WEBVTT 1 00:00:00.600 --> 00:00:02.919 This is Gary and Shannon and you're listening to KFI 2 00:00:03.080 --> 00:00:05.839 AM six forty, the Gary and Shannon Show on demand 3 00:00:05.919 --> 00:00:10.839 on the iHeartRadio app. That's just said, that's not sad, 4 00:00:10.960 --> 00:00:14.279 it's a data think about her and to celebrate her life. 5 00:00:14.880 --> 00:00:16.719 But I didn't think about it until I just looked 6 00:00:16.760 --> 00:00:19.079 at the rundown. Oh. 7 00:00:19.160 --> 00:00:21.320 I mean, well, you haven't thought about twenty eight. I 8 00:00:21.480 --> 00:00:24.440 haven't thought about the between twenty eight, right. It'd be 9 00:00:24.440 --> 00:00:27.480 different if you saw it earlier and just forgot. 10 00:00:27.440 --> 00:00:29.640 We're finally doing a burial. If I told you that, 11 00:00:30.039 --> 00:00:32.600 Oh good, when is that? That's gonna be next month? 12 00:00:33.280 --> 00:00:35.759 That's nice? Hey, what's the rush? 13 00:00:36.439 --> 00:00:36.880 I know? 14 00:00:36.960 --> 00:00:40.039 And that seems to be the families. That's that's the 15 00:00:40.039 --> 00:00:42.479 way we've always done it. Come on, we don't have 16 00:00:42.520 --> 00:00:45.640 any religious obligations to bury people within twenty four hours, 17 00:00:45.640 --> 00:00:47.920 so we wait, you know, fourteen months or in some 18 00:00:48.039 --> 00:00:49.320 cases two years. 19 00:00:49.759 --> 00:00:53.439 Are why did you wait so long? It's not like 20 00:00:53.479 --> 00:00:55.200 you're planning a party or anything. 21 00:00:55.359 --> 00:00:57.280 No, and it's just going to be it's just going 22 00:00:57.320 --> 00:00:59.600 to be family. Yeah, So I don't know, I don't 23 00:00:59.600 --> 00:01:02.479 know why. How do you make those decisions? I mean, 24 00:01:02.520 --> 00:01:05.560 that's first of all we had to figure out where 25 00:01:05.640 --> 00:01:09.640 because there's two family there's kind of two family cemetery things, 26 00:01:09.680 --> 00:01:12.640 and one is mom's generally mom's side, the other one 27 00:01:12.680 --> 00:01:14.319 is generally Dad's side. 28 00:01:14.400 --> 00:01:16.159 I can see if this was a you thing where 29 00:01:16.159 --> 00:01:19.120 you're kind of like, eh, whatever, But your sisters seem. 30 00:01:18.920 --> 00:01:21.359 Kind of type A to me. Yeah, I mean there 31 00:01:21.439 --> 00:01:21.760 is that. 32 00:01:22.120 --> 00:01:24.239 So that's kind of different to me that they would 33 00:01:24.280 --> 00:01:27.280 wait so long. Yeah, well, but there's been a lot 34 00:01:27.280 --> 00:01:29.079 of there's been a lot going on, a lot, there's 35 00:01:29.079 --> 00:01:31.599 been a lot going on, so it's probably fine. It's 36 00:01:31.719 --> 00:01:33.840 actually kind of nice that you can do it with 37 00:01:33.920 --> 00:01:36.920 all the other crap to get done with the house 38 00:01:36.959 --> 00:01:38.640 and the property and all that. Now that that's been 39 00:01:40.000 --> 00:01:40.319 on that. 40 00:01:40.319 --> 00:01:42.640 True, Yeah, that is true. So anyway, not to start 41 00:01:42.680 --> 00:01:44.079 on a negative note. 42 00:01:44.000 --> 00:01:46.799 Wow, well you want to do more dead stuff? 43 00:01:46.840 --> 00:01:50.599 We could. I found out that if if a bible 44 00:01:50.879 --> 00:01:55.879 becomes damaged, the way that you dispose of it is 45 00:01:56.040 --> 00:01:57.840 however you want to dispose of it. 46 00:01:58.000 --> 00:02:00.000 Why would you dispose of it just because it was damaged. 47 00:02:00.359 --> 00:02:04.319 Well, there was a time in the most recent past, 48 00:02:04.359 --> 00:02:08.080 I'd say within the last week that my dog shredded 49 00:02:08.800 --> 00:02:09.400 a bible. 