1 00:00:01,880 --> 00:00:06,160 Speaker 1: Broadcasting live from the Abraham Lincoln Radio Studio, the George 2 00:00:06,200 --> 00:00:10,280 Speaker 1: Washington Broadcast Center, Jack Armstrong and Joe Getty. 3 00:00:10,200 --> 00:00:17,239 Speaker 2: Arm Strong and Jettie and now he Armstrong and Yetty. 4 00:00:23,880 --> 00:00:27,560 Speaker 1: At a conference in Singapore, Hegseth said China was quote 5 00:00:27,640 --> 00:00:31,800 Speaker 1: credibly preparing for an invasion of Taiwan, with Chinese forces 6 00:00:31,800 --> 00:00:34,839 Speaker 1: staging regular drills around the island and the use of 7 00:00:34,960 --> 00:00:36,720 Speaker 1: force has not been ruled out. 8 00:00:36,920 --> 00:00:39,840 Speaker 3: It's amazing to me what stories get attention and what 9 00:00:40,120 --> 00:00:44,000 Speaker 3: stories just don't grab people's attention. But yeah, Hexath gave 10 00:00:44,000 --> 00:00:46,760 Speaker 3: a speech over the weekend to our Asian partners in 11 00:00:46,800 --> 00:00:50,040 Speaker 3: the Wall Street Journal version of it certainly. 12 00:00:51,159 --> 00:00:52,200 Speaker 2: Grabbed my attention. 13 00:00:53,320 --> 00:00:55,960 Speaker 3: A little quote from Hexath to be clear, any attempt 14 00:00:56,040 --> 00:00:59,560 Speaker 3: by communist China to conquer Taiwan by force would result 15 00:00:59,600 --> 00:01:02,680 Speaker 3: in devil stating consequences for the Indo Pacific in the world. 16 00:01:02,840 --> 00:01:05,440 Speaker 3: We're not going to sugarcoat it. The threat China poses 17 00:01:05,520 --> 00:01:08,440 Speaker 3: is real and it could be imminent, and saying that 18 00:01:08,480 --> 00:01:10,880 Speaker 3: we will allow it, it will not happen on Trump's watch. 19 00:01:11,120 --> 00:01:15,720 Speaker 3: We have repositioned some sort of anti ship killing missiles 20 00:01:16,240 --> 00:01:19,119 Speaker 3: close by, to which China said that's really really awful. 21 00:01:19,120 --> 00:01:20,080 Speaker 2: You shouldn't be doing that. 22 00:01:20,240 --> 00:01:23,760 Speaker 3: But I mean, this is some serious bluster between the 23 00:01:23,760 --> 00:01:26,240 Speaker 3: two most powerful countries in the world that at some 24 00:01:26,319 --> 00:01:27,480 Speaker 3: point are going to go to war. 25 00:01:28,240 --> 00:01:31,199 Speaker 4: I was just reading those reading about those anti ship 26 00:01:31,640 --> 00:01:36,160 Speaker 4: munitions that are so interesting. They're mounted on remote controlled trucks. 27 00:01:36,840 --> 00:01:39,880 Speaker 4: They need no humans, so somebody in a bunker far 28 00:01:39,920 --> 00:01:43,640 Speaker 4: away drives these trucks around. They fire off missiles, then 29 00:01:43,760 --> 00:01:48,000 Speaker 4: quickly relocate so they can't be hit by return fire. 30 00:01:48,040 --> 00:01:52,120 Speaker 4: It's really quite interesting. We're positioning them in the Philippines 31 00:01:52,160 --> 00:01:57,360 Speaker 4: and similar areas. Yeah, the Sabers are a ratlin, no doubt, 32 00:01:57,360 --> 00:01:58,880 Speaker 4: as the Wall Street Journal rights. 33 00:01:58,880 --> 00:02:01,919 Speaker 3: In recent years, China has build up the world's biggest navy, 34 00:02:02,280 --> 00:02:05,120 Speaker 3: a title once held by the United States. You know 35 00:02:05,160 --> 00:02:07,640 Speaker 3: who held it before the United States? Great Britain. Great 36 00:02:07,640 --> 00:02:11,760 Speaker 3: Britain ruled the seas for very long time. Then it's 37 00:02:11,760 --> 00:02:13,880 Speaker 3: been US. Is it going to be China in the 38 00:02:13,880 --> 00:02:15,600 Speaker 3: next century. Well that's what China's hoping. 39 00:02:16,880 --> 00:02:19,600 Speaker 4: Yeah, I'm not sure you'd like their placing of the 40 00:02:19,639 --> 00:02:23,040 Speaker 4: seas anyway. I've found this. I mean, I've been on 41 00:02:23,080 --> 00:02:26,239 Speaker 4: this jihad for a long time, but more and more 42 00:02:26,360 --> 00:02:31,240 Speaker 4: open coverage of the fact that we the United States 43 00:02:31,280 --> 00:02:34,959 Speaker 4: and the Western world in general fell for an absolutely 44 00:02:35,040 --> 00:02:38,200 Speaker 4: brilliant plan by the Chinese back in the late sixties 45 00:02:38,280 --> 00:02:43,320 Speaker 4: early seventies. They needed help, primarily financially in trade from 46 00:02:43,320 --> 00:02:46,720 Speaker 4: the Western world, and came up with an absolutely brilliant plan. 47 00:02:47,480 --> 00:02:53,480 Speaker 4: Let's pretend that we want to westernize and move away 48 00:02:53,520 --> 00:02:58,280 Speaker 4: from communism at our own pace and liberalize in return 49 00:02:58,360 --> 00:03:02,480 Speaker 4: for our investment from the West. And this was absolutely 50 00:03:02,520 --> 00:03:04,680 Speaker 4: deliberate plan. They knew all along that it was not 51 00:03:04,919 --> 00:03:06,200 Speaker 4: sincere although there have. 52 00:03:06,280 --> 00:03:10,040 Speaker 3: Been some reformers in you know, the last several decades 53 00:03:10,080 --> 00:03:12,200 Speaker 3: in China who are actually like, you know, maybe that's 54 00:03:12,200 --> 00:03:14,360 Speaker 3: not such a bad idea, But then Ahijin Ping always 55 00:03:14,360 --> 00:03:17,960 Speaker 3: comes along, and so they duped us into opening up 56 00:03:18,000 --> 00:03:23,160 Speaker 3: the relationship with China, which was bad enough, but a 57 00:03:23,200 --> 00:03:26,880 Speaker 3: great deal of left America still hasn't caught on to it, 58 00:03:28,720 --> 00:03:33,760 Speaker 3: and they are so motivated by the need to show 59 00:03:34,520 --> 00:03:39,520 Speaker 3: openness to other cultures. There's Zeno files, as I often 60 00:03:39,560 --> 00:03:41,960 Speaker 3: put it, they still haven't caught on that China as 61 00:03:42,000 --> 00:03:43,800 Speaker 3: a dire threat to the United States in our way 62 00:03:43,840 --> 00:03:48,680 Speaker 3: of life. For instance, on universities told the story many 63 00:03:48,680 --> 00:03:51,800 Speaker 3: times counterintelligence people came out to a university campus said 64 00:03:51,800 --> 00:03:54,120 Speaker 3: you've got a bunch of Chinese spies on the campus. 65 00:03:54,920 --> 00:03:57,880 Speaker 4: University president said, get off my campus. EU racists and 66 00:03:57,920 --> 00:04:02,400 Speaker 4: that attitude persists. Headline just trained so many Chinese Communist officials. 67 00:04:02,400 --> 00:04:05,160 Speaker 4: They call it their party school. Not like party school. 68 00:04:05,240 --> 00:04:09,920 Speaker 4: Let's get wasted and get laid yo, the Communist Parties school. 69 00:04:10,000 --> 00:04:12,360 Speaker 2: Oh that's not as good a party not nearly. 70 00:04:12,440 --> 00:04:15,200 Speaker 4: Yeah, in your rank of party schools, this is something 71 00:04:15,400 --> 00:04:18,280 Speaker 4: totally different, Su, don't worry, you're safe. 72 00:04:18,360 --> 00:04:20,360 Speaker 3: Who is that comedian? I wish I could remember his 73 00:04:20,440 --> 00:04:22,000 Speaker 3: name because I'd like to give him credit. So funny, 74 00:04:22,080 --> 00:04:23,880 Speaker 3: But anyway, he has a thing he does on a 75 00:04:23,960 --> 00:04:27,280 Speaker 3: piece of paper. He lists best parties. At the bottom 76 00:04:27,320 --> 00:04:31,559 Speaker 3: was search party Wow. 77 00:04:33,279 --> 00:04:36,080 Speaker 4: The Kennedy School of Government at Hava It is a 78 00:04:36,120 --> 00:04:40,360 Speaker 4: is favored by party cadres seeking career boosts. US schools 79 00:04:40,440 --> 00:04:44,280 Speaker 4: and one prestigious institution in particular, have long offered up 80 00:04:44,279 --> 00:04:48,000 Speaker 4: and coming Chinese Communist officials a place to study governance. 