WEBVTT 1 00:00:01.480 --> 00:00:07.280 NBC News on CACAA Lomelada, sponsored by Teamsters Local nineteen thirty two, Protecting 2 00:00:07.320 --> 00:00:17.839 the Future of working Families Teamsters nineteen thirty two dot org. The information economy 3 00:00:17.960 --> 00:00:22.640 has a ride. The world is teeming with innovation as new business models reinvent 4 00:00:22.879 --> 00:00:28.839 every industry industry. Inside Analysis is your source of information and insight about how 5 00:00:28.839 --> 00:00:33.479 to make the most of this exciting new era. Learn more at Inside analysis 6 00:00:33.560 --> 00:00:43.399 dot Comsideanalysis dot com. And now here's your host, Eric Kavanaugh. All 7 00:00:43.520 --> 00:00:47.200 Right, folks, hello, and welcome to the only coast to coast radio 8 00:00:47.240 --> 00:00:50.799 show in the US of A that's all about the information economy. It's time 9 00:00:50.840 --> 00:00:55.039 for Inside Analysis, or truly Eric Kavanaugh is here part of the dm Radio 10 00:00:55.159 --> 00:00:58.679 Broadcasting Network, and folks, I'm very excited to have my good buddy and 11 00:00:58.840 --> 00:01:06.879 broadcasting legend Jim Harris on the line. He is quite the social media influencer 12 00:01:07.000 --> 00:01:08.599 these days, has been for a while. He's been going to Davos for 13 00:01:08.719 --> 00:01:11.879 years, he's been going to the World Economic Forum, he's been going to 14 00:01:12.000 --> 00:01:15.920 CEES, the Consumer Electronics Show for years, and in fact, he is 15 00:01:17.040 --> 00:01:23.480 almost always one of the top voices individual voices for social media amplification at these 16 00:01:23.519 --> 00:01:26.079 events. I get his tweets all the time, and they're always very carefully 17 00:01:26.079 --> 00:01:32.159 crafted. He always finds really interesting trends to talk about, really cool stuff 18 00:01:32.280 --> 00:01:36.560 like solar panels or new technologies that are coming out. Of course, drones 19 00:01:36.560 --> 00:01:40.799 are everywhere, and today we wanted to talk about what we saw at CEES 20 00:01:40.879 --> 00:01:44.079 and what he saw at DeVos. I've never been the DeVos, at least 21 00:01:44.079 --> 00:01:49.200 not yet, but maybe next year I'll be rubbing elbows with the world shakers 22 00:01:49.200 --> 00:01:52.840 and movers out there. But Jim, welcome to the show. Thanks so 23 00:01:52.920 --> 00:01:57.439 much for doing our program today, And real quick thoughts on CS. What 24 00:01:57.519 --> 00:02:01.519 were your impressions from the show show this year? Well, Eric, great 25 00:02:01.560 --> 00:02:07.199 to be on the show. And for both CEES and for DAVOS, the 26 00:02:07.240 --> 00:02:15.319 top topic was AI. And every year Google has its Google Io Developer Conference, 27 00:02:16.000 --> 00:02:23.120 and you know, it's a multi day event. The CEO has a 28 00:02:23.199 --> 00:02:34.520 couple hour kenote and somebody summarized the entire keynote and it's into about twelve seconds 29 00:02:34.560 --> 00:02:39.840 and it goes something like this AI AI, we're embedding AI, taking generative 30 00:02:39.879 --> 00:02:46.960 AI, AI, AI AI AI. And really that is what came out 31 00:02:46.000 --> 00:02:54.240 of CEES and out of Davos. The top trending hashtag for both events was 32 00:02:54.280 --> 00:02:59.599 AI, and I'm sure it will be for Mobile World Congress at the end 33 00:02:59.639 --> 00:03:05.560 of I Burian start of March in Barcelona, which is the top global mobile 34 00:03:05.919 --> 00:03:12.120 connectivity event. So AI is the hottest topic, and it has been since 35 00:03:12.240 --> 00:03:19.560 chat gptt launched November thirtieth, twenty twenty two. Now, I've been talking 36 00:03:19.599 --> 00:03:25.759 about AI for a decade and people just didn't get it. But once chat 37 00:03:25.759 --> 00:03:32.919 GPT had this conversational interface, people understood it, and it is the hottest 38 00:03:32.960 --> 00:03:38.159 topic. Yeah, I mean, it absolutely is. And it's funny because 39 00:03:38.879 --> 00:03:42.360 chat GPT was not new. There were earlier versions of it that some people 40 00:03:42.360 --> 00:03:45.719 were already playing around with, so it had been there. But I think 41 00:03:45.759 --> 00:03:50.159 it's it's just the fact that they did this rollout and so many people jumped 42 00:03:50.199 --> 00:03:53.120 on to check it out, and their minds were just blown, right, 43 00:03:53.159 --> 00:03:57.120 I mean, their minds were absolutely blown. And even if Sam Altman says 44 00:03:57.159 --> 00:04:00.879 strange things now and again, like you said, the year of large language 45 00:04:00.879 --> 00:04:03.159 models is over, and everyone was like, uh, are you are you 46 00:04:03.199 --> 00:04:08.759 talking? Who are you again? I think he was talking about small models, 47 00:04:08.800 --> 00:04:12.759 which is very interesting. There's mistral, there's this sort of they call 48 00:04:12.800 --> 00:04:15.399 it a combination of experts or mixture of experts or something, which are smaller 49 00:04:15.439 --> 00:04:21.319 models. And my buddies from the Northeastern University, the Institute for Experiential AI, 50 00:04:21.519 --> 00:04:26.519 are saying that you don't have to be bigger to be better with these 51 00:04:26.639 --> 00:04:30.360 language models, right, and so there's a tremendous amount of innovation in that 52 00:04:30.439 --> 00:04:32.399 space. I mean, it's really it's kind of blowing my mind. It 53 00:04:32.480 --> 00:04:35.800 reminds me really of the had Dupe days. For those of you who did 54 00:04:35.839 --> 00:04:40.000 not know about had Dupe. It came out of Yahoo. It was their 55 00:04:40.040 --> 00:04:44.079 engine for indexing the web. It was turned over to Google. They open 56 00:04:44.160 --> 00:04:47.360 sourced it, and then companies like Cloudera jumped on and Horton Works and map 57 00:04:47.480 --> 00:04:53.160 r and at one point there were like seven distributions of headdoup of the headdoup 58 00:04:53.240 --> 00:04:56.920 distributed file system hdfs, and then there were six, and there were five 59 00:04:56.959 --> 00:04:58.959 of four and three and two and one, and they're like, Okay, 60 00:04:59.000 --> 00:05:00.720 that was too much. I feel that way about these models, all these 61 00:05:01.000 --> 00:05:05.720 foundational models. But if you look at where it's all going, I mean, 62 00:05:05.879 --> 00:05:13.959 these AI models will absolutely revolutionize consumer software, enterprise software. They're being 63 00:05:14.000 --> 00:05:16.879 baked into everything. I mean when I walked around at CEES, many of 64 00:05:16.879 --> 00:05:21.279 the booth were talking about AI and their smart devices and their smart systems. 65 00:05:21.360 --> 00:05:29.240 Right, So it's already permeated the electronics industry, right. Yeah. AI 66 00:05:29.439 --> 00:05:35.160 is being embedded in hardware, in software, in internal company processes, in 67 00:05:35.279 --> 00:05:42.160 external customer facing applications, so it's being embedded everywhere. And I go to 68 00:05:42.439 --> 00:05:47.560 more than sixty conferences a year ERIC and at everyone, AI is the top 69 00:05:47.680 --> 00:05:54.759 trending topic. And if you're a startup seeking to raise money from a VC 70 00:05:55.480 --> 00:06:00.959 and you don't have the words AI jen ai in your pitch deck, you're 71 00:06:00.959 --> 00:06:05.279 not getting funded. So it is this relentless drum beat. And I believe 72 00:06:05.480 --> 00:06:13.040 that the innovation we're seeing right now around AI and jen ai is like the 73 00:06:13.079 --> 00:06:16.279 birth of the Web in ninety three. And in nineteen ninety three, we 74 00:06:16.399 --> 00:06:23.240 had the Internet beforehand, but it was hard to use, it was cryptic, 75 00:06:23.360 --> 00:06:28.639 it was user vicious, and you know, you had to remember Eunuch's 76 00:06:28.759 --> 00:06:32.079 command. So it was there, but nobody in the general public knew about 77 00:06:32.079 --> 00:06:39.519 it. Researchers in academ and in the military knew about it, but the 78 00:06:39.680 --> 00:06:44.720 people didn't know about it. And once the web was born and you could 79 00:06:45.040 --> 00:06:49.160 surf around just by clicking with your mouse, it exploded. And so I 80 00:06:49.319 --> 00:06:53.680 was telling my clients in ninety three, this is going to change everything, 81 00:06:54.120 --> 00:06:57.360 and they looked at me, some of them like I had two heads, 82 00:06:57.480 --> 00:07:02.720 you know. But here today we can see that it is baked into everything 83 00:07:02.879 --> 00:07:09.279 thirty years later, and the same is going to happen with AI and generative 84 00:07:09.319 --> 00:07:16.120 AI. But the pace of change is infinitely faster. We're turbocharging change right 85 00:07:16.160 --> 00:07:20.920 now. Yeah, yeah, that's exactly right. And I should say there 86 00:07:20.920 --> 00:07:25.680 are lots of other traditional kinds of technologies that were at the show. I 87 00:07:25.720 --> 00:07:30.120 stopped by the Logitech suite and I checked out some of their new stuff. 88 00:07:30.120 --> 00:07:33.079 It's very interesting the Logitech Reach. I did want to give some time to 89 00:07:33.120 --> 00:07:39.800 those folks. Very simple product. The announcement date was, I guess back 90 00:07:39.800 --> 00:07:43.079 in September, and what's cool with this? So it's basically it's just it's 91 00:07:43.079 --> 00:07:46.279 called the Logitech Reach. It's just a boom device for your camera, basically. 