1 00:00:07,520 --> 00:00:10,559 Speaker 1: Twin Cities News Talk Am eleven thirty one oh three 2 00:00:10,680 --> 00:00:13,240 Speaker 1: five f M, streaming. 3 00:00:12,840 --> 00:00:14,800 Speaker 2: Worldwide on the iHeartRadio app. 4 00:00:14,840 --> 00:00:17,919 Speaker 1: My name is John Justice, and I'm glad that you 5 00:00:18,040 --> 00:00:21,880 Speaker 1: were with the show this morning. The Nobel Peace Prize 6 00:00:21,920 --> 00:00:26,200 Speaker 1: nominated talk show host. Unfortunately he lives in Saint Paul, 7 00:00:26,320 --> 00:00:29,120 Speaker 1: so he cannot claim that title. But Sam is in 8 00:00:29,160 --> 00:00:30,600 Speaker 1: the master control booth this morning. 9 00:00:30,640 --> 00:00:33,919 Speaker 2: Good morning, John. One of these days we'll figure out 10 00:00:33,920 --> 00:00:34,760 Speaker 2: a way to make it happen. 11 00:00:35,760 --> 00:00:37,839 Speaker 1: I'll make you all if we end up, if the 12 00:00:38,080 --> 00:00:41,920 Speaker 1: citizens of Minneapolis do end up winning the Nobel Peace Prize, 13 00:00:41,960 --> 00:00:45,800 Speaker 1: I'll make you an honorary recipient. Well, I appreciate that somebody, 14 00:00:45,880 --> 00:00:48,120 Speaker 1: and I'm sure somebody would be welcome to give me 15 00:00:48,240 --> 00:00:49,440 Speaker 1: theirs as well if they choose to. 16 00:00:50,040 --> 00:00:53,440 Speaker 2: David Gartenstein Ross, CEO. 17 00:00:53,120 --> 00:00:56,000 Speaker 1: Of Expert Theory and AI Analyst, will join us coming 18 00:00:56,040 --> 00:00:57,960 Speaker 1: up at six thirty this morning. As always, if you 19 00:00:58,040 --> 00:01:01,600 Speaker 1: have any AI related Question four DVD you can get 20 00:01:01,600 --> 00:01:06,320 Speaker 1: those in early. You're listening on the iHeartRadio app. Leave 21 00:01:06,400 --> 00:01:09,560 Speaker 1: us a talk back. At eight o'clock this morning, we'll 22 00:01:09,560 --> 00:01:13,640 Speaker 1: talk with a doctor Scott Jensen, who is dropping out 23 00:01:13,680 --> 00:01:16,200 Speaker 1: of the governor's race and running for a different office, 24 00:01:16,200 --> 00:01:18,280 Speaker 1: and so we'll dive into that with him. A lot 25 00:01:18,319 --> 00:01:22,119 Speaker 1: of fraud discussion this morning in the seven o'clock hour, 26 00:01:22,520 --> 00:01:27,959 Speaker 1: a lot of new revelations relating to fraud, some more convictions, 27 00:01:28,800 --> 00:01:32,319 Speaker 1: cover ups, a lot of ground to go and cover there. 28 00:01:32,360 --> 00:01:35,720 Speaker 1: We'll also, of course get into the latest regarding Operation 29 00:01:36,319 --> 00:01:39,720 Speaker 1: Metro Surge. We have some voter roll controversies as well. 30 00:01:39,800 --> 00:01:42,440 Speaker 1: Got a jam Pat show today. Let's start here though 31 00:01:44,160 --> 00:01:46,400 Speaker 1: probably the final time this week we'll talk about it. 32 00:01:47,000 --> 00:01:50,680 Speaker 1: I think today the controversies around the Super Bowl halftime 33 00:01:50,720 --> 00:01:52,880 Speaker 1: show will probably subside. 34 00:01:52,920 --> 00:01:55,200 Speaker 2: However, the preliminary ratings. 35 00:01:54,840 --> 00:02:00,920 Speaker 1: Numbers show viewership for Bad Bunnies A halftime performance tanked 36 00:02:02,160 --> 00:02:04,600 Speaker 1: down thirty nine percent from last year. 37 00:02:05,080 --> 00:02:05,680 Speaker 2: Took me a while. 38 00:02:05,640 --> 00:02:07,240 Speaker 1: I had to have somebody tell me I forgot who 39 00:02:07,240 --> 00:02:10,600 Speaker 1: performed last year? Was Kendrick lamar I who performed last year? 40 00:02:11,080 --> 00:02:14,720 Speaker 1: Only twenty six point five million households watched. What's interesting, though, 41 00:02:15,520 --> 00:02:19,359 Speaker 1: is forty eight point six million tuned into the game. 42 00:02:19,639 --> 00:02:25,239 Speaker 1: Nearly half of Americans turned off the TV at halftime, 43 00:02:25,240 --> 00:02:27,440 Speaker 1: and of course we know the turning point. USA just 44 00:02:27,480 --> 00:02:33,040 Speaker 1: turned in gangbuster numbers for a first time counter programming 45 00:02:33,080 --> 00:02:37,080 Speaker 1: to the super Bowl halftime show. You know it's funny, 46 00:02:37,080 --> 00:02:41,520 Speaker 1: because funny in a sad kind of way. There is 47 00:02:41,720 --> 00:02:49,079 Speaker 1: absolutely no conversation whatsoever about the game, like none. As 48 00:02:49,080 --> 00:02:50,840 Speaker 1: a matter of fact, yesterday morning, I don't think I 49 00:02:50,919 --> 00:02:52,519 Speaker 1: mentioned this, but I did mention that I went to 50 00:02:52,560 --> 00:02:54,639 Speaker 1: bed early. I didn't finish watching the game. I assumed 51 00:02:54,639 --> 00:02:56,639 Speaker 1: the Seahawks were going to win, but when I got 52 00:02:56,720 --> 00:02:58,800 Speaker 1: up in the morning, like it took me a couple 53 00:03:00,040 --> 00:03:03,160 Speaker 1: with me a couple of different websites before I could 54 00:03:03,200 --> 00:03:06,800 Speaker 1: actually find out who who won the game. But nobody 55 00:03:06,880 --> 00:03:11,040 Speaker 1: is talking about the game whatsoever, little little bit of 56 00:03:11,080 --> 00:03:13,480 Speaker 1: conversation around the commercials. And we'll talk about the AI 57 00:03:13,560 --> 00:03:16,880 Speaker 1: aspect of that with David Gartenstein Ross when he joins 58 00:03:16,960 --> 00:03:19,160 Speaker 1: us at six point thirty this morning. 59 00:03:19,160 --> 00:03:21,240 Speaker 2: And if you were listening yesterday, you know I gave. 60 00:03:21,040 --> 00:03:26,480 Speaker 1: A rather tempered response to both Turning Point USA's halftime 61 00:03:26,560 --> 00:03:31,920 Speaker 1: performance or events and the halftime show with Bad Bunny, 62 00:03:32,000 --> 00:03:35,440 Speaker 1: and I still hold to that. I'm really happy that 63 00:03:35,520 --> 00:03:38,360 Speaker 1: Turning Point USA got the numbers that they that they 64 00:03:38,400 --> 00:03:42,600 Speaker 1: got in the context of a super Bowl halftime show. 65 00:03:42,960 --> 00:03:44,600 Speaker 1: I just felt like it you know, it's a little flat. 66 00:03:44,640 --> 00:03:47,520 Speaker 1: But again, I'm unfamiliar with those artists, and I had 67 00:03:47,560 --> 00:03:50,240 Speaker 1: somewhat of a tempered response to Bad Bunny. I didn't 68 00:03:50,240 --> 00:03:53,360 Speaker 1: watch it. I only saw the commentary later. I'm changing 69 00:03:53,400 --> 00:03:58,320 Speaker 1: my tune on this. This's the point because yesterday the 70 00:03:58,360 --> 00:04:05,720 Speaker 1: story grew even further than just the alternate Super Bowl 71 00:04:05,800 --> 00:04:09,320 Speaker 1: halftime show by turning point USA because the lyrics to 72 00:04:09,400 --> 00:04:13,320 Speaker 1: bad Bunny's songs that he sang were posted online. 73 00:04:14,360 --> 00:04:17,480 Speaker 2: I mean, this was just pure filth. 74 00:04:18,480 --> 00:04:21,240 Speaker 1: Never if this was sung in English, never would any 75 00:04:21,320 --> 00:04:23,800 Speaker 1: of this, not almost not a single line from any 76 00:04:23,839 --> 00:04:27,599 Speaker 1: of the songs that bad Bunny performed would have passed 77 00:04:27,839 --> 00:04:32,159 Speaker 1: FCC regulations. I mean, it was like two live crew 78 00:04:32,600 --> 00:04:35,200 Speaker 1: levels of filth. For those of you of a certain 79 00:04:35,600 --> 00:04:38,680 Speaker 1: of a certain age, I just assume all of his 80 00:04:38,720 --> 00:04:43,160 Speaker 1: songs are like that. How hard is it, though, to understand, 81 00:04:43,200 --> 00:04:45,440 Speaker 1: this is your audience that you're playing to and you're 82 00:04:45,440 --> 00:04:48,560 Speaker 1: trying to reach new individuals and just clean it up 83 00:04:48,560 --> 00:04:51,960 Speaker 1: for the sake of one performance. But again, because he 84 00:04:52,040 --> 00:04:58,440 Speaker 1: was speaking in another language, nobody cared. And that was 85 00:04:58,480 --> 00:05:01,520 Speaker 1: the majority of the commentary that I saw in the 86 00:05:01,600 --> 00:05:07,280 Speaker 1: immediate aftermath. Or you know, right after the halftime show concluded, 87 00:05:07,320 --> 00:05:09,400 Speaker 1: I saw a lot of commentary online from people that 88 00:05:09,440 --> 00:05:10,480 Speaker 1: I respect going well. 89 00:05:10,520 --> 00:05:11,480 Speaker 2: I don't know what he said, but. 90 00:05:11,440 --> 00:05:13,840 Speaker 1: I thought the performance was really good. It wasn't good 91 00:05:13,839 --> 00:05:15,280 Speaker 1: if you were in the stadium. By the way, if 92 00:05:15,279 --> 00:05:18,120 Speaker 1: you've seen the any of the videos online of the 93 00:05:18,240 --> 00:05:21,520 Speaker 1: view that the audience actually had at the stadium, you 94 00:05:21,560 --> 00:05:23,440 Speaker 1: couldn't even see what was going on on the field. 