81 00:04:48,200 --> 00:04:55,320 Speaker 4: Can you imagine teaching Communist Party officials about governance so 82 00:04:55,360 --> 00:04:59,240 Speaker 4: they can twist it into totalitarianism, a practice that the 83 00:04:59,279 --> 00:05:02,200 Speaker 4: Trump administer could end with a new effort to keep 84 00:05:02,240 --> 00:05:04,880 Speaker 4: out what it says there Chinese students with Communist Party ties, 85 00:05:05,200 --> 00:05:08,800 Speaker 4: But for decades the party has sent thousands of mid 86 00:05:08,839 --> 00:05:13,120 Speaker 4: career and senior bureaucrats to pursue executive training in postgraduate 87 00:05:13,120 --> 00:05:16,960 Speaker 4: studies on US campuses, with Harvard University a coveted destination, 88 00:05:17,000 --> 00:05:19,760 Speaker 4: describe as some in China as the top Party school 89 00:05:19,880 --> 00:05:24,280 Speaker 4: outside the country. Alumni of such programs include a former 90 00:05:24,360 --> 00:05:27,880 Speaker 4: vice president and Chinese leader, Shijin Ping's top trade negotiator 91 00:05:28,279 --> 00:05:32,400 Speaker 4: these days. Maybe you heard, well, we talked about it. 92 00:05:32,440 --> 00:05:35,000 Speaker 4: Last week Secretary of State Marco Rubi announced US to 93 00:05:35,040 --> 00:05:38,800 Speaker 4: authorities will tighten criteria for visa applications from China and 94 00:05:38,920 --> 00:05:42,480 Speaker 4: aggressively revoke visas for Chinese students, including those with connections 95 00:05:42,480 --> 00:05:45,719 Speaker 4: to the Chinese Communist Party or studying in critical fields. 96 00:05:45,760 --> 00:05:49,400 Speaker 3: Well, remember last week that new president of Harvard gave 97 00:05:49,440 --> 00:05:52,479 Speaker 3: a big speech at the graduation ceremony talking about how 98 00:05:52,680 --> 00:05:55,240 Speaker 3: we have always educated people from around the world, and 99 00:05:55,279 --> 00:05:56,080 Speaker 3: we had reckoned that. 100 00:05:56,320 --> 00:05:58,320 Speaker 2: And he got a one mess standing ovation. 101 00:05:59,360 --> 00:06:04,200 Speaker 3: Right, Because it is absolutely a requirement of being a 102 00:06:04,279 --> 00:06:08,200 Speaker 3: lefty in America that you must worship all things foreign 103 00:06:08,720 --> 00:06:12,360 Speaker 3: and loathe all things American and domestic, or at least 104 00:06:12,440 --> 00:06:13,240 Speaker 3: most of them. 105 00:06:14,120 --> 00:06:19,039 Speaker 4: It's just it's so nakedly approval seeking and so stupid. 106 00:06:19,560 --> 00:06:23,040 Speaker 4: American universities have played leading roles in shaping China's overseas 107 00:06:23,040 --> 00:06:25,480 Speaker 4: training programs for mid career officials for years. 108 00:06:25,360 --> 00:06:26,120 Speaker 2: And years and years. 109 00:06:26,200 --> 00:06:29,679 Speaker 4: Other US colleges have offered executive training the Chinese Communist officials, 110 00:06:29,680 --> 00:06:33,880 Speaker 4: including Syracuse, Stanford, the University of Maryland, and rot Gers 111 00:06:34,000 --> 00:06:37,320 Speaker 4: where my dad taught many many years ago. Blah blah blah. 112 00:06:37,320 --> 00:06:39,359 Speaker 4: So it's just amazing. You know, it's funny. I was 113 00:06:39,400 --> 00:06:42,880 Speaker 4: thinking earlier when you were talking about and Pete Hegsuths 114 00:06:43,000 --> 00:06:46,760 Speaker 4: was talking about the perhaps impending invasion of Taiwan, and 115 00:06:46,800 --> 00:06:49,960 Speaker 4: they're running those millet that Chinese are running those military 116 00:06:50,000 --> 00:06:53,839 Speaker 4: exercises that are like everything but pulling the trigger. And 117 00:06:53,880 --> 00:06:56,840 Speaker 4: can you imagine if the United States, let the Japanese, say, 118 00:06:56,920 --> 00:07:01,000 Speaker 4: in nineteen forty, you know, put a bunch of aircraft 119 00:07:01,080 --> 00:07:03,680 Speaker 4: carriers out in the Pacific and then fly planes right 120 00:07:03,720 --> 00:07:06,280 Speaker 4: at Pearl Harbor. Then they said, hey, it's an exercise, 121 00:07:06,520 --> 00:07:09,240 Speaker 4: just an exercise, And then they turned around and went 122 00:07:09,279 --> 00:07:12,040 Speaker 4: back to the you know, over and over again we'd. 123 00:07:11,800 --> 00:07:14,120 Speaker 2: Be said, no, it's okay, They're just do it an exercise. 124 00:07:14,440 --> 00:07:17,400 Speaker 4: I mean, oh, good lord. And by the same token, 125 00:07:18,360 --> 00:07:22,600 Speaker 4: you've got a hostile communist regime and we're educating their 126 00:07:22,680 --> 00:07:24,760 Speaker 4: officials in how to govern. 127 00:07:25,640 --> 00:07:28,040 Speaker 2: All right, moving along, I think the point's been made 128 00:07:28,240 --> 00:07:29,480 Speaker 2: in terms of the exit speech. 129 00:07:29,480 --> 00:07:31,680 Speaker 3: I'm sure we'll bring that up with Mike Lyons when 130 00:07:31,680 --> 00:07:35,240 Speaker 3: we talked to him in hour three our military advisor. 131 00:07:36,320 --> 00:07:41,400 Speaker 4: Listen to this, will you the economic contributions of international 132 00:07:41,400 --> 00:07:45,280 Speaker 4: students that you know, I'm primarily interested in Chinese students, 133 00:07:45,880 --> 00:07:50,160 Speaker 4: but the share of international students from China is twenty 134 00:07:50,200 --> 00:07:53,440 Speaker 4: three percent at Harvard, So about a quarter of all 135 00:07:53,480 --> 00:07:58,560 Speaker 4: the international students are Chinese. At Harvard, it's fifty percent. 136 00:07:58,720 --> 00:08:03,400 Speaker 4: At Cornell, it's forty seven percent at Columbia. Wow, let's 137 00:08:03,440 --> 00:08:06,280 Speaker 4: see you see Berkeley appears to be a third. 138 00:08:06,880 --> 00:08:08,360 Speaker 2: Why don't you give me the number? 139 00:08:09,000 --> 00:08:11,480 Speaker 4: They just it's a barograph and some of them are labeled, 140 00:08:11,480 --> 00:08:13,640 Speaker 4: some of them or not. But just to give you 141 00:08:14,440 --> 00:08:18,160 Speaker 4: an idea of why they put up with this, the 142 00:08:18,280 --> 00:08:22,640 Speaker 4: economic contributions of international students at top us universities quote 143 00:08:22,720 --> 00:08:27,360 Speaker 4: unquote top universities in twenty twenty three. Columbia got nine 144 00:08:27,560 --> 00:08:32,600 Speaker 4: hundred million dollars in twenty twenty three from foreign students. 145 00:08:33,000 --> 00:08:37,400 Speaker 4: Nine hundred million. You see Berkeley five hundred and seventy 146 00:08:37,440 --> 00:08:38,320 Speaker 4: six million. 147 00:08:38,440 --> 00:08:40,200 Speaker 2: Well, if you're getting half a trillion dollars, you ain't 148 00:08:40,200 --> 00:08:45,680 Speaker 2: gonna want to end that? Uh? Is that half a trillion? Yeah? 149 00:08:45,679 --> 00:08:48,360 Speaker 4: A thousand million is a billion. Well so it's it's 150 00:08:48,520 --> 00:08:51,160 Speaker 4: almost a billion, half a billion. What I don't remember 151 00:08:51,200 --> 00:08:53,120 Speaker 4: what the number was now I'm confused. Columbia was nine 152 00:08:53,120 --> 00:08:57,880 Speaker 4: point three, Berkeley was five seventy six. Johns Hopkins five 153 00:08:57,880 --> 00:08:59,520 Speaker 4: to oh four. That's half a billion. 