92 00:07:46.639 --> 00:07:49.600 But they did a couple of really clever things, one of which is 93 00:07:51.439 --> 00:07:56.399 a kind of handle you can grab onto such that the perspective doesn't change when 94 00:07:56.399 --> 00:07:59.759 you go around a table, like if you can imagine going around like this 95 00:08:00.160 --> 00:08:03.680 showing different objects on the table, it will keep the perspective the same instead 96 00:08:03.680 --> 00:08:07.120 of turning things upside down. Right, you have to kind of see it 97 00:08:07.120 --> 00:08:09.839 to appreciate it. But you think about how many creators there are out there 98 00:08:09.879 --> 00:08:13.000 for this whole creator economy. You think how many people are trying to do 99 00:08:13.079 --> 00:08:16.959 training online all this stuff. And what's really cool about this is how they 100 00:08:18.079 --> 00:08:22.360 launched it. So Logitech went out and they did a go fundbe campaign when 101 00:08:22.360 --> 00:08:26.759 they basically said, hey, this is a new product we're going to develop. 102 00:08:26.000 --> 00:08:28.519 If we can raise this money, this is what it'll do, and 103 00:08:28.560 --> 00:08:31.159 if you want to be the first one to get it, just go ahead, 104 00:08:31.240 --> 00:08:35.399 donate some money to this gofundb page and we'll have you on the top 105 00:08:35.440 --> 00:08:39.519 of the list. They sawd out in like five minutes. They raised all 106 00:08:39.519 --> 00:08:43.159 the money in five minutes. I'm like, what, that's amazing, What 107 00:08:43.279 --> 00:08:48.399 a clever plan. Now, this is really interesting. This is innovating in 108 00:08:48.519 --> 00:08:54.279 how we innovate. Historically, companies spent two million dollars to develop a new 109 00:08:54.320 --> 00:08:58.639 product and then launched it and it might go well, it might flop or 110 00:08:58.639 --> 00:09:05.919 whatever. With a GoFundMe campaign with crowd sourcing, hey do you really want 111 00:09:05.919 --> 00:09:11.360 this? They totally de risk innovation. That's right. And I was on 112 00:09:11.519 --> 00:09:18.519 a program with a global leader from Lego. You know, Lego's pretty classic. 113 00:09:18.679 --> 00:09:24.279 It's made out of plastic. It comes in little squares and shapes and 114 00:09:24.320 --> 00:09:28.320 as pegs and you walk it together. We all played with it as kids. 115 00:09:28.840 --> 00:09:31.840 Well, what they've done is they've said to kids, hey, if 116 00:09:31.879 --> 00:09:37.639 you want a certain type of Lego set, you tell us what it is, 117 00:09:37.840 --> 00:09:41.120 and you get ten thousand of your friends to say they want it to 118 00:09:41.919 --> 00:09:46.840 and you do the marketing, like show us what a campaign would look like. 119 00:09:46.639 --> 00:09:52.799 They make all the little children do the work, and then they once 120 00:09:52.159 --> 00:09:58.399 if they get ten thousand kids excited about it, the product launches. That's 121 00:09:58.399 --> 00:10:05.159 amazing. Any guesses on how big this crowdsource business is? Yeah, I 122 00:10:05.240 --> 00:10:11.200 feel like Doctor Evil. A billion dollars a year, A billion dollars a 123 00:10:11.279 --> 00:10:13.480 year off of child labor, right, yeah, right, that's that's pretty 124 00:10:13.519 --> 00:10:16.000 funny. Well, and they're not the only ones, right, So this 125 00:10:16.159 --> 00:10:20.759 is a trend. And I noticed that the Consumer Electronics Show at CS and 126 00:10:22.120 --> 00:10:24.120 it was so cool to see you there. And folks, we had our 127 00:10:24.120 --> 00:10:26.480 own booth, our own recording studio for DM Radio. So we'll be back 128 00:10:26.519 --> 00:10:31.559 next year if you want to be recorded. Eric who for the other media 129 00:10:31.639 --> 00:10:37.360 that we're in our block, Oh, Reuters and CNN and Bloomberg and the 130 00:10:37.399 --> 00:10:43.120 Associated Press and DM Radio. Yeah, baby hit the big time. But 131 00:10:43.240 --> 00:10:48.000 one of the other companies we looked at was LG Life's Good and they gave 132 00:10:48.039 --> 00:10:50.639 me a whole tour of their technologies and they did something very similar to what 133 00:10:50.679 --> 00:10:54.720 the folks at Logitech are doing, and that they bring out all kinds of 134 00:10:54.759 --> 00:10:58.440 prototype technologies and put them on the show floor and then they watch and see 135 00:10:58.480 --> 00:11:03.559 who likes this, who likes that. It's a brilliant strategy for knowing what 136 00:11:03.759 --> 00:11:07.200 to work on right And these days, man, when margins are tight, 137 00:11:07.519 --> 00:11:11.360 when the economy is you know, it's not in recession, but it's teetering. 138 00:11:11.440 --> 00:11:15.720 It's just strange. It's a very strange economy right now. You have 139 00:11:15.759 --> 00:11:18.600 to be careful about stuff. And I think you nailed it when you said 140 00:11:18.600 --> 00:11:24.120 they have d risked innovation and that is a big deal, right, big 141 00:11:24.159 --> 00:11:30.600 big deal. Yeah, you know, because knowing where to place your bets 142 00:11:30.879 --> 00:11:35.480 is so important, especially in manufacturing because it's got such a long process from 143 00:11:35.559 --> 00:11:39.440 concept, ideation all the way through manufacturing, production, distribution, et cetera. 144 00:11:39.879 --> 00:11:43.279 Well, Logitech with this deal, they expert. They greased those tracks, 145 00:11:43.799 --> 00:11:48.000 just greased tracks when right into production. Bang this thing out and it's 146 00:11:48.039 --> 00:11:52.559 great. It's very simple, but it's something that wasn't out there before, 147 00:11:52.879 --> 00:11:58.080 and they're catering to what the audience wants, right, So I just want 148 00:11:58.120 --> 00:12:03.919 to stress how de risk is. Imagine I went to any executive, any 149 00:12:03.000 --> 00:12:07.919 CEO, and said, hey, if I could take your innovation pipeline and 150 00:12:09.120 --> 00:12:15.679 guarantee you one hundred percent that every new product would be successful out of the 151 00:12:15.720 --> 00:12:20.279 gate, well, right, and it would make you a billion dollars? 152 00:12:20.840 --> 00:12:28.600 Would you pay attention? You should? We should? We really look at 153 00:12:28.600 --> 00:12:33.000 how do we innovate, and how do we innovate and how we innovate? 154 00:12:33.679 --> 00:12:39.240 Yeah, innovation innovation. You gotta love that. It's very meta. It 155 00:12:39.320 --> 00:12:43.559 is very meta innovation innovation. It says crazy and the who else did we 156 00:12:43.600 --> 00:12:46.840 see? I mean I walked around quite a bit on the show floor. 157 00:12:46.879 --> 00:12:50.519 There's LG of course, well MasterCard, right if you look at the interesting 158 00:12:50.559 --> 00:12:56.120 stuff MasterCard was working on. We interviewed Raja Raja Manar from MasterCard. He 159 00:12:56.399 --> 00:13:01.120 is the communications VP, as ever called, and they've done something very interesting 160 00:13:01.159 --> 00:13:05.639 which I think you're going to see in other iterations in other ways where they 161 00:13:05.679 --> 00:13:11.120 got together with a bunch of trusted content creators and they used existing content to 162 00:13:11.240 --> 00:13:18.080 train a model to be a small to mid sized business AI assistant to help 163 00:13:18.120 --> 00:13:20.519 you launch your business. How do you launch a business, how do you 164 00:13:22.000 --> 00:13:24.759 set up your books, how do you get clients? All these little things 165 00:13:26.080 --> 00:13:30.200 that are very useful. And what's cool about this is again like people have 166 00:13:30.240 --> 00:13:35.120 to understand these publicly trained models like chat, GPT, there's a ton of 167 00:13:35.159 --> 00:13:37.519 information in there. I mean, it's going to take us years just to 168 00:13:37.559 --> 00:13:41.720 figure out what's in there. That's how much is in there, right, 169 00:13:41.000 --> 00:13:46.960 But the problem from a business perspective is that's too obtuse. There's too much 170 00:13:46.000 --> 00:13:50.879 information, it's too broad. Now these systems use waitings basically, that's what 171 00:13:50.919 --> 00:13:54.000 they're really doing. And there are tokens, which token is like I think 172 00:13:54.080 --> 00:13:58.840 half a word they say. And in the early days, the AI could 173 00:13:58.840 --> 00:14:05.039 see one token behind it and one token ahead and that's how it actually cranks 174 00:14:05.080 --> 00:14:07.159 out the right. It might blow people's minds to know that. And then 175 00:14:07.200 --> 00:14:11.360 these transformer models came out where now you have almost like a Greek chorus of 176 00:14:13.399 --> 00:14:18.480 recommendation components, little bots, and then there's still one decision maker that goes, 177 00:14:18.519 --> 00:14:20.399 Okay, I'm gonna go with this token decative. But the point is 178 00:14:20.399 --> 00:14:24.639 now they can see like five six tokens right, five six tokens left, 179 00:14:24.720 --> 00:14:28.799 and that's how they develop things. But the point is the more focused you 180 00:14:28.840 --> 00:14:31.919 can be. And this is what they talk about when they talk about embeddings, 181 00:14:31.919 --> 00:14:35.039 for example, and training your model. You can really do it one 182 00:14:35.039 --> 00:14:37.600 of two ways. You can fine tune the model itself on the data you 183 00:14:37.639 --> 00:14:41.440 train it on, or you can put all your embeddings into a vector database 184 00:14:41.480 --> 00:14:48.159 and get this RAG model. What they call regenerative augmented or a retrieval augmented 185 00:14:48.200 --> 00:14:50.639 generation, that's what it stands for, came out of Facebook. And those 186 00:14:50.720 --> 