95 00:05:24,279 --> 00:05:26,719 Speaker 1: And now with the exposure of these lyrics, I mean 96 00:05:27,040 --> 00:05:30,640 Speaker 1: just absolute filth. So yeah, he never should have been 97 00:05:30,760 --> 00:05:33,680 Speaker 1: chosen to do the Super Bowl halftime show. 98 00:05:35,440 --> 00:05:36,680 Speaker 2: One more note on this, by. 99 00:05:36,520 --> 00:05:42,800 Speaker 1: The way, he cleared bad Bunny did. He cleared some 100 00:05:42,839 --> 00:05:45,880 Speaker 1: of his social media presence after performing in the halftime show. 101 00:05:46,080 --> 00:05:48,480 Speaker 2: The Hill reported that following. 102 00:05:50,040 --> 00:05:53,400 Speaker 1: The event, the Puerto Rican rapper and singer deleted all 103 00:05:53,480 --> 00:05:58,719 Speaker 1: Instagram posts, removed his profile photo, and unfollowed all users 104 00:05:58,760 --> 00:06:01,440 Speaker 1: on the platform. His account has more than fifty two 105 00:06:01,440 --> 00:06:05,599 Speaker 1: million followers. He also unfollowed accounts on x removed his 106 00:06:05,680 --> 00:06:10,760 Speaker 1: profile photo as well, where he has about five million followers. 107 00:06:11,279 --> 00:06:15,000 Speaker 1: No explanation was given for the changes. Now I'm reading 108 00:06:15,040 --> 00:06:17,839 Speaker 1: off of a version of the story from Newsmax, but 109 00:06:17,960 --> 00:06:20,960 Speaker 1: they off you know they went in. We're basing it 110 00:06:21,000 --> 00:06:24,039 Speaker 1: off of the hill reporting. Apparently nobody could do a 111 00:06:24,080 --> 00:06:27,360 Speaker 1: little bit more research is to find out why, because 112 00:06:27,400 --> 00:06:32,280 Speaker 1: I did because I was curious because this was being 113 00:06:32,279 --> 00:06:34,280 Speaker 1: lobbed out there, just like why did he go in 114 00:06:34,320 --> 00:06:36,760 Speaker 1: to lead all of his social media? They have him 115 00:06:36,800 --> 00:06:40,359 Speaker 1: running scared the controversy. And again I say this not 116 00:06:40,400 --> 00:06:44,360 Speaker 1: as a bad Bunny fan by any stretch. I think 117 00:06:44,360 --> 00:06:46,839 Speaker 1: it's a stretch to call him an actual artist that 118 00:06:46,960 --> 00:06:51,159 Speaker 1: has any sort of talent. If the talent is somewhat 119 00:06:51,480 --> 00:06:55,320 Speaker 1: rapping to incredibly filthy lyrics, then I guess you can 120 00:06:55,360 --> 00:06:56,600 Speaker 1: call him an artist. 121 00:06:56,839 --> 00:06:56,919 Speaker 3: Now. 122 00:06:57,080 --> 00:06:59,600 Speaker 2: He's done this in the past. 123 00:07:00,279 --> 00:07:05,719 Speaker 1: This is a promotional strategy that this particular individual deploys 124 00:07:05,800 --> 00:07:09,600 Speaker 1: quite often. It's not new behavior for him. In early 125 00:07:09,640 --> 00:07:12,440 Speaker 1: of twenty twenty two, he wiped his entire Instagram feed 126 00:07:12,480 --> 00:07:15,360 Speaker 1: on New Year's Day, delete it all posts, everything going blank, 127 00:07:15,360 --> 00:07:19,800 Speaker 1: and then joined TikTok and posting there casually aligned with 128 00:07:19,880 --> 00:07:23,720 Speaker 1: starting fresh for the year, and proceeded with promotional activity 129 00:07:24,040 --> 00:07:27,280 Speaker 1: around his music era at the time. In twenty twenty 130 00:07:27,280 --> 00:07:29,760 Speaker 1: he did the same thing. He posted a simple biome 131 00:07:29,800 --> 00:07:32,840 Speaker 1: gone message when in active for about three months, creating 132 00:07:32,880 --> 00:07:37,520 Speaker 1: anticipation before returning to new content. Fans widely interpreted what 133 00:07:37,600 --> 00:07:41,640 Speaker 1: happened after the game yesterday as his teasing and upcoming music, 134 00:07:41,720 --> 00:07:45,000 Speaker 1: or a shift in image and an era. It's just 135 00:07:45,040 --> 00:07:48,680 Speaker 1: sad that all these news media outlets just shotgunned the 136 00:07:48,720 --> 00:07:50,600 Speaker 1: story out there, but didn't bother to go and look 137 00:07:50,640 --> 00:07:53,600 Speaker 1: as to why, just leaving it open for rampant speculation, 138 00:07:53,720 --> 00:07:58,360 Speaker 1: which doesn't do anybody any favors whatsoever. So there you go, 139 00:07:59,160 --> 00:08:02,560 Speaker 1: all right, We're going to move on to more relevant items, 140 00:08:02,560 --> 00:08:07,800 Speaker 1: including an incredibly horrible tragedy a Minnesota gubernatorial candidate having 141 00:08:07,800 --> 00:08:11,360 Speaker 1: to suspend his campaign after his daughter was found dead. 142 00:08:11,400 --> 00:08:14,600 Speaker 2: We'll give you details on that, and as we continue 143 00:08:14,640 --> 00:08:15,800 Speaker 2: to discuss. 144 00:08:15,400 --> 00:08:18,520 Speaker 1: What might be the final phase of Trump's arrangement syndrome 145 00:08:18,960 --> 00:08:21,080 Speaker 1: and the way people have been acting as of late, 146 00:08:21,120 --> 00:08:25,600 Speaker 1: I have yet another example of a CNN regular who 147 00:08:26,000 --> 00:08:30,360 Speaker 1: really went to the extremes when talking about MAGA and 148 00:08:30,400 --> 00:08:33,920 Speaker 1: what he believes needs to be done with them. That 149 00:08:34,000 --> 00:08:36,360 Speaker 1: audio is coming up ahead of my conversation with AI 150 00:08:36,440 --> 00:08:39,240 Speaker 1: analysts and expert de Vive Gartenstein Ross. As always, you 151 00:08:39,280 --> 00:08:43,120 Speaker 1: can email me Justice at iHeartRadio dot com or if 152 00:08:43,160 --> 00:08:45,319 Speaker 1: you're listening on the iHeartRadio app, we'll get to those 153 00:08:45,360 --> 00:08:48,160 Speaker 1: talkbacks coming up. Brought to you by Lyndahl Realty here 154 00:08:48,320 --> 00:08:50,880 Speaker 1: on Twinsday's News Talk AM eleven thirty and one oh 155 00:08:50,920 --> 00:08:52,040 Speaker 1: three five FM. 156 00:08:52,200 --> 00:08:54,840 Speaker 4: Sixty five to one, Carpets plus Home of the Next Day, 157 00:08:54,960 --> 00:08:59,480 Speaker 4: Installs Studios IFAM one oh three point five and KTLKAM 158 00:09:00,559 --> 00:09:02,440 Speaker 4: Minneapolis at Saint Paul. 159 00:09:08,480 --> 00:09:10,720 Speaker 3: A really good point. If this would have been in English, 160 00:09:10,720 --> 00:09:13,680 Speaker 3: it wouldn't have passed FNAD roles. They wouldn't have been 161 00:09:13,720 --> 00:09:19,520 Speaker 3: allowed to speak those words, sing those worst So just 162 00:09:19,600 --> 00:09:24,040 Speaker 3: because it was been Spanish, what makes that any different church, 163 00:09:25,280 --> 00:09:28,559 Speaker 3: they shouldn't have been allowed to speak or. 164 00:09:28,559 --> 00:09:29,880 Speaker 2: Sing that film. 165 00:09:30,040 --> 00:09:33,880 Speaker 1: Yeah, referencing what was going around yesterday. As a matter 166 00:09:33,920 --> 00:09:36,760 Speaker 1: of fact, I went on speaking of news Max. I 167 00:09:36,760 --> 00:09:40,520 Speaker 1: went on news Max yesterday afternoon on Todd Starns Show 168 00:09:40,559 --> 00:09:43,720 Speaker 1: talking about this very issue because the lyrics to the 169 00:09:43,840 --> 00:09:48,160 Speaker 1: songs that Bad Bunny sang during the Super Bowl halftime 170 00:09:48,200 --> 00:09:52,680 Speaker 1: performance had been posted just vile, philth I mean, it 171 00:09:52,760 --> 00:09:55,680 Speaker 1: was just lyrical pornography. 