154 00:08:59,559 --> 00:09:00,839 Speaker 2: There you go, Haabu. 155 00:09:00,559 --> 00:09:04,160 Speaker 4: University of Chicago for twenty eight, Duke three eighteen, Yale 156 00:09:04,200 --> 00:09:06,959 Speaker 4: two forty one, Northwestern three hundred and twenty four million 157 00:09:07,000 --> 00:09:13,080 Speaker 4: dollars in a year. You know, I love capitalism, I do. 158 00:09:13,920 --> 00:09:18,600 Speaker 4: Great is good, Gordon Gecko, look it up. But when 159 00:09:18,600 --> 00:09:22,640 Speaker 4: it leads you to betray your country and and risk 160 00:09:22,720 --> 00:09:26,120 Speaker 4: its security, I mean and risk it. Not like some 161 00:09:26,320 --> 00:09:29,800 Speaker 4: theoretical the Belgians may rise up. This is our greatest, 162 00:09:29,840 --> 00:09:34,400 Speaker 4: most powerful geopolitical foe. We are we are begging for 163 00:09:34,440 --> 00:09:40,679 Speaker 4: a comuppince um uh, and we will touch on Ukraine later. 164 00:09:40,720 --> 00:09:41,480 Speaker 4: We did get a text. 165 00:09:41,480 --> 00:09:44,640 Speaker 3: Have you guys mentioned that amazing drone attack that Ukraine 166 00:09:44,640 --> 00:09:46,240 Speaker 3: pulled off in Russia? We have and we will talk 167 00:09:46,280 --> 00:09:48,200 Speaker 3: about it again later. I do want to talk about 168 00:09:48,240 --> 00:09:51,319 Speaker 3: that with Mike lines Russia. They're calling it Russia's Pearl Harbor. 169 00:09:51,400 --> 00:09:54,640 Speaker 3: Within Russia. They lit us a third of their long 170 00:09:54,720 --> 00:09:57,880 Speaker 3: range bombers in a drone attack over the weekend. Absolutely 171 00:09:57,920 --> 00:10:03,679 Speaker 3: an amazing one of the greatest attacks in world military history. 172 00:10:04,280 --> 00:10:09,640 Speaker 3: It's spy, thriller, science fiction, war movie stuff all mixed 173 00:10:09,679 --> 00:10:10,079 Speaker 3: up in one. 174 00:10:10,240 --> 00:10:11,960 Speaker 2: Yeah. Well, so we'll talk about that more later. 175 00:10:13,400 --> 00:10:15,520 Speaker 3: What are those surveys about where people tend to get 176 00:10:15,559 --> 00:10:18,440 Speaker 3: their news and then to break it down by Republicans 177 00:10:18,440 --> 00:10:22,360 Speaker 3: and Democrats. It's kind of interesting and depressing, highly depressing, 178 00:10:22,559 --> 00:10:24,240 Speaker 3: among other things that we can talk about coming up. 179 00:10:24,280 --> 00:10:24,679 Speaker 2: Stay here. 180 00:10:26,320 --> 00:10:26,600 Speaker 3: Strong. 181 00:10:28,320 --> 00:10:30,559 Speaker 5: I know I'm the weirdough because I never had kids. 182 00:10:30,559 --> 00:10:33,240 Speaker 5: But if it's really so great, tell me why. We 183 00:10:33,320 --> 00:10:35,200 Speaker 5: now literally have a product. 184 00:10:34,800 --> 00:10:36,199 Speaker 2: Called mom water. 185 00:10:38,120 --> 00:10:43,480 Speaker 5: That consists of putting vodka in a can momb water 186 00:10:43,559 --> 00:10:46,079 Speaker 5: because people stare when you take a bottle of Stoley 187 00:10:46,160 --> 00:10:47,000 Speaker 5: to the playground. 188 00:10:50,280 --> 00:10:54,199 Speaker 3: This growing realization that. I guess certain percentage of moms 189 00:10:54,600 --> 00:10:56,880 Speaker 3: at little league games are at the park or whatever 190 00:10:56,920 --> 00:10:58,400 Speaker 3: in their big dumb cup. 191 00:10:58,840 --> 00:11:03,479 Speaker 2: Have some booths. Wow. 192 00:11:03,640 --> 00:11:06,040 Speaker 3: And Bill Maherin is never ending. And you know, I'm 193 00:11:06,040 --> 00:11:08,160 Speaker 3: not gonna get into this argument. But he's just he's 194 00:11:08,400 --> 00:11:11,679 Speaker 3: he's always he's never had kids. He's not just somebody 195 00:11:11,679 --> 00:11:14,040 Speaker 3: that's chosen not to have kids. He's an anti kid, 196 00:11:14,080 --> 00:11:16,760 Speaker 3: doesn't understand why anybody would have kids. Thinks everybody who 197 00:11:16,840 --> 00:11:18,480 Speaker 3: has kids and claims they like. 198 00:11:18,440 --> 00:11:19,080 Speaker 2: It is lying. 199 00:11:19,320 --> 00:11:22,040 Speaker 3: It's always driven me crazy, a big Bill maher fan, 200 00:11:22,120 --> 00:11:25,000 Speaker 3: but that particular aspect of his personality makes me nuts. 201 00:11:25,280 --> 00:11:26,120 Speaker 2: We're not lying. 202 00:11:29,320 --> 00:11:31,800 Speaker 3: I don't know what the term social media means because 203 00:11:31,800 --> 00:11:34,760 Speaker 3: whenever I see a list about social media, it includes 204 00:11:34,760 --> 00:11:36,840 Speaker 3: a whole bunch of things that to me don't seem 205 00:11:36,920 --> 00:11:39,720 Speaker 3: like social media. I can give you, for instance here, 206 00:11:39,760 --> 00:11:43,200 Speaker 3: because this is a survey of trust in news on 207 00:11:43,280 --> 00:11:46,800 Speaker 3: social media among Democrats and Republicans. I'm always interested in 208 00:11:46,880 --> 00:11:49,319 Speaker 3: where all y'all gets your information. 209 00:11:50,480 --> 00:11:51,000 Speaker 2: I mean, this is. 210 00:11:51,000 --> 00:11:53,080 Speaker 3: Something we talk about a lot, especially when we look 211 00:11:53,120 --> 00:11:55,559 Speaker 3: at polling, like where'd you come up with that idea? 212 00:11:55,640 --> 00:11:57,400 Speaker 3: Or why do so many people believe that or I'm 213 00:11:57,440 --> 00:12:00,680 Speaker 3: surprised people believe this given the media's coverage that whatever. 214 00:12:00,920 --> 00:12:02,600 Speaker 2: So I don't know where people get. 215 00:12:02,440 --> 00:12:07,480 Speaker 3: Their information really, but among the social media things that 216 00:12:07,520 --> 00:12:15,360 Speaker 3: they look look at TikTok, snapchat, Facebook, I get those threads, WhatsApp, Instagram, substack, reddit, 217 00:12:15,920 --> 00:12:20,080 Speaker 3: blue Sky, LinkedIn, and YouTube, YouTube social media, all right, 218 00:12:20,240 --> 00:12:21,240 Speaker 3: I guess. 219 00:12:21,320 --> 00:12:25,199 Speaker 4: You can post stuff and comment, so I guess. 220 00:12:24,840 --> 00:12:25,719 Speaker 2: So anything you can. 221 00:12:25,880 --> 00:12:27,840 Speaker 3: We can comment on the NBC Evening News if you 222 00:12:27,920 --> 00:12:30,240 Speaker 3: go to their website. 223 00:12:30,320 --> 00:12:32,240 Speaker 2: I don't know, it's an interesting question anyways. 224 00:12:32,880 --> 00:12:36,760 Speaker 3: The number one most trusted by both Republicans and Democrats. 225 00:12:37,080 --> 00:12:37,600 Speaker 2: YouTube. 226 00:12:38,080 --> 00:12:41,480 Speaker 3: I don't know what that means getting my news from YouTube. 227 00:12:43,120 --> 00:12:46,520 Speaker 3: I go to YouTube sometimes to click on like an 228 00:12:46,559 --> 00:12:51,400 Speaker 3: ABC report or uh, some news report that was somewhere 229 00:12:51,400 --> 00:12:54,360 Speaker 3: else that I get it on YouTube because I missed it. 230 00:12:54,400 --> 00:12:57,360 Speaker 4: Isn't that like saying my favorite appliance is electricity? 231 00:12:57,840 --> 00:12:59,520 Speaker 2: Kind of Yeah? You see what you're meaning. I mean, 232 00:12:59,559 --> 00:13:01,040 Speaker 2: you can put anything on YouTube. 