absolutely fantastic. A lot of senior level people, a lot of practitioners out 35 00:02:38.919 --> 00:02:42.879 there learning about different technologies. What's going on in AI and what is not 36 00:02:43.120 --> 00:02:46.159 going on in AI? My goodness, it's everywhere these days between these large 37 00:02:46.240 --> 00:02:51.520 language models, so called foundational models, and the transformers. Of course, 38 00:02:51.960 --> 00:02:54.599 I've got a thread going with Jurgen schmid Huber. He's the guy who wrote 39 00:02:54.639 --> 00:03:00.039 the papers on the transformers, and you'll hear about that in this interview with 40 00:03:00.080 --> 00:03:01.719 DJ Murphy, and you hear a lot more. But if you want to 41 00:03:01.719 --> 00:03:07.199 find out more about Data Universe, send me an email info at Insideanalysis dot 42 00:03:07.199 --> 00:03:10.719 com. And with that, here's my interview with DJ Murphy. Take it 43 00:03:10.719 --> 00:03:16.639 away. So I'll give you a quick background on DM Radio. I call 44 00:03:16.680 --> 00:03:21.680 it the longest running show in the world about data. We've been broadcasting since 45 00:03:21.680 --> 00:03:25.719 two thousand and eight. It was first just online and then about nine years 46 00:03:25.759 --> 00:03:30.360 ago, I guess we cracked onto the actual airwaves and brought the show to 47 00:03:30.479 --> 00:03:35.639 radio stations. It's throw at about twelve or fourteen markets around the country, 48 00:03:36.000 --> 00:03:39.400 some big one San Francisco, La, Chicago, Atlanta, et cetera, 49 00:03:39.479 --> 00:03:44.520 a bunch of different markets, and the whole point is to just talk about 50 00:03:44.599 --> 00:03:49.560 issues related to data as a resource. So we talk about data, warehousing, 51 00:03:49.639 --> 00:03:53.639 analytics, data ingestion, data movement, data cleansing, data quality, 52 00:03:53.680 --> 00:03:59.240 anything to do with using data at some significant scale. And we typically get 53 00:03:59.240 --> 00:04:03.360 three guests per show, and the whole idea was to create this roundtable style 54 00:04:03.479 --> 00:04:08.960 discussion where each guest gets their own segment of about eight to nine minutes to 55 00:04:09.000 --> 00:04:11.520 talk about what they're working on, what they see in the marketplace, and 56 00:04:11.560 --> 00:04:15.039 then we all do the roundtables at the end, so it allows everyone to 57 00:04:15.039 --> 00:04:19.279 get their perspective out within the context of this conversation, and then the roundtables 58 00:04:19.279 --> 00:04:25.920 allow us to kind of tie it all together. So we're in year sixteen. 59 00:04:26.079 --> 00:04:29.079 Next year is going to be our seventeenth year, so we talked to 60 00:04:29.120 --> 00:04:32.399 lots and lots of folks, a lot of things have changed obviously over the 61 00:04:32.480 --> 00:04:35.639 years. And then we have another show inside Analysis, which is really all 62 00:04:35.680 --> 00:04:41.319 about the information economy, new jobs, new roles, new titles, new 63 00:04:41.360 --> 00:04:46.680 business models, new ways of doing things. Anything along those lines qualifies for 64 00:04:46.160 --> 00:04:50.720 that show. So it's a bit more broad and focus. And then about 65 00:04:50.759 --> 00:04:56.319 four years ago, I guess I got nudged into creating a TV show out 66 00:04:56.319 --> 00:05:00.000 of this stuff. We have a show called future Proof, which is really 67 00:05:00.040 --> 00:05:03.920 it just repurposed clips from the radio shows. So we have a student in 68 00:05:04.000 --> 00:05:08.759 Chicago the records, they do all sorts of different camera angles to make it 69 00:05:08.759 --> 00:05:12.439 feel like a TV show. Let me just carve that out and put it 70 00:05:12.480 --> 00:05:16.000 onto about six or seven markets now around the country, including Washington, d 71 00:05:16.120 --> 00:05:19.720 C. So we just got our own time slot in DC on Channel ten, 72 00:05:19.759 --> 00:05:23.519 which is like one of the cable companies there at Channel ten. We 73 00:05:23.600 --> 00:05:26.680 got our own time slots. So we're pretty excited about that. That's crazy. 74 00:05:26.920 --> 00:05:30.040 Yeah, great, These are all things that we can use to promote 75 00:05:30.079 --> 00:05:36.000 the show and obviously get speakers, get luminaries, advisory board members, really 76 00:05:36.040 --> 00:05:41.199 whoever. That it's fantastic, what a platform it is. And I'm we're 77 00:05:41.240 --> 00:05:45.720 so really so happy that you're able to help out on this. Yes, 78 00:05:45.720 --> 00:05:48.560 it's great. Are you based in Chicago, No, I'm from Chicago, 79 00:05:48.680 --> 00:05:53.439 but these days I'm based in Pittsburgh. Okay, I asked. We were 80 00:05:53.480 --> 00:05:57.000 just my family was just out there last weekend. My son is a freshman 81 00:05:57.040 --> 00:05:59.920 at De Paul, So oh, congratulations. Yeah, I know, it's 82 00:06:00.160 --> 00:06:01.560 right where I am. We're in New England, so it was kind of 83 00:06:01.560 --> 00:06:05.439 a trip for us. But I love it. I went to school in 84 00:06:05.480 --> 00:06:10.839 the Midwest, so familiar with Chicago and loved and I was really happy when 85 00:06:10.839 --> 00:06:15.560 he decided to let's agree that. Yeah, it's great. It's an awesome 86 00:06:15.560 --> 00:06:20.399 neighborhood, love the area and so forth. So yeah, so you kind 87 00:06:20.399 --> 00:06:24.800 of got into this just a little bit there. Do you do you have 88 00:06:24.959 --> 00:06:29.360 sort of a basic message or fundamental themes that you hit when you you know, 89 00:06:29.439 --> 00:06:31.600 each time you get on the air, whether it be radio or TV. 90 00:06:32.199 --> 00:06:35.879 I mean, what, uh, yeah, okay, go ahead. 91 00:06:36.279 --> 00:06:41.439 The standard approach I'd take is that we do topics so I can send you 92 00:06:41.439 --> 00:06:44.319 if I haven't already, I'll give you a list to the the ed cal 93 00:06:44.920 --> 00:06:47.480 And we actually just folded the last five years a bet Oftertal calendars into one 94 00:06:47.480 --> 00:06:51.120 Google sheet, so you can kind of go back in time and see what 95 00:06:51.199 --> 00:06:54.639 was last year, what was two years ago, three years ago, all 96 00:06:54.639 --> 00:06:57.639 that stuff and all the guests. So if there's anyone you ever want introductions 97 00:06:57.680 --> 00:07:00.920 to, just let me know. Connected to probably almost all the people who've 98 00:07:00.920 --> 00:07:06.319 been on these shows. But we do specific topics, so we did yesterday. 99 00:07:06.439 --> 00:07:12.240 We did a show on the modern data stack versus hyperscale data warehousing. 100 00:07:12.759 --> 00:07:15.240 Right, we had the guy from OC and Tom we had an analyst on. 101 00:07:15.600 --> 00:07:17.519 I know a lot of these independent analysts too, so we can get 102 00:07:17.519 --> 00:07:21.240 those guys involved however you want. I know, Stephanie said that you're not 103 00:07:21.279 --> 00:07:24.959 doing a call for speakers this time. It's all going to be curated, 104 00:07:25.199 --> 00:07:28.879 which is probably smart. You know, go with people you know and people 105 00:07:28.920 --> 00:07:30.959 you can trust to do a good job. So I'm happy to make some 106 00:07:31.639 --> 00:07:38.519 introductions there. But basically the shows all have roughly the same calculus, if 107 00:07:38.519 --> 00:07:42.360 you will, in that we define a topic, we go out find speakers 108 00:07:42.360 --> 00:07:45.800 who can speak to that topic, and then I just tell guests think of 109 00:07:45.839 --> 00:07:48.240 two or three key points you want to make on this topic. We'll discuss 110 00:07:48.240 --> 00:07:54.040 them in the pre call, which is about fifteen minutes, and then I'll 111 00:07:54.160 --> 00:07:59.199 use those as guideposts and navigate through a natural conversation. So we never want 112 00:07:59.199 --> 00:08:01.160 a thing scripted, obviously, but we do want to know what we're going 113 00:08:01.240 --> 00:08:05.920 to talk about. And then what we find is that there's always something interesting. 114 00:08:05.079 --> 00:08:09.519 So a lot of times we'll pivot halfway through because we just pick up 115 00:08:09.560 --> 00:08:13.759 on this interesting thread and what I figured out and this I was hoping this 116 00:08:13.800 --> 00:08:16.040 would be true and it's been very true. Is that when you get several 117 00:08:16.079 --> 00:08:20.279 experts in the same domain on the same show, they get excited about that 118 00:08:20.439 --> 00:08:22.839 because usually they know each other. I mean, I can't tell you how 119 00:08:22.879 --> 00:08:26.839 many times I've had people go, oh, I've always loved your work so 120 00:08:26.920 --> 00:08:28.199 and so, and it's great to be on the show with you and stuff. 