00:14:56.200 embeddings are your moorings basically that those are your anchors of truth that allow that 187 00:14:58.159 --> 00:15:01.840 optimize the quality of the the content that you get back. But that's how 188 00:15:01.879 --> 00:15:07.519 these things work, right, And as you get more and more narrowly focused 189 00:15:07.559 --> 00:15:09.320 on things. That's when you get tremendous value. That's what my buddy, 190 00:15:09.320 --> 00:15:15.720 doctor Sama Fayad says from the Institute for Experiential AI Northeastern University. The narrow 191 00:15:15.840 --> 00:15:18.600 the better. In fact, one guy said, narrow AI is the only 192 00:15:18.679 --> 00:15:22.919 AI. Is that what you're hearing as well? Well, I want to 193 00:15:22.960 --> 00:15:26.960 take it and comment on something else first, to go to a big picture, 194 00:15:26.120 --> 00:15:31.000 which is many people don't know this, but more than ninety nine percent 195 00:15:31.080 --> 00:15:39.000 of companies in the US are SMEs, small and medium sized enterprises small under 196 00:15:39.039 --> 00:15:43.879 one hundred people, medium under five hundred people more than ninety nine percent, 197 00:15:45.720 --> 00:15:56.120 and SMEs by definition can't afford an AI data scientist on staff. So MasterCard 198 00:15:56.559 --> 00:16:00.960 is aiming at a market that really is the engine of our economy globally and 199 00:16:02.000 --> 00:16:10.960 in North America definitely to help them. So a friend of mine is engaged 200 00:16:11.000 --> 00:16:15.960 in a startup. He works, he's working with a hospital and he's used 201 00:16:15.000 --> 00:16:19.840 chat ept the developer model to ingest all their policy manuals, you know, 202 00:16:21.879 --> 00:16:29.519 and so it's a very tightly defined base of knowledge. And you're a frontline 203 00:16:29.600 --> 00:16:34.000 manager in the hospital and you have somebody who's got poor performance. You don't 204 00:16:34.039 --> 00:16:38.039 have to go read sixty pages of manuals on how to you know, do 205 00:16:38.120 --> 00:16:44.919 a performance improvement program. You just say I have a poorly performing employee, 206 00:16:44.919 --> 00:16:49.960 what do I do? And the model will just spit out one You write 207 00:16:49.960 --> 00:16:56.919 them a letter saying your performance is poorer to you develop a PIP performance improvement 208 00:16:56.000 --> 00:17:00.559 program. You know, you do this this, and if after three warnings 209 00:17:00.600 --> 00:17:04.960 they're still bad, you fire them. So but who has time to try 210 00:17:06.000 --> 00:17:11.440 and figure out what to do? Or imagine all government documents for the US 211 00:17:11.559 --> 00:17:17.599 government being ingested, so you don't have to figure out right about how to 212 00:17:17.759 --> 00:17:22.559 fill out this form or do I need that form and this form? So 213 00:17:22.759 --> 00:17:29.119 imagine just making everything simple for people. So this is an amazing set of 214 00:17:29.160 --> 00:17:34.119 outcomes and MasterCard is aiming to do this for the SME market, which I 215 00:17:34.160 --> 00:17:38.240 think is really cool. Yeah, and we're I said, We're going to 216 00:17:38.319 --> 00:17:42.319 see more things like this, And you know, Frankly, I'll be giving 217 00:17:42.319 --> 00:17:48.039 a speech a talk at the Data Universe conference in New York April tenth and 218 00:17:48.079 --> 00:17:52.519 eleventh, and I'll be talking about the death of journalism. And what I'm 219 00:17:52.559 --> 00:18:00.240 talking about is how traditional journalism models organizational structures are out the window. It's 220 00:18:00.240 --> 00:18:07.200 gonna be impossible for them to outflank these large language models. They're gonna have 221 00:18:07.279 --> 00:18:08.720 to join them and they can't beat him. You got to join them. 222 00:18:08.759 --> 00:18:11.960 But folks, don't touch out. Dog'll be right back. You're listening to 223 00:18:11.000 --> 00:18:25.720 Inside Analysis. Welcome back to Inside Analysis. Here's your host, Eric Tabanac. 224 00:18:27.440 --> 00:18:30.160 All right, folks, back here on Inside Analysis with my good buddy 225 00:18:30.200 --> 00:18:37.319 Jim Harris, broadcaster, extraordinary social media master. He's he's very good at 226 00:18:37.319 --> 00:18:41.720 that kind of stuff. It's spreading the word amplification. Reach out to us 227 00:18:41.759 --> 00:18:45.720 if you want your event amplified, send me an email info at inside analysis 228 00:18:45.880 --> 00:18:49.119 dot com. And we're talking about AI as the top topic at the Consumer 229 00:18:49.119 --> 00:18:53.200 Electronics Show and it was also the top topic at Davos. And you know, 230 00:18:53.240 --> 00:18:56.680 in the last segment, Jim, you were just outlining some of the 231 00:18:56.720 --> 00:19:02.680 most awesome use cases for AI, which is you feed these models. And 232 00:19:02.720 --> 00:19:07.480 I think most companies are going to want their own private model to not have 233 00:19:07.559 --> 00:19:14.880 to run the risk of putting their PII or their IP intellectual property out into 234 00:19:14.880 --> 00:19:18.160 some public model because we really don't know what happens. I mean, chat 235 00:19:18.200 --> 00:19:22.039 GPT is a black box at this point in time. Now, Facebook, 236 00:19:22.079 --> 00:19:25.880 I think Lama is open source. So people were screaming buddy murder about that. 237 00:19:25.960 --> 00:19:27.319 But hey, man, I say, hats off for going open source 238 00:19:27.319 --> 00:19:33.519 with this stuff. But you were giving on this really fascinating model use case 239 00:19:33.599 --> 00:19:38.119 for AI, which is take volumes of policy documents. So think how governments 240 00:19:38.160 --> 00:19:42.559 actually run the laws, the regulations, the policies, all these things. 241 00:19:42.880 --> 00:19:47.759 You can feed them into a large language model and then you don't need to 242 00:19:47.839 --> 00:19:52.319 go pouring over details to try to understand things anymore. To get the policies, 243 00:19:52.640 --> 00:19:55.720 you just ask your little interactive chatbot, your little buddies, say hey, 244 00:19:55.759 --> 00:19:59.559 how do I fire this person? Or even simple stuff like how do 245 00:19:59.599 --> 00:20:02.680 I registered to be a foreign agent? That's a big thing in the US 246 00:20:02.720 --> 00:20:06.319 these days. Ooh did you be a foreign agent or not? What does 247 00:20:06.359 --> 00:20:11.319 the process look like? Well, hitherto that is something that only the highest 248 00:20:11.359 --> 00:20:15.680 paid lawyers in the country could help you do. Now you don't need that. 249 00:20:15.119 --> 00:20:18.839 Now I talk to someone who put together a legal document and memorandum just 250 00:20:18.920 --> 00:20:22.160 using chat GPT brought in and like showed it to a layer. It's like, 251 00:20:22.200 --> 00:20:23.039 wow, did you become a lawyer he's like, no, I just 252 00:20:23.160 --> 00:20:27.359 used chat GPT. And yes, there was that story of the one foolish 253 00:20:27.440 --> 00:20:32.839 lawyer who wrote something up and didn't bother to check his sources. But I 254 00:20:32.880 --> 00:20:36.960 almost wonder if that wasn't sort of amplified as fud, you know, fear, 255 00:20:37.039 --> 00:20:38.319 uncertainty and dread, like, oh, you don't want to use this 256 00:20:38.400 --> 00:20:42.519 technology, it'll be dangerous. Like so people making five hundred dollars an hour 257 00:20:44.079 --> 00:20:48.240 to do basic law work, right, that's dangerous to their business model. 258 00:20:48.839 --> 00:20:52.839 But goodness gracious, you can learn so. And from a government perspective, 259 00:20:52.160 --> 00:20:56.480 that is just the most amazing thing ever, because most people want to follow 260 00:20:56.480 --> 00:21:00.000 the rules, they just may not understand the rule. And now they have 261 00:21:00.039 --> 00:21:03.480 an easy on rep to understanding what that stuff does. What do you think, 262 00:21:04.640 --> 00:21:11.440 absolutely we have to pay attention to this. At the Consumer Electronics Show, 263 00:21:11.519 --> 00:21:18.319 the CEO of Intel had a keynote where he was being interviewed by the 264 00:21:18.440 --> 00:21:25.640 Nasdaq reporter for a media outlet, and he said, we're seeing ten x 265 00:21:26.359 --> 00:21:34.039 one hundred x one thousand x ten thousand x productivity increases use cases. And 266 00:21:34.119 --> 00:21:40.720 I thought, is this just hyperbole or is it reality? So I went 267 00:21:40.839 --> 00:21:45.160 back and I thought about it, and here's my testing of his statement. 268 00:21:45.680 --> 00:21:52.160 So Kathy Wood, who's the CEO of our invest on Wall Street, who's 269 00:21:52.200 --> 00:21:57.880 brilliant. I love what she says. She focuses on exponential change and disruption. 270 00:22:00.359 --> 00:22:08.200 Her team has said that using generative AI, teams that do coding programming 271 00:22:08.519 --> 00:22:15.000 will be ten x more productive than teams that don't use AI by twenty thirty. 272 00:22:15.279 --> 00:22:18.440 And it's not in either or. It's an end. It's you can't 273 00:22:18.480 --> 00:22:25.920 just give your programming to AI. Just like anything I get back from GPT, 274 00:22:26.079 --> 00:22:30.119 I don't put out in the world. It's a first draft for me. 275 00:22:30.599 --> 00:22:33.960 It gets me sixty to ninety percent of the way there. So the 276 00:22:33.000 --> 00:22:38.359 base case of tenx is Kathy Wood. And then in Davos, I put 277 00:22:38.400 --> 00:22:45.680 this statement to a panel and a woman who was a marketer said, you 278 00:22:45.680 --> 00:22:49.759 know, back in the nineteen eighties in the nineties in advertising, when we 279 00:22:49.839 --> 00:22:55.880 had a client, we'd sketch out all these boards by hand for an advertising 280 00:22:55.920 --> 00:22:59.839 campaign. It would take us two weeks. And I thought about Madmin, 281 00:23:00.119 --> 00:23:06.559 you know, very mad man, right, But today she said, using 282 00:23:06.640 --> 00:23:11.400 chat, GPT and mid journey, it takes us two hours. And I 283 00:23:11.440 --> 00:23:17.039 did the math on that, and that's a five hundred x increase in productivity. 