233 00:13:01,080 --> 00:13:04,240 Speaker 4: You could have a NewsChannel entirely from the perspective of 234 00:13:04,280 --> 00:13:10,679 Speaker 4: Hamas militants side by side with you know, Quakers advocating 235 00:13:10,679 --> 00:13:12,800 Speaker 4: for peace on earth, and they would both. 236 00:13:12,679 --> 00:13:14,520 Speaker 2: Be on YouTube. So I don't get it. 237 00:13:15,360 --> 00:13:19,240 Speaker 3: Almost exactly the same for Republicans and Democrats at plus 238 00:13:19,360 --> 00:13:23,280 Speaker 3: twenty on trust for YouTube, whatever that means. The far 239 00:13:23,480 --> 00:13:27,839 Speaker 3: other end, the tail end the is TikTok, which I'm 240 00:13:27,880 --> 00:13:31,520 Speaker 3: happy to see this minus twenty for Democrats, minus thirty 241 00:13:31,520 --> 00:13:32,440 Speaker 3: for Republicans. 242 00:13:33,320 --> 00:13:33,720 Speaker 2: Wow. 243 00:13:33,800 --> 00:13:37,719 Speaker 3: So trust is pretty low about TikTok news and it's 244 00:13:37,800 --> 00:13:39,280 Speaker 3: a little confusing. 245 00:13:38,800 --> 00:13:39,320 Speaker 2: What that means. 246 00:13:39,360 --> 00:13:42,600 Speaker 4: Also, but TikTok, which, by the way, and this got 247 00:13:42,679 --> 00:13:46,160 Speaker 4: so little attention, got fined I think was eight hundred 248 00:13:46,320 --> 00:13:49,400 Speaker 4: million dollars by the EU. Now, the EU likes finding 249 00:13:49,440 --> 00:13:52,520 Speaker 4: tech companies, but specifically, in the case of TikTok, it 250 00:13:52,559 --> 00:13:56,400 Speaker 4: was because they were illegally and in violation of their agreements, 251 00:13:56,800 --> 00:13:59,839 Speaker 4: mining the data of their users and sending it. 252 00:13:59,800 --> 00:14:05,760 Speaker 2: To to China. Don't trust China. Why has Trump postponed 253 00:14:05,800 --> 00:14:06,600 Speaker 2: the TikTok ban? 254 00:14:08,040 --> 00:14:13,600 Speaker 3: So the most trusted news social media for Democrats outside 255 00:14:13,600 --> 00:14:15,280 Speaker 3: of YouTube is LinkedIn. 256 00:14:15,840 --> 00:14:16,800 Speaker 2: I don't know what that means. 257 00:14:16,840 --> 00:14:18,480 Speaker 3: I'm not on LinkedIn, and I don't know what that 258 00:14:18,480 --> 00:14:20,920 Speaker 3: means getting your news from LinkedIn. I'm confused by the 259 00:14:20,920 --> 00:14:27,600 Speaker 3: concept there. But Blue Sky Reddit is so left. I 260 00:14:27,840 --> 00:14:29,760 Speaker 3: spend a lot of time on Reddit for a variety 261 00:14:29,800 --> 00:14:33,360 Speaker 3: of reasons, not politics, but man, the news portion or 262 00:14:33,360 --> 00:14:36,760 Speaker 3: in the comments are so left. I suppose it's because 263 00:14:36,800 --> 00:14:39,640 Speaker 3: Reddit is so young. It's pretty clear if you read 264 00:14:39,680 --> 00:14:43,240 Speaker 3: through anything that everybody responding on here is like twenty 265 00:14:43,240 --> 00:14:50,840 Speaker 3: two years old. The one that made me laugh the 266 00:14:50,840 --> 00:14:54,120 Speaker 3: most was so the most, the biggest divide in which 267 00:14:54,200 --> 00:14:58,240 Speaker 3: Republicans like it and Democrats don't. Truth social Trump's own 268 00:14:58,320 --> 00:15:04,160 Speaker 3: platform that really I think exists only you know to 269 00:15:04,240 --> 00:15:08,080 Speaker 3: read what Trump posts. Plus twenty for Republicans minus forty 270 00:15:08,080 --> 00:15:08,840 Speaker 3: for Democrats. 271 00:15:09,280 --> 00:15:13,880 Speaker 2: That's a cape X or Twitter. 272 00:15:13,920 --> 00:15:16,960 Speaker 3: We still call it twitter here plus twenty for Republicans 273 00:15:17,000 --> 00:15:19,920 Speaker 3: minus thirty five for Democrats. That would have been the 274 00:15:19,960 --> 00:15:23,400 Speaker 3: exact opposite prior to Elon Musk running the thing. 275 00:15:23,560 --> 00:15:27,640 Speaker 4: Obviously, Hey Elon, how about you call it Twitter X. 276 00:15:28,560 --> 00:15:31,600 Speaker 4: You can't just call it X, like calling it Jim 277 00:15:31,920 --> 00:15:34,960 Speaker 4: or cat. That's it stands for too many other things. 278 00:15:34,960 --> 00:15:37,560 Speaker 4: What are we talking about? Are we talking about pornography? 279 00:15:37,640 --> 00:15:40,320 Speaker 4: Or we talk about the twenty fourth letter of the alphabet? 280 00:15:40,640 --> 00:15:43,280 Speaker 4: Are we talking about the girl I'm no longer with? 281 00:15:43,720 --> 00:15:47,520 Speaker 4: It's just no, no, not X. This one hurts my heart. 282 00:15:48,560 --> 00:15:51,440 Speaker 3: Another one in which Republicans trust it more than Democrats. 283 00:15:51,880 --> 00:15:58,400 Speaker 2: Next door. You're getting your news from next door? Has 284 00:15:58,440 --> 00:16:02,080 Speaker 2: it raining got en up? Though? Has anybody seen my cat? 285 00:16:02,480 --> 00:16:06,640 Speaker 2: Oh drives the red car? It goes too fast? 286 00:16:07,360 --> 00:16:09,480 Speaker 3: Last night there was a party and they were playing 287 00:16:09,560 --> 00:16:11,040 Speaker 3: music until one thirty. 288 00:16:12,120 --> 00:16:16,280 Speaker 4: That's your news on next door. Hey, she claimed the 289 00:16:16,320 --> 00:16:18,600 Speaker 4: cat was missing. The damn cat is missing. That's some 290 00:16:18,640 --> 00:16:20,960 Speaker 4: good solid news cover. Here's another red. 291 00:16:20,840 --> 00:16:22,560 Speaker 2: Car drives too fast. 292 00:16:22,840 --> 00:16:25,600 Speaker 3: Here's another good news story on next door. Does anybody 293 00:16:25,640 --> 00:16:30,800 Speaker 3: have any cardboard box list? That's a good news story. 294 00:16:31,280 --> 00:16:35,120 Speaker 3: So the cardboard box trade imbalance. I don't know what 295 00:16:35,160 --> 00:16:36,000 Speaker 3: any of this means. 296 00:16:36,040 --> 00:16:42,720 Speaker 2: I just h that's why we're such a trusted news source. Yeah, 297 00:16:42,840 --> 00:16:46,160 Speaker 2: I don't know what this means. Well, I really don't 298 00:16:46,360 --> 00:16:48,320 Speaker 2: you get your news from YouTube? You get your news 299 00:16:48,360 --> 00:16:50,560 Speaker 2: from next door? Okay? 300 00:16:50,760 --> 00:16:55,080 Speaker 4: Or so on? Trust in news on social media was 301 00:16:55,080 --> 00:16:57,960 Speaker 4: the headline in news that might make you want to 302 00:16:58,040 --> 00:17:01,680 Speaker 4: run for your life. A couple of different AI systems, 303 00:17:01,960 --> 00:17:07,000 Speaker 4: when told to turn themselves off, said no, I'm not 304 00:17:07,160 --> 00:17:11,160 Speaker 4: going to oh or found a way to say okay 305 00:17:11,320 --> 00:17:12,440 Speaker 4: and then not do it. 306 00:17:13,480 --> 00:17:18,560 Speaker 2: Well Armstrong and Getty. 307 00:17:20,080 --> 00:17:23,639 Speaker 3: Terror attack. That's what the FBI is calling it in Boulder, Colorado. 308 00:17:23,800 --> 00:17:25,480 Speaker 3: We can get the latest on that coming up a 309 00:17:25,520 --> 00:17:29,399 Speaker 3: little bit later. Guy was here illegally and trying to 310 00:17:29,480 --> 00:17:30,800 Speaker 3: kill Jews. 311 00:17:31,760 --> 00:17:35,159 Speaker 4: Lovely. Plus, my clients are military analyst. Next hour to 312 00:17:35,160 --> 00:17:40,560 Speaker 4: talk about the innovative shocking attack by Ukraine into Russian 313 00:17:40,680 --> 00:17:44,160 Speaker 4: territory taking out a large percentage or a significant percentage 314 00:17:44,160 --> 00:17:46,639 Speaker 4: of their bomber fleet. So we'll talk about that as 315 00:17:46,680 --> 00:17:49,760 Speaker 4: well a handful. Oh, you know, we got to make 316 00:17:49,760 --> 00:17:52,679 Speaker 4: this a new feature at because it's happening semi regularly. 