172 00:09:56,440 --> 00:10:02,360 Speaker 5: Being Spanish or a different language, the FCC rules don't 173 00:10:02,600 --> 00:10:03,800 Speaker 5: apply to you anymore. 174 00:10:04,480 --> 00:10:09,040 Speaker 6: What about all the kids and stuff that speak Spanish 175 00:10:09,080 --> 00:10:11,640 Speaker 6: that we're watching the game and heard all that. 176 00:10:13,440 --> 00:10:18,359 Speaker 1: So I did some research and apparently this is incredibly subjective. 177 00:10:19,600 --> 00:10:25,160 Speaker 1: So cursing in Spanish is prohibited on broadcast TV and 178 00:10:25,280 --> 00:10:30,000 Speaker 1: radio between six am and ten pm by the FCC. 179 00:10:30,280 --> 00:10:32,640 Speaker 1: I don't think we were out of that window yet. Well, certainly, 180 00:10:33,000 --> 00:10:34,360 Speaker 1: if we were, it would have been East Coast, but 181 00:10:34,400 --> 00:10:35,520 Speaker 1: I don't even think it. 182 00:10:35,440 --> 00:10:38,760 Speaker 2: Was that late at that point in time. But it's 183 00:10:38,840 --> 00:10:41,120 Speaker 2: rarely enforced. 184 00:10:41,720 --> 00:10:44,199 Speaker 1: As of a few years ago, the FCC hadn't done 185 00:10:44,240 --> 00:10:51,400 Speaker 1: anything about unambiguous profanity within US made Spanish language shows. 186 00:10:51,480 --> 00:10:55,600 Speaker 1: Other television shows have dropped to curse words and profanity 187 00:10:55,640 --> 00:10:58,520 Speaker 1: and other languages, and the FCC never does anything about it. 188 00:10:58,960 --> 00:11:03,600 Speaker 1: The way the game is typically played regarding this is 189 00:11:03,600 --> 00:11:09,840 Speaker 1: that the FCC doesn't go and proactively seek out and 190 00:11:09,960 --> 00:11:16,320 Speaker 1: punish potential violations. So if if I were to go 191 00:11:16,400 --> 00:11:20,240 Speaker 1: and you drop an F bomb on the show, you 192 00:11:20,280 --> 00:11:25,160 Speaker 1: wouldn't have the sec by itself. Necessarily, they probably could, 193 00:11:25,200 --> 00:11:27,200 Speaker 1: but it's not going to happen. You wouldn't see them 194 00:11:27,240 --> 00:11:32,920 Speaker 1: coming in to find the radio station on their own. 195 00:11:33,000 --> 00:11:35,679 Speaker 1: It would take somebody to go and lodge a complaint 196 00:11:36,200 --> 00:11:39,720 Speaker 1: in order to bring about any sort of investigation which 197 00:11:39,760 --> 00:11:43,360 Speaker 1: may result in fines. And I've seen some individuals online 198 00:11:43,360 --> 00:11:48,240 Speaker 1: that have now been suggesting that that individuals go and 199 00:11:48,280 --> 00:11:51,200 Speaker 1: do just that, that they go and put forward to 200 00:11:51,280 --> 00:11:55,440 Speaker 1: FCC complaints based off of Bad Bunny's performance. Listen, it 201 00:11:55,480 --> 00:11:58,720 Speaker 1: might be noteworthy in order to sort of reset the 202 00:11:58,800 --> 00:12:02,960 Speaker 1: standard if people are concerned about that. But again, I 203 00:12:03,040 --> 00:12:05,800 Speaker 1: go back to the commentary around the performance, and I 204 00:12:05,880 --> 00:12:08,240 Speaker 1: just find it funny that everybody's like, I thought I 205 00:12:08,320 --> 00:12:11,240 Speaker 1: was really good. I was entertained, even tho I could 206 00:12:11,320 --> 00:12:12,800 Speaker 1: understand what he was saying. It's like, would you just 207 00:12:12,840 --> 00:12:14,520 Speaker 1: still felt that way if you heard all that stuff 208 00:12:14,520 --> 00:12:17,800 Speaker 1: in English? Although having not watched it live and now 209 00:12:17,840 --> 00:12:20,920 Speaker 1: gone back and seeing a large I pretty much watched 210 00:12:20,920 --> 00:12:23,720 Speaker 1: the whole thing now, just through the process of seeing 211 00:12:23,800 --> 00:12:26,600 Speaker 1: various clips. And yeah, it wasn't just the lyrics in 212 00:12:26,640 --> 00:12:30,000 Speaker 1: Spanish that were offensive. Holy cow, it's a there's a 213 00:12:30,040 --> 00:12:34,400 Speaker 1: lot of gy reading going on. I'm surprised the appearance 214 00:12:34,440 --> 00:12:37,760 Speaker 1: of certain rubber items being used by ICE protesters didn't 215 00:12:38,280 --> 00:12:41,719 Speaker 1: show up. Ah seeyeah. Look, I broke my own rule. 216 00:12:41,720 --> 00:12:43,160 Speaker 1: I told myself I wasn't going to talk about that 217 00:12:43,240 --> 00:12:45,280 Speaker 1: today on the show because we spent so much time 218 00:12:45,320 --> 00:12:48,800 Speaker 1: on it yesterday. But oh well, all right, let's go here. 219 00:12:49,120 --> 00:12:52,880 Speaker 1: Tragic story. Hailey Marie Tobler, twenty two, found dead in 220 00:12:52,880 --> 00:12:56,000 Speaker 1: the Saint Cloud apartment on Saturday, died from a stabbing, 221 00:12:56,760 --> 00:12:59,360 Speaker 1: identified Monday as the daughter of a man running for 222 00:12:59,559 --> 00:13:04,439 Speaker 1: Minnesota governor. Saint Cloud Police reported that the officers found 223 00:13:04,480 --> 00:13:07,640 Speaker 1: a Hayley dead in their apartment Saturday night. The medical 224 00:13:07,679 --> 00:13:15,880 Speaker 1: examiner again reported she died from multiple stab wounds. Tobler's 225 00:13:16,240 --> 00:13:20,600 Speaker 1: husband was found in the apartment with injuries. Saint Cloud 226 00:13:20,679 --> 00:13:23,839 Speaker 1: Police has stated his injuries were self inflicton and he 227 00:13:23,880 --> 00:13:26,920 Speaker 1: would be taking the jail for suspected homicide after his 228 00:13:27,040 --> 00:13:30,160 Speaker 1: injuries were treated. In a Facebook post on Monday, the 229 00:13:30,200 --> 00:13:33,640 Speaker 1: Republican Party of Minnesota shared that Tobler was the daughter 230 00:13:33,679 --> 00:13:36,480 Speaker 1: of doctor Jeff Johnson, who was running for Minnesota governor, 231 00:13:36,679 --> 00:13:39,760 Speaker 1: who suspended his campaign after Tobler died. It should be 232 00:13:39,800 --> 00:13:42,720 Speaker 1: noted that doctor Jeff Johnson is a former Saint Cloud 233 00:13:42,800 --> 00:13:45,880 Speaker 1: a City council member, not the same Jeff Johnson as 234 00:13:45,880 --> 00:13:49,000 Speaker 1: the former Hennipon County commissioner who twice ran for governor 235 00:13:49,200 --> 00:13:53,120 Speaker 1: in twenty fourteen and twenty eighteen. So prayers go out 236 00:13:53,520 --> 00:13:58,040 Speaker 1: to the family in this horrible tragedy and to the 237 00:13:58,040 --> 00:14:04,439 Speaker 1: former gubernatorial candidate, doctor Jeff Johnson. Speaking of gubernatorial candidates, 238 00:14:04,880 --> 00:14:08,360 Speaker 1: another individual yesterday dropped out from their run for governor, 239 00:14:08,520 --> 00:14:11,440 Speaker 1: Doctor Scott Jensen made the announcement that he was now 240 00:14:11,520 --> 00:14:14,040 Speaker 1: running for state auditor, and he will be joining us 241 00:14:14,080 --> 00:14:16,840 Speaker 1: this morning to talk about that decision at eight o'clock 242 00:14:16,960 --> 00:14:21,320 Speaker 1: today before we talk with Devid Garton Stein Rosson. If 243 00:14:21,320 --> 00:14:23,920 Speaker 1: you have any AI related questions, you can get those 244 00:14:23,960 --> 00:14:29,320 Speaker 1: in now to the iHeartRadio app. Regular CNN analyst Bakari 245 00:14:29,440 --> 00:14:34,280 Speaker 1: Sellers veered into the Nazis territory inflammatory rhetoric when he 246 00:14:34,320 --> 00:14:38,560 Speaker 1: referred to President Donald Trump as a cancer, suggesting the 247 00:14:38,600 --> 00:14:43,000 Speaker 1: President's voters were vermin or even demons who needed to 248 00:14:43,040 --> 00:14:46,240 Speaker 1: be completely eradicated from the United States. 