50 00:02:09.719 --> 00:02:12.520 Really, so it wasn't just the yoga blocks. 51 00:02:12.520 --> 00:02:14.439 It was not just the yoga blocks. In fact, this 52 00:02:14.599 --> 00:02:19.280 was pre yoga block. He went after and just destroyed 53 00:02:19.840 --> 00:02:21.680 a Bible that was sitting on the couch. What do 54 00:02:21.719 --> 00:02:24.080 you think that means? I do not know, but I 55 00:02:24.159 --> 00:02:27.400 did ask for pastoral advice. I texted a friend of 56 00:02:27.439 --> 00:02:29.159 mine who was a pastor, and I asked him what, 57 00:02:29.319 --> 00:02:33.759 first of all, what I should do with dog? And well, 58 00:02:33.800 --> 00:02:36.280 clearly the dog is not redeemed. Put the dog down. 59 00:02:36.840 --> 00:02:38.560 And then I said, well, how do I dispose of 60 00:02:38.599 --> 00:02:41.560 a Bible? Is there like is there. 61 00:02:42.919 --> 00:02:44.439 Is there like some sort of turn it in? 62 00:02:44.759 --> 00:02:44.919 Well? 63 00:02:45.319 --> 00:02:49.960 Ak, is it like an American flag? Like there's there's 64 00:02:50.000 --> 00:02:53.879 procedures and policies and physicians. And he said, yeah, put 65 00:02:53.919 --> 00:02:56.879 it in the recycle bin. Was a jump dum. 66 00:02:58.159 --> 00:03:02.520 It's a book. So it was sitting on the couch. 67 00:03:02.599 --> 00:03:05.319 Huh. That does that mean that you regularly read it? 68 00:03:05.439 --> 00:03:08.080 My wife was reading it? Oh good, Yeah, that's good. 69 00:03:08.400 --> 00:03:10.479 At least there's one of you in that house that 70 00:03:11.479 --> 00:03:12.599 is God fearing and. 71 00:03:13.560 --> 00:03:16.599 I have a Bible. Yeah, but you don't read it. 72 00:03:17.439 --> 00:03:20.919 In the other room behind behind the gate. Why is 73 00:03:20.960 --> 00:03:22.199 the Bible behind the. 74 00:03:22.159 --> 00:03:24.800 Gate so that the dog doesn't chew it up. When's 75 00:03:24.800 --> 00:03:27.879 the last time you picked up your Bible? You mean 76 00:03:27.879 --> 00:03:30.080 the book or in my phone? I have an app 77 00:03:30.120 --> 00:03:30.599 on my phone. 78 00:03:30.639 --> 00:03:30.960 Oh you do? 79 00:03:31.400 --> 00:03:36.800 Okay, I bring up right there, Matthew fifteen is what 80 00:03:36.840 --> 00:03:43.439 it was. Oh okay. Some ps two Psalm two, Verse one, 81 00:03:44.360 --> 00:03:47.240 Why do the nations rage and the people's plot in vain? 82 00:03:47.520 --> 00:03:49.960 The kings of the earth set themselves and the rulers 83 00:03:50.000 --> 00:03:52.960 take counsel together against the Lord and against his anointed. 84 00:03:53.120 --> 00:03:55.159 That's exactly what I was thinking of when I was 85 00:03:55.159 --> 00:03:59.960 thinking about this Ai fight between well, China and everybody else. 86 00:04:00.560 --> 00:04:03.840 This because the kings of the world of Ai, they're 87 00:04:03.840 --> 00:04:05.599 doing what the Psalm says. 88 00:04:05.479 --> 00:04:10.599 Well, and they're doing a kind of behind curtains. I 89 00:04:10.639 --> 00:04:13.240 don't want to say behind closed doors, because there's some 90 00:04:13.280 --> 00:04:16.560 information that we know, but there's so little that we 91 00:04:16.680 --> 00:04:21.560 know about the generation of deep seek Ai. Why I 92 00:04:21.600 --> 00:04:24.439 mean this it is, it's not just chess. 