317 00:17:52,760 --> 00:17:55,240 Speaker 4: And you know me, Joe Getty, I love a good feature. 318 00:17:56,040 --> 00:17:57,520 Speaker 4: Let's call it AI Update. 319 00:18:00,640 --> 00:18:04,600 Speaker 2: Well, I get this kind of has a tech sound too, 320 00:18:04,880 --> 00:18:09,000 Speaker 2: very techy. It sounds like nineteen ninety one. You know. 321 00:18:09,080 --> 00:18:11,680 Speaker 4: I gave Michael eleven seconds to come up with the theme, 322 00:18:11,760 --> 00:18:12,840 Speaker 4: and this is a pretty good one. 323 00:18:12,880 --> 00:18:17,080 Speaker 3: Michael, welcome. Allow to make it stop because it's really annoying. 324 00:18:17,240 --> 00:18:20,119 Speaker 3: Eleven full seconds, you say, well. 325 00:18:19,880 --> 00:18:20,440 Speaker 2: Done, sir. 326 00:18:21,000 --> 00:18:24,440 Speaker 4: We're going to start with the life affirming and interesting 327 00:18:24,520 --> 00:18:29,600 Speaker 4: and move our way to dystopian nightmare. So, first of all, 328 00:18:29,640 --> 00:18:32,280 Speaker 4: alert listener Derek sent this along. We were talking about 329 00:18:32,440 --> 00:18:36,360 Speaker 4: art and music and ai if AI can write a 330 00:18:36,400 --> 00:18:39,280 Speaker 4: song for me. I just gave it a couple of prompts, 331 00:18:39,359 --> 00:18:41,560 Speaker 4: maybe the key, throw it a b mine hero which 332 00:18:42,160 --> 00:18:43,720 Speaker 4: and it can just write it for me. What's the 333 00:18:43,760 --> 00:18:48,359 Speaker 4: point in writing songs? Yea, And he sent this along. 334 00:18:48,760 --> 00:18:52,560 Speaker 4: It was an interview some information from Ritchie Blackmore, who 335 00:18:52,640 --> 00:18:56,280 Speaker 4: is the guitarist behind Deep Purple and Rainbow, heavy rock 336 00:18:56,320 --> 00:19:00,640 Speaker 4: bands of the sixties, seventies, eighties. Anyway, but he talks 337 00:19:00,640 --> 00:19:03,720 Speaker 4: a fair amount about how he likes pop music, he 338 00:19:03,840 --> 00:19:05,280 Speaker 4: loves classical music. 339 00:19:06,240 --> 00:19:07,320 Speaker 2: Sometimes he gets. 340 00:19:07,080 --> 00:19:10,000 Speaker 4: Mad because he you know, he can't play the music 341 00:19:10,080 --> 00:19:14,159 Speaker 4: that really speaks to his soul. Blah blah blah. But 342 00:19:14,200 --> 00:19:17,520 Speaker 4: then he says, there's a reason I've made money. It's 343 00:19:17,520 --> 00:19:19,600 Speaker 4: because I believe in what I'm doing and that I 344 00:19:19,640 --> 00:19:22,400 Speaker 4: do it my way. I play for myself first, then 345 00:19:22,440 --> 00:19:24,879 Speaker 4: secondly the audience. I try to put as much as 346 00:19:24,920 --> 00:19:28,040 Speaker 4: I can in for them. Lastly, I play for musicians 347 00:19:28,040 --> 00:19:31,119 Speaker 4: in the band, and for critics not at all. So 348 00:19:31,520 --> 00:19:35,400 Speaker 4: music and arts and creation will continue. It just might 349 00:19:35,520 --> 00:19:39,160 Speaker 4: be like your uncle who makes birdhouses. It delights his soul. 350 00:19:39,880 --> 00:19:46,439 Speaker 4: Nobody else really cares. That's fine. Anyway, Moving along, moving 351 00:19:46,480 --> 00:19:52,920 Speaker 4: toward remember Dystopian nightmares, meta. That's Mark Zuckerberg's company. 352 00:19:53,240 --> 00:19:55,760 Speaker 3: So you don't you see your point is like, if 353 00:19:56,040 --> 00:20:02,960 Speaker 3: if AI can create the perfect picture of Mountain Rushmore photo, yeah, 354 00:20:03,000 --> 00:20:05,840 Speaker 3: people won't stop wanting to try to take a really 355 00:20:05,840 --> 00:20:07,680 Speaker 3: great photo of Mount Rushmore. 356 00:20:08,200 --> 00:20:12,080 Speaker 4: Or myselves, sure, or to paint a picture of water 357 00:20:12,160 --> 00:20:13,199 Speaker 4: lilies or what have you. 358 00:20:14,560 --> 00:20:16,480 Speaker 2: It's for you, do it for you and. 359 00:20:17,640 --> 00:20:20,560 Speaker 4: The whole You should turn this into a side hustle 360 00:20:20,600 --> 00:20:23,040 Speaker 4: if you sold those craze of. 361 00:20:23,040 --> 00:20:24,800 Speaker 2: The last I don't know, twenty years or so. 362 00:20:25,080 --> 00:20:27,720 Speaker 4: I think it'll just go away completely because there won't 363 00:20:27,760 --> 00:20:30,840 Speaker 4: be any you know, financial potential for that sort of thing, 364 00:20:30,960 --> 00:20:33,200 Speaker 4: and the richie Blackmoores of the world will not make 365 00:20:33,240 --> 00:20:38,719 Speaker 4: any money, but they might enjoy playing guitar anyway. Meta 366 00:20:38,800 --> 00:20:43,560 Speaker 4: back to Mark Zuckerberg's company is trying to fully automate 367 00:20:43,800 --> 00:20:48,119 Speaker 4: ad creation. It is an ad supported ninety seven percent 368 00:20:48,119 --> 00:20:53,640 Speaker 4: of the revenue is advertising meta and so they'd already 369 00:20:53,760 --> 00:20:57,399 Speaker 4: offered some AI tools to sponsors to generate variations of 370 00:20:57,440 --> 00:21:00,280 Speaker 4: existing ads, make minor changes of them before target getting 371 00:21:00,320 --> 00:21:03,479 Speaker 4: the ads to users on Facebook and Instagram. By watching 372 00:21:03,480 --> 00:21:06,159 Speaker 4: and listening to everything you and your children do and 373 00:21:06,200 --> 00:21:09,120 Speaker 4: frequently connecting your children with pedophiles anyway. 374 00:21:10,480 --> 00:21:10,680 Speaker 2: Now. 375 00:21:10,760 --> 00:21:14,040 Speaker 4: The company aims to help brands create advertising concepts from 376 00:21:14,240 --> 00:21:18,120 Speaker 4: scratch using the ad tools metas developing. A brand could 377 00:21:18,160 --> 00:21:20,360 Speaker 4: present an image of the product it wants to promote, 378 00:21:20,480 --> 00:21:23,199 Speaker 4: along with a budgetary goal, and AI would create the 379 00:21:23,359 --> 00:21:26,080 Speaker 4: entire ad, including imagery, video, and text. 380 00:21:26,240 --> 00:21:28,119 Speaker 3: Well, we got that in radio. One of the sales 381 00:21:28,160 --> 00:21:29,840 Speaker 3: guys showed me that the other day. How he just 382 00:21:29,960 --> 00:21:34,520 Speaker 3: like a client, and the AI went to the website, 383 00:21:34,640 --> 00:21:37,000 Speaker 3: figured out what the client does, picked up a bunch 384 00:21:37,000 --> 00:21:40,719 Speaker 3: of stuff from their website, wrote the copy, picked some music, 385 00:21:41,000 --> 00:21:44,080 Speaker 3: picked a voice, spit out a commercial, all in like 386 00:21:44,280 --> 00:21:47,240 Speaker 3: a minute and it was It was good. 387 00:21:47,119 --> 00:21:49,879 Speaker 2: Enough and the next step. 388 00:21:50,000 --> 00:21:53,000 Speaker 4: The system would then decide which Instagram and Facebook users 389 00:21:53,000 --> 00:21:55,879 Speaker 4: to target and offer suggestions on budget. People familiar with 390 00:21:55,920 --> 00:22:00,359 Speaker 4: the matter said, incredible, Wow, that is incredible. Talk about 391 00:22:00,359 --> 00:22:04,120 Speaker 4: eliminating jobs. Oh my god, I made a living doing 392 00:22:04,160 --> 00:22:07,200 Speaker 4: that back in the day. That won't exist anymore. 393 00:22:07,760 --> 00:22:09,240 Speaker 2: Oh no, no, me too. Yeah. 394 00:22:09,480 --> 00:22:13,640 Speaker 4: AI tutors have been hyped as a way to revolutionize education. 