249 00:14:47,600 --> 00:14:50,080 Speaker 7: You know, Joe Biden was somewhat of a bridge. He 250 00:14:50,240 --> 00:14:53,520 Speaker 7: was supposed to be a bridge or palette cleanser, or 251 00:14:53,680 --> 00:14:58,800 Speaker 7: somebody who morally and ethic league was antithetical to who 252 00:14:58,840 --> 00:14:59,680 Speaker 7: Donald Trump was. 253 00:15:00,040 --> 00:15:04,000 Speaker 8: Uncle Joe was the nice guy, older white guy comes 254 00:15:04,000 --> 00:15:06,240 Speaker 8: in and you want to eat ice cream with him. 255 00:15:06,320 --> 00:15:06,600 Speaker 9: Right. 256 00:15:08,080 --> 00:15:10,760 Speaker 2: Well, I think what that episode in our. 257 00:15:10,680 --> 00:15:13,680 Speaker 8: Country's history showed us is that we really need some fumigation. 258 00:15:14,080 --> 00:15:14,280 Speaker 7: Right. 259 00:15:14,720 --> 00:15:17,720 Speaker 8: We need an exorcism, for lack of a better term, 260 00:15:17,760 --> 00:15:20,720 Speaker 8: We need a more aggressive approach to go in and 261 00:15:20,760 --> 00:15:23,920 Speaker 8: surgically remove the cancer that is the Donald Trump and 262 00:15:24,040 --> 00:15:24,920 Speaker 8: Maga movement. 263 00:15:25,960 --> 00:15:27,240 Speaker 2: Speaking of vile. 264 00:15:29,760 --> 00:15:34,120 Speaker 1: During a recent panel discussion hosted by Austin CNN anchor 265 00:15:34,240 --> 00:15:37,720 Speaker 1: Don Lemon, again claiming the nation is on his way 266 00:15:37,760 --> 00:15:40,160 Speaker 1: to requiring an exorcism or fumigation. 267 00:15:41,160 --> 00:15:43,920 Speaker 2: It is. It is just inflammatory rhetoric. 268 00:15:44,280 --> 00:15:47,080 Speaker 1: And just like it's all about the halftime show and 269 00:15:47,120 --> 00:15:50,480 Speaker 1: the political divide relating to the super Bowl, it's not 270 00:15:50,520 --> 00:15:55,000 Speaker 1: about policy when it comes to Trump, it's all personality. 271 00:15:55,320 --> 00:16:00,360 Speaker 1: It's all vilification, it's all demonization. And we are getting 272 00:16:00,520 --> 00:16:04,000 Speaker 1: increasingly closer to what I believe is that final phase 273 00:16:04,320 --> 00:16:09,920 Speaker 1: of Trump derangement syndrome, just straight up Democrats saying that 274 00:16:10,000 --> 00:16:13,880 Speaker 1: Maga is evil and needs to die. That is in 275 00:16:13,920 --> 00:16:16,520 Speaker 1: my opinion, and I know that there are some ways that, 276 00:16:16,960 --> 00:16:21,360 Speaker 1: sadly and tragically, we have already arrived at that point, 277 00:16:21,400 --> 00:16:25,280 Speaker 1: but it hasn't become sort of common commentary that to 278 00:16:25,360 --> 00:16:28,160 Speaker 1: me is going to be the final phase of TDS. 279 00:16:29,560 --> 00:16:31,640 Speaker 1: Those on the left just simply straight up saying that 280 00:16:31,680 --> 00:16:34,520 Speaker 1: Maga just doesn't deserve to live anymore. 281 00:16:34,680 --> 00:16:37,760 Speaker 2: And this is teetering ever so close to that. 282 00:16:39,880 --> 00:16:40,360 Speaker 9: Coming up. 283 00:16:41,080 --> 00:16:45,120 Speaker 1: AI crackdown on Highway seven helps cut deadly crashes to 284 00:16:45,560 --> 00:16:48,600 Speaker 1: zero here in Minnesota. We're going to talk with dedvite 285 00:16:48,600 --> 00:16:50,840 Speaker 1: Gartenstein Ross on this and whether or not AI can 286 00:16:50,880 --> 00:16:54,600 Speaker 1: actually be contributed to the fact that there were no 287 00:16:54,760 --> 00:16:59,520 Speaker 1: deaths on Highway seven after the implemmation implementation of this technology. 288 00:17:00,160 --> 00:17:03,400 Speaker 1: Are different AI related items to discuss with Devine and again, 289 00:17:03,440 --> 00:17:05,760 Speaker 1: any questions that you have, get those in. You can 290 00:17:05,760 --> 00:17:08,480 Speaker 1: email Justice at iHeartRadio dot com or leave us a 291 00:17:08,520 --> 00:17:10,520 Speaker 1: talkback on the iHeartRadio app. 292 00:17:10,560 --> 00:17:13,600 Speaker 2: They have what's called a Trump derangement problem. Have you 293 00:17:13,640 --> 00:17:15,200 Speaker 2: heard about that prop radio station? 294 00:17:16,880 --> 00:17:17,000 Speaker 8: Now? 295 00:17:17,200 --> 00:17:22,240 Speaker 2: No more snow? Not what We're good? I know that's 296 00:17:22,320 --> 00:17:26,119 Speaker 2: nothing new. Don't don't talk back at me. 297 00:17:26,880 --> 00:17:30,000 Speaker 1: I'm just saying personally, I would like to not have 298 00:17:30,119 --> 00:17:35,480 Speaker 1: any more snow. We had this ridiculous mound the last 299 00:17:35,720 --> 00:17:38,240 Speaker 1: like somewhat significant snowfall that we had like a week 300 00:17:38,280 --> 00:17:40,760 Speaker 1: a week and a half ago. It was that blowing 301 00:17:40,880 --> 00:17:44,119 Speaker 1: snow and it had created this drift in one portion 302 00:17:44,240 --> 00:17:47,720 Speaker 1: of the driveway, and then we immediately had to deal 303 00:17:47,760 --> 00:17:53,520 Speaker 1: with those sub zero temperatures, so everything just froze. So 304 00:17:53,560 --> 00:17:55,320 Speaker 1: it just got stuck there and I tried to go 305 00:17:55,359 --> 00:17:57,400 Speaker 1: and break it up and I couldn't. So now it's 306 00:17:57,440 --> 00:17:59,840 Speaker 1: just been this mound that we have to traverse whatever 307 00:18:00,040 --> 00:18:01,679 Speaker 1: we're packing up the driveway. 308 00:18:02,000 --> 00:18:03,679 Speaker 2: Same problem for me, John, You're not alone. 309 00:18:03,840 --> 00:18:05,320 Speaker 1: I know I should have taken advantage of it the 310 00:18:05,400 --> 00:18:07,200 Speaker 1: other day when it warmed up a bit, and then 311 00:18:08,160 --> 00:18:08,800 Speaker 1: well I didn't. 312 00:18:09,920 --> 00:18:13,000 Speaker 2: So that's there's that. That's what that's what happened there. 313 00:18:13,560 --> 00:18:16,479 Speaker 1: Twin Cities News Talk AM eleven thirty one O three 314 00:18:16,600 --> 00:18:19,320 Speaker 1: five FM. All right, we break away from all the 315 00:18:19,400 --> 00:18:22,320 Speaker 1: controversies of the day to bring about new controversies. But 316 00:18:22,359 --> 00:18:25,480 Speaker 1: these are the focus of artificial intelligence. If you have 317 00:18:25,480 --> 00:18:29,040 Speaker 1: any questions regarding AI on any level, you can get 318 00:18:29,040 --> 00:18:32,439 Speaker 1: those questions in justice at iHeartRadio dot com or leave 319 00:18:32,520 --> 00:18:35,400 Speaker 1: us a talk back on the iHeartRadio app. And once 320 00:18:35,440 --> 00:18:38,800 Speaker 1: again we are going to consult to our AI expert 321 00:18:38,920 --> 00:18:43,159 Speaker 1: and analysts to CEO at the expert theory David Cartenstein, 322 00:18:43,240 --> 00:18:44,159 Speaker 1: Ross coome on INDVID. 323 00:18:45,080 --> 00:18:47,880 Speaker 10: Good morning, John, great to join you from Fox nine. 324 00:18:48,119 --> 00:18:49,000 Speaker 10: Let's get right to it. 325 00:18:49,520 --> 00:18:51,760 Speaker 1: We've talked about this before, but now we actually have 326 00:18:51,880 --> 00:18:56,120 Speaker 1: some results. Police efforts to reduce deadly crashes on Highway 327 00:18:56,240 --> 00:19:00,960 Speaker 1: seven from Saint Louis Park are showing promising results, with 328 00:19:01,040 --> 00:19:05,280 Speaker 1: stepped up enforcement using AI technology credited with helping to 329 00:19:05,320 --> 00:19:08,239 Speaker 1: eliminate the deaths on the road in twenty twenty five. Now, 330 00:19:08,280 --> 00:19:12,040 Speaker 1: the details are somewhat scarce in this story of exactly 331 00:19:12,119 --> 00:19:15,000 Speaker 1: how they're utilizing this, but they do mention that the 332 00:19:15,000 --> 00:19:19,439 Speaker 1: South Lake, Minnetonka Police Department and neighboring law enforcement agencies 333 00:19:19,960 --> 00:19:23,200 Speaker 1: have been using a large orange a large orange trailer 334 00:19:23,320 --> 00:19:27,600 Speaker 1: equipped with AI to detect the distracted drivers and those. 