121 00:08:28.639 --> 00:08:33.159 And what's interesting is that, to me, one of the real values 122 00:08:33.360 --> 00:08:39.039 is just the experience itself. So we're kind of in the experience management business, 123 00:08:39.039 --> 00:08:43.720 and that we just give really cool audio visual experiences to experts in data, 124 00:08:43.799 --> 00:08:48.480 to experts in technology. We get them all excited about stuff and then 125 00:08:48.600 --> 00:08:52.799 go on about our day, right right, Why data, Way back when 126 00:08:52.879 --> 00:08:56.440 when you were sort of putting this idea together, why did you land on 127 00:08:56.519 --> 00:09:01.360 data or you know, what was your experience at the time. Sure, 128 00:09:01.519 --> 00:09:05.559 so I worked for a company called Damon Consulting back in two thousand and one, 129 00:09:05.000 --> 00:09:09.960 and they were at data Warehousing consultancy, so doing ETL jobs, extract, 130 00:09:11.000 --> 00:09:13.559 transform, load, filling your data warehouse, et cetera. And the 131 00:09:13.639 --> 00:09:18.279 analyst in me just thought, I mean, this is going to be huge 132 00:09:18.720 --> 00:09:22.519 because there's so much information. And it's funny because I knew it would be 133 00:09:22.519 --> 00:09:24.039 big, and I've been blown away by how big it got. I mean, 134 00:09:24.080 --> 00:09:30.559 I didn't even realize because I didn't back then see where the nexus would 135 00:09:30.559 --> 00:09:35.080 occur between data and AI, right, And that's really the bottom line is 136 00:09:35.080 --> 00:09:37.519 that to train models, whatever model you want to train, whatever you want 137 00:09:37.519 --> 00:09:41.159 to do with AI, you need some kind of data as the foundation to 138 00:09:41.279 --> 00:09:45.639 train it, to use it, to execute it. So data is going 139 00:09:45.679 --> 00:09:48.720 to be crucial. Well how do you get to how do you understand that 140 00:09:48.759 --> 00:09:50.440 data? How do you get to data literacy? So I saw that back 141 00:09:50.440 --> 00:09:54.519 in two thousand and one at Damon Consulting, and then I went on my 142 00:09:54.559 --> 00:09:58.519 own for a couple of years and I figured what I wanted to do is 143 00:09:58.679 --> 00:10:05.480 repor a consulting firms in data because the consultants are the intersection between the clients 144 00:10:05.480 --> 00:10:09.720 and the technologies. Right. So they have to understand business speak, the 145 00:10:09.720 --> 00:10:15.159 business problems, the challenges, also the data, and then also know which 146 00:10:15.200 --> 00:10:20.120 technology is to use and when. And that obviously is a very challenging situation 147 00:10:20.360 --> 00:10:24.559 because you can look at a vendor's website and still have very little idea what 148 00:10:24.600 --> 00:10:28.440 they do. I mean, I've had many experiences where I'm like, oh, 149 00:10:28.600 --> 00:10:31.080 you sound like these guys like are you competitors? Like no, we're 150 00:10:31.080 --> 00:10:33.600 partners. I'm like, all right, did not get that from looking at 151 00:10:33.639 --> 00:10:37.799 your website. So you have to get into the weeds a bit to understand. 152 00:10:37.919 --> 00:10:39.840 So I focused on doing that for a couple of years, and then 153 00:10:39.879 --> 00:10:43.240 I wound up getting the Data Warehousing Institute as a client, and then I 154 00:10:43.279 --> 00:10:48.080 wound up working for the Data Warehousing Institute as an employee and help them launch 155 00:10:48.159 --> 00:10:52.879 their webinar series back in two thousand and five, which has been going ever 156 00:10:52.000 --> 00:10:56.799 since. And with that we took the same formula, right. I love 157 00:10:56.879 --> 00:11:03.720 having formulae for creating content because that way it's consistent and people know what to 158 00:11:03.799 --> 00:11:07.440 expect. And so the form of the back then was, Hey, you 159 00:11:07.480 --> 00:11:11.919 guys have this whole roster of consultants who come and teach at your shows. 160 00:11:11.919 --> 00:11:15.919 They're not employees, but they're partners. I said, let's get abstracts from 161 00:11:15.960 --> 00:11:18.679 them of what they're working on, what they want to talk about, and 162 00:11:18.679 --> 00:11:24.519 then we'll shop those abstracts to the vendor community to to matchmake basically, because 163 00:11:24.559 --> 00:11:28.080 TWI is a big list and these vendors they all want demand gen. Right, 164 00:11:28.120 --> 00:11:31.720 that's when they want content and they want demand gen. It's pretty much 165 00:11:31.759 --> 00:11:35.600 the two things they're always going to want. So we just created this beautiful 166 00:11:37.240 --> 00:11:41.600 marshaling area, if you will, for conversations around data and technology and that's 167 00:11:41.639 --> 00:11:45.320 been rocking ever since. And then after that, I actually pitched DM radio 168 00:11:45.360 --> 00:11:46.279 to them. They didn't want it, so I was like, wow, 169 00:11:46.639 --> 00:11:48.679 I'm going to do it, you know, whether or not you want it. 170 00:11:50.519 --> 00:11:54.639 So I jumped ship and wound up partnering with Source Media, which is 171 00:11:54.679 --> 00:11:58.279 based in New York. Media company you know, I know, Source Media, 172 00:11:58.360 --> 00:12:01.639 Yepica, Banker, all these guys. They were my partner for the 173 00:12:01.679 --> 00:12:05.960 first I don't know, four or five years or so, and then that 174 00:12:07.039 --> 00:12:09.360 went away, and I worked with the Diversity for a while and now we're 175 00:12:09.399 --> 00:12:11.279 just kind of on our own. But that's where the data journey began, 176 00:12:11.519 --> 00:12:18.320 was like twenty two and a half years ago and it's just going full full 177 00:12:18.360 --> 00:12:24.919 speed ahead. Well you were way ahead of that. I mean, it 178 00:12:24.000 --> 00:12:31.200 is all about what you foresaw at that point. So let me and you 179 00:12:31.279 --> 00:12:35.919 talked about AI and how they relate is that. I actually had two questions 180 00:12:35.919 --> 00:12:39.399 here. One was like, what's the hottest topic you're talking about on the 181 00:12:39.440 --> 00:12:43.600 show right now? I get the feeling that it's AI, right, Is 182 00:12:43.639 --> 00:12:46.879 AI sucking all the oxygen away from other important topics or is it just that 183 00:12:46.960 --> 00:12:54.799 important? Or how are you managing you know, those discussions both. Yes, 184 00:12:54.120 --> 00:12:58.080 it is absolutely sucking the oxygen out of the room, without question, 185 00:12:58.639 --> 00:13:03.320 also sucking the money out of the right. Really, because I was talking 186 00:13:03.360 --> 00:13:07.159 to a guy who is a really sharp guy I would recommend as a speaker, 187 00:13:07.279 --> 00:13:11.200 Brian Raymond from Unstructured Io. And what they do. They call it 188 00:13:11.279 --> 00:13:16.919 ETL for LMS, so extract transform for large language models, and all this 189 00:13:16.960 --> 00:13:20.799 stuff is really geared around embeddings. And embeddings, as you may know, 190 00:13:22.080 --> 00:13:26.519 are in human terms, they're like your memories, They're like what you've learned, 191 00:13:26.879 --> 00:13:31.200 right, So we all walk around with this whole stack of embeddings in 192 00:13:31.200 --> 00:13:35.519 her mind, and as you can imagine the embeddings from your earliest days are 193 00:13:35.519 --> 00:13:39.320 the most powerful because they're formative. Right as you're born into this world. 194 00:13:39.679 --> 00:13:45.919 You just have all this capacity to learn and figure stuff out. And your 195 00:13:45.919 --> 00:13:50.759 embeddings it's like your schooling, Right, those are embeddings, Like children go 196 00:13:50.799 --> 00:13:52.720 to school and they get embeddings all day. And I'm kind of driving this 197 00:13:52.799 --> 00:13:56.919 point home. So with AI, it's the same kind of thing. They 198 00:13:56.960 --> 00:14:01.639 have these large language models, which are apps literally fascinating. I mean there's 199 00:14:01.720 --> 00:14:07.759 like fifteen or sixteen of them now, and they are these modern era deep 200 00:14:07.840 --> 00:14:13.159 learning neural networks basically, and so what they can do is fabricate text. 201 00:14:13.240 --> 00:14:16.600 And you probably know this, but all they're really doing is predicting what text 202 00:14:18.279 --> 00:14:22.360 they think you want based upon your prompts. Right. The problem that they 203 00:14:22.399 --> 00:14:24.799 have is that they will make stuff up. And they make stuff up because 204 00:14:24.840 --> 00:14:30.159 that's what they were designed to do. They're basically designed to fuse vectors of 205 00:14:30.240 --> 00:14:33.799 information into a string of text or into an image according to your instruction. 206 00:14:35.440 --> 00:14:39.200 Well, how do you avoid the hallucinations is you train the models on your 207 00:14:39.279 --> 00:14:45.440 data, on your trusted corporate data. So one of my fun topics here 208 00:14:45.440 --> 00:14:52.000 and I'll even throw this out as the potential session is large language models represent 209 00:14:52.080 --> 00:14:56.320 a second chance for data. And what I mean by that is we have 210 00:14:56.399 --> 00:15:03.519 spent forty fifty years now doing all sorts of calisthenics around trying to organize data, 211 00:15:03.639 --> 00:15:07.240 analyze data, integrate it, curate it, do all these things in 212 00:15:07.320 --> 00:15:11.679 lots of systems. And there are, of course databases, there are analytical 213 00:15:11.720 --> 00:15:15.639 systems. There are data catalogs, which are a fairly recent entrance to the 214 00:15:15.679 --> 00:15:18.200 market, well fairly recent fifteen years or so, twelve years, and there 215 00:15:18.200 --> 00:15:22.759 are all these wonderful things. But it's a mess out there, and there 216 00:15:22.799 --> 00:15:26.759 are all these different ways of doing things. And if you were to look 217 00:15:26.840 --> 00:15:31.279 at the if you were to appreciate the corpus of data for any organization, 218 00:15:31.000 --> 00:15:33.639 it's a mess. It's all over the map, it's in all sorts of 219 00:15:33.639 --> 00:15:37.960 different formats. It's in your email, it's in your powerpoints, all these 220 00:15:37.960 --> 00:15:41.799 different things. Well, Brian, what unstructured does is they will take your 221 00:15:41.799 --> 00:15:45.759 powerpoints, your PDFs, your word documents, your email, whatever, They'll 222 00:15:45.799 --> 00:15:50.840 extract the semantic meaning out of that and then publish it as an embedding in 223 00:15:50.879 --> 00:15:54.480 a large language model or in a vector database basically, so they are tackling 224 00:15:54.840 --> 00:15:58.480 that side of the equation. Right. This is not a perfect metaphor, 225 00:15:58.519 --> 00:16:02.000 but if you think left brain right brain, let's say, let's say the 226 00:16:02.080 --> 00:16:06.799 left brain is your ability to talk and to have vocabulary into no words. 