284 00:23:17.599 --> 00:23:19.759 Now, her next point was, that's not end to end in the 285 00:23:19.799 --> 00:23:26.039 process, you still have to meet the client. There's no productivity gain there. 286 00:23:26.160 --> 00:23:30.680 You still have to negotiate the final campaign and pricing and everything else. 287 00:23:32.160 --> 00:23:38.079 But for part of the process, there's a five hundred x productivity improvement. 288 00:23:38.599 --> 00:23:42.400 So that was kind of the mid use case. And then I began thinking 289 00:23:42.400 --> 00:23:48.359 about the ten thousand X statement. And every year I go to a conference, 290 00:23:48.400 --> 00:23:52.000 it's my favorite conference every year. It's called Nicks med It's looking at 291 00:23:52.000 --> 00:23:59.799 the future of medicine, and a PhD student will spend a whole year studying 292 00:23:59.839 --> 00:24:04.640 the interaction of a single protein with all other proteins in the human body to 293 00:24:04.759 --> 00:24:08.599 try and understand the emergence of cancer. Takes a whole year to do this 294 00:24:08.759 --> 00:24:17.720 study. Well, they gave this process to AI and got back six hundred 295 00:24:17.799 --> 00:24:26.680 and forty four million years of PhD equivalent work. So there's a use case 296 00:24:26.880 --> 00:24:33.559 that six hundred and forty four million X productivity increase, which is way bigger 297 00:24:33.559 --> 00:24:37.519 than ten thousand. So in the end I agreed with the CEO of Intel 298 00:24:37.640 --> 00:24:45.960 statement ten to ten thousand x productivity increases, but it won't necessarily be across 299 00:24:45.039 --> 00:24:52.640 an entire chain of process. It may just be part of it. And 300 00:24:52.720 --> 00:24:57.000 the other thing that the panelist in Davos said is, you know, when 301 00:24:57.039 --> 00:25:02.279 we get these games, what do we get with them? So rather than 302 00:25:03.960 --> 00:25:10.319 just having a couple of boards to show a client, we might actually develop 303 00:25:11.759 --> 00:25:18.599 five full commercials with AI. In other words, we get five hundred x 304 00:25:18.680 --> 00:25:23.240 time back, but we'll spend half of that giving them more choice, richer 305 00:25:23.440 --> 00:25:29.759 choices. So all these productivity gains don't go to the bottom line. Some 306 00:25:29.799 --> 00:25:34.440 of them go to more enhanced outputs. Yeah, and that makes a lot 307 00:25:34.480 --> 00:25:41.160 of sense, and again it's such a game changer. I had a company 308 00:25:41.160 --> 00:25:45.319 called alter X on a webinar a few months ago, and they're a business 309 00:25:45.319 --> 00:25:48.839 intelligence analytics kind of company and they have something called Magic Documents, which is 310 00:25:48.880 --> 00:25:52.759 their use of JENNYI and it's absolutely brilliant. I was blown away by this. 311 00:25:53.160 --> 00:25:56.599 What they do is they allow you to take a report, let's say 312 00:25:56.599 --> 00:26:00.880 from your ERP system or your P and L or something, some report that 313 00:26:00.960 --> 00:26:06.119 is generated by a trusted system that you use a source of record, A 314 00:26:06.119 --> 00:26:08.359 system of record, I should say, like your ERP for example, and 315 00:26:08.440 --> 00:26:12.279 you generate your report, it's all numbers of how many products were sold in 316 00:26:12.319 --> 00:26:17.799 things of this nature. Then you load that into their system and you give 317 00:26:17.839 --> 00:26:21.839 it the target that you want, like PowerPoint for example, or an article. 318 00:26:22.160 --> 00:26:26.400 But the PowerPoint one is quite impressive. And say, generate a fifteen 319 00:26:26.480 --> 00:26:30.079 slide presentation based on this report, and it goes z and just bangs it 320 00:26:30.079 --> 00:26:33.880 out, complete with graphics and bar charts and pie charts and all this kind 321 00:26:33.920 --> 00:26:37.440 of stuff. And then you just go through and make sure that it got 322 00:26:37.480 --> 00:26:41.319 things right. But the point is you don't have to write that thing from 323 00:26:41.559 --> 00:26:45.039 scratch. Can I tell you how much time that just saved? Not just 324 00:26:45.240 --> 00:26:51.480 time, but painful, unpleasant time. I mean to be honest admission here, 325 00:26:51.480 --> 00:26:53.519 I've never been a fan of PowerPoint. I don't like how it was 326 00:26:53.559 --> 00:26:56.359 set up, the structure. I get where it came from. I understand, 327 00:26:56.400 --> 00:27:00.119 I just don't like it. It just kind of annoys me. And 328 00:27:00.200 --> 00:27:03.519 here just steam rolls right through that process. And then what you could do 329 00:27:03.640 --> 00:27:07.160 is say, oh, now I want to write this for the IT team, 330 00:27:07.240 --> 00:27:08.319 or I want to write this for the marketing team. I want to 331 00:27:08.319 --> 00:27:12.000 write this for the Board of Directors, and it changes the tone and it 332 00:27:12.079 --> 00:27:17.799 changes what goes into it such that you can have five different versions of the 333 00:27:17.839 --> 00:27:22.039 same report for five different groups that you banged out in like two three hours 334 00:27:22.119 --> 00:27:30.119 instead of two to three weeks. That's like forty fifty x improvements and it 335 00:27:30.359 --> 00:27:33.400 just, you know, it just solves these painful problems. Now you can 336 00:27:33.440 --> 00:27:37.279 spend your time absorbing the information, talking to the board, talking to the 337 00:27:37.279 --> 00:27:40.160 marketing team to figure stuff out, because that's still going to be a human 338 00:27:40.200 --> 00:27:45.880 process. Right. But you said before the show we were talking dull, 339 00:27:45.039 --> 00:27:48.319 dirty and dangerous, like all that kind of stuff is out the window. 340 00:27:48.640 --> 00:27:55.279 Those are amazing stories of improvements. And then you look at Davos and what 341 00:27:55.319 --> 00:27:59.279 these folks you're talking about. These are CEOs, These are government leaders. 342 00:27:59.440 --> 00:28:03.319 These are people who are responsible for the well being of thousands and sometimes millions 343 00:28:03.319 --> 00:28:07.880 of people. Guess what that needs to be top of mind. I'm guessing 344 00:28:07.920 --> 00:28:14.359 it was. Tell us it was. And you can apply AI to some 345 00:28:14.680 --> 00:28:22.400 challenges that you might not think of. So we need to explode exponentially the 346 00:28:22.680 --> 00:28:30.119 amount of batteries that we have globally to electrify transportation, for instance, and 347 00:28:30.200 --> 00:28:37.079 get away from fossil fuels, and some people criticize EV batteries because, for 348 00:28:37.079 --> 00:28:41.839 instance, some of the old ones have cobalt in them, right, which 349 00:28:41.880 --> 00:28:48.160 is a conflict mineral or their child miners. Well, you can use AI 350 00:28:48.799 --> 00:28:56.000 to look at new battery chemistries which completely eliminate the need for cobalt at all. 351 00:28:56.400 --> 00:29:03.119 Right, And LFP lithium phosphate is lithium iron phosphate is an example of 352 00:29:03.160 --> 00:29:07.480 that there's no cobalt in it. So now we had a problem with cobalt, 353 00:29:07.519 --> 00:29:15.079 and it's cheaper and you can scale it faster, So we can use 354 00:29:15.119 --> 00:29:18.960 AI to focus on problems. But again it's going to be the scientists and 355 00:29:19.720 --> 00:29:26.559 AI working collaboratively. You can't just outsource problems to AI. People are always 356 00:29:26.559 --> 00:29:30.640 going to be part of this, but they're unintended or surprising use cases. 357 00:29:30.720 --> 00:29:37.000 So in Douvilas, I was talking to one executive who is really quite nervous 358 00:29:37.039 --> 00:29:45.000 about speaking in front of groups, and his parents were having their fiftieth winning 359 00:29:45.039 --> 00:29:51.079 anniversary and he hates like he's not a comedian or a humorist. He wanted 360 00:29:51.079 --> 00:29:56.839 something that was poignant, that was funny, so he had chat gp T 361 00:29:56.680 --> 00:30:02.799 right. The first draft his wedding toast for his parents, and then he 362 00:30:02.960 --> 00:30:07.359 threw an iterative process. He refined it. Well. He gave this talk 363 00:30:07.599 --> 00:30:11.359 at his parents' fiftieth and his mom came up and said, honey, that 364 00:30:11.599 --> 00:30:18.440 was fantastic, and his son, who's eleven, says, yeah, Granny, 365 00:30:18.480 --> 00:30:22.319 it was chat GPT that did it. Of course, Granny doesn't know 366 00:30:22.359 --> 00:30:29.400 what the hell the eleven years talking about. We can use this. This 367 00:30:29.680 --> 00:30:33.839 is a way of taking everyone and leveling up their skills in all sorts of 368 00:30:33.920 --> 00:30:38.960 areas that they're not skilled at well. And you know, let's think just 369 00:30:40.039 --> 00:30:45.359 real quick about the medical field where right now in the United States. I 370 00:30:45.400 --> 00:30:48.079 don't know what it's like in Canada and the United States. It is so 371 00:30:48.680 --> 00:30:52.720 segmented and compartmentalized, and you have different groups responsible for different things. I 372 00:30:52.720 --> 00:30:56.079 mean, I've seen this up clost and personal and unpleasant ways. But the 373 00:30:56.119 --> 00:31:00.000 point is you have to get disappointed with this person. They say, okay, 374 00:31:00.039 --> 00:31:02.279 we should talk to that person. That takes three weeks to get to 375 00:31:02.319 --> 00:31:04.319 that person. Then it takes another three weeks to get to some other person. 