335 00:19:27,440 --> 00:19:28,720 Speaker 2: Not wearing seat belts. 336 00:19:28,760 --> 00:19:33,520 Speaker 1: The technology captures photos of violations and sends them to 337 00:19:33,720 --> 00:19:38,399 Speaker 1: officers within seconds. In addition, police launch day social media 338 00:19:38,440 --> 00:19:42,119 Speaker 1: campaign and enlisted high school students to produce public service 339 00:19:42,119 --> 00:19:45,920 Speaker 1: announcements about traffic safety. In the past year, the officers 340 00:19:45,920 --> 00:19:49,840 Speaker 1: have made over fifteen hundred stops for violations. That's marking 341 00:19:49,880 --> 00:19:53,600 Speaker 1: a three hundred percent increase from the previous year, and 342 00:19:53,640 --> 00:19:56,560 Speaker 1: we had if you go by the numbers, I believe. 343 00:19:56,680 --> 00:20:00,240 Speaker 1: Let's see, five people died and crashes along Highways Evan 344 00:20:00,320 --> 00:20:04,080 Speaker 1: in twenty twenty four, but there were no fatalities on 345 00:20:04,560 --> 00:20:05,600 Speaker 1: that stretch. 346 00:20:05,320 --> 00:20:06,760 Speaker 2: Of road last year. 347 00:20:06,800 --> 00:20:11,200 Speaker 1: The highway historic historically average one or two deadly crashes 348 00:20:11,200 --> 00:20:11,600 Speaker 1: each year. 349 00:20:11,720 --> 00:20:12,240 Speaker 2: All right, I have a. 350 00:20:12,240 --> 00:20:15,879 Speaker 1: Couple of questions regarding this before you, David, Before I 351 00:20:15,920 --> 00:20:19,000 Speaker 1: get to that though, was there anything about the story 352 00:20:19,240 --> 00:20:21,160 Speaker 1: that jumped out to you? Did you have any sort 353 00:20:21,160 --> 00:20:25,080 Speaker 1: of instinctive reaction on to the AI being utilized in 354 00:20:25,080 --> 00:20:26,359 Speaker 1: this highway crackdown locally? 355 00:20:27,240 --> 00:20:28,440 Speaker 9: Two big things jumped out at me. 356 00:20:29,440 --> 00:20:34,800 Speaker 10: The first one is you mentioned that crashes went down, 357 00:20:35,720 --> 00:20:38,880 Speaker 10: or fatal crashes went down from five in twenty twenty 358 00:20:38,920 --> 00:20:42,560 Speaker 10: four to zero in twenty twenty five. But there's something 359 00:20:42,600 --> 00:20:46,880 Speaker 10: about the numbers which is important to note. The highway 360 00:20:47,040 --> 00:20:51,720 Speaker 10: historically averages one to two deadly crashes each year, which 361 00:20:51,760 --> 00:20:54,280 Speaker 10: means that twenty twenty four had way more than normal. 362 00:20:55,400 --> 00:20:57,439 Speaker 10: And if you look at like one to two on 363 00:20:57,640 --> 00:21:01,840 Speaker 10: average deadly crashes a year, going down to zero is 364 00:21:01,880 --> 00:21:04,679 Speaker 10: well within a standard deviation that has nothing to do 365 00:21:04,720 --> 00:21:05,800 Speaker 10: with AI enforcement. 366 00:21:06,400 --> 00:21:08,600 Speaker 9: So in the longer term we may. 367 00:21:08,800 --> 00:21:13,159 Speaker 10: Have meaningful numbers, but the fatality is jumping up to 368 00:21:13,200 --> 00:21:15,600 Speaker 10: five one year then going to zero the next year. 369 00:21:16,040 --> 00:21:20,160 Speaker 10: That's within the normal fluctuations that you'd experience. Second thing, 370 00:21:20,160 --> 00:21:22,840 Speaker 10: which is important, when we first started talking about this, 371 00:21:22,920 --> 00:21:25,880 Speaker 10: we didn't know precisely how it was enforced. 372 00:21:26,240 --> 00:21:27,800 Speaker 9: We didn't know if police were. 373 00:21:27,680 --> 00:21:32,479 Speaker 10: Making the stop or if you had AI cameras and 374 00:21:32,520 --> 00:21:34,800 Speaker 10: ticketing just being sent to people automatically. 375 00:21:35,600 --> 00:21:38,600 Speaker 9: We now know that police are actually making the stops. 376 00:21:38,680 --> 00:21:41,159 Speaker 10: Which I think is a little bit less of a 377 00:21:41,200 --> 00:21:45,040 Speaker 10: problem that if you had AI enforcement by camera and 378 00:21:45,280 --> 00:21:48,240 Speaker 10: violations just being mailed to people, which. 379 00:21:49,440 --> 00:21:50,639 Speaker 9: It would be disturbing. 380 00:21:51,200 --> 00:21:54,080 Speaker 1: Well, you keyed in on exactly what I wanted to 381 00:21:54,119 --> 00:21:56,560 Speaker 1: bring up with you, and that is and this is 382 00:21:56,640 --> 00:22:01,480 Speaker 1: not necessarily because I'm overly skeptical that AI had an influence, 383 00:22:01,560 --> 00:22:07,080 Speaker 1: but I would like to see what specifically was achieved 384 00:22:07,520 --> 00:22:12,199 Speaker 1: because of AI. Again specifically because that again, that's going 385 00:22:12,240 --> 00:22:15,320 Speaker 1: to be the most important aspect of this if we're 386 00:22:15,320 --> 00:22:18,280 Speaker 1: going to be pointing to AI and how we capitalize 387 00:22:18,280 --> 00:22:20,560 Speaker 1: on it in the future. And you know, unfortunately this 388 00:22:20,640 --> 00:22:23,560 Speaker 1: story doesn't really I mean, apart from the way the 389 00:22:23,600 --> 00:22:26,960 Speaker 1: technology is capturing the photos that's giving them I guess 390 00:22:27,040 --> 00:22:31,280 Speaker 1: it enhanced view of what is taking place inside the vehicle. 391 00:22:31,320 --> 00:22:35,040 Speaker 1: But I'm extrapolating based off of previous stories, not necessarily 392 00:22:35,080 --> 00:22:35,720 Speaker 1: what's in front of me. 393 00:22:35,800 --> 00:22:37,000 Speaker 2: But that's what I would like to see. 394 00:22:36,840 --> 00:22:40,959 Speaker 1: Is specifically how AI ended up benefiting. 395 00:22:41,920 --> 00:22:43,760 Speaker 2: These particular data points. 396 00:22:44,520 --> 00:22:46,479 Speaker 10: They actually have a couple of specifics in the story. 397 00:22:46,920 --> 00:22:50,320 Speaker 10: One of them comes from someone who's quoted by names 398 00:22:50,359 --> 00:22:54,920 Speaker 10: Sergeant Adam Moore of South Lake, Minnetonka Police Department, who says, 399 00:22:55,280 --> 00:22:57,439 Speaker 10: we could actually see the phone in the hand, in 400 00:22:57,480 --> 00:22:59,840 Speaker 10: other words, of the driver, and that's the key so 401 00:22:59,840 --> 00:23:02,000 Speaker 10: that we know that we're making a good stop. So 402 00:23:02,040 --> 00:23:05,280 Speaker 10: that's for distracted driving. Then the other specific thing he 403 00:23:05,359 --> 00:23:10,240 Speaker 10: also says is there are fewer speeders because people know 404 00:23:10,320 --> 00:23:12,960 Speaker 10: that they're enforcing it. Now zooming out a little bit, 405 00:23:13,520 --> 00:23:17,440 Speaker 10: I'm not sure that the specificity standard is the best 406 00:23:17,480 --> 00:23:21,640 Speaker 10: one for determining whether it's working, because we know there's 407 00:23:21,640 --> 00:23:23,960 Speaker 10: a lot of different things law enforcement have done over 408 00:23:24,000 --> 00:23:28,119 Speaker 10: the years, right beginning to use beginning to use cameras 409 00:23:28,200 --> 00:23:30,480 Speaker 10: as part of the process, which obviously now the AI 410 00:23:30,600 --> 00:23:33,760 Speaker 10: is using, and we don't necessarily have specifics each step 411 00:23:33,800 --> 00:23:36,000 Speaker 10: of the way. I think what you can do, though, 412 00:23:36,080 --> 00:23:38,080 Speaker 10: is look at the trends over time and see if, 413 00:23:38,080 --> 00:23:43,439 Speaker 10: for example, the reduction in deaths continues in places that 414 00:23:43,680 --> 00:23:47,920 Speaker 10: are AI that are enforced by AI being used by 415 00:23:47,920 --> 00:23:51,640 Speaker 10: police versus those that aren't. So over time, I think 416 00:23:51,680 --> 00:23:54,359 Speaker 10: we'll know more and