227 00:16:07.039 --> 00:16:11.519 Your right brain is everything you remember. It's the base gates, the repository 228 00:16:11.519 --> 00:16:15.440 of information you use to answer questions. And that's what we're seeing here with 229 00:16:15.519 --> 00:16:19.679 these large language models. So what you want to do is use the LLM 230 00:16:19.879 --> 00:16:25.559 for its text generative abilities, for its image generative abilities in that case, 231 00:16:26.159 --> 00:16:32.200 but you really want to make sure that it's leveraging your trusted data. So 232 00:16:32.240 --> 00:16:34.320 what does that mean. It means all these companies have to find a way 233 00:16:34.559 --> 00:16:40.200 to get their data into embeddings, and what I'd suggest is only put the 234 00:16:40.240 --> 00:16:42.039 good stuff in. It's like, you know, I'll break my own rule 235 00:16:42.080 --> 00:16:45.759 and letting our kid watch TikTok. We try to make sure we know what 236 00:16:45.799 --> 00:16:48.480 he sees. But those are all little embeddings that are coming in. They're 237 00:16:48.480 --> 00:16:52.240 picking up all sorts of little things, and it's like, you want to 238 00:16:52.240 --> 00:16:57.080 be careful about what goes into that that matrix. Right, So he is 239 00:16:57.120 --> 00:17:00.159 a good guest, and he used this term that I just absolutely love. 240 00:17:00.559 --> 00:17:06.200 He said, these embeddings are anchors of truce. I was like, oh 241 00:17:06.240 --> 00:17:10.920 man, yeah, that's good, right, because you want they're basically moorings. 242 00:17:10.960 --> 00:17:15.319 They're gonna they're gonna bend the wave, like bend the light wave toward 243 00:17:15.400 --> 00:17:18.759 them such that it'll pick up the interesting things and give you good content on 244 00:17:18.799 --> 00:17:22.759 the other side. Right, so and then other you know. So, 245 00:17:23.000 --> 00:17:26.880 just the other day on a radio show, I learned half of what I 246 00:17:26.920 --> 00:17:30.519 know on these radio shows. We had a guy, Bill French, who 247 00:17:30.640 --> 00:17:36.759 talked about how Google's next thing is they have built a vectorization layer underneath their 248 00:17:36.880 --> 00:17:41.880 entire architecture such that if you're using Bard it's not it's not in everyone's account 249 00:17:41.960 --> 00:17:45.799 yet, but I think it's coming. You can type at drive and then 250 00:17:45.920 --> 00:17:52.559 at Gmail and it'll automatically point to your corpus of pre vectorized data in your 251 00:17:52.599 --> 00:17:57.039 g drive and we'll pick that up right after the break. You are listening 252 00:17:57.119 --> 00:18:11.400 to Inside Analysis, expected Welcome back to Inside Analysis. Here's your host, 253 00:18:11.960 --> 00:18:19.240 Eric Tabanac. Okay, back here Inside Analysis, and here is the continuation 254 00:18:19.319 --> 00:18:30.720 of my interview with DJ Murphy of RX Global and Data universe. So here 255 00:18:30.240 --> 00:18:34.200 unstructured is trying to make their money telling you, hey, we can vectorize 256 00:18:34.200 --> 00:18:37.960 all your content. Google is like, if you're on GCP, we already 257 00:18:37.960 --> 00:18:41.680 did it for you. Don't worry about it. Like what I mean, 258 00:18:41.400 --> 00:18:47.559 think about how crazy that is? So long story short, Yes, LLLMS 259 00:18:47.640 --> 00:18:52.839 and AI and the foundational models are absolutely sucking in all the energy. And 260 00:18:52.880 --> 00:18:56.000 it was Brian Ramy who told me that there's like thirty billion dollars in investment 261 00:18:56.240 --> 00:19:02.839 in AI related to LLLMS, like the last nine months or something. So 262 00:19:03.119 --> 00:19:07.400 that's clearly because you know, the vcs are always trying to predict the future, 263 00:19:07.960 --> 00:19:10.920 right, and they always they always move in packs. There's usually a 264 00:19:10.960 --> 00:19:14.519 first mover and then all the rest go in behind. And you just saw 265 00:19:14.559 --> 00:19:21.000 this with Amazon throwing four billion at Nthropic. How much that's a lot of 266 00:19:21.000 --> 00:19:23.680 money to throw at something. So they all see it, and I see 267 00:19:23.720 --> 00:19:27.640 it too, I'll tell you honestly. Even though the hallucinations are a problem, 268 00:19:27.799 --> 00:19:32.079 and again that's when you don't have good enough moorings, don't have enough 269 00:19:32.079 --> 00:19:37.400 angers of truth, that's when you get the hallucinations. My personal opinion, 270 00:19:37.440 --> 00:19:40.519 there will probably be people who disagree with me on this. I don't think 271 00:19:40.559 --> 00:19:47.480 there's anything that's not in the crosshairs right now what gets done because when you 272 00:19:47.480 --> 00:19:52.359 get down to brass tacks, access to data. So data retrieval, data 273 00:19:52.440 --> 00:19:56.920 persistence are crucial components of every application. No matter what you do, any 274 00:19:56.960 --> 00:20:00.680 software you use must have data Otherwise what is it doing. You know, 275 00:20:00.799 --> 00:20:04.279 it needs data as it's fuel. It has some sort of form of data 276 00:20:04.319 --> 00:20:08.680 as its output. So what's not in the crosshairs. You can use these 277 00:20:08.880 --> 00:20:11.759 llms to write your emails, you can use it to write articles. I 278 00:20:11.799 --> 00:20:15.440 mean you have to fine tune it. I think if you're going to use 279 00:20:15.480 --> 00:20:19.480 that responsibly, you have to be careful about that. But even in my 280 00:20:19.640 --> 00:20:25.960 opinion, and this is very controversial, I think a lot of the value 281 00:20:26.039 --> 00:20:30.559 that's being wrought from data warehousing environments, which is very expensive stuff, can 282 00:20:30.680 --> 00:20:34.480 be delivered via these new foundational models. So, in other words, in 283 00:20:34.519 --> 00:20:40.839 dead warehousing, they put tremendous amounts of effort into extracting data from source systems, 284 00:20:41.200 --> 00:20:45.599 transforming it into a schema that can fit into the warehouse, then managing 285 00:20:45.599 --> 00:20:51.160 the warehouse doing analysis on that aggregated data. Well, you know that's very 286 00:20:51.200 --> 00:20:55.440 expensive, it's very time consuming, and the bottom line is you're in terms 287 00:20:55.480 --> 00:21:00.000 of what it's actually used, what actually generates value for the business is probably 288 00:21:00.000 --> 00:21:06.039 I'll be five percent of that data. So in other words, you're theoretically 289 00:21:06.160 --> 00:21:11.279 wasting ninety five percent of your time in this old model. Now, the 290 00:21:11.319 --> 00:21:14.599 one person if you look at the big companies, you got Terra Data, 291 00:21:14.640 --> 00:21:17.640 you've got Alterics, you've got Data Bricks, you've got Snowflake. These are 292 00:21:17.640 --> 00:21:22.680 all the sort of big data warehouse, data analytics vendors. Of those, 293 00:21:22.799 --> 00:21:26.480 the one who seems to really know what's happening here is Ali Gatzi of Data 294 00:21:26.519 --> 00:21:30.839 Bricks, because he's the guy who threw one point two billion to buy Mosaic 295 00:21:32.400 --> 00:21:37.400 whatever it was, six seven, eight months ago. And the week that 296 00:21:38.359 --> 00:21:42.759 deal cleared, I had Navien Raw on Damn Radio, so I interviewed the 297 00:21:42.799 --> 00:21:47.240 CEO of Mosaic like that week and he's very smart guy, as you can 298 00:21:47.279 --> 00:21:49.559 imagine, and he was saying that you know, right now, it is 299 00:21:49.599 --> 00:21:53.359 the marketing is the content creation stuff. But I'm here to tell you that 300 00:21:53.440 --> 00:21:57.880 that is just the beginning. And then interactive AI is what they're really talking 301 00:21:57.920 --> 00:22:00.720 about. Because I'll throw out at you and then I'll be quiet for a 302 00:22:00.759 --> 00:22:08.640 minute. If you think about the complexity of these models learning to reflect back 303 00:22:10.359 --> 00:22:18.400 fully syntactically correct, accurate pros, it's pretty complex, well machine to machine 304 00:22:18.400 --> 00:22:22.240 conversations, even though it's very difficult for most people to understand what they're doing. 305 00:22:22.559 --> 00:22:29.519 They're much less complex than that, which means a foundational model should be 306 00:22:29.960 --> 00:22:37.960 very powerful in terms of optimizing your current information and systems architecture in my opinion. 307 00:22:37.400 --> 00:22:40.240 So what does that mean. I mean you're already seeing a couple of 308 00:22:40.279 --> 00:22:44.160 companies do interesting stuff here. Hammer Space is one. I was on the 309 00:22:44.240 --> 00:22:48.400 Data Unchained podcast just earlier this week, and they're doing They are the company 310 00:22:48.400 --> 00:22:53.440 that is training Lama too, So the new model that's coming out of Facebook, 311 00:22:55.279 --> 00:22:59.119 they're actually doing the training of the model with their data, and they 312 00:22:59.200 --> 00:23:03.759 do this hyper efficient data movement where they only move this piece of data or 313 00:23:03.759 --> 00:23:08.359 that piece of data, which is a stark opposition to ETL. In the 314 00:23:08.400 --> 00:23:12.559 ETL world, you extract, transform, and load huge sums of data into 315 00:23:12.559 --> 00:23:17.200 your data warehouse. And the point I'm making here is that I think we're 316 00:23:17.200 --> 00:23:22.759 going to see crazy efficiency gains using foundational models in AI to get to the 317 00:23:22.839 --> 00:23:26.359 nuts that you want, so you don't have to shape the whole tree, 318 00:23:26.440 --> 00:23:30.319 gather all the nuts, move them across town to a warehouse, then package 319 00:23:30.359 --> 00:23:32.640 them and then ship them out to people. No, you're just going to 320 00:23:32.720 --> 00:23:37.559 use teeny tiny little little drones to deliver eight acorns to this address in ten 321 00:23:37.640 --> 00:23:41.839 acorns to that address. Wow, it is moving. Well, it's certainly 322 00:23:41.839 --> 00:23:47.759 moving faster than I have any sort of conception of I'm glad there are folks 323 00:23:47.839 --> 00:23:52.119 out there like you that have some sort of conception of it. And that 324 00:23:52.200 --> 00:23:56.519 actually leads me to kind of what I wanted to ask next about business users 325 00:23:57.000 --> 00:24:04.000 and kind of the state of the expertise or literacy on average, Like what 326 00:24:04.079 --> 00:24:12.319 you see in across businesses. Are people ready to use this stuff and leverage 327 00:24:12.400 --> 00:24:18.559 these advancements? Yeah? Yeah, even the ones that are available now, 328 00:24:19.000 --> 00:24:22.759 and not even what's coming down the pike. I mean, just rough estimate. 