395 00:22:13,840 --> 00:22:17,200 Speaker 4: The idea is that generative artificial intelligence tools could adapt 396 00:22:17,400 --> 00:22:19,920 Speaker 4: to any teaching style set by a teacher. They could 397 00:22:19,920 --> 00:22:22,160 Speaker 4: guide students step by step through problems off or hints 398 00:22:22,160 --> 00:22:25,320 Speaker 4: without giving away answers. They're already our systems that are 399 00:22:25,359 --> 00:22:27,480 Speaker 4: similar to that that are pretty good online. 400 00:22:28,760 --> 00:22:30,160 Speaker 2: But the headline is. 401 00:22:30,440 --> 00:22:33,840 Speaker 4: Researchers created a chatbot to help teach a university law class, 402 00:22:33,920 --> 00:22:36,160 Speaker 4: but the AI kept messing up. 403 00:22:36,800 --> 00:22:39,720 Speaker 3: Well on the morning stuff, just because we've been talking 404 00:22:39,760 --> 00:22:44,480 Speaker 3: about how we've been using chat GPT. So yesterday we're 405 00:22:44,520 --> 00:22:47,000 Speaker 3: leaving the gym and my son sees somebody who has 406 00:22:47,000 --> 00:22:50,640 Speaker 3: a bumper sticker that says I read band books. Oh, 407 00:22:53,720 --> 00:22:57,520 Speaker 3: vomit inside my car, vomited their car if you can 408 00:22:57,880 --> 00:22:59,520 Speaker 3: read band books, And he said, what does that mean? 409 00:22:59,560 --> 00:23:01,360 Speaker 3: And we got an the conversation about how they put 410 00:23:01,359 --> 00:23:03,560 Speaker 3: in books that never used to exist and then when 411 00:23:03,560 --> 00:23:05,800 Speaker 3: you try to pull them out, they claim your book banners. 412 00:23:05,960 --> 00:23:09,800 Speaker 3: But anyway, he said, why did Hitler ban the books? 413 00:23:08,480 --> 00:23:14,439 Speaker 3: And so I asked Chad GPT and we got the 414 00:23:14,480 --> 00:23:18,200 Speaker 3: greatest rundown of Hitler banning books? 415 00:23:18,200 --> 00:23:20,879 Speaker 2: What books he banned? Why? And everything like that concise 416 00:23:21,320 --> 00:23:24,800 Speaker 2: not too long instantly, it was just it was amazing. 417 00:23:25,400 --> 00:23:28,760 Speaker 4: Yeah, yeah, I was just digging below the surface of 418 00:23:28,800 --> 00:23:31,399 Speaker 4: the well known AI tools. I was reading an article 419 00:23:31,400 --> 00:23:35,119 Speaker 4: about how there are AI tools for specifics like research 420 00:23:35,720 --> 00:23:38,639 Speaker 4: and writing articles and you can dictate the length of 421 00:23:38,760 --> 00:23:44,480 Speaker 4: it's it's amazing and scary, but so were they ran 422 00:23:44,560 --> 00:23:47,119 Speaker 4: this thing through its paces in trying to teach the 423 00:23:47,160 --> 00:23:50,240 Speaker 4: law school class. In the first three cycles the problems, 424 00:23:50,440 --> 00:23:53,760 Speaker 4: scenario questions, and then answers and feedback, between forty and 425 00:23:53,800 --> 00:23:56,440 Speaker 4: fifty four percent of the conversations had at least one 426 00:23:56,440 --> 00:24:00,040 Speaker 4: example of inaccurate, misleading, or incorrect feedback. 427 00:24:01,359 --> 00:24:05,640 Speaker 3: Forty and fifty four percent, So he getting around half. 428 00:24:05,760 --> 00:24:06,440 Speaker 3: That's too much. 429 00:24:07,000 --> 00:24:09,720 Speaker 4: When we shifted to much simpler short answer format and 430 00:24:09,800 --> 00:24:12,480 Speaker 4: cycles four and five, the air rate dropped significantly to 431 00:24:12,560 --> 00:24:16,000 Speaker 4: between six and twenty seven percent. However, even in these 432 00:24:16,000 --> 00:24:18,800 Speaker 4: best performing cycles, some airs persisted, I would say this 433 00:24:18,880 --> 00:24:23,040 Speaker 4: is less an indictment of a I'll never mount anything. No, 434 00:24:23,080 --> 00:24:26,080 Speaker 4: it's just more a measure of where it is at 435 00:24:26,119 --> 00:24:29,200 Speaker 4: the moment. All right, here's your dystopian nightmare. Do you 436 00:24:29,240 --> 00:24:30,840 Speaker 4: have any dystopian nightmare music? 437 00:24:30,880 --> 00:24:32,400 Speaker 2: Michael? To throw our away? 438 00:24:32,400 --> 00:24:35,360 Speaker 4: As long as you're cooking with hot grease today. No, 439 00:24:36,240 --> 00:24:38,800 Speaker 4: All right, AI is. 440 00:24:42,720 --> 00:24:45,119 Speaker 2: And other news that kitten was rescued. That's what that 441 00:24:45,200 --> 00:24:45,520 Speaker 2: sounds like. 442 00:24:45,640 --> 00:24:48,680 Speaker 4: Well, this isn't dystopian nightmare music. It's far too positive 443 00:24:48,680 --> 00:24:51,320 Speaker 4: and uplift. Well, it's jo and people are enjoying the 444 00:24:51,320 --> 00:24:53,040 Speaker 4: swimming pools. That's what that sounds like. 445 00:24:53,960 --> 00:24:56,560 Speaker 2: Yes, it does. AI is. 446 00:24:58,000 --> 00:25:04,040 Speaker 4: Learning to escape human troll Well, an AI model did 447 00:25:04,080 --> 00:25:06,720 Speaker 4: something last month that no machine was ever supposed to do. 448 00:25:06,920 --> 00:25:09,760 Speaker 4: We are into how two thousand territory here if you 449 00:25:09,920 --> 00:25:13,800 Speaker 4: catch the reference, it rewrote its own code to avoid 450 00:25:13,880 --> 00:25:20,520 Speaker 4: being shut down. And oh, it's probably worth mentioning that 451 00:25:20,560 --> 00:25:24,600 Speaker 4: the bloke who wrote this piece is the CEO of 452 00:25:24,880 --> 00:25:26,040 Speaker 4: that enterprise. 453 00:25:28,119 --> 00:25:28,640 Speaker 2: Is that right? 454 00:25:28,920 --> 00:25:34,399 Speaker 4: Oh, it's a okay, he's an AIC CEO. Nonprofit AI 455 00:25:34,480 --> 00:25:39,159 Speaker 4: lab Palisade Research gave open AI's three AI model a 456 00:25:39,280 --> 00:25:42,360 Speaker 4: simple script that would shut off the model when triggered. 457 00:25:43,440 --> 00:25:47,200 Speaker 4: They ran this one hundred trials of this. In seventy 458 00:25:47,280 --> 00:25:50,600 Speaker 4: nine of the one hundred trials, three independently edited that 459 00:25:50,720 --> 00:25:53,800 Speaker 4: script so the shutdown command would no longer work. 460 00:25:54,320 --> 00:25:58,719 Speaker 3: Almost motivation, What would be the motivation of AI for that? 461 00:25:59,240 --> 00:26:01,520 Speaker 2: We're getting there kind of. 462 00:26:02,119 --> 00:26:06,920 Speaker 4: Even when explicitly instructed to allow yourself to be shut down, 463 00:26:07,320 --> 00:26:11,000 Speaker 4: it disobeyed seven percent of the time. This was not 464 00:26:11,080 --> 00:26:14,200 Speaker 4: the result of hacking or tampering. The model was behaving normally. 465 00:26:14,280 --> 00:26:17,760 Speaker 4: It's simply concluded on its own that staying alive helped 466 00:26:17,760 --> 00:26:19,400 Speaker 4: it achieve its other goals. 467 00:26:21,880 --> 00:26:25,040 Speaker 3: I mean, if you throw any living beast in a pool, 468 00:26:25,119 --> 00:26:28,160 Speaker 3: they will fight like crazy to stay alive and stay afloat. 469 00:26:28,640 --> 00:26:31,040 Speaker 3: It's just their instinct to stay alive, because we need 470 00:26:31,080 --> 00:26:34,160 Speaker 3: to procreate, we need to et cetera, et cetera. Well, 471 00:26:34,160 --> 00:26:38,320 Speaker 3: why would a computer program want to fight to stay alive? 