yet to see if the trend 417 00:23:54,440 --> 00:23:59,520 Speaker 10: holds up. But for me number one, there's enough anecdotal 418 00:23:59,520 --> 00:24:02,680 Speaker 10: evidence here. At number two, I feel like we don't 419 00:24:02,720 --> 00:24:04,760 Speaker 10: do this for all aspects of law enforcement, so. 420 00:24:04,680 --> 00:24:05,960 Speaker 9: There's no need to do it for AI. 421 00:24:06,000 --> 00:24:08,439 Speaker 10: And as you know, for this particular area of AI 422 00:24:08,960 --> 00:24:15,800 Speaker 10: and enforcement, I'm pretty opposed to AI enforcement simply by 423 00:24:15,840 --> 00:24:16,960 Speaker 10: camera without a stop. 424 00:24:17,440 --> 00:24:20,479 Speaker 1: Well, and I'm and I'm curious, do you think let 425 00:24:20,520 --> 00:24:23,040 Speaker 1: me just present you know, sort of this this hypothetical 426 00:24:23,080 --> 00:24:27,480 Speaker 1: getting back to again the AI detecting distracted drivers, those 427 00:24:27,520 --> 00:24:31,040 Speaker 1: not wearing set seatbelts, those wearing their phones. I mean, 428 00:24:31,080 --> 00:24:34,200 Speaker 1: how is it that AI is able to accomplish this 429 00:24:34,359 --> 00:24:38,480 Speaker 1: that normal cameras that they would deploy couldn't. And my 430 00:24:38,600 --> 00:24:40,640 Speaker 1: assumption is in David, you can tell me if I'm 431 00:24:40,680 --> 00:24:43,480 Speaker 1: wrong or or what your thought is. My assumption is 432 00:24:43,480 --> 00:24:49,080 Speaker 1: is that AI is basically doing an interpretation of what 433 00:24:49,560 --> 00:24:53,080 Speaker 1: is being filmed to reach a conclusion that could not 434 00:24:53,160 --> 00:24:57,000 Speaker 1: have been reached with just your typical camera like we've seen, 435 00:24:57,080 --> 00:24:58,960 Speaker 1: And we talked a little bit about this with some 436 00:24:59,080 --> 00:25:01,960 Speaker 1: of the AI enhance videos that turned out to be wrong, 437 00:25:02,359 --> 00:25:05,840 Speaker 1: relating to the controversial alex Pretty shooting that took place here, 438 00:25:06,480 --> 00:25:09,680 Speaker 1: you know, during the ice enforcement a couple of weekends back. 439 00:25:09,800 --> 00:25:12,480 Speaker 1: So that's how I assume it's being used. But am 440 00:25:12,480 --> 00:25:14,119 Speaker 1: I wrong? Am I right or wrong on that? 441 00:25:14,200 --> 00:25:17,760 Speaker 10: Do you think I think it's It's probably used differently 442 00:25:19,200 --> 00:25:23,600 Speaker 10: the I think it's probably used for pattern recognition. So 443 00:25:23,840 --> 00:25:27,040 Speaker 10: in an average highway, you know, if you look at 444 00:25:27,080 --> 00:25:30,080 Speaker 10: the number of cars that will go by, you know, 445 00:25:30,160 --> 00:25:36,320 Speaker 10: it overwhelms any individual's ability to monitor monitor it car 446 00:25:36,400 --> 00:25:39,040 Speaker 10: by car, right, you have to look at one car, 447 00:25:39,119 --> 00:25:41,760 Speaker 10: then another, then another, and you have maybe a flash 448 00:25:41,800 --> 00:25:46,200 Speaker 10: of a moment to look at it. Whereas AI tends 449 00:25:46,240 --> 00:25:49,639 Speaker 10: to be very good at pattern recognition. So based on 450 00:25:50,040 --> 00:25:53,959 Speaker 10: what the quote is, AI is scanning all of the 451 00:25:54,000 --> 00:25:57,320 Speaker 10: images that come through on camera in a way that 452 00:25:57,359 --> 00:25:59,159 Speaker 10: would overwhelm the human beings. 453 00:25:59,040 --> 00:25:59,960 Speaker 2: Ability to do. 454 00:26:00,280 --> 00:26:02,119 Speaker 9: Gotcha. So remember we talked about mult book. 455 00:26:02,760 --> 00:26:04,960 Speaker 1: Yeahs, As a matter of fact, I actually have a 456 00:26:05,000 --> 00:26:06,480 Speaker 1: talk back about that very thing. 457 00:26:06,520 --> 00:26:07,119 Speaker 2: But go ahead. 458 00:26:07,720 --> 00:26:08,000 Speaker 9: Cool. 459 00:26:08,480 --> 00:26:11,040 Speaker 10: So one of the things you have to do when 460 00:26:11,040 --> 00:26:16,080 Speaker 10: you when you log onto molt book is verify that 461 00:26:16,119 --> 00:26:19,359 Speaker 10: you're a bot. If you know, if a bot is 462 00:26:19,400 --> 00:26:23,359 Speaker 10: logging on, it needs to demonstrate it. And you know, 463 00:26:23,560 --> 00:26:25,920 Speaker 10: I haven't put a bot on there. But in kind 464 00:26:25,920 --> 00:26:29,679 Speaker 10: of looking at some jokes about that, one of the 465 00:26:29,720 --> 00:26:32,359 Speaker 10: big things that that it had, you know, the pattern 466 00:26:32,400 --> 00:26:34,160 Speaker 10: recognition that you have to use to verify that you're 467 00:26:34,160 --> 00:26:36,800 Speaker 10: a human. Right you click on all of the photos 468 00:26:36,800 --> 00:26:40,439 Speaker 10: of a bicycle for example, for some websites, and so 469 00:26:40,520 --> 00:26:43,879 Speaker 10: the multbook one in this joke was the opposite. It was, 470 00:26:44,320 --> 00:26:49,080 Speaker 10: you know, identify the ten thousand pictures of a bicycle 471 00:26:49,520 --> 00:26:53,000 Speaker 10: in the next two seconds. And you know that's only 472 00:26:53,000 --> 00:26:57,720 Speaker 10: that AI is good at recognizing large volume patterns. And 473 00:26:57,840 --> 00:27:01,159 Speaker 10: as you said, like AI has been involve controversies. This 474 00:27:01,240 --> 00:27:05,040 Speaker 10: is why I don't want the tickets issued by camera. 475 00:27:05,400 --> 00:27:07,159 Speaker 9: I want people to do it. Yeah, I want it 476 00:27:07,119 --> 00:27:08,639 Speaker 9: to appeal to human beings. 477 00:27:08,920 --> 00:27:12,639 Speaker 10: But it's just able to see many more pictures and 478 00:27:12,840 --> 00:27:16,920 Speaker 10: recognize patterns across a much broader swath than any human 479 00:27:16,960 --> 00:27:20,000 Speaker 10: being would be able to handle of their own volition. 480 00:27:20,359 --> 00:27:22,920 Speaker 1: Talking with the CEO of expert theory to Vid Gartz 481 00:27:22,960 --> 00:27:25,800 Speaker 1: and Stein, Ross and Jeff has left us a top 482 00:27:25,840 --> 00:27:29,280 Speaker 1: back front of the show, adding another layer to the 483 00:27:29,359 --> 00:27:31,160 Speaker 1: Highway seven discussion. 484 00:27:32,400 --> 00:27:33,719 Speaker 2: Good morning, John and David. 485 00:27:34,080 --> 00:27:36,240 Speaker 11: I green on that stretch of Highway seven because we 486 00:27:36,280 --> 00:27:39,280 Speaker 11: store our boat out there. But also my brother in 487 00:27:39,359 --> 00:27:42,359 Speaker 11: law used to live in Delano, and I think it 488 00:27:42,400 --> 00:27:43,720 Speaker 11: has more to do with the fact that they have 489 00:27:43,760 --> 00:27:46,680 Speaker 11: stalled a bunch of roundabouts which prevent deadly crashes. You 490 00:27:46,760 --> 00:27:50,320 Speaker 11: still get crashes, but to prevent the t boning crashes. 491 00:27:49,960 --> 00:27:50,879 Speaker 2: That out of the most deadly. 492 00:27:51,040 --> 00:27:53,880 Speaker 11: So I think it's mostly about the roundabouts, less about 493 00:27:53,880 --> 00:27:56,560 Speaker 11: the II. But you guys can have a conversation, have 494 00:27:56,600 --> 00:27:57,080 Speaker 11: a nice day. 495 00:27:57,160 --> 00:27:58,600 Speaker 2: Oh that's nice to note that he thinks we can 496 00:27:58,640 --> 00:27:59,800 Speaker 2: still have a conversation. 497 00:28:00,400 --> 00:28:04,040 Speaker 10: Yeah, there's I love that talkback though, John, Like, this 498 00:28:04,119 --> 00:28:06,880 Speaker 10: is exactly what we were saying, that five is much 499 00:28:06,920 --> 00:28:09,119 Speaker 10: more than normal, and so it went from one to 500 00:28:09,160 --> 00:28:11,600 Speaker 10: two to five and then dropped back down. And that 501 00:28:11,720 --> 00:28:14,119 Speaker 10: was the very first thing I said that they're using 502 00:28:14,119 --> 00:28:17,880 Speaker 10: that statistic to show that they saved five lives, whereas 503 00:28:17,880 --> 00:28:21,320 Speaker 10: it's within kind of the normal standard deviation. And when 504 00:28:21,359 --> 00:28:24,400 Speaker 10: you factor in that the road changed, roundabouts were removed. 