329 00:24:22.799 --> 00:24:26.039 I'm a big fan of the Peretto principle and right the eighty twenty rule 330 00:24:26.720 --> 00:24:32.559 probably, and there are always breakdowns from there. Right, it's like the 331 00:24:32.559 --> 00:24:37.720 twenty percent of the twenty percent, so probably four percent of workers are all 332 00:24:37.759 --> 00:24:41.039 in now. They totally get it. They're working on it, they're focused 333 00:24:41.079 --> 00:24:45.519 on it, and they're they're excelling. Overall, I'd say twenty percent are 334 00:24:45.519 --> 00:24:48.359 open to it, they've kind of tested things out, and the other eighty 335 00:24:48.400 --> 00:24:52.920 percent are probably just too busy doing other stuff. Right, And that's the 336 00:24:52.960 --> 00:24:55.880 problem, right, is Like, one of my favorite people in the world 337 00:24:56.000 --> 00:25:00.000 is a guy who got me in fact, a guy who signed the form 338 00:25:00.160 --> 00:25:03.920 to get dm radio launched financially back in two thousand and steven. I guess 339 00:25:03.960 --> 00:25:08.079 it was David Greeno. He had this term, and I just love these 340 00:25:08.119 --> 00:25:12.920 expressions because they communicate so much. They're so pithy. He referred to the 341 00:25:12.960 --> 00:25:19.759 tyranny of urgency. I was like, it's painful to hear that, you 342 00:25:19.799 --> 00:25:23.200 know, And that's what most people are dealing with, is the tyranny emergency. 343 00:25:23.240 --> 00:25:26.200 They got to get through their emails, they got to get through all 344 00:25:26.200 --> 00:25:27.519 this stuff they have to do today, you know. Then they got to 345 00:25:27.519 --> 00:25:30.519 get the kids to school and all this stuff. And so that's what's sort 346 00:25:30.519 --> 00:25:37.160 of driving the reluctance or the inertia on being able to leverage these technologies. 347 00:25:37.200 --> 00:25:40.680 But I think that is that's going to change pretty quickly. I think it's 348 00:25:40.680 --> 00:25:44.480 going to be you know, thirty seventy within three months. And you know, 349 00:25:44.519 --> 00:25:47.440 Microsoft is a reason for that. Google is a reason for that. 350 00:25:47.839 --> 00:25:49.440 You know, the fact that you don't have to go out and like buy 351 00:25:49.519 --> 00:25:55.039 some software and install it, well, that's big news. I mean, 352 00:25:55.079 --> 00:26:00.680 honestly, if you if you want my honest opinion on one of the most 353 00:26:00.720 --> 00:26:04.160 important things to have right now in the software world for survival. It's called 354 00:26:04.279 --> 00:26:10.880 install base. Like just people who are already using your stuff. Right. 355 00:26:11.079 --> 00:26:14.440 So I got an email from some girl at Google yesterday saying, Hi, 356 00:26:14.480 --> 00:26:18.480 I'm your account executive. I've never heard from any account executive at Google first 357 00:26:18.519 --> 00:26:21.000 time ever, and she's like, oh, prices are going up and this 358 00:26:21.079 --> 00:26:23.119 is why, et cetera, et cetera. But she pointed out that if 359 00:26:23.119 --> 00:26:26.079 you're logged into your Google account, you're just going to get all this new 360 00:26:26.119 --> 00:26:30.640 functionality, right. So the fact that you don't have to go out and 361 00:26:30.759 --> 00:26:34.839 buy something like you used to in the on prem days, that's a big 362 00:26:34.920 --> 00:26:38.839 deal. So that's going to reach the tracks to usage. Now the question 363 00:26:38.920 --> 00:26:44.599 becomes, you know, how useful is its? Have they sorted out the 364 00:26:44.640 --> 00:26:47.680 guardrails? I mean, that's one of the issues is the guardrails on this 365 00:26:47.720 --> 00:26:51.400 stuff, and so what are those Well, to a certain extent, those 366 00:26:51.400 --> 00:26:55.079 are embeddings. You know, you can do things different ways. There's this 367 00:26:55.279 --> 00:27:02.680 RAG model that came out of Facebook and it has to do with how you 368 00:27:02.759 --> 00:27:07.119 actually persisted. So retrieval augmented generation is what it's called. It's a strategy 369 00:27:07.160 --> 00:27:12.559 that helps address both LM hallucinations and out of date training data. So basically, 370 00:27:14.039 --> 00:27:15.480 you use, you go into the language model, you call out your 371 00:27:15.480 --> 00:27:18.240 stuff, and then at the last minute you're like, all right, let's 372 00:27:18.279 --> 00:27:22.359 check and make sure that's right, and you know, let's throw another guardrail, 373 00:27:22.400 --> 00:27:25.200 if you will, into the picture. And so the guardrails are going 374 00:27:25.279 --> 00:27:27.960 to be really important. But the problem is everything is moving so fast, 375 00:27:29.039 --> 00:27:30.680 so it's like people just want to go, go, go and use this 376 00:27:30.759 --> 00:27:33.599 stuff. You do have to be careful that you know it's not hallucinating, 377 00:27:33.599 --> 00:27:37.960 it's not making stuff up because you know, for example, chat GPT. 378 00:27:37.920 --> 00:27:41.400 The first time someone searched me that I saw, they showed me what it 379 00:27:41.440 --> 00:27:45.400 said and it was very accurate. I was like, wow, that's pretty 380 00:27:45.400 --> 00:27:48.480 impressive. The second time I did it, it said I'd written three books. 381 00:27:48.519 --> 00:27:53.680 I was like, are you sure, they usually don't write a book 382 00:27:53.680 --> 00:27:59.359 and then black out, but you know, yeah, so it's just you 383 00:27:59.359 --> 00:28:04.319 know it that is going to be a real serious issue going forward because you 384 00:28:04.359 --> 00:28:07.880 know, think about it, We've had to deal with fake news and now 385 00:28:07.920 --> 00:28:14.279 we have this issue of hallucinations where these powerful engineers just make it stuff up. 386 00:28:14.839 --> 00:28:18.480 And when I was watching the Data Bricks conference, there was a girl 387 00:28:18.519 --> 00:28:22.000 doing a great demo about how they are allowing the training of these models. 388 00:28:22.400 --> 00:28:26.079 And that's what you do, is you train it basically. And here's one 389 00:28:26.079 --> 00:28:30.359 of the sort of dirty secrets, if you will, of old deep learning 390 00:28:30.400 --> 00:28:33.759 modules and old AI and new as well. I had a guy in the 391 00:28:33.759 --> 00:28:37.039 show. I'd have to look up his name. He's a very smart guy 392 00:28:37.119 --> 00:28:41.079 to Boston somewhere, and he's one of the experts on deep learning and AI. 393 00:28:41.599 --> 00:28:42.880 And he told me a stat that blew my mind. He said, 394 00:28:44.240 --> 00:28:47.480 when you're training this stuff, and that's when you have human beings just saying 395 00:28:47.559 --> 00:28:51.039 yes or no. You may have seen the thing is as a blueberry muffin 396 00:28:51.119 --> 00:28:52.880 or a dog. That whole thing if you come across that, because like 397 00:28:52.920 --> 00:28:57.279 the AA engines have a hard time person figuring it out. He goes. 398 00:28:57.319 --> 00:29:03.079 As it turns out, very few people are good at that task because it 399 00:29:03.160 --> 00:29:07.799 requires you to be very focused and to always answer correctly. It's like those 400 00:29:07.839 --> 00:29:11.160 stupid captures that you'll get when you're trying to log into a system and like, 401 00:29:11.200 --> 00:29:15.000 Okay, which of these pictures has a motorcycle in it? Like? 402 00:29:15.839 --> 00:29:18.799 Really, when do I And that's a fun story too. Did you know 403 00:29:18.839 --> 00:29:22.119 that you're actually training models when you do that? I did not know that. 404 00:29:22.319 --> 00:29:26.319 Yeah, that's Google and they're doing they they had this crowdsourced effort basically 405 00:29:26.400 --> 00:29:30.720 to train models to identify stuff. So you're when you do the capture, 406 00:29:30.759 --> 00:29:37.000 you're actually helping train a bit model for the future. Do you think that's 407 00:29:37.000 --> 00:29:41.599 become its primary purpose and not not the security of whatever? You know? 408 00:29:41.799 --> 00:29:45.160 That's a very very good question. Honestly, I absolutely detest those captures. 409 00:29:45.200 --> 00:29:48.839 I mean, I understand why they're necessary, but they always forced me back. 410 00:29:48.880 --> 00:29:52.839 I don't know's it's a pain, and they're they're vague sometimes to the 411 00:29:52.960 --> 00:29:56.160 to the point of of you know, them using it to maybe that's what 412 00:29:56.200 --> 00:30:02.599 they're experimenting with as far as the training goes, right, But yeah, 413 00:30:02.839 --> 00:30:08.240 so that's kind of interesting. So like aside from AI, where are do 414 00:30:08.279 --> 00:30:14.640 you see trends or anything where people are investing their time as far as professional 415 00:30:14.640 --> 00:30:19.599 growth goes for businesses and employees within these businesses, what they're kind of looking 416 00:30:19.640 --> 00:30:25.440 at, well, clearly clearly they want to educate themselves on AI. That's 417 00:30:25.519 --> 00:30:29.079 that's going to be the biggest thing. But well, data is data again 418 00:30:29.200 --> 00:30:33.559 is the important thing. So data lineage, data quality, data movement, 419 00:30:34.160 --> 00:30:41.759 data pipelines. That's a very very hot topic because data pipelines are these as 420 00:30:41.759 --> 00:30:45.400 the name would suggest, connections from sources to targets. So what used to 421 00:30:45.440 --> 00:30:51.960 be just et L extract transform loads is now much more sophisticated. Where you'll 422 00:30:52.000 --> 00:30:56.519 pull from this system and what they do is they'll use a graphic interface where 423 00:30:56.559 --> 00:31:02.359 you have different nodes on a pipeline and one not will be to access the 424 00:31:02.480 --> 00:31:04.559 data. One no will be to check the quality of the data. One 425 00:31:04.599 --> 00:31:10.000 note will be to transform the data into a different scheme of for example, 426 00:31:10.400 --> 00:31:14.079 one note will be to enrich the data, so add some information from another 427 00:31:14.160 --> 00:31:18.440 source. Basically, data pipelines are really really important. Real time is really 428 00:31:18.480 --> 00:31:23.599 important. You know, that's a big deal. There are other technologies too 429 00:31:23.599 --> 00:31:26.559 that are still important. Even though they're kind of on a back burner, 430 00:31:26.799 --> 00:31:34.759 blockchain is still interesting, interesting things going on in that space around durability but 431 00:31:34.839 --> 00:31:40.759 also immutability is a big one. But the blockchain has had trouble, you 432 00:31:40.759 --> 00:31:45.440 know, finding its relevant use cases. But I think it's it's still out 433 00:31:45.440 --> 00:31:48.880 there. It's still an important thing to have, and we'll probably come in 434 00:31:48.960 --> 00:31:52.039 handy. And then just analytics in general, being able to number crunch. 