376 00:31:04.319 --> 00:31:08.119 And this just takes time, and you're driving all around and waiting and 377 00:31:08.160 --> 00:31:12.000 trying to get stuff done. Well, if you think about again, you 378 00:31:12.000 --> 00:31:15.119 have to do this responsibly. But if you start, and I guarantee it's 379 00:31:15.119 --> 00:31:19.559 happening already, you get these models and you start training fairly narrow focused models 380 00:31:19.960 --> 00:31:25.799 on all the research that's available, and then any doctor turns into a ten 381 00:31:26.000 --> 00:31:32.720 X doctor. Any generalist can understand what the specialists know based upon information that's 382 00:31:32.759 --> 00:31:37.000 in the record, that's been understood, that's been research that's been shared, 383 00:31:37.480 --> 00:31:40.960 frood vetted, And that's really going to be where the heavy lifting is done 384 00:31:41.000 --> 00:31:45.400 now is in training the models and constantly giving them feedback. And this I'll 385 00:31:45.440 --> 00:31:49.839 get back to the Institute for Experiential AI. Doctor Sama Faiad says, all 386 00:31:49.839 --> 00:31:53.519 the time human in the loop, there's always going to be a human in 387 00:31:53.559 --> 00:31:56.680 the loop. Don't think you're going to turn this thing on walk away and 388 00:31:56.720 --> 00:32:00.079 have it do your job for you know, it is going to give you 389 00:32:00.240 --> 00:32:05.000 all this additional information, all this additional context you're able to see in a 390 00:32:05.039 --> 00:32:08.559 particular case. Well, actually, eighty percent of the time that we give 391 00:32:08.599 --> 00:32:14.400 this medication, this happens unless there's this other medication that's contraindicated and you have 392 00:32:14.480 --> 00:32:16.480 this problem. It used to be you had to sit around and read journals 393 00:32:16.480 --> 00:32:20.640 to get all that stuff. Now you'll be able to just interact with your 394 00:32:20.640 --> 00:32:23.839 little chatbot and you're in the hospital and say, what are the contrainteications for 395 00:32:23.880 --> 00:32:25.480 this? Up? Up, up, up up up. You get them 396 00:32:25.519 --> 00:32:28.720 all on the screen for you. What about that? What about for this 397 00:32:28.759 --> 00:32:30.480 blood type? But up up It just all this stuff is going to come 398 00:32:30.519 --> 00:32:36.160 flooding out and we'll finally be able to use this before the person had to 399 00:32:36.200 --> 00:32:40.880 either remember all this stuff or have manuals at their side. And you know, 400 00:32:42.359 --> 00:32:45.119 have you ever seen doctors looking at manuals when you're in the hospital. 401 00:32:45.200 --> 00:32:47.799 No, they're never doing that. They're just walking around talking to people doing 402 00:32:47.839 --> 00:32:54.480 their thing. Point being these technologies, this AI stuff is a force multiplier 403 00:32:55.039 --> 00:32:59.960 in just about any category. And I'm glad. I'm very encouraged to hear 404 00:33:00.079 --> 00:33:01.519 that. Folks at DABOS. We're talking about that one minute till the next 405 00:33:01.559 --> 00:33:09.880 break, go ahead. So seven thousand new medical journal articles came out today. 406 00:33:10.400 --> 00:33:15.119 So the question is, has your general practitioner as your doctor read them 407 00:33:15.160 --> 00:33:21.839 all this morning? You know, So we're gonna have to use AI if 408 00:33:21.880 --> 00:33:25.440 we want the best outcomes and it's not gonna be okay, the AI is 409 00:33:25.480 --> 00:33:30.759 going to be your doctor. It's gonna be your doctor. Plus AI are 410 00:33:30.799 --> 00:33:37.039 gonna wigh out perform doctors who don't use AI, and you'll have better health 411 00:33:37.079 --> 00:33:39.880 outcomes. Yeah. No, that's exactly right. It's gonna be your doctor 412 00:33:40.240 --> 00:33:44.960 plus AI. Your doctor is not going away, but you're gonna hope that 413 00:33:45.000 --> 00:33:47.640 your doctor uses this stuff because they're gonna be way, way, way informed. 414 00:33:47.640 --> 00:33:50.559 Folks, don't touch up that. I'll be right back. You're listening 415 00:33:50.640 --> 00:34:02.079 to Inside Analysis. Expect Welcome back to Analysis. Here's your host, Eric 416 00:34:02.240 --> 00:34:08.039 Tavanaugh. All right, folks, back here on Inside Analysis talking to the 417 00:34:08.280 --> 00:34:14.840 legend himself, Jim Harris, broadcaster, a social media expert, world traveler, 418 00:34:15.039 --> 00:34:17.480 sixty conferences a year and he does it every year. I don't know 419 00:34:17.559 --> 00:34:22.280 how he does that. He jumps all off with the planet. Mobile World 420 00:34:22.320 --> 00:34:24.519 Congress is coming next, folks, So if you want some amplification at Mobile 421 00:34:24.519 --> 00:34:29.679 World Congress, send me an email info at Inside Analysis dot com. Email's 422 00:34:29.679 --> 00:34:30.760 not going away, by the way, I just want to say everyone keeps 423 00:34:31.079 --> 00:34:36.280 talking about the NB not going to happen. Email is still king for communication 424 00:34:36.360 --> 00:34:38.599 for transactional from important stuff, send him an email, even if it's just 425 00:34:38.599 --> 00:34:42.360 got to link in it to the zoom, like, where's he just got 426 00:34:42.400 --> 00:34:46.480 my email? But everything is changing. It's changing at like rapid speed. 427 00:34:47.039 --> 00:34:51.519 And I'm glad to hear that. At Davos and Cees, these were very, 428 00:34:51.639 --> 00:34:53.480 very hot topics, and you know, we have to look on the 429 00:34:53.480 --> 00:34:57.880 bright side. It's going to be really fun and interesting to dive into this. 430 00:34:57.920 --> 00:35:01.239 And I'll give you my theory about how media is going to change dramatically. 431 00:35:01.800 --> 00:35:07.880 You're going to see smart news organizations figure out that they need to leverage 432 00:35:07.920 --> 00:35:12.960 their voice and train their own models on their content and their style, and 433 00:35:13.000 --> 00:35:17.039 then start using this stuff to generate content. And then again the role is 434 00:35:17.079 --> 00:35:22.400 going to be not so much the reporter but the curator. So you still 435 00:35:22.400 --> 00:35:24.280 will want to check facts, You'll still want to want to check dates and 436 00:35:24.320 --> 00:35:28.719 specifics and things of that nature, but you're going to I mean, what 437 00:35:28.800 --> 00:35:30.639 I think is going to happen here is you're going to have systems of record 438 00:35:31.239 --> 00:35:37.039 connected to llms spinning out useful bits of information or really persisting it in a 439 00:35:37.079 --> 00:35:42.840 model somewhere, and then the reader like you, me, whoever else, 440 00:35:43.239 --> 00:35:47.119 will have a much more personalized, engaging experience with the journalist, which is 441 00:35:47.199 --> 00:35:50.639 going to be their own personal journalist. So you'd be like, let's say 442 00:35:50.679 --> 00:35:52.960 I name my journalist Bob, Like, hey, Bob, what happened at 443 00:35:52.960 --> 00:35:55.880 the city council, meaning last night with respect to the new tax cuts or 444 00:35:55.880 --> 00:35:59.920 something? And I'll be like, well, Congressman so and so said, 445 00:35:59.920 --> 00:36:02.320 that'll just tell you all this stuff. And then you're going to have this 446 00:36:02.400 --> 00:36:07.880 interactive process where you're just asking it what's going on relevant to you. Think 447 00:36:07.880 --> 00:36:12.519 about how much better that is for an individual. I think if you look 448 00:36:12.559 --> 00:36:15.719 at traditional broadcast meeting, you've spent a lot of time in that world. 449 00:36:15.119 --> 00:36:20.199 It was this one size fits all model where everyone who's watching the show is 450 00:36:20.239 --> 00:36:22.880 watching it at the same time, stamp at the same continuums, getting the 451 00:36:22.880 --> 00:36:28.920 same exact content. That's not good enough anymore. People need personalized content that 452 00:36:29.000 --> 00:36:31.079 is relevant for them. And that's why I think broadcast is going to have 453 00:36:31.119 --> 00:36:35.559 some trouble dealing with all this stuff. But how do you see I mean, 454 00:36:35.599 --> 00:36:37.280 you're a broadcaster, how do you see the future of media panning out 455 00:36:37.280 --> 00:36:42.800 here? Well, media is going to have to change. I mean, 456 00:36:42.840 --> 00:36:46.559 I've been a journalist for forty years, and I've seen huge amount of change. 457 00:36:47.119 --> 00:36:52.639 When I got into television, we used to have these million dollar edit 458 00:36:52.719 --> 00:36:59.639 suites where we'd actually have tape to tape, the tape was edited and cut 459 00:36:59.679 --> 00:37:04.800 with it zach donives and you know, with the scotch tape put back together. 