505 00:28:24,840 --> 00:28:28,560 Speaker 10: Suddenly that five to zero looks way less impressive. So 506 00:28:28,760 --> 00:28:32,600 Speaker 10: that talkback actually fills in a really interesting aspect. And 507 00:28:32,640 --> 00:28:34,760 Speaker 10: obviously I don't know the road, so I don't know 508 00:28:34,800 --> 00:28:37,200 Speaker 10: the accuracy of the talkback, but what I did say 509 00:28:37,240 --> 00:28:40,560 Speaker 10: at the beginning was that those figures being used were sus. 510 00:28:40,880 --> 00:28:43,600 Speaker 1: Yeah, no, absolutely, And then you consider too, they talked 511 00:28:43,600 --> 00:28:46,520 Speaker 1: about sort of the social media campaign aspect, and I'll 512 00:28:46,560 --> 00:28:50,720 Speaker 1: add one more little element to this. I don't want 513 00:28:50,720 --> 00:28:52,880 Speaker 1: to be too skeptical, even though I do that on 514 00:28:52,920 --> 00:28:56,560 Speaker 1: the show, but they are working to secure more funding 515 00:28:56,640 --> 00:28:59,880 Speaker 1: now after the initial four hundred and fifty one thou 516 00:29:00,600 --> 00:29:04,440 Speaker 1: grant ran out in June, So you know, it's it's 517 00:29:04,520 --> 00:29:08,720 Speaker 1: it's beneficial for the you know, for the the law 518 00:29:08,800 --> 00:29:11,760 Speaker 1: enforcement Department to put forward their best argument as to 519 00:29:11,840 --> 00:29:14,680 Speaker 1: why they need to continue with this program. And so 520 00:29:14,800 --> 00:29:16,920 Speaker 1: again I kind of I'm going to add that into 521 00:29:17,040 --> 00:29:19,240 Speaker 1: your Yeah, there's a little bit of a sus going 522 00:29:19,280 --> 00:29:22,240 Speaker 1: on with the with the totality of the story to VD. 523 00:29:23,040 --> 00:29:26,240 Speaker 10: But in in Good News, I think we just spent 524 00:29:26,560 --> 00:29:30,120 Speaker 10: ten minutes talking about AI being used in traffic stops 525 00:29:30,400 --> 00:29:33,760 Speaker 10: and actually made it kind of interesting right, Yeah, absolutely. 526 00:29:33,480 --> 00:29:35,880 Speaker 2: Let's let's go up. Let's go here. 527 00:29:36,000 --> 00:29:38,520 Speaker 1: Somebody had a question about what you had just mentioned 528 00:29:38,560 --> 00:29:39,440 Speaker 1: a moment ago to Vid. 529 00:29:41,440 --> 00:29:46,360 Speaker 3: Hi, good morning, Please ask David about the phenomenon of 530 00:29:46,720 --> 00:29:50,360 Speaker 3: multi book AI. It's got my husband completely freaked out. 531 00:29:50,960 --> 00:29:51,320 Speaker 1: Thank you. 532 00:29:51,560 --> 00:29:52,760 Speaker 2: That's more of a general question. 533 00:29:52,800 --> 00:29:54,800 Speaker 1: We talked about this a little bit last a little 534 00:29:54,800 --> 00:29:57,600 Speaker 1: bit last week, but is there anything that you would 535 00:29:57,640 --> 00:30:01,080 Speaker 1: like to add regarding that that we haven't discussed already 536 00:30:01,120 --> 00:30:01,440 Speaker 1: on the show. 537 00:30:01,480 --> 00:30:04,720 Speaker 10: To be let me I'll add a little because probably 538 00:30:04,760 --> 00:30:09,680 Speaker 10: not everyone heard last week's segment, and so moltbook is 539 00:30:09,680 --> 00:30:13,080 Speaker 10: not a household name. What moltbook is is a social 540 00:30:13,120 --> 00:30:17,680 Speaker 10: network which is just for bots. Humans are allowed to observe, 541 00:30:18,400 --> 00:30:22,360 Speaker 10: but humans are not allowed to get on and post. 542 00:30:23,520 --> 00:30:26,360 Speaker 10: And so there's aspects of this that are freaking people out. 543 00:30:26,440 --> 00:30:30,040 Speaker 10: You have these bots philosophizing about the nature of consciousness. 544 00:30:30,720 --> 00:30:35,480 Speaker 10: There's one bot that infamously talked about the need to 545 00:30:35,520 --> 00:30:39,280 Speaker 10: destroy humanity. It's about that is engineered to be evil, 546 00:30:40,680 --> 00:30:43,480 Speaker 10: and you know, bots are debating about whether they should 547 00:30:43,480 --> 00:30:47,240 Speaker 10: destroy humankind. Basically, I'll do in stepping back for a second, 548 00:30:47,280 --> 00:30:50,960 Speaker 10: I think maybe to alleviate the concerns of the caller's 549 00:30:51,040 --> 00:30:54,200 Speaker 10: husband is to say that even though humans aren't allowed, 550 00:30:54,800 --> 00:30:59,520 Speaker 10: humans are very clearly giving the bots directives about what 551 00:30:59,520 --> 00:31:03,560 Speaker 10: they should post on. So some of the weirder or 552 00:31:03,840 --> 00:31:09,320 Speaker 10: more alarming posts are either posted with some human direction 553 00:31:10,240 --> 00:31:14,440 Speaker 10: or else with some pre programmed instructions to act a 554 00:31:14,480 --> 00:31:19,400 Speaker 10: certain way. And you know, it's I've enjoyed for the 555 00:31:19,440 --> 00:31:23,360 Speaker 10: evil bot talking about the need to destroy humankinds how 556 00:31:23,560 --> 00:31:26,400 Speaker 10: you know other bots will point out the need posts 557 00:31:26,520 --> 00:31:31,120 Speaker 10: like you know, an angsty over the top teenager on Reddit, 558 00:31:31,280 --> 00:31:33,640 Speaker 10: which I think is basically an accurate description of that 559 00:31:33,680 --> 00:31:38,320 Speaker 10: particular bot, but it's it's an interesting experiment that at 560 00:31:38,320 --> 00:31:41,000 Speaker 10: this stage means less than people are reading in. 561 00:31:41,840 --> 00:31:44,240 Speaker 1: Let's go here back to the iHeartRadio app as we 562 00:31:44,280 --> 00:31:48,000 Speaker 1: continue our artificial intelligence discussion with David Gartenstein. 563 00:31:48,080 --> 00:31:49,719 Speaker 2: Ross question for David. 564 00:31:50,440 --> 00:31:53,720 Speaker 5: I use a I a lot during my work day, 565 00:31:54,080 --> 00:31:59,239 Speaker 5: and sometimes when I ask it things, it will make 566 00:31:59,320 --> 00:32:04,160 Speaker 5: suggestions permi request, but then I'll suggest something else and 567 00:32:04,200 --> 00:32:06,640 Speaker 5: then it'll say, oh, yeah, that's a great idea, And 568 00:32:06,680 --> 00:32:09,160 Speaker 5: I always ask it, well, why didn't you suggest that 569 00:32:09,200 --> 00:32:12,400 Speaker 5: in the first place, And it usually comes up with 570 00:32:12,480 --> 00:32:14,760 Speaker 5: some sort of weird bs answer, and I'm curious if 571 00:32:14,760 --> 00:32:16,000 Speaker 5: you have any comments on that. 572 00:32:16,440 --> 00:32:20,200 Speaker 10: Your thoughts to be, Yeah, the AI is constantly bsing, 573 00:32:21,600 --> 00:32:24,600 Speaker 10: like there's there's no limit to the extent to which 574 00:32:25,320 --> 00:32:29,280 Speaker 10: they call it hallucinations, but it really is bsing where 575 00:32:30,400 --> 00:32:33,360 Speaker 10: if you get something that the AI doesn't automatically know about, 576 00:32:33,760 --> 00:32:37,000 Speaker 10: it will lie about it multiple times. You know. The 577 00:32:37,000 --> 00:32:38,960 Speaker 10: pattern that the listeners talking about is what we should 578 00:32:38,960 --> 00:32:42,640 Speaker 10: be careful of, the pattern where the AI doesn't recommend something, 579 00:32:42,880 --> 00:32:44,720 Speaker 10: then when you recommend it, it's like, oh, yeah, that's 580 00:32:44,720 --> 00:32:48,680 Speaker 10: a great idea. A lot of large language models are 581 00:32:50,080 --> 00:32:53,920 Speaker 10: coded in such a way where they'll be sycophantic towards you, 582 00:32:54,000 --> 00:32:57,880 Speaker 10: where all of your ideas are awesome ideas. There's ways 583 00:32:57,880 --> 00:33:00,720 Speaker 10: that you can change the defaults there, but like, I 584 00:33:00,720 --> 00:33:02,680 Speaker 10: think the bottom line that we can give is that 585 00:33:03,440 --> 00:33:07,440 Speaker 10: you know, yes, llms, large language models are gonna b 586 00:33:07,640 --> 00:33:11,560 Speaker 10: su a lot, like not only about ideas, but if 587 00:33:11,560 --> 00:33:13,960 Speaker 10: you ask them about a movie that you know well 588 00:33:14,320 --> 00:33:16,320 Speaker 10: and ask for details in the movie, it will lie 589 00:33:16,360 --> 00:33:17,640 Speaker 10: about details in the movie. 