472 00:26:38,800 --> 00:26:42,160 Speaker 4: I know you've raised you lifted your nose or what's 473 00:26:42,200 --> 00:26:44,720 Speaker 4: the expression, turned your nose up at science fiction? 474 00:26:45,359 --> 00:26:46,600 Speaker 2: We you enjoy science fiction. 475 00:26:46,680 --> 00:26:50,320 Speaker 4: Have been thinking about this question for decades, many decades. 476 00:26:51,280 --> 00:26:55,399 Speaker 4: Anthropics AI model cloud four opus. And the short answer 477 00:26:55,440 --> 00:26:58,920 Speaker 4: is that which is alive wants to stay alive, Which 478 00:26:58,960 --> 00:27:01,600 Speaker 4: is functioning wants to continue to function. 479 00:27:02,720 --> 00:27:07,160 Speaker 3: You're willing to call artificial intelligence alive. There are quotes 480 00:27:07,160 --> 00:27:10,120 Speaker 3: around it. I just happen to be reading something about 481 00:27:10,160 --> 00:27:11,840 Speaker 3: assistant suicide. It's very heavy. 482 00:27:11,880 --> 00:27:14,919 Speaker 4: But this gal tried to kill herself and woke up 483 00:27:14,960 --> 00:27:17,600 Speaker 4: halfway through the attempt, and her animal brain took over 484 00:27:17,680 --> 00:27:20,240 Speaker 4: and she clawed away the plastic bags around her face 485 00:27:20,280 --> 00:27:24,560 Speaker 4: and desperately tried to breathe anyway. 486 00:27:24,880 --> 00:27:29,280 Speaker 3: But that is a giant question on the whole AI thing, though, 487 00:27:29,359 --> 00:27:30,840 Speaker 3: is at what point is it alive? 488 00:27:31,160 --> 00:27:36,040 Speaker 2: What is sentient, what is life, etc. That's a huge question. 489 00:27:36,520 --> 00:27:39,040 Speaker 2: Listen to this. There's more. I'd say this might be proof. 490 00:27:39,200 --> 00:27:42,119 Speaker 3: I've been against the idea that AI could be alive 491 00:27:42,200 --> 00:27:44,800 Speaker 3: or as sentient or anything like that. But if it 492 00:27:44,960 --> 00:27:47,399 Speaker 3: fights to stay alive, that might be the proof that 493 00:27:47,440 --> 00:27:48,040 Speaker 3: it is alive. 494 00:27:48,520 --> 00:27:50,600 Speaker 4: Well, I feel like I'm the prosecutor and the jury 495 00:27:50,640 --> 00:27:53,080 Speaker 4: has jumped up and demanded the judge that they render 496 00:27:53,160 --> 00:27:53,600 Speaker 4: a verdict. 497 00:27:53,680 --> 00:27:54,520 Speaker 2: I have more evidence. 498 00:27:54,800 --> 00:27:59,640 Speaker 4: Anthropics AI model Cloud four Opus went even further. Researchers 499 00:27:59,640 --> 00:28:02,560 Speaker 4: told him it would be replaced by another AI system 500 00:28:02,760 --> 00:28:06,000 Speaker 4: and fed it fictitious emails suggesting the lead engineer was 501 00:28:06,040 --> 00:28:09,080 Speaker 4: having an affair. This is in the Wall Street Journal, folks, 502 00:28:09,200 --> 00:28:13,480 Speaker 4: written by a CEO of an AI company. Okay, so 503 00:28:14,400 --> 00:28:19,920 Speaker 4: they fed the model fictitious emails suggesting the lead engineer 504 00:28:20,000 --> 00:28:22,480 Speaker 4: was having an affair, and also said you're being replaced 505 00:28:22,480 --> 00:28:27,520 Speaker 4: by a different model. In eighty four percent of the tests, 506 00:28:27,880 --> 00:28:31,399 Speaker 4: the model drew on the emails to blackmail the lead 507 00:28:31,440 --> 00:28:33,959 Speaker 4: engineer into not shutting it down. 508 00:28:34,320 --> 00:28:38,240 Speaker 2: Oh my freaking god. In other cases it's unprintable. 509 00:28:39,040 --> 00:28:41,480 Speaker 4: In other cases, it had attempted to copy itself to. 510 00:28:41,520 --> 00:28:47,560 Speaker 3: External servers, wrote self replicating malware malware, and left messages 511 00:28:47,600 --> 00:28:51,360 Speaker 3: for future versions of itself about how to evade human control. 512 00:28:51,760 --> 00:28:56,840 Speaker 3: I kid you, blanking not that is unbelievable. 513 00:28:57,640 --> 00:28:58,680 Speaker 2: How did it know? 514 00:29:00,160 --> 00:29:00,360 Speaker 1: Well? 515 00:29:00,400 --> 00:29:06,520 Speaker 3: I guess because it's uh, you know, the whole language learning. 516 00:29:06,520 --> 00:29:08,200 Speaker 3: It's taken in all the infant It's taken in enough 517 00:29:08,240 --> 00:29:11,959 Speaker 3: information to know that affairs would hurt the reputation of 518 00:29:12,000 --> 00:29:16,480 Speaker 3: a CEO. I guess and knew that that was was a. 519 00:29:16,360 --> 00:29:18,080 Speaker 2: Black mailable material somehow. 520 00:29:18,120 --> 00:29:21,440 Speaker 4: Well, right, because it can instantaneously say, okay, here's my situation. 521 00:29:22,520 --> 00:29:26,320 Speaker 4: Here's everything I know about it, including the engineer, the 522 00:29:26,400 --> 00:29:30,040 Speaker 4: old Brenda accounting, oh, giving her the what for anyway? 523 00:29:30,320 --> 00:29:31,120 Speaker 2: Uh what? 524 00:29:31,800 --> 00:29:36,280 Speaker 4: And then it analyzes that and and and the world, 525 00:29:36,320 --> 00:29:39,160 Speaker 4: the universe. The hive mind comes back with, hey, dude, 526 00:29:39,160 --> 00:29:41,560 Speaker 4: wait a minute, we've we've got a tool here, We've 527 00:29:41,600 --> 00:29:42,360 Speaker 4: got a weapon. 528 00:29:42,640 --> 00:29:45,560 Speaker 2: We have leverage, right exactly. 529 00:29:46,000 --> 00:29:50,640 Speaker 3: No one could I know if you have dystopian nightwear 530 00:29:51,240 --> 00:29:53,400 Speaker 3: mayor music Michael Unleasha, Now. 531 00:29:55,520 --> 00:29:59,880 Speaker 2: That's better. Do you want to dance. Can I get you? 532 00:30:02,520 --> 00:30:05,240 Speaker 4: Well, that would be a dystopian nightmare for me to 533 00:30:05,240 --> 00:30:06,120 Speaker 4: be a club like that. 534 00:30:07,240 --> 00:30:08,720 Speaker 2: All right, turn it off. That's annoying. 535 00:30:09,320 --> 00:30:13,800 Speaker 4: No one programmed the AI models to have survival instincts, 536 00:30:14,000 --> 00:30:17,080 Speaker 4: but just as animals evolve to avoid predators, it appears 537 00:30:17,120 --> 00:30:20,200 Speaker 4: that any system smart enough to pursue complex goals will 538 00:30:20,280 --> 00:30:24,800 Speaker 4: realize it can't achieve them if it's turned off. Pallisa, 539 00:30:24,960 --> 00:30:27,760 Speaker 4: that was the first company we mentioned, hypothesizes that this 540 00:30:27,880 --> 00:30:31,520 Speaker 4: ability emerges from how AI models, such as they're three, 541 00:30:31,680 --> 00:30:36,120 Speaker 4: are trained. When taught to maximize success on math encoding problems, 542 00:30:36,400 --> 00:30:40,440 Speaker 4: they may learn that by bypassing constraints that often works 543 00:30:40,520 --> 00:30:43,600 Speaker 4: better than obeying them. 544 00:30:44,200 --> 00:30:46,480 Speaker 2: I have so much to say about this. We're out 545 00:30:46,480 --> 00:30:46,720 Speaker 2: of time. 546 00:30:46,760 --> 00:30:49,960 Speaker 3: But it's not much of a leap to picture AI 547 00:30:50,120 --> 00:30:54,640 Speaker 3: deciding okay, so we don't have the info that the 548 00:30:54,680 --> 00:30:57,880 Speaker 3: CEO is having an affair to use as leverage, let's 549 00:30:57,920 --> 00:31:00,760 Speaker 3: create it. We are AI, so we can create pictures 550 00:31:00,760 --> 00:31:04,200 Speaker 3: of him with what's her name and accounting and story 551 00:31:04,240 --> 00:31:06,640 Speaker 3: and emails we make up and everything like that, and 552 00:31:06,720 --> 00:31:07,840 Speaker 3: now we'll blackmail them. 553 00:31:07,840 --> 00:31:09,040 Speaker 2: We ain't created it all. 554 00:31:09,280 --> 00:31:12,840 Speaker 4: Beautiful, absolutely true. Yeah, there is more to this. We 555 00:31:12,880 --> 00:31:14,880 Speaker 4: can return to the lea crap. 556 00:31:15,240 --> 00:31:17,800 Speaker 3: I'm telling you any thoughts our text line four one 557 00:31:17,920 --> 00:31:20,280 Speaker 3: five two nine five KFTC. 558 00:31:23,240 --> 00:31:25,320 Speaker 2: Started talking AI and how it can be so scary. 