460 00:37:05.400 --> 00:37:09.960 And now you can do all the editing on a laptop or on your 461 00:37:10.000 --> 00:37:17.639 even your mobile phone with titling and everything else. So I've seen head spinning 462 00:37:17.840 --> 00:37:22.880 change over the forty years i've been a journalist. I want to give another 463 00:37:23.000 --> 00:37:30.519 example that people might not think of, how chat GPT and Generative AI can 464 00:37:30.639 --> 00:37:37.360 save up to three hundred and sixty billion dollars in the US healthcare system. 465 00:37:37.719 --> 00:37:42.280 So if you go to your doctor and the doctor wants to order a test, 466 00:37:42.360 --> 00:37:46.159 but the insurance company in the US denis paying for it, that doctor 467 00:37:46.239 --> 00:37:52.559 has to write an appeal letter why I deserve to have this test done. 468 00:37:52.159 --> 00:37:58.199 So there's this urologist who was writing about five appeal letters a week, taking 469 00:37:58.320 --> 00:38:04.559 twenty to thirty minutes for each one. Now he just uses chat GPT without 470 00:38:04.679 --> 00:38:09.440 using the patient's name or the insurance company. Basically instantly, it spits out 471 00:38:09.480 --> 00:38:15.400 a letter as to why this particular condition needs this test, including citations from 472 00:38:15.559 --> 00:38:22.159 medical journals. Wow, So take five appeal letters a week at thirty minutes, 473 00:38:22.199 --> 00:38:27.159 and what is one hundred and fifty minutes every single week of a urologious 474 00:38:27.199 --> 00:38:30.679 time worth who's on a quarter million dollar salary. You do the math. 475 00:38:30.760 --> 00:38:38.800 It's big numbers. So we're going to see AI eliminating wasteful, inefficient administrative 476 00:38:38.880 --> 00:38:46.159 work in the medical system. And that can be worth by one analysis of 477 00:38:46.719 --> 00:38:52.880 a major consulting firm three hundred and sixty billion dollars a year. That's cool, 478 00:38:52.920 --> 00:38:57.280 blows me away. I'll give you another example that comes from next Med, 479 00:38:57.360 --> 00:39:00.079 which I mentioned in the first segment. It's my favorite conference. Well, 480 00:39:01.320 --> 00:39:07.599 there was a doctor from the NHS in the UK National Health Service. 481 00:39:07.719 --> 00:39:10.960 It is the largest single payer provider in the world. They do one hundred 482 00:39:10.960 --> 00:39:19.400 and six fifty six billion pounds of healthcare funding every single year, and during 483 00:39:19.400 --> 00:39:24.079 the pandemic, twenty five million people a year missed their appointments. Now this 484 00:39:24.199 --> 00:39:29.320 is problematic because you know you have weight times to get an appointment with a 485 00:39:29.360 --> 00:39:32.800 specialist, and then somebody misses their appointment. That's terrible. And there are 486 00:39:34.000 --> 00:39:39.400 health consequences. If you're a woman and you miss two mammograms, you increase 487 00:39:39.519 --> 00:39:45.639 the chance of stage three breast cancer by three hundred percent. Or if you 488 00:39:45.840 --> 00:39:54.440 have a retina problem and you miss a single appointment, you can increase your 489 00:39:54.519 --> 00:40:00.800 chance of blindness five thousand percent. So it's not just the twenty five missed 490 00:40:00.880 --> 00:40:06.760 appointments, it's the medical outcomes of not catching things early enough. So this 491 00:40:06.960 --> 00:40:12.880 doctor, very frustrated, looked at the twenty five million missed appointments and without 492 00:40:13.000 --> 00:40:17.360 using any medical data, so he was only using databases with age, income, 493 00:40:17.559 --> 00:40:22.800 geography, those kind of things. He was able to predict based on 494 00:40:23.039 --> 00:40:31.079 history, who would miss their appointment for upcoming and then they do this eight 495 00:40:31.119 --> 00:40:36.599 weeks before assigning appointments. They can then say, now here are some of 496 00:40:36.599 --> 00:40:39.760 the reasons. It might be an elderly man who doesn't want to get on 497 00:40:39.840 --> 00:40:44.599 the tube late at night because his eyesight is bad, right, so you've 498 00:40:44.639 --> 00:40:47.960 given him a nine pm appointment and he misses it. Or it might be 499 00:40:49.039 --> 00:40:52.719 a single mum. You give an appointment during the day, but she can't 500 00:40:52.719 --> 00:40:58.400 afford to take time off work, so she misses it. So the system, 501 00:40:58.599 --> 00:41:04.079 the AI predicts who will miss their appointment and then they get busy in 502 00:41:04.239 --> 00:41:07.840 trying to mitigate that. In other words, give them the first choice of 503 00:41:07.920 --> 00:41:13.599 appointments where they're going to actually show up and where this system has been tested. 504 00:41:14.239 --> 00:41:20.360 It has increased the capacity of the NHS by six percent. Wow. 505 00:41:20.519 --> 00:41:25.400 Now when you do the math on that, that's nine billion pounds of value 506 00:41:27.000 --> 00:41:36.199 being created with no medical data, just normal parameters, so nothing confidential. 507 00:41:36.760 --> 00:41:43.960 So this blows me away. We're going to be using AI for profound outcomes, 508 00:41:44.639 --> 00:41:49.039 and if we don't use it, we're disadvantaging ourselves. Like, imagine 509 00:41:49.039 --> 00:41:53.760 increasing the capacity of the healthcare system by six percent with zero cost. That's 510 00:41:53.800 --> 00:41:57.519 just that's crazy. And you know, you reminded me of something. We 511 00:41:57.599 --> 00:42:00.840 had Dan Bodner on the show a couple months ago. I guess he is 512 00:42:00.880 --> 00:42:05.400 the CEO of Varrant v er I n T. Look those folks up online 513 00:42:05.760 --> 00:42:07.880 and they've got thirty five bots now, And it was interesting. I was 514 00:42:07.920 --> 00:42:12.239 talking with them about bots and you know, maybe they'll be richer, more 515 00:42:12.239 --> 00:42:15.559 interesting, et cetera. And he goes, oh, actually, what our 516 00:42:15.679 --> 00:42:19.159 perspective is that with a bot, you want it to be very very good 517 00:42:19.159 --> 00:42:22.119 at one thing and then that's what it does. It does that one thing, 518 00:42:22.159 --> 00:42:25.920 and it does it extremely well. And he was talking about scheduling and 519 00:42:27.000 --> 00:42:30.679 what a nightmarre scheduling is for someone who runs a hospital, for the nurses, 520 00:42:30.679 --> 00:42:34.599 for the doctors on call, for all this kind of stuff, because 521 00:42:34.679 --> 00:42:39.559 hitherto the scheduling problem is highly manual, and the person in charge of scheduling 522 00:42:39.599 --> 00:42:43.280 has to be like, okay, Bob is called in sick, Now who 523 00:42:43.280 --> 00:42:45.800 can I call? And you have to go look up phone numbers and kind 524 00:42:45.800 --> 00:42:47.760 of scratch your head, who can I get? It's painful, it's a 525 00:42:47.840 --> 00:42:52.760 very unpleasant process, he said. What they did is they put a program 526 00:42:52.800 --> 00:42:58.239 together. It's like a points system where if you agree to take someone's role 527 00:42:58.280 --> 00:43:00.000 when they're sick, you get a few x your points. If it's a 528 00:43:00.000 --> 00:43:04.199 really nasty shift, you get even extra points. And they keep track of 529 00:43:04.199 --> 00:43:07.000 all this stuff. And now anytime someone's out, you just hit a button 530 00:43:07.079 --> 00:43:10.159 brip, it goes ask him if you don't want to ask him. All 531 00:43:10.159 --> 00:43:15.320 that stuff is automated. So now there's like that scheduling probably it's problems solve 532 00:43:15.599 --> 00:43:17.920 done. You're like, what like that used to be one of the hardest 533 00:43:19.239 --> 00:43:23.360 jobs for people who are an admin in hospitals and different facilities and like, 534 00:43:23.400 --> 00:43:27.400 problem solved. Next, what else you got? Why don't you go talk 535 00:43:27.440 --> 00:43:29.760 to the patient for a few minutes? Right? Oh, yeah, that's 536 00:43:29.880 --> 00:43:34.119 right, the patients. So I mean, these are amazing examples of how 537 00:43:34.920 --> 00:43:42.039 these technologies, these algorithms can tackle very complex challenges very easily for almost no 538 00:43:42.159 --> 00:43:44.360 money. I mean, you have to pay for the software and get things 539 00:43:44.360 --> 00:43:49.519 installed, but once it's there, that is a massive game changer. One 540 00:43:49.559 --> 00:43:52.840 minute lefts what's your what's your advice to the world, Jim, Well, 541 00:43:53.320 --> 00:43:59.360 we have to embrace this technology. If you look at a one hundred and 542 00:43:59.480 --> 00:44:02.679 fifty years years ago, eighty percent of jobs in the US were in agriculture, 543 00:44:04.599 --> 00:44:08.440 and there were huge protests when farm automation came along, tractors and combines, 544 00:44:08.480 --> 00:44:12.159 like, what will happen to all the horse people, the horse people, 545 00:44:12.199 --> 00:44:15.519 the horse people, the horse people. You know, here we are 546 00:44:15.159 --> 00:44:21.320 one hundred and fifty, one hundred and sixty years later and agriculture employment is 547 00:44:21.400 --> 00:44:25.119 less than two percent. So we survived farm automation, right, But what 548 00:44:25.239 --> 00:44:31.559 is certain is that jobs changed. People had to learn and reskill, society 549 00:44:31.679 --> 00:44:36.199 change. Companies had to change. So this is really a story not about 550 00:44:36.199 --> 00:44:43.280 the technology, but about change, about reskilling. That's right, reskilling, 551 00:44:43.320 --> 00:44:46.039 folks. Get out there, get out start using this stuff. Folks, 552 00:44:46.119 --> 00:44:54.079 you've been listening to Inside Analysis. Respect. Okay, folks, time for 553 00:44:54.119 --> 00:44:59.280 the podcast bonus segment here on Inside Analysis with our good buddy Jim Harris. 