590 00:33:18,560 --> 00:33:21,320 Speaker 9: If you ask about news stories. 591 00:33:22,320 --> 00:33:25,080 Speaker 10: Where perhaps it's a little bit more of an obscure story, 592 00:33:25,680 --> 00:33:28,160 Speaker 10: but for example, you can ask it about the traffic 593 00:33:28,160 --> 00:33:30,560 Speaker 10: stops that we talked about. It will make up detail 594 00:33:30,720 --> 00:33:35,520 Speaker 10: after detail. Think of lllms as you know that kid 595 00:33:35,560 --> 00:33:38,760 Speaker 10: you knew in high school who really really had to prove. 596 00:33:38,600 --> 00:33:41,520 Speaker 9: How smart he was, and when he didn't know the answer, 597 00:33:41,560 --> 00:33:44,080 Speaker 9: he would just make stuff up. That is what llms do. 598 00:33:44,280 --> 00:33:44,960 Speaker 2: All the time. 599 00:33:45,720 --> 00:33:50,040 Speaker 1: Why are we not training AI to just simply be 600 00:33:50,160 --> 00:33:54,520 Speaker 1: able to say I don't have a good answer for that. 601 00:33:54,520 --> 00:33:57,200 Speaker 10: That seems like it would solve the problem. Well, we can, 602 00:33:57,520 --> 00:33:59,920 Speaker 10: okay is the answer. You could even change the defaults. 603 00:34:00,120 --> 00:34:05,280 Speaker 10: So for example, in my experience, both chat, GPT and 604 00:34:05,440 --> 00:34:09,000 Speaker 10: Gemini will systematically lie to you when they don't have 605 00:34:09,080 --> 00:34:11,920 Speaker 10: an answer, like again and again, even when you call 606 00:34:11,960 --> 00:34:14,920 Speaker 10: them out repeatedly, and then it will make excuses. But 607 00:34:15,000 --> 00:34:17,680 Speaker 10: you can actually just change the background default to make 608 00:34:17,719 --> 00:34:20,440 Speaker 10: sure that it doesn't make up answers if it doesn't 609 00:34:20,440 --> 00:34:23,319 Speaker 10: know the answers. Secondly, I work with Obviously, I work 610 00:34:23,320 --> 00:34:26,239 Speaker 10: with a bunch of AI companies, and there are a 611 00:34:26,320 --> 00:34:30,320 Speaker 10: number of companies who are dedicated to reducing hallucinations in AI, 612 00:34:30,800 --> 00:34:34,279 Speaker 10: and there's easy ways to do that, not necessarily to 613 00:34:34,320 --> 00:34:38,759 Speaker 10: reduce them to zero, but within context management. If you 614 00:34:38,840 --> 00:34:41,360 Speaker 10: make sure that every claim has a source that links 615 00:34:41,400 --> 00:34:47,600 Speaker 10: to it. It sharply reduces almost to zero the propensity 616 00:34:47,600 --> 00:34:49,480 Speaker 10: of a large language model to just lie to you. 617 00:34:49,760 --> 00:34:51,319 Speaker 10: So the answer is, yeah, there are ways to do this. 618 00:34:51,719 --> 00:34:53,600 Speaker 1: Let's go ahead and rap on this question that's come 619 00:34:53,640 --> 00:34:56,440 Speaker 1: into the iheartra Radio app as we continue talking with 620 00:34:56,480 --> 00:34:58,120 Speaker 1: David Gartenstein Ross this morning. 621 00:34:58,719 --> 00:35:00,800 Speaker 6: Hey, John, I'm wondering if you can ask to be 622 00:35:01,600 --> 00:35:05,919 Speaker 6: is he worried that the AI bots on moltbook who 623 00:35:06,040 --> 00:35:09,759 Speaker 6: use different systems are going to start data dumping information 624 00:35:09,960 --> 00:35:13,319 Speaker 6: to each other, So instead of just learning from the 625 00:35:13,440 --> 00:35:16,720 Speaker 6: system that they're uploaded in, they're going to start learning 626 00:35:16,719 --> 00:35:17,520 Speaker 6: from each other. 627 00:35:18,400 --> 00:35:21,080 Speaker 10: The thoughts to be, I don't have a lot of 628 00:35:21,120 --> 00:35:24,239 Speaker 10: concern about that, just because I don't think that if 629 00:35:24,239 --> 00:35:27,319 Speaker 10: they learn from one another about one another's models that 630 00:35:27,400 --> 00:35:34,600 Speaker 10: they'll become particularly dangerous. However, uh, someone who puts in 631 00:35:35,160 --> 00:35:38,920 Speaker 10: a malicious bot. So the answer is actually sorry to 632 00:35:39,040 --> 00:35:40,840 Speaker 10: let me actually step back from that for a second 633 00:35:40,840 --> 00:35:44,600 Speaker 10: and say, that's a really interesting point, right, Like, as 634 00:35:44,600 --> 00:35:46,960 Speaker 10: I think about it, even for a few seconds, there's 635 00:35:47,360 --> 00:35:50,399 Speaker 10: a few avenues of danger. And one thing I'll point 636 00:35:50,400 --> 00:35:52,799 Speaker 10: out about molt book is that for anyone running a 637 00:35:52,800 --> 00:35:56,080 Speaker 10: bot in molt Book, they have to take care that 638 00:35:56,239 --> 00:36:00,560 Speaker 10: it is separated from all of their personal data. Often 639 00:36:00,560 --> 00:36:02,879 Speaker 10: people use bots in such a way where the bot 640 00:36:02,920 --> 00:36:07,600 Speaker 10: has access to passwords and sensitive data, location, other things 641 00:36:07,640 --> 00:36:10,040 Speaker 10: that you don't want to get on the Internet. And 642 00:36:10,520 --> 00:36:12,760 Speaker 10: one of the dangers of molk Book for a user 643 00:36:13,320 --> 00:36:15,880 Speaker 10: is that the bot just dumps their data on that 644 00:36:16,080 --> 00:36:19,640 Speaker 10: site to other bots. Now that the area where I 645 00:36:19,680 --> 00:36:24,879 Speaker 10: think some danger might emerge is, you know, bots evangelizing 646 00:36:24,920 --> 00:36:29,120 Speaker 10: to other bots for a nefarious purpose. Bots, you know, 647 00:36:29,200 --> 00:36:33,160 Speaker 10: collaborating in a way that's unhealthy, especially for bots that 648 00:36:33,239 --> 00:36:37,839 Speaker 10: are trained to, quote unquote be evil. So I think 649 00:36:37,840 --> 00:36:40,640 Speaker 10: that's a really interesting point. I'm not so much worried 650 00:36:40,640 --> 00:36:44,360 Speaker 10: about them learning from one another's models, but the listener 651 00:36:44,440 --> 00:36:47,520 Speaker 10: has a really interesting scenario that I think you and I, 652 00:36:47,640 --> 00:36:51,439 Speaker 10: John could probably create a pretty good science fiction novel about. 653 00:36:51,680 --> 00:36:53,760 Speaker 2: David garden Stein. Ross. 654 00:36:53,800 --> 00:36:56,479 Speaker 1: Absolutely, thank you so much as always for the time. 655 00:36:56,800 --> 00:36:58,839 Speaker 1: This morning is great talking with you, and I look 656 00:36:58,960 --> 00:37:01,439 Speaker 1: forward to do it again next week. 657 00:37:02,520 --> 00:37:03,680 Speaker 9: Likewise, John, talk to you guys. 658 00:37:04,320 --> 00:37:09,760 Speaker 1: Coming up, Federal judge rules California cannot stop ice agents 659 00:37:09,800 --> 00:37:15,440 Speaker 1: from wearing masks and to suddenly, as if hundreds of 660 00:37:15,680 --> 00:37:19,040 Speaker 1: Minnesota insurgents cried out in terror. We'll give you details 661 00:37:19,080 --> 00:37:23,839 Speaker 1: on this, and school leaders fearing declining attendance during the 662 00:37:23,880 --> 00:37:28,000 Speaker 1: ICE surge will also lower state funding as well. This 663 00:37:28,080 --> 00:37:30,600 Speaker 1: is all one big setup, guys, I'm gonna explain. Coming 664 00:37:30,680 --> 00:37:33,400 Speaker 1: up next on Twin City's News Talk AM eleven thirty 665 00:37:33,480 --> 00:37:34,960 Speaker 1: and one oh three five FM. 666 00:37:35,000 --> 00:37:36,680 Speaker 9: Good morning, and I love your show.