93 00:04:24.600 --> 00:04:27.879 Here's China playing chess with our checkers. You know, Biden 94 00:04:28.079 --> 00:04:32.000 going out and you know, banning companies here in the 95 00:04:32.120 --> 00:04:36.120 United States from exporting AI chips to China was kind 96 00:04:36.120 --> 00:04:38.360 of like, we're not going to give you our super powerful, 97 00:04:38.480 --> 00:04:43.879 super expensive chips, and China's like f us, no f u, 98 00:04:44.199 --> 00:04:46.720 We'll do it on our own for cheaper. We don't 99 00:04:46.759 --> 00:04:49.720 need your chips. We look like fools in this, like 100 00:04:49.759 --> 00:04:53.639 we thought that everyone needed our super expensive chips to 101 00:04:53.759 --> 00:04:57.120 go and make their AI advances and that's not the case. 102 00:04:57.279 --> 00:04:59.920 We look like the dumb dums on the global state 103 00:05:00.279 --> 00:05:03.319 right now. And Trump is right that this should be 104 00:05:03.360 --> 00:05:04.920 a wake up call. 105 00:05:05.160 --> 00:05:07.720 And then all it takes is a Chinese mcguiver to go, 106 00:05:07.879 --> 00:05:11.360 oh well, I don't have the greatest, the fastest, the 107 00:05:11.360 --> 00:05:16.319 biggest chips. Give me a ballpoint pen, ballpoint pen, some 108 00:05:16.439 --> 00:05:17.839 duct tape, and then all. 109 00:05:17.920 --> 00:05:21.519 The trans transmission it used to be a transmission. Tranny 110 00:05:21.600 --> 00:05:24.279 was a word for a transmission before I do any other, 111 00:05:24.360 --> 00:05:25.480 meaning let's. 112 00:05:25.199 --> 00:05:28.399 Bring that back. So yeah, there's a lot that's that's 113 00:05:28.439 --> 00:05:30.720 going on. Also, I have a question that I just 114 00:05:30.759 --> 00:05:34.120 thought of, who is using this on a regular basis? 115 00:05:35.000 --> 00:05:38.040 Who uses chat GPT? 116 00:05:38.680 --> 00:05:42.040 I think ki, I think it's gonna be everyone's using 117 00:05:42.079 --> 00:05:43.839 it in the future. 118 00:05:43.600 --> 00:05:46.319 Right, Well, I mean because to me. Okay, so there's 119 00:05:46.319 --> 00:05:48.639 a couple there's a couple of things. This morning I 120 00:05:48.680 --> 00:05:53.560 realized Wall Street Journal now summarizes their articles. There's three 121 00:05:53.639 --> 00:05:56.000 bullet points, and I think a lot of news organizations 122 00:05:56.079 --> 00:05:58.839 do this now that I recognize it. Three bullet points 123 00:05:58.839 --> 00:06:00.639 at the top of a long article. You can read 124 00:06:00.639 --> 00:06:02.240 those three bullet points and get an idea of what 125 00:06:02.279 --> 00:06:04.000 the article is. And there's a little thing in the 126 00:06:04.000 --> 00:06:07.279 corner that says, what's this. I clicked on that this morning, 127 00:06:07.279 --> 00:06:11.560 and it says, this is an AI generated summary of 128 00:06:11.839 --> 00:06:14.279 the article that you're about to read. Well, no one's 129 00:06:14.319 --> 00:06:16.399 going to read it if AI already just summed it 130 00:06:16.480 --> 00:06:18.959 up in three bullet points. That's one thing. The other 131 00:06:19.000 --> 00:06:23.319 thing is Google, when you search for something now automatically 132 00:06:23.399 --> 00:06:28.680 gives you the AI version compendium of everything it's searched for. 133 00:06:29.199 --> 00:06:31.879 Apple is trying to get you to update all of 134 00:06:31.920 --> 00:06:37.000 your iOS and operating system stuff with its AI version. 135 00:06:37.480 --> 00:06:38.839 It jammed down your throat. 