554 00:44:59.280 --> 00:45:01.679 He's always fun to talk to, and I have to say, Jim, 555 00:45:01.679 --> 00:45:07.920 there has been some unpleasant news recently. I saw and apparently you've seen it 556 00:45:07.960 --> 00:45:12.079 as well. I saw City Group plans to lay off twenty thousand people. 557 00:45:12.639 --> 00:45:16.840 SAP announced the layff of eight thousand employees. Eight thousand, and I think 558 00:45:16.880 --> 00:45:21.039 they've got about eighty or ninety thousand employees. And I looked it up and 559 00:45:21.079 --> 00:45:22.480 it said as of twenty twenty one, it's like one hundred and seven. 560 00:45:22.639 --> 00:45:27.239 I think it's a little bit inflated. I'm guessing it's about ten percent of 561 00:45:27.280 --> 00:45:30.159 their workforce that they're laying off. And they said it's a retooling for JENAI, 562 00:45:30.480 --> 00:45:35.199 all right. I don't know if they buy that argument, but it 563 00:45:35.280 --> 00:45:38.440 is certainly big news. Microsoft thing off nineteen hundred, so these layoffs are 564 00:45:38.719 --> 00:45:43.440 coming fast and furious. What are your thoughts on that, and what does 565 00:45:43.440 --> 00:45:47.159 it say about where we're going. Well, a couple things. One is 566 00:45:47.599 --> 00:45:54.039 we're in a time that people have been predicting a recession for about two and 567 00:45:54.119 --> 00:46:00.639 a half years now, so this is the longest time ever of a predicted 568 00:46:00.679 --> 00:46:06.360 recession that has never happened. So one of the insights around this is, 569 00:46:06.960 --> 00:46:15.360 yes, large companies historically are the largest net job losers, but recessions happen 570 00:46:15.480 --> 00:46:21.960 when unemployment is high, and all these people who are let go have jobs 571 00:46:22.079 --> 00:46:28.280 tomorrow. They're in high demand high tech work. So we have not seen 572 00:46:28.639 --> 00:46:34.239 this recession come to fruition that has been predicted by economists for two and a 573 00:46:34.320 --> 00:46:38.880 half years. So that's the good news. Now, if you are one 574 00:46:38.920 --> 00:46:46.400 of these eight thousand people at SAP or twenty thousand or nineteen hundred at Microsoft, 575 00:46:49.320 --> 00:46:54.440 what is the important message here for you? And job security today is 576 00:46:54.559 --> 00:47:04.559 based on learning and changing and accepting on certainty, So people have to continually 577 00:47:05.400 --> 00:47:10.719 be retooling, reskilling. That's what creates job security. Mm hmmm, yeah, 578 00:47:10.760 --> 00:47:14.320 I think that's yeah, go ahead, no, no, you go 579 00:47:14.360 --> 00:47:19.280 ahead. I think that's an excellent point. I do think that these new 580 00:47:19.360 --> 00:47:24.119 technologies, this GENAI stuff, these are force multiplying technologies. I mean, 581 00:47:24.280 --> 00:47:30.320 used correctly, one person can now be as productive as two or three or 582 00:47:30.360 --> 00:47:35.440 four people just because of the time that they save and the creative juice that 583 00:47:35.480 --> 00:47:39.639 they inject into the process. You look at chat ept bard is the one 584 00:47:39.639 --> 00:47:45.760 that I typically use. There's clawed by anthropic. These are generative tools and 585 00:47:45.840 --> 00:47:49.960 they are very powerful. I mean, it's amazing. They're very good at 586 00:47:49.960 --> 00:47:52.280 poetry. By the way, if anyone is interested in writing some poetry, 587 00:47:52.280 --> 00:47:54.760 they're actually quite good at that kind of stuff because all they're doing is reflecting 588 00:47:54.840 --> 00:47:59.400 back what they've been trained on. Right. So, I think I was 589 00:47:59.400 --> 00:48:01.119 talking to high On Park, a good buddy mind from Amalgam Insights, and 590 00:48:01.159 --> 00:48:06.599 he was saying he thinks that these layoffs are prematurely announced and that these companies 591 00:48:06.639 --> 00:48:09.760 really haven't yet figured out what they're going to be doing with these technologies, 592 00:48:09.840 --> 00:48:15.320 and so you know what exactly is going on. It feels to me like 593 00:48:15.360 --> 00:48:19.679 a cost savings measure, which is not new. But you know, his 594 00:48:19.880 --> 00:48:23.800 argument was maybe they should have waited and to see how their organizations can embrace 595 00:48:23.840 --> 00:48:28.519 these new technologies before going and laying people off. I mean, I think 596 00:48:28.519 --> 00:48:31.480 he's right, But what do you think I would agree. One of the 597 00:48:31.519 --> 00:48:43.599 things that caught Silicon Valley by surprise was Elon Musk buying Twitter and then laying 598 00:48:43.639 --> 00:48:49.000 off seventy five percent, eighty percent of the workforce at Twitter, right, 599 00:48:49.000 --> 00:48:52.639 and they that was head spinning and people were thinking, well, Twitter's going 600 00:48:52.679 --> 00:48:57.320 to go down. It just won't be able to operate. But it's been 601 00:48:57.400 --> 00:49:00.719 up the now. It has had a noted sharing there, but it had 602 00:49:00.760 --> 00:49:06.559 an outage here and there before Eli bought it. But the point of this 603 00:49:06.920 --> 00:49:14.800 is that we really need to look at what does an organization need in terms 604 00:49:14.840 --> 00:49:20.679 of people, how does it operate optimally? And in fact, the amount 605 00:49:20.679 --> 00:49:27.480 of innovation coming out of Twitter has been greater in the last six months than 606 00:49:27.519 --> 00:49:32.880 the prior six years. There have been more features, and so this doesn't 607 00:49:32.880 --> 00:49:38.599 seem to compute. How do you end up innovating more with seventy five to 608 00:49:38.599 --> 00:49:49.159 eighty percent fewer people. So really it challenges everyone to say, our real 609 00:49:49.280 --> 00:49:53.159 job is to innovate, Our real job is to keep our skills current. 610 00:49:53.960 --> 00:49:59.480 And one of the insights from I know we're going to be talking about Davos 611 00:49:59.519 --> 00:50:07.079 and ce Yes, but one insight was from DeVos is that when you talk 612 00:50:07.159 --> 00:50:12.239 to CEOs, forty percent of them believe we're going to have to reskill forty 613 00:50:12.280 --> 00:50:16.559 percent of our people within five years. But when you talk to people in 614 00:50:16.599 --> 00:50:22.960 the organization on the frontline and frontline management, ninety seven percent of them say 615 00:50:23.039 --> 00:50:30.440 we need training and development and reskilling right now, not in five years. 616 00:50:30.519 --> 00:50:37.280 So there's this huge disconnect between what companies think they need to do in terms 617 00:50:37.360 --> 00:50:43.880 of training, education and reskilling and what people on the ground are saying. 618 00:50:44.559 --> 00:50:52.639 And given this earlier discussion that ultimate security is based on relearning and on continually 619 00:50:52.760 --> 00:50:59.719 keeping up your skill set, we need to significantly invest more in training, 620 00:51:00.119 --> 00:51:04.400 education, reskilling. Yeah, I think that's true, and I think you're 621 00:51:04.480 --> 00:51:07.119 right about the fact that it is now, it's not in three to five 622 00:51:07.199 --> 00:51:13.480 years. I mean, this JENAI stuff is, to be clear, focused 623 00:51:13.519 --> 00:51:19.239 on a couple key segments of the enterprise software landscape, namely content creation, 624 00:51:19.960 --> 00:51:25.280 namely chat bots for example, hr basic stuff that a model can be trained 625 00:51:25.320 --> 00:51:31.000 on very easily or relatively easily. But there are lots of other spaces where 626 00:51:31.320 --> 00:51:38.360 AI will in the future attack traditional enterprise software applications. So it's not everywhere 627 00:51:38.440 --> 00:51:42.800 yet, but it's gonna be. I mean you have to think that these 628 00:51:42.880 --> 00:51:47.480 new foundational models are going to subsume most of what enterprise software has been doing 629 00:51:47.519 --> 00:51:51.719 for the last thirty forty years. And you look at an SAP with eight 630 00:51:51.760 --> 00:51:57.719 thousand layoffs, Well, SAP's core business, that's EERP enterprise resource planning, 631 00:51:57.760 --> 00:52:02.320 it's human resources with few glass you know, it's concur with expenses. I 632 00:52:02.360 --> 00:52:07.880 mean, just real quick think about you see on TV now, I saw 633 00:52:07.880 --> 00:52:09.800 an ad for one of these tax firms saying, oh, with AI, 634 00:52:09.920 --> 00:52:13.920 now we can do your taxes so much faster. And that is true. 635 00:52:14.320 --> 00:52:16.400 And I can tell you this, my friend, like pretty soon it's going 636 00:52:16.440 --> 00:52:20.599 to be so much easier to do your taxes. You're gonna say, go 637 00:52:21.280 --> 00:52:23.320 and just generate what you think is the case, and then I'll go through 638 00:52:23.440 --> 00:52:28.440 and check and correct you on stuff. But then if your business doesn't really 639 00:52:28.519 --> 00:52:32.239 change next year, it's going to be even easier. So the accounting world 640 00:52:32.480 --> 00:52:37.079 is in the cross airs of jen Ai. I promise you final thoughts from 641 00:52:37.119 --> 00:52:45.599 you, jib, I'd agree completely white collar work, there's profound efficiencies that 642 00:52:45.760 --> 00:52:52.159 can be wrought out of our historical processes. And if you think about anything 643 00:52:52.360 --> 00:52:58.320 that's dull, dirty, or dangerous, it's gonna be automated. Right. 