435 00:31:52.160 --> 00:31:59.319 Hyper scale is an interesting topic in and of itself. We had Chris gladmin 436 00:31:59.400 --> 00:32:02.119 of oc AND on our show yesterday, and they're based out of Chicago. 437 00:32:02.119 --> 00:32:07.480 In fact, he's from lamonts where I used to be the reporter and editor 438 00:32:07.519 --> 00:32:09.759 for the Lamont Metropolitan newspaper. That was my first job right out of college. 439 00:32:09.759 --> 00:32:15.599 They stumbled into a newspaper job. But they're doing hyper scale data warehousing 440 00:32:15.640 --> 00:32:21.359 where we're talking, and they're using these new envme cards which are so much 441 00:32:21.480 --> 00:32:23.920 more powerful than the old ones. So he was talking about how when you 442 00:32:24.000 --> 00:32:28.319 retrieve data it's in packets, it's like, you know, four bits basically, 443 00:32:28.400 --> 00:32:31.119 well, these they can do two thousand, so you're talking about five 444 00:32:31.200 --> 00:32:37.400 hundred x. You're talking about orders of magnitude greater and they're just doing crazy 445 00:32:37.440 --> 00:32:42.480 stuff, these huge, huge data warehouses, petabytes of information, you know, 446 00:32:42.519 --> 00:32:45.759 so hyper scale is an issue. Cloud is an issue. Of course, 447 00:32:45.759 --> 00:32:50.799 security is always going to be an issue, you know that. And 448 00:32:50.839 --> 00:32:52.880 with that, will take a quick break and we'll pick up the conversation once 449 00:32:52.920 --> 00:33:07.680 again. Right after these messages, you were listening to Inside Analysis. Welcome 450 00:33:07.720 --> 00:33:15.920 back to Inside Analysis. Here's your host, Eric Tavanaugh. Okay, folks 451 00:33:15.960 --> 00:33:21.279 are back here on Inside Analysis in Media Race as it were talking to DJ 452 00:33:21.400 --> 00:33:25.480 Murphy about hot trends like hyperscale and cloud and security. And here are my 453 00:33:25.519 --> 00:33:30.599 thoughts on security and passwords. I think passwords are the biggest pain in the 454 00:33:30.640 --> 00:33:36.039 freaking universe and I'm still not entirely sure why they haven't solved them. They're 455 00:33:36.079 --> 00:33:38.920 starting to with the biometric stuff. You started to see more apps on your 456 00:33:38.960 --> 00:33:44.279 phone leverage that, so, you know, security, even though the audience 457 00:33:44.359 --> 00:33:49.000 is pretty small, it's always important. And governance, governance at the other 458 00:33:49.480 --> 00:33:54.400 really big important topic. How do you govern access to these data assets because 459 00:33:54.400 --> 00:33:58.480 in the old days you had two options. Basically, you could govern it 460 00:33:58.640 --> 00:34:01.400 at the database layer, meaning you're allowed to access the database or you're not, 461 00:34:02.079 --> 00:34:06.480 or you could govern it at the application layer, meaning okay, user, 462 00:34:06.720 --> 00:34:10.039 your your role is X, you're a manager or whatever, I'll govern 463 00:34:10.079 --> 00:34:13.840 it that way. Both of those are very inefficient ways to do it. 464 00:34:13.920 --> 00:34:17.519 Certainly, the database one is incredibly inefficient because you only want your database administrators 465 00:34:17.559 --> 00:34:22.559 touching that. And you know, if you have four thousand people who are 466 00:34:22.599 --> 00:34:25.400 allowed to use this analytics software, you know, having Bob go in there 467 00:34:25.440 --> 00:34:29.559 every day and take take out Susie, Okay, she will go out. 468 00:34:29.559 --> 00:34:34.280 I mean, you know, automating that, that process of governance is huge. 469 00:34:36.599 --> 00:34:39.800 So I want to turn to the event a little bit. You're a 470 00:34:39.840 --> 00:34:44.119 broadcaster. You want to get your message out to as many people as possible. 471 00:34:45.880 --> 00:34:52.639 How important do you think like a micro sort of situation like a face 472 00:34:52.679 --> 00:34:58.480 to face event is, or like the propagation of knowledge or or whatever. 473 00:34:58.599 --> 00:35:00.320 I mean, you're obviously in this you want to you're helping us out and 474 00:35:00.360 --> 00:35:04.679 you're advising us on this. How important are the are the face to face 475 00:35:04.679 --> 00:35:07.320 events? Oh? They're hugely important, you know, and we're seeing that 476 00:35:07.440 --> 00:35:13.599 come back now, roaring back. Obviously with COVID, everyone went into hiding 477 00:35:13.639 --> 00:35:16.440 for a period of time and for lots of different reasons for good reasons, 478 00:35:16.519 --> 00:35:21.719 obviously, but virtual will only get you so far. And I'm a virtual 479 00:35:21.719 --> 00:35:24.199 guy, so I do virtual stuff all the time. I love virtual it's 480 00:35:24.239 --> 00:35:29.519 really important. But person to person, shaking hands, talking to people, 481 00:35:29.559 --> 00:35:34.039 sitting down, it's always important. It's always going to be important. And 482 00:35:34.480 --> 00:35:38.800 these these focused events are are really crucial, you know, because I've gone 483 00:35:38.880 --> 00:35:43.840 to a bunch of different conferences over the years. I've been involved in promoting 484 00:35:43.880 --> 00:35:47.480 conferences and speaking of conferences, et cetera. And the sort of broader consumer 485 00:35:47.480 --> 00:35:52.559 technology ones aren't so interesting to me because I'm a data guy. So you 486 00:35:52.719 --> 00:35:55.920 have to have a very focused conversation. And this is what trade press is 487 00:35:55.920 --> 00:36:00.840 all about. And you know, frankly, trade press in my is dying. 488 00:36:00.880 --> 00:36:04.519 It's been dying for a number of years now because largely of the Internet. 489 00:36:04.639 --> 00:36:08.679 Because trade publications used to make all their money on print dads in magazines 490 00:36:08.719 --> 00:36:14.519 and now those are gone. So take an industry and just rip the financial 491 00:36:14.599 --> 00:36:17.079 carpet out from under their feet. That's a problem. So you know, 492 00:36:17.119 --> 00:36:21.519 a handful of being able to pivot to webinars, which are one of the 493 00:36:21.519 --> 00:36:28.039 biggest ones, and obviously events, but absolutely events are huge They're very important. 494 00:36:28.400 --> 00:36:30.280 The session design is always important, you know, making sure you've covered 495 00:36:30.320 --> 00:36:34.480 all the bases of you know, who wants to learn about what? But 496 00:36:34.719 --> 00:36:38.440 I mean no doubt in person events are awesome. Do you think there's a 497 00:36:38.559 --> 00:36:43.360 takeaway for this particular event that's going to be important for the people that attend 498 00:36:45.800 --> 00:36:52.320 granular knowledge, you know, workshops like have a combination of sessions. Round 499 00:36:52.320 --> 00:36:55.800 Tables are always fun. I love roundtables, and I'll have to moderate round 500 00:36:55.800 --> 00:37:02.679 tables in part because you get the chemistry of the group involved, right and 501 00:37:04.039 --> 00:37:07.159 with a single presenter, it's just one person or even two, But when 502 00:37:07.159 --> 00:37:10.039 you have a round table, especially if they're you know, open minded to 503 00:37:10.199 --> 00:37:15.400 how that gets managed, you can sort of build up a momentum and a 504 00:37:15.480 --> 00:37:20.440 chemistry that you can't get in other sessions. Now you have to have people 505 00:37:20.440 --> 00:37:24.320 who moderated effectively obviously really know what they're talking about. But that's that would 506 00:37:24.320 --> 00:37:30.559 be one of my suggestions is have numerous roundtables that are really interactive and dynamic 507 00:37:30.800 --> 00:37:34.920 right where you go in And I always tell people, I always joke we'll 508 00:37:34.920 --> 00:37:36.960 talk about we're going to talk about, but once the bell rings, all 509 00:37:37.000 --> 00:37:40.159 bets are off Yeah, it's live conversation, so you know, we'll see 510 00:37:40.519 --> 00:37:43.920 where things go. So I'd be sure to do that. And then you 511 00:37:43.920 --> 00:37:50.239 know, meet and great sessions and pre cocktails. That's always that's the best 512 00:37:50.239 --> 00:37:52.760 part of a live event, isn't it. That's what I tell my wife. 513 00:37:52.800 --> 00:37:54.119 I go to these events for the free food and the free cocktails and 514 00:37:54.239 --> 00:37:59.360 like joking, half joking, but it does help as people, you know, 515 00:37:59.559 --> 00:38:00.760 especially at the end of the day, like the six o'clock thing, 516 00:38:00.760 --> 00:38:05.760 the seven o'clock thing, you know, that's when everyone has been talking all 517 00:38:05.840 --> 00:38:08.000 day business. They want to go, you know whatever, drink wine or 518 00:38:08.000 --> 00:38:12.360 even a coke or something and just chat about what's going on. All