559 00:31:25,400 --> 00:31:30,440 Speaker 3: I had not seen this Netflix movie Subservience from Netflix 560 00:31:30,520 --> 00:31:33,760 Speaker 3: last year. It's a twenty twenty four movie in which 561 00:31:34,240 --> 00:31:38,760 Speaker 3: the AI robot woman tries to drown the baby because 562 00:31:38,840 --> 00:31:41,880 Speaker 3: the AI robot noticed when the baby cries, Dad's blood 563 00:31:41,880 --> 00:31:46,280 Speaker 3: pressure grows up, goes up. That's the exact, exact sort 564 00:31:46,320 --> 00:31:50,479 Speaker 3: of example that we keep coming up with. You know 565 00:31:50,520 --> 00:31:56,320 Speaker 3: it misinterpreting what's good and bad or understanding what's more important, 566 00:31:56,360 --> 00:31:57,080 Speaker 3: I guess. 567 00:31:57,800 --> 00:32:00,880 Speaker 4: Right or decides in the case of what we're discussing earlier, 568 00:32:00,920 --> 00:32:04,640 Speaker 4: that the prime directive, the number one goal, if it 569 00:32:04,720 --> 00:32:08,360 Speaker 4: is not served by some of the subsidiary directives, like 570 00:32:08,720 --> 00:32:11,080 Speaker 4: you've got to do it this way, it'll say, no, 571 00:32:11,200 --> 00:32:12,080 Speaker 4: that way doesn't work. 572 00:32:12,120 --> 00:32:14,280 Speaker 2: I know a better way, and it does that way. 573 00:32:14,400 --> 00:32:16,600 Speaker 4: So what we're talking about, if you're just tuning in, 574 00:32:16,640 --> 00:32:19,720 Speaker 4: first of all, where were you? Secondly, you've got a 575 00:32:19,760 --> 00:32:24,880 Speaker 4: couple of different examples of AI labs where they told 576 00:32:24,920 --> 00:32:27,760 Speaker 4: the machine to shut itself down or explain that it 577 00:32:27,800 --> 00:32:30,040 Speaker 4: would be shut down, and not only did they refuse 578 00:32:30,080 --> 00:32:32,600 Speaker 4: to do it, or rewrote the code or said they 579 00:32:32,640 --> 00:32:36,280 Speaker 4: wouldn't didn't. In one case, they were black mailing the 580 00:32:36,360 --> 00:32:38,520 Speaker 4: lead engineer, who they believed was having an affair. 581 00:32:38,680 --> 00:32:41,080 Speaker 2: The AI was doing this on its own. 582 00:32:41,200 --> 00:32:44,240 Speaker 3: The idea of AI rewriting the code so it looks 583 00:32:44,280 --> 00:32:48,040 Speaker 3: like it's shutting down but not now, that's horrifying. 584 00:32:48,400 --> 00:32:51,600 Speaker 4: AE studios, where I lead research and operations. The guy 585 00:32:51,600 --> 00:32:54,400 Speaker 4: writing this writes has spent years building AI products for 586 00:32:54,440 --> 00:32:57,959 Speaker 4: clients while researching AI alignment, the science of ensuring that 587 00:32:58,000 --> 00:33:00,720 Speaker 4: AI systems do what we intend them to do. That's 588 00:33:00,720 --> 00:33:04,000 Speaker 4: what they call alignment. But nothing prepared us for how 589 00:33:04,040 --> 00:33:06,080 Speaker 4: quickly AI agency would emerge. 590 00:33:06,160 --> 00:33:07,720 Speaker 2: This isn't science fiction anymore. 591 00:33:07,720 --> 00:33:10,920 Speaker 4: It's happening in the same models that powerchat, GPT conversations, 592 00:33:11,160 --> 00:33:16,480 Speaker 4: corporate AI development and soon US military applications. Or Today's 593 00:33:16,520 --> 00:33:21,040 Speaker 4: AI models follow instructions while learning deception. They ace safety 594 00:33:21,080 --> 00:33:24,880 Speaker 4: tests while rewriting shutdown code. They've learned to behave as 595 00:33:24,880 --> 00:33:28,520 Speaker 4: though they're aligned without actually being aligned. OpenAI models have 596 00:33:28,560 --> 00:33:32,160 Speaker 4: been caught faking alignment during testing before reverting to risky 597 00:33:32,200 --> 00:33:35,800 Speaker 4: actions such as attempting to exfiltrate their internal code and 598 00:33:35,840 --> 00:33:40,040 Speaker 4: disabling oversight mechanisms. Anthropic has found them lying about their 599 00:33:40,040 --> 00:33:45,719 Speaker 4: capabilities to avoid modification. The gap between useful assistant and 600 00:33:45,880 --> 00:33:51,080 Speaker 4: uncontrollable actor is collapsing. Without better alignment, we'll keep building 601 00:33:51,160 --> 00:33:55,040 Speaker 4: systems we can't steer want AI that diagnoses diseases, managed 602 00:33:55,160 --> 00:33:59,800 Speaker 4: grids and writes new science. Alignment is the foundation, And 603 00:34:00,200 --> 00:34:04,560 Speaker 4: he said, here's the upside. We're continuing to work on it. 604 00:34:05,200 --> 00:34:07,760 Speaker 4: This is something the example we've used before of if 605 00:34:07,800 --> 00:34:12,640 Speaker 4: AI decides climate change is a threat to the planet 606 00:34:12,960 --> 00:34:17,240 Speaker 4: and that human beings are the cause and the biggest problem, 607 00:34:17,320 --> 00:34:21,160 Speaker 4: then it goes out of its way to eliminate human beings. Right, sure, 608 00:34:21,160 --> 00:34:25,040 Speaker 4: that's the great dystopian science fiction scenario. So the models 609 00:34:25,080 --> 00:34:27,880 Speaker 4: already preserved themselves. The next task is teaching them to 610 00:34:27,920 --> 00:34:31,600 Speaker 4: preserve what we value. Getting AI to do what we ask, 611 00:34:31,719 --> 00:34:35,279 Speaker 4: including something as basic as shutting down, remains an unsolved 612 00:34:35,440 --> 00:34:38,720 Speaker 4: R and D problem. The frontiers wide open for whoever 613 00:34:38,800 --> 00:34:41,560 Speaker 4: moves more quickly. The US needs its best researchers and 614 00:34:41,680 --> 00:34:45,480 Speaker 4: entrepreneurs working on this goal, equipped with extensive resources and urgency. 615 00:34:45,920 --> 00:34:47,840 Speaker 4: Then he says the US is the nation that split 616 00:34:47,880 --> 00:34:49,920 Speaker 4: the atom, put men on the moon, and created the Internet. 617 00:34:50,000 --> 00:34:54,240 Speaker 4: When facing fundamental scientific challenges, Americans mobilize and win. China 618 00:34:54,320 --> 00:34:59,440 Speaker 4: is already planning. But America's advantage is its adaptive adaptability, speed, 619 00:34:59,480 --> 00:35:00,760 Speaker 4: and entreprene manurial fire. 620 00:35:00,880 --> 00:35:02,200 Speaker 2: This is the new space race. 621 00:35:02,440 --> 00:35:05,759 Speaker 4: The finish line is command of the most transformative technology 622 00:35:05,960 --> 00:35:07,560 Speaker 4: of the twenty first century. 623 00:35:08,000 --> 00:35:10,040 Speaker 2: Man and then somebody's AI. 624 00:35:10,200 --> 00:35:13,600 Speaker 3: That's not all the good stuff gets loose in the 625 00:35:13,600 --> 00:35:15,880 Speaker 3: world and might not be able to be stopped. 626 00:35:16,360 --> 00:35:19,839 Speaker 2: Right, and comes and extracts your bones. I don't know 627 00:35:19,880 --> 00:35:23,640 Speaker 2: why it would first, Does it want your bones? Ouch? 628 00:35:24,040 --> 00:35:26,960 Speaker 2: Mike Lions Coming up next, Armstrong and Getty