644 00:52:58.440 --> 00:53:01.719 Yeah, Well, if you just think about eighty percent of some phone calls 645 00:53:01.719 --> 00:53:07.920 into call centers, it call centers are people simply resetting their password? Well, 646 00:53:07.039 --> 00:53:13.079 couldn't an AI do that just as well with three point factor authentication, 647 00:53:13.239 --> 00:53:16.679 your voice, the phone number you're calling in on, and then some pin 648 00:53:17.320 --> 00:53:22.480 just as well as a human. Right boom, that's right, and I 649 00:53:22.480 --> 00:53:27.199 think we'll close on that. Folks, Dull, dirty, or dangerous, 650 00:53:27.519 --> 00:53:30.320 that's all in the crosshairs of AI. Folks, you have been listening to 651 00:53:30.400 --> 00:53:37.239 Inside Analysis, don't miss admitted of the action. Check out the podcast. 652 00:53:37.280 --> 00:53:43.440 It's at www. KCAA radio dot com, the station that leaves no listener 653 00:53:43.480 --> 00:53:51.039 behind KCAA. And now the voices of KCAA with an exciting announcement. Want 654 00:53:51.039 --> 00:53:54.199 to hear NBC News or KCAA anywhere you go, Well, now there's a 655 00:53:54.239 --> 00:53:59.840 nap for that. KCAA is celebrating twenty five years and our silver anniversary with 656 00:54:00.159 --> 00:54:04.760 a brand new app. The new KCAA App is now available on your smart 657 00:54:04.760 --> 00:54:07.400 device, cell phone, in your car, or any place. Just search 658 00:54:07.480 --> 00:54:12.679 KCAA on Google Play or in the Apple Store. One touch and you can 659 00:54:12.719 --> 00:54:15.679 listen on your car radio, Bluetooth device, Android Auto, or Apple Car 660 00:54:15.760 --> 00:54:21.119 plug. Catch the KCAA buzz in your earbuds or on the streets. Celebrating 661 00:54:21.159 --> 00:54:24.639 twenty five years of talk news and excellence with our new KCAA app. Just 662 00:54:24.800 --> 00:54:34.320 do it and download it. KCAA celebrating twenty five years. T Hebot Club's 663 00:54:34.360 --> 00:54:37.280 original pure power to Rco super Ta comes from the only tree in the world 664 00:54:37.360 --> 00:54:42.440 that fungus does not grow on. As a result, it naturally has antifungal, 665 00:54:42.599 --> 00:54:47.559 anti infection, anti viral, antibacterial, anti inflammation, and anti parasite 666 00:54:47.559 --> 00:54:51.920 properties. So the t is great for healthy people because it helps build the 667 00:54:51.920 --> 00:54:55.519 immune system, and it can be truly miraculous for someone fighting a potentially life 668 00:54:55.599 --> 00:55:00.599 threatening disease due to an infection, diabetes, or cancer. The tea is 669 00:55:00.639 --> 00:55:04.920 also organic and naturally caffeine free. A one pound package of tea is forty 670 00:55:05.000 --> 00:55:07.079 nine to ninety five, which includes shipping. To order, please visit to 671 00:55:07.159 --> 00:55:12.559 Hebota club dot com. T Hebow is spelled tea like tom a, h 672 00:55:12.719 --> 00:55:15.840 ee b like boy o. They continue with the word t and then the 673 00:55:15.880 --> 00:55:21.239 word club. The complete website is to Hebot Club dot com or call us 674 00:55:21.320 --> 00:55:25.159 at eight one eight six one zero eight zero eight eight Monday through Saturday nine 675 00:55:25.159 --> 00:55:30.679 am to five pm California time. That's eight one eight six one zero eight 676 00:55:30.800 --> 00:55:36.840 zero eight eight to Hebot club dot com. With sixty years of fascinating facts. 677 00:55:37.199 --> 00:55:42.960 This is the man from yesterday and back in time we go to this 678 00:55:43.079 --> 00:55:47.400 time in nineteen sixty one, Eastman Kodak says it is dropping sponsorship of The 679 00:55:47.559 --> 00:55:52.000 Ed Sullivan Show on CBS TV and The Ozzy and Harriet Show on ABCTV to 680 00:55:52.079 --> 00:55:57.360 make room for a brand new show called Walt Disney's Wonderful World of Color on 681 00:55:57.480 --> 00:56:15.559 NBCTV this fall Walt Disney Presents and more from this time in nineteen sixty three, 682 00:56:15.880 --> 00:56:21.719 CBS announces it's adding comedies petticoat Junction and My Favorite Martian to the schedule 683 00:56:21.719 --> 00:56:24.119 of this fall. My Favorite Martian will air Sunday nights, since Star as 684 00:56:24.199 --> 00:56:30.559 Bill Bixby and Ray Lawston as the Martian. Who are You Well? I 685 00:56:30.559 --> 00:56:34.480 could give you some wild tale of being a Russian astronaut or the designer of 686 00:56:34.480 --> 00:56:38.159 an experimental spacecraft, but the plain fact of the matter is I'm from Mars 687 00:56:38.519 --> 00:56:44.360 Mars Mars and from this time in nineteen seventy six. Rock cal Welch is 688 00:56:44.360 --> 00:56:46.719 set to co star in a new movie called Mother, Jugs and Speed, 689 00:56:47.000 --> 00:56:51.239 and she's set to guest star on Saturday Night Live. Yeah. I think 690 00:56:51.280 --> 00:56:54.199 I am basically a kind of private person, but I'm also kind of a 691 00:56:54.280 --> 00:56:58.840 kid at heart, and I guess that's why still, you know, get 692 00:56:58.840 --> 00:57:07.280 excited and nervous with more at Man from yesterday dot Com. It's that time 693 00:57:07.320 --> 00:57:12.719 of year again, No, not the holidays. Medicare open enrollment and if 694 00:57:12.760 --> 00:57:15.719 you have questions about Medicare, you should talk to the local experts. Paul 695 00:57:15.719 --> 00:57:21.119 Barratte and Associates. All of his agents are certified with plans that are accepted 696 00:57:21.159 --> 00:57:24.039 by most of the medical groups in our area. Call nine oh nine seven 697 00:57:24.159 --> 00:57:29.559 nine three oh three eight five. Their service is free and after forty two 698 00:57:29.679 --> 00:57:32.039 years of the business, their agents are trained to help you pick the plan 699 00:57:32.159 --> 00:57:38.119 that's right for you. NBC News Radio, I'm Chris Karragio. A strong 700 00:57:38.159 --> 00:57:43.039 Pacific storm is lashing the West coast, with parts of southern California now under 701 00:57:43.119 --> 00:57:46.559 an extreme flash flood risk. The Weather Prediction Center is warning of life threatening 702 00:57:46.599 --> 00:57:51.719 flash flooding from Monterey to Los Angeles, and it's lil US update the National 703 00:57:51.719 --> 00:57:54.880 Weather Service as the ongoing atmospheric river event will continue to produce heavy rain, 704 00:57:54.960 --> 00:57:59.800 strong winds, high surf, and heavy snow and the higher elevations across central 705 00:58:00.039 --> 00:58:04.760 in southern California over the next few days. More retaliatory air strikes are expected 706 00:58:04.760 --> 00:58:07.920 as the US and the UK target Hooti rebels in Yemen. Pentagon officials say 707 00:58:07.920 --> 00:58:12.840 the lightest air strikes hit thirty six sites, including underground weapons facilities, in 708 00:58:12.880 --> 00:58:16.119 response to attacks on commercial ships in the Red Sea. Appearing on NBC's Meet 709 00:58:16.159 --> 00:58:21.320 the Press, National Security Advisor Jake Sullivan says US forces are carrying out President 710 00:58:21.320 --> 00:58:24.199 Biden's order for a serious response, which is now under way. Meanwhile, 711 00:58:24.320 --> 00:58:29.440 Secretary of State Anthony Blincoln is headed back to the Middle East to rally you 712 00:58:29.519 --> 00:58:34.760 a support among America's Arab allies in the region. The National Security Council spokesman 713 00:58:34.800 --> 00:58:38.320 says the US gave Iraq sufficient advanced notice about the air strikes against Iran back 714 00:58:38.400 --> 00:58:44.519 militias. We also want to see the Iraqi government move with more alacrity to 715 00:58:44.559 --> 00:58:47.920 help US read the threat of these militia groups on Iraqi soil. Speaking on 716 00:58:49.000 --> 00:58:52.480 Fox News Sunday, John Kirby said there were appropriate notifications and discussions with the 717 00:58:52.519 --> 00:58:57.800 Iraqi government prior to Friday's air strikes in Iraq and Syria. Iraq condemned the 718 00:58:57.840 --> 00:59:02.639 attacks and called US claims of Gordon with the Iraqi government unfounded. GOP presidential 719 00:59:02.639 --> 00:59:07.639 hopeful and former UN Ambassador NICKI Haley says the deadly drone attack on the US 720 00:59:07.639 --> 00:59:10.840 outpost in Jordan could have been prevented. Speaking on CNN State of the Union, 721 00:59:10.920 --> 00:59:15.039 Haley said President Biden's lifting of sanctions against Iran freed up money that's being 722 00:59:15.119 --> 00:59:19.760 used to fund attacks such as the one on January twenty eighth. She also 723 00:59:19.840 --> 00:59:23.880 criticized the president for not responding to previous strikes against American soldiers since the start 724 00:59:23.920 --> 00:59:29.239 of the Israel homass War and waiting until three service members were killed in the 725 00:59:29.280 --> 00:59:35.559 most recent strike. I'm Chris Karagio, NBC News Radio, NBC News on 726 00:59:35.719 --> 00:59:42.119 CACAA Lomlada sponsored by Teamsters Local nineteen thirty two Protecting the Future of Working Families 727 00:59:42.159 --> 00:59:52.159 Teamsters nineteen thirty two, dot Org, Dutch Bakery and Variety of Foods in 728 00:59:52.199 --> 00:59:55.039 Ontario is reminding everyone to please give blood. By doing so, you may 729 00:59:55.079 --> 00:59:59.679 save the life of a child, surgery patient or an accident victim. So 730 00:59:59.719 --> 01:00:02.360 give blood. It's safe, it's simple, and it's needed. This message 731 01:00:02.480 --> 01:00:07.760 courtesy of your friends at Dutch Bakery and Variety Foods in Ontario for a great 732 01:00:07.760 --> 01:00:08.400 selection you eat imported food from all