that 519 00:38:12.400 --> 00:38:16.679 stuff is really important. As someone advising on the program, what do you 520 00:38:16.800 --> 00:38:23.000 hope that this event will deliver to the community that you communicate to on a 521 00:38:23.599 --> 00:38:29.960 daily or weekly basis that may they may not be getting Elsewhere's hard, that's 522 00:38:30.000 --> 00:38:34.280 a hard fact, you know. So Mike Ferguson, he's the guy who 523 00:38:34.840 --> 00:38:37.679 suggested that you all talk to me, who's incredibly knowledgeable. Guy like getting 524 00:38:37.719 --> 00:38:44.480 a dozen folks of that caliber too. I won't call them fact checkers because 525 00:38:44.480 --> 00:38:47.400 I don't like all fact checking thing. But they what's really cool is that 526 00:38:47.480 --> 00:38:53.000 when someone when these vendors know that there's someone in the room who knows all 527 00:38:53.039 --> 00:39:00.360 their weaknesses, boy do they behave they just know? We learn that from 528 00:39:00.400 --> 00:39:01.599 experience that you know, if you get a round table. When I get 529 00:39:01.599 --> 00:39:04.639 the end, I always want an analyst on the call if I can. 530 00:39:05.119 --> 00:39:10.800 And three vendors like, it's amazing how constructive the conversation becomes because number one, 531 00:39:10.840 --> 00:39:14.119 they don't take pot shots at each other. Number Two, they don't 532 00:39:14.119 --> 00:39:16.480 talk about stuff they don't have, you know. I mean I often joke 533 00:39:16.519 --> 00:39:22.679 that talking to especially a bigger enterprise software executive, is like talk to a 534 00:39:22.719 --> 00:39:24.639 politician. I mean, they have their messaging down, they have their talking 535 00:39:24.719 --> 00:39:29.039 points, and you got to find some way to get them off with that 536 00:39:29.119 --> 00:39:32.920 message, to get some gritty detail. And that's usually a curve ball of 537 00:39:32.960 --> 00:39:37.400 some kind or you know, some question that pops up and you have to 538 00:39:37.440 --> 00:39:40.679 wait and let it kind of flow out, as opposed to you can't badger 539 00:39:40.760 --> 00:39:45.159 someone, right, this is all it's all friendly media. I don't do 540 00:39:45.239 --> 00:39:47.519 sixty minutes type stuff, right, I'm always doing friendly media. But I 541 00:39:47.599 --> 00:39:51.679 want to get to the meet. I want to get to the hard facts 542 00:39:51.679 --> 00:39:53.239 that people need to know, like, oh, you want to use this 543 00:39:53.280 --> 00:39:57.599 technology that that'shed out knowledge. You know, they don't actually work well or 544 00:39:57.639 --> 00:40:01.159 they are maybe they're very complimentary. You don't know that stuff until you've talked 545 00:40:01.159 --> 00:40:04.880 to someone who has used it, who has built the solution, you know, 546 00:40:04.880 --> 00:40:07.039 who's worked with it. That's where you find the bugs and the sort 547 00:40:07.039 --> 00:40:10.639 of rough edges on things. And that's that's what I would love people to 548 00:40:10.639 --> 00:40:14.719 get more of, is you know, what are the rough edges in fact 549 00:40:15.079 --> 00:40:20.840 that could even be an interesting show or an interesting conversation, is you know, 550 00:40:21.559 --> 00:40:24.039 what do people think is the case versus what really is the case? 551 00:40:25.000 --> 00:40:30.480 AI? And it's like we learned that with that New York Times guy who 552 00:40:30.679 --> 00:40:36.000 interviewed chat GBT when it first came out. And I try to tell people 553 00:40:36.800 --> 00:40:42.400 he is responsible for that conversation going sideways because he misunderstood what the tech is. 554 00:40:42.440 --> 00:40:45.280 He's treating it like a sentient being, trying to, you know, 555 00:40:45.800 --> 00:40:51.039 get it to say colorful things, which it clearly did. But again, 556 00:40:51.119 --> 00:40:54.880 all it's doing is reflecting back from the corpus of techs that it was trained 557 00:40:54.920 --> 00:40:59.920 on what it thinks you want to hear. So you know, prompt engineer, 558 00:41:00.239 --> 00:41:04.679 for example, is absolutely do workshops on that all day long on prompt 559 00:41:04.760 --> 00:41:09.280 engineering, because that is it's really crucial to get what you wanted or something, 560 00:41:09.320 --> 00:41:14.880 to give very precise instruction, and it will take that instruction like. 561 00:41:14.960 --> 00:41:17.159 That's the thing is that it's like a little assistant who will do everything you 562 00:41:17.199 --> 00:41:21.840 ask it to do as best as they can do it, you know, 563 00:41:21.960 --> 00:41:25.239 And I always tell it please and thank you always. It won't filter out 564 00:41:25.280 --> 00:41:30.800 your worst impulses. It'll just always be nice to the machines. You know, 565 00:41:30.840 --> 00:41:31.719 they may wake up one day, I don't know. There you go, 566 00:41:31.840 --> 00:41:35.880 There you go, Eric, thank you very very much. What I'm 567 00:41:35.880 --> 00:41:39.519 going to do is I'm going to go through and do this up sort of 568 00:41:39.559 --> 00:41:45.880 as a Q and A. I will shorten your most of your answers tragically 569 00:41:45.159 --> 00:41:50.119 unfortunately, but I am going to send it to you before we publish, 570 00:41:50.239 --> 00:41:53.239 just in case I've edited something the wrong way or whatever. And let's take 571 00:41:53.280 --> 00:41:58.039 for shut. Okay, thank you so much. I appreciate your recording it. 572 00:41:58.480 --> 00:42:00.519 Send me the transcript when you have a chance. And this has been 573 00:42:00.559 --> 00:42:08.559 great, and hope to have a chance to meet too. And that concludes 574 00:42:08.559 --> 00:42:15.079 my conversation with DJ Murphy from our ex Global Read exhibitions. To folks who 575 00:42:15.159 --> 00:42:21.280 are putting on the Data Universe conference in New York City April eleventh and twelfth, 576 00:42:21.840 --> 00:42:24.159 be sure to check that out. We're working on the speakers right now, 577 00:42:24.239 --> 00:42:27.800 working on the themes, working on all that kind of fun stuff. 578 00:42:28.039 --> 00:42:30.800 Should be a fantastic show, and look for yours truly to be walking around. 579 00:42:30.800 --> 00:42:34.880 If you want to offer some advice on what you think the show should 580 00:42:34.880 --> 00:42:38.559 cover, by all means, send me an email info at insideanalysis dot com. 581 00:42:38.880 --> 00:42:42.840 You know, I'm always keen to find out what folks are looking to 582 00:42:42.960 --> 00:42:46.039 learn, and these days it's going to be a lot about those foundational models. 583 00:42:46.039 --> 00:42:50.480 It's going to be a lot about how to train large language models. 584 00:42:50.519 --> 00:42:53.000 It's going to be a lot about what else is in the crosshairs. And 585 00:42:53.079 --> 00:42:58.039 to be honest, folks, I don't think anything is not in the crosshairs 586 00:42:58.400 --> 00:43:00.719 of these foundational models. So we're going to find out. But at the 587 00:43:00.719 --> 00:43:06.519 moment, according to Navin Raw of Mosaic mL and several other experts I've talked 588 00:43:06.519 --> 00:43:10.320 to, yes, it's mostly marketing copy. It's mostly copy for blogs and 589 00:43:10.440 --> 00:43:15.719 articles and emails and all sorts of other things. So a lot of that 590 00:43:15.719 --> 00:43:17.760 stuff's going to be going on, but I'm telling you other things are in 591 00:43:17.800 --> 00:43:22.840 the cross heirs. Interactive AI is going to be huge. Interactive AI is 592 00:43:22.840 --> 00:43:27.760 where we're going to get much deeper in terms of letting AI run systems and 593 00:43:27.880 --> 00:43:32.760 optimize information architectures, optimize how things get done in the real world. That's 594 00:43:32.800 --> 00:43:45.760 coming soon to an enterprise near you. Respect the legacy. Southern California's KCAA, 595 00:43:46.360 --> 00:43:53.800 the number one talk radio station in the Inlet Empire. KCAA Radio has 596 00:43:53.840 --> 00:43:58.400 openings for one hour talk shows. If you want to host a radio show, 597 00:43:58.559 --> 00:44:01.639 now is the time. Make ACA your flangship station. Our rates are 598 00:44:01.679 --> 00:44:07.079 affordable and our services are second to none. We broadcast to a population of 599 00:44:07.199 --> 00:44:10.639 five million people plus. We stream and podcast on all major online audio and 600 00:44:10.719 --> 00:44:15.679 video systems. If you've been thinking about broadcasting a weekly radio program on real 601 00:44:15.880 --> 00:44:21.880 radio plus the internet, contact our CEO at two eight one five nine nine 602 00:44:22.119 --> 00:44:25.239 ninety eight hundred two eight one five nine nine ninety eight hundred. You can 603 00:44:25.280 --> 00:44:29.880 skype your show from your home to our Redlands, California studio, where our 604 00:44:29.960 --> 00:44:34.639 live producers and engineers are ready to work with you personally. A radio program 605 00:44:34.679 --> 00:44:38.840 on KCIA is the perfect work from home advocation in these stressful times. Just 606 00:44:38.920 --> 00:44:44.559 type Kcaaradio dot com into your browser to learn more about hosting a show on 607 00:44:44.639 --> 00:44:47.480 the best station in the nation, or call our CEO for details. Two 608 00:44:47.519 --> 00:44:54.239 eight one, five nine, nine ninety eight hundred open for takeout and delivery. 609 00:44:54.280 --> 00:44:59.440 Hel Tapiac, Mexican food restaurant in the Tri City Center of Redlands, 610 00:44:59.760 --> 00:45:02.760 is back. Their entire family is on hand to serve up their delicious burritos, 611 00:45:02.800 --> 00:45:07.079 but chaka charizo, wavos, Ranchero's steak and eggs just part of their 612 00:45:07.239 --> 00:45:12.280 mouth watering great food. Since fifteen thirty one, people have marveled at the 613 00:45:12.280 --> 00:45:15.760 miracle of El Tapiac, and now you can marvel at the great food the 614 00:45:15.880 --> 00:45:19.800 Lugo family has been serving up for over two decades, Nestled quietly in the 615 00:45:19.800 --> 00:45:23.079 corner of the Tri City Center shopping mall next to Burlington Code Factory. Support 616 00:45:23.119 --> 00:45:27.599 them. They can't wait to serve you some of their delectable, authentic South 617 00:45:27.639 --> 00:45:30.320 of the Border Mexican fair at great prices, served up with love. Support 618 00:45:30.320 --> 00:45:35.400 the area's best loved Mexican food restaurant in these tough times. Order up a 619 00:45:35.440 --> 00:45:38.559 tasty meal on the phone for delivery or takeout for breakfast, lunch or dinner 620 00:45:38.639 --> 00:45:43.559 from ten am to six pm called nine to nine three oh seven zero zero 621 00:45:43.559 --> 00:45:47.519 one seven. That's nine o nine three oh seven zero zero seventeen or google 622 00:45:47.519 --> 00:45:54.519 el Tapiac Redlands and treat yourself. An auto dealer says he's saving all the 623 00:45:54.519 --> 00:45:59.960 good cars for himself. It's the Onion Radio News. This is Doyle Redland 624 00:46:00.119 --> 00:46:06.239 Reporting. Owner Jim Gannetti of Gannetti Chevrolet admitted to reporters today that his slogan 625 00:46:06.440 --> 00:46:10.079 the best cars for the best prices is a lie. Instead of selling the 626 00:46:10.119 --> 00:46:15.719 best cars, Gonnetti has secretly been keeping all the best cars for himself. 627 00:46:15.079 --> 00:46:20.400 He now says the guilt and storage costs have finally forced him to come clean. 628 00:46:21.039 --> 00:46:23.119 It was too much. Every time I heard my own jingle, I 629 00:46:23.920 --> 00:46:28.519 just felt like a fraud. Gunnetti added that he was afraid of becoming like 630 00:46:28.599 --> 00:46:32.360 his father, legendary auto dealer Carl Gannetti, who died alone in a house 631 00:46:32.440 --> 00:46:37.840 filled with two hundred eighty three Sedans. Doyle Redland for The Onion Radio News 632 00:46:37.960 --> 00:46:44.519 Online at the onion dot com to Hebo Tea Club's original pure POWD D'Arco super 633 00:46:44.519 --> 00:46:47.280 Tea helps build red corpusos in the blood, which carry oxygen to organs and 634 00:46:47.360 --> 00:46:52.840 cells. Our organs and cells need oxygen to regenerate themselves. The immune system 635 00:46:52.880 --> 00:46:55.519 needs oxygen to develop, and cancer dies in oxygen. So the tea is 636 00:46:55.559 --> 00:47:00.159 great for healthy people because it helps build the immune system and be miraculous for 637 00:47:00.199 --> 00:47:05.400 some inviting a potentially life threatening disease due to an infection, diabetes, or 638 00:47:05.440 --> 00:47:08.599 cancer. The ta is also organic and naturally caffeine free. A one pound 639 00:47:08.639 --> 00:47:13.559 package of tea is forty nine ninety five, which includes shipping. To order, 640 00:47:13.599 --> 00:47:16.239 please visit to Hebot Club dot com. T hebow is spelled tea like 641 00:47:16.360 --> 00:47:22.119 tom, a h ee b like boyoh, then continue with the word t 642 00:47:22.559 --> 00:47:25.719 and then the word club. The complete website is to Hebot Club dot com 643 00:47:25.840 --> 00:47:30.360 or call us at eight one eight six one zero eight zero eight eight Monday 644 00:47:30.400 --> 00:47:36.880 through Saturday nine am to five pm California time. That's eight one eight six 645 00:47:37.000 --> 00:47:44.159 one zero eight zero eight eight to Hebot Club dot com. Our society has 646 00:47:44.199 --> 00:47:50.880 coined expressions like philanthropist to encourage and hail people's charitable spirit. Look on the 647 00:47:50.880 --> 00:47:53.800 flip side of that shiny coin of generosity, however, and you will find 648 00:47:53.840 --> 00:48:00.639 that its base substance is societal selfishness. After all, the knee for charity 649 00:48:00.719 --> 00:48:07.280 only exists because we're tolerating intentional injustices and widespread inequality created by power elites. 650 00:48:07.960 --> 00:48:14.599 A society as supremely wealthy as ours ought not be relegating needy families and essential 651 00:48:14.599 --> 00:48:19.760 components of the common good to the whims of a few rich philanthropists. Yes, 652 00:48:20.119 --> 00:48:23.159 corporate and individual donations can help at the margins, but they don't fix 653 00:48:23.239 --> 00:48:29.639 anything. Thus, food banks, health clinics, etc. Must constantly scrounge 654 00:48:29.639 --> 00:48:35.119 for more charity, while big donors have their charitable spirits subsidized with tax breaks 655 00:48:35.119 --> 00:48:39.760 that siphon money from our public treasury. Especially offensive to me is the common 656 00:48:39.800 --> 00:48:45.119 grandiose assertion by fat cat donors that charity is their way of quote giving back 657 00:48:45.199 --> 00:48:51.239 to society. Hello, if they can give so much, it's probably because 658 00:48:51.280 --> 00:48:55.559 they've been taking too much. As business columnists Andrew Ross Sorkin points out, 659 00:48:55.880 --> 00:49:00.079 all too often, charitable gifts are used to make up for the failure of 660 00:49:00.119 --> 00:49:05.960 companies to pay people a living wage and treat their workers with dignity. It's 661 00:49:06.000 --> 00:49:09.400 not just the unemployed who rely on food banks, but janitors, nannies, 662 00:49:09.559 --> 00:49:14.880 uber drivers, check out clerks, and others who work full time but are 663 00:49:14.880 --> 00:49:19.519 so poorly paid they can't make ends meet. That's not a sad charity case, 664 00:49:19.880 --> 00:49:23.679 but a matter of criminal exploitation by wealthy elites, and the charitable thing 665 00:49:23.719 --> 00:49:29.440 to do is to outlaw it and require a living wage for all. This 666 00:49:29.599 --> 00:49:32.599 is Jim Hitier saying, as Sorkin puts it, the aim should be to 667 00:49:32.719 --> 00:49:37.199 create a society where we don't need places like food banks. We should be 668 00:49:37.239 --> 00:49:43.079 trying to put the food banks out of business. It's that time of year 669 00:49:43.159 --> 00:49:46.400 again, No, not the holidays. Medicare open enrollment and if you have 670 00:49:46.519 --> 00:49:51.639 questions about Medicare, you should talk to the local experts. Paul Barrett and 671 00:49:51.719 --> 00:49:54.960 Associates. All of his agents are certified with plans that are accepted by most 672 00:49:55.039 --> 00:50:00.480 of the medical groups in our area. Call nine oh nine, seven oh 673 00:50:00.519 --> 00:50:04.559 three eighty five. There are services free and after forty two years of the 674 00:50:04.599 --> 00:50:07.280 business, their agents are trained to help you pick the plan that's right for 675 00:50:07.360 --> 00:50:14.039 you. For several years KCAA has been marketing the Youngevity brand of nutritional and 676 00:50:14.079 --> 00:50:19.320 personal care products. Our experience with Youngevity has been one hundred percent positive, 677 00:50:19.719 --> 00:50:22.920 so we are pleased to recommend them to you. Regarding nutritional supplements, we 678 00:50:23.000 --> 00:50:28.960 recommend pollen Burst in the berry flavor and Tangy Tangerine two point zero in the 679 00:50:29.000 --> 00:50:32.960 tablet form. For regularity issues, we recommend three day cleanse, and for 680 00:50:34.039 --> 00:50:38.519 personal care, we recommend morning hydration cream. You can shop online for Youngevity 681 00:50:38.599 --> 00:50:45.000 at www dot KCAA team dot com, or you can order by phone by 682 00:50:45.039 --> 00:50:49.119 calling eight hundred ninet eighty two three one ninety seven and tell customer support that 683 00:50:49.199 --> 00:50:53.079 you are part of the KCAA team. Youngevity is an American company based in 684 00:50:53.119 --> 00:50:59.000 San Diego. Call Youngevity at eight hundred ninety two three one ninety seven and 685 00:50:59.039 --> 00:51:04.280 ask about monthly autoship that allows you to buy you longevity products at wholesale prices. 686 00:51:04.639 --> 00:51:09.480 That number again eight hundred nine eighty two three one ninety seven. Tune 687 00:51:09.519 --> 00:51:15.360 into the Faran Dozier Show Music Marks Place in Time, the soundtrack to Life. 688 00:51:15.599 --> 00:51:20.679 Sunday nights at eight pm are KCAA Radio playing the hottest hits and the 689 00:51:20.719 --> 00:51:25.400 coolest conversations Sunday nights at APM on The Feran Dozier Show within the array of 690 00:51:25.519 --> 00:51:31.480 music, talk, sports, community outreach and veteran resources with hits from the 691 00:51:31.599 --> 00:51:37.440 sixties, seventies, eighties, nineties, and today's hits. The Faran Dozier 692 00:51:37.559 --> 00:51:45.400 Show on KCAA Radio, on all available streaming platforms and ONLO six point five 693 00:51:45.599 --> 00:52:08.960 FM and ten fifty Am. The Farando Zier Show on KCAA Radio. Was 694 00:52:09.039 --> 00:52:14.360 your car involved in an accident or just need help with dense All Magic Paint 695 00:52:14.400 --> 00:52:19.119 and Body Collision Centers in business for over thirty years. They're highly trained staff 696 00:52:19.239 --> 00:52:22.719 and certified technicians and friendly staff are the best in the business and treat each 697 00:52:22.800 --> 00:52:27.360 car as if it was their own. All Magic Paint and Body Collision Centers 698 00:52:27.400 --> 00:52:30.920 are family owned and offer state of the art equipment and tools to ensure optimum 699 00:52:30.920 --> 00:52:36.840 results. They use the latest technology in computerized color matching and specialize in frame 700 00:52:36.960 --> 00:52:42.320 repairs with their modern laser measuring systems. They're OEM certified and they have four 701 00:52:42.360 --> 00:52:45.519 locations to serve you. All Magic Paint and Body Collision Centers offer rental car 702 00:52:45.559 --> 00:52:51.239 assistance with free dropoff and pickup services too, and their work has a lifetime 703 00:52:51.320 --> 00:52:54.480 guarantee. All Magic Paint and Body Collision Centers are in Narco, East Vale, 704 00:52:54.559 --> 00:52:59.519 Marino Valley and in Fontana. Call them at one eight hundred and sixty 705 00:52:59.519 --> 00:53:02.920 one Magic. That's one eight hundred sixty one Magic All Magic Paint and Body 706 00:53:02.920 --> 00:53:07.719 Collision Centers one eight hundred sixty one Magic, All Magic Paint and All Bodies 707 00:53:07.760 --> 00:53:16.320 says drive carefully. You like to safely leverage bank money to earn double digit 708 00:53:16.400 --> 00:53:22.639 returns income tax free, with guarantees and no downside market risk. How can 709 00:53:22.679 --> 00:53:25.000 you do this? This is Farence, host of the Your Personal Bank Show. 710 00:53:25.199 --> 00:53:30.880 One You fund a high cash value policy one time to earn dividends and 711 00:53:30.960 --> 00:53:35.239 interest. Two establish a bank line of credit using the cash in your policy 712 00:53:35.280 --> 00:53:38.280 as collateral. When you earn more in dividends from your policy than the interest 713 00:53:38.360 --> 00:53:43.199 the bank charges, you keep the difference, and the difference is average two 714 00:53:43.199 --> 00:53:46.679 to five percent annually in your favor for the past forty plus years. Three, 715 00:53:46.760 --> 00:53:52.119 the bank funds contributions years two to twenty plus. Each year the bank 716 00:53:52.159 --> 00:53:54.639 adds funds, your rate of return increases. Your average rate of return can 717 00:53:54.679 --> 00:54:00.519 grow too strong double digits annually within a few years. Contact me your personal 718 00:54:00.559 --> 00:54:04.639 bank dot com, your personal bank dot com or eight sixty six two six 719 00:54:04.760 --> 00:54:07.920 eight four four two two eight sixty six two six eight four four two two 720 00:54:08.079 --> 00:54:14.159 for more info, or tune into your Personal Bank Show. Your Personal Bank 721 00:54:14.159 --> 00:54:17.960 Show airs Tuesdays at four pm right here on CASEAA ten fifty AM and one 722 00:54:19.000 --> 00:54:23.480 oh six point five FM, the station that leaves no listeners behind Bob Vila 723 00:54:23.519 --> 00:54:27.760 here with my home improvement tip of the day, how much snow on the 724 00:54:27.840 --> 00:54:30.039 roof is too much? That depends a lot on the way your roof was 725 00:54:30.039 --> 00:54:35.960 constructed. Steep and smooth roofs tend to shed snow easily, while roofs that 726 00:54:36.000 --> 00:54:39.599 are only slightly pitched or flat tend to collect big drifts. Another important factor 727 00:54:39.639 --> 00:54:44.119 is the weight of the snow. Removing a heavy snowload can be tricky. 728 00:54:44.400 --> 00:54:46.519 If you have a multi story house, you'd best not be climbing up and 729 00:54:46.559 --> 00:54:52.239 down icy cold ladders to dizzying heights. Better to leave that to licensed, 730 00:54:52.280 --> 00:54:54.679 insured pros who have the right equipment to get the job done right. On 731 00:54:54.760 --> 00:54:58.400 the other hand, if you have a single story home, you can use 732 00:54:58.400 --> 00:55:01.679 a long telescoping snow rate to pull snow off the roof. One caution, 733 00:55:01.800 --> 00:55:06.800 though, Rakes that come into contact with shingles can do a lot of damage, 734 00:55:06.880 --> 00:55:09.119 so look for sturdy models with small rollers that keep the edge of the 735 00:55:09.199 --> 00:55:14.119 rake away from the shingles. Finally, before you start pulling snow off the 736 00:55:14.159 --> 00:55:16.320 roof, put some thought into where the snow's going to land. You don't 737 00:55:16.320 --> 00:55:20.760 want to damage your plants. Get more info at Bob Vila dot com and 738 00:55:20.920 --> 00:55:25.960 right here at home with me, Bob Vila. It's a bird, it's 739 00:55:27.000 --> 00:55:31.519 a bird. No, it's super raw. Okay, a gimmicky opening for 740 00:55:31.559 --> 00:55:36.719 a commercial about super Roth universal life insurance, but I'm sure it got your 741 00:55:36.719 --> 00:55:39.559 attention. Now. What is a super Roth, you ask. It's a 742 00:55:39.559 --> 00:55:45.519 permanent indexed universal life insurance that's totally liquid and easily accessible once it matures. 743 00:55:45.599 --> 00:55:50.519 Can be used to supplement retirement savings or a death benefit, or both. 744 00:55:50.760 --> 00:55:54.599 Has no income or contribution limit, has no five year rule like rawth iras, 745 00:55:54.960 --> 00:56:00.239 has no ten percent penalty for accessing the funds before age fifty nine nine 746 00:56:00.239 --> 00:56:02.760 and a half. Oh, and the average historical returns are five to seven 747 00:56:02.800 --> 00:56:08.079 percent annually tax free super rocks also lock in gains, which means you don't 748 00:56:08.079 --> 00:56:12.360 lose your money when the market is down. It sounds incredible, right, 749 00:56:12.760 --> 00:56:16.880 sounds super super Rocks are the way of the future, specifically your future. 750 00:56:17.159 --> 00:56:22.400 To see if you qualify for a super Roth, go online to the Superwroth 751 00:56:22.519 --> 00:56:27.519 dot Com. For over a century, AM radio has evolved to meet the 752 00:56:27.559 --> 00:56:30.840 needs of our community. More than eighty million listeners depend on AM radio each 753 00:56:30.880 --> 00:56:35.760 month. It's also the backbone of the emergency alert system, keeping us safe 754 00:56:35.760 --> 00:56:38.800 in dangerous times. A new bill in Congress would ensure this free, reliable 755 00:56:38.840 --> 00:56:44.039 service remains in cars. Text AM to five two eight eighty six and tell 756 00:56:44.079 --> 00:56:47.440 Congress to support the AM Radio for every vehicle act message data rates may play. 757 00:56:47.599 --> 00:56:50.639 You may receive up to four messages a month, and you may text 758 00:56:50.639 --> 00:56:54.880 stop to stop This message Furnished by the National Association of Broadcasters. Did you 759 00:56:54.920 --> 00:57:00.480 know here at KCAA ten fifty AM that we developed an app for all your 760 00:57:00.519 --> 00:57:05.039 Android devices. We're talking about your smartphone, your tablets, you name it. 761 00:57:05.280 --> 00:57:09.039 You have an Android format. You can take KCAA with you everywhere you 762 00:57:09.119 --> 00:57:14.199 go. We're talking about our audio stream, our video stream, and even 763 00:57:14.239 --> 00:57:22.280 our podcast go to KCAA Express dot com. That's CASECAA Express dot com KCAA 764 00:57:22.440 --> 00:57:28.800 Express dot com. This segment is sponsored by press Print, Southern California's best 765 00:57:28.800 --> 00:57:32.280 full service union printer and mailhouse. They offer the lowest prices around with unmatched 766 00:57:32.280 --> 00:57:37.960 service and reliability and free delivery throughout Socount. Press Print can print anything from 767 00:57:37.000 --> 00:57:42.119 letterhead, business cards, and campaign literature to mailers of any size, lawn 768 00:57:42.159 --> 00:57:45.639 signs, banners, doorhangers, or just about anything you might want. Press 769 00:57:45.679 --> 00:57:49.119 Print promises to save money for you, your business, or your campaign. 770 00:57:49.159 --> 00:57:52.079 If you'd like to learn more, contact Mike Kroumbrint at press Print seven one, 771 00:57:52.199 --> 00:57:57.119 four, three nine nine eight seven zero eight. Get the Union Bug 772 00:58:00.400 --> 00:58:04.440 NBC News Radio. I'm Chris Karrazio. Secretary of Saint Anthony Blincoln made a 773 00:58:04.440 --> 00:58:07.920 pair of unannounced visits in the Middle East. Today, he met with Palestinian 774 00:58:07.960 --> 00:58:12.679 Authority President Makmudabas during a surprise stop on the occupied West Bank. Then later 775 00:58:12.880 --> 00:58:15.840 Blincoln traveled to Baghdad and said he believes progress is being made as the US 776 00:58:15.920 --> 00:58:21.079 works to keep the conflict from spreading. Two US senators who are back from 777 00:58:21.119 --> 00:58:25.280 a recent trip to Israel are introducing a bipartisan resolution to deter Iran from escalating 778 00:58:25.280 --> 00:58:30.760 the Israel Hamas War. If Hezblah opens up a second front in the North 779 00:58:30.800 --> 00:58:35.519 against Israel in a substantial way, then we should hit the Islamic Republic of 780 00:58:35.559 --> 00:58:39.920 Iran. South Carolina Republican Lindsay Graham and Connecticut Democrat Richard Blumenthal set on CNN 781 00:58:39.960 --> 00:58:44.440 State of the Union. The non binding resolution calls for putting i Ran on 782 00:58:44.519 --> 00:58:49.079 notice that the US will come after them if any other militias backed by Iran 783 00:58:49.159 --> 00:58:52.880 try to expand the war. Bloomenthal called the resolution aggressive but absolutely necessary. 784 00:58:53.039 --> 00:58:58.440 Both senators believe that along with receiving bipartisan support in the Senate, the resolution 785 00:58:58.519 --> 00:59:01.920 would receive support in the Middle East from other countries besides Israel, They said. 786 00:59:01.960 --> 00:59:07.519 Egypt and Saudi Arabia both loathe and fear hamasen has Blaw as much as 787 00:59:07.599 --> 00:59:10.679 Israel does. Former President Trump is scheduled to take the stand in the Trump 788 00:59:10.760 --> 00:59:15.800 organization's civil fraud trial in New York tomorrow. Scott Carr has More. Trump's 789 00:59:15.800 --> 00:59:21.320 sons Rik and Donald Trump Junior, testified last week in the trial that accuses 790 00:59:21.360 --> 00:59:25.880 the Trump family organization of inflating their wealth to obtain better loans. Daughter Ivanka 791 00:59:25.960 --> 00:59:30.159 Trump will testify Wednesday, despite her attempt to legally block it, telling the 792 00:59:30.199 --> 00:59:35.239 court she would suffer undue hardship if forced to testify in the middle of a 793 00:59:35.280 --> 00:59:39.599 school week. Last week, Presiding Judge Arthur Engern prohibited Trump's attorneys from making 794 00:59:39.639 --> 00:59:45.199 any further comments about confidential communications between the judge and his staff. Judge Ngaran 795 00:59:45.280 --> 00:59:51.239 says three of Trump's lawyers falsely accused his law clerk of bias and of improperly 796 00:59:51.320 --> 00:59:55.960 influencing the trial. I'm Scott Carr, I'm Chris Karagio, NBC News Radiohilistinea 797 00:59:57.039 --> 01:00:00.400 CACAA, Lomalinda and one O six point five m K two ninety three c 798 01:00:00.519 --> 01:00:08.000 F Burrito Valley from the Bureau of Economic Geology. This is Earth date.