136 00:06:39.040 --> 00:06:44.879 I was shocked, and not so shocked with the quickness 137 00:06:44.879 --> 00:06:48.639 that I adopted Google's AI search results. When they first 138 00:06:48.680 --> 00:06:50.519 started doing that, I said, what the F is this? 139 00:06:50.959 --> 00:06:53.639 Screw this, I'm not paying attention to that. I would 140 00:06:53.639 --> 00:06:56.319 scroll down, I would click on the various different things 141 00:06:56.319 --> 00:06:59.879 that popped up to reach the summation that the AI 142 00:07:00.079 --> 00:07:02.000 bullet points would have given me right off the bat. 143 00:07:02.399 --> 00:07:05.879 So now, I mean it was probably a matter of 144 00:07:06.000 --> 00:07:08.879 days where I was putting up a strong front of 145 00:07:09.120 --> 00:07:11.839 I'm not using AI, I'm not using Google's AI, the 146 00:07:12.160 --> 00:07:14.639 AI bullet points are feeding me, I'm not using that. 147 00:07:15.040 --> 00:07:17.720 It was probably a matter of like eight to twelve 148 00:07:17.839 --> 00:07:21.399 days between that and me gobbling it all up. 149 00:07:22.079 --> 00:07:25.279 Can't stop it because it's easy and we're lazy people, Okay, 150 00:07:25.319 --> 00:07:30.560 but I mean, who's For example, this deepseak ai has 151 00:07:30.839 --> 00:07:33.519 skyrocketed over the last three or four days to the 152 00:07:33.560 --> 00:07:38.399 top of Apple's App Store downloads. Why what are people 153 00:07:38.639 --> 00:07:42.120 just curious about it? Because when chat gpt came out, 154 00:07:42.439 --> 00:07:46.199 people were like, make a poem about a monkey, hump 155 00:07:46.240 --> 00:07:47.120 and a deer, and. 156 00:07:47.079 --> 00:07:50.800 Make a poem about penis, and people went with that 157 00:07:50.879 --> 00:07:53.759 because it was a curiosity, right. I think a lot 158 00:07:53.800 --> 00:07:56.879 of people use this for work related purposes. 159 00:07:57.199 --> 00:08:02.000 Oh yeah, school, I will say this. I did ask 160 00:08:02.199 --> 00:08:07.000 chat gpt to write a radio play for our holiday 161 00:08:07.839 --> 00:08:10.319 show and it was awful. 162 00:08:10.839 --> 00:08:13.879 Okay, so here's something that you should know. We don't 163 00:08:13.879 --> 00:08:18.120 have real jobs, right, so we can't fathom how people 164 00:08:18.120 --> 00:08:22.439 would use this in a practical and time saving work 165 00:08:22.519 --> 00:08:23.720 smarter not harder manner. 166 00:08:23.759 --> 00:08:25.839 I'm hoping this is the I want to learn. I 167 00:08:25.879 --> 00:08:30.160 want if you use AI on a regular basis, whether 168 00:08:30.160 --> 00:08:32.759 it's Chat, GPT or deep seek or whatever it is, 169 00:08:33.080 --> 00:08:35.399 if you use it on a regular basis, I'd like 170 00:08:35.480 --> 00:08:37.080 to know how you can leave us a message on 171 00:08:37.120 --> 00:08:40.960 the talkback feature on the iHeart app. Wall Street is 172 00:08:41.519 --> 00:08:45.919 some you know, turning to some sense of normalcy after 173 00:08:46.000 --> 00:08:49.200 what happened yesterday when, for example, Nvideo, one of the 174 00:08:49.240 --> 00:08:55.480 big AI chip makers, lost more than five hundred billion 175 00:08:55.679 --> 00:08:59.879 dollars in market value in one day, five hundred bit. 176 00:09:00.120 --> 00:09:03.080 It had five hundred billion to lose, and it lost 177 00:09:03.120 --> 00:09:04.480 five hundred billion dollars. 178 00:09:04.639 --> 00:09:08.480 A couple bits of feedback. Look at David Muir's muscles, 179 00:09:08.759 --> 00:09:12.960 My goodness, is that necessary to put those out? 180 00:09:13.440 --> 00:09:16.799 It's a hot day in that helicopter. That's a G 181 00:09:16.960 --> 00:09:20.120 I Joe commercial, right there, wasn't it? That's exactly what 182 00:09:20.159 --> 00:09:21.559 it was. Okay. 183 00:09:22.519 --> 00:09:24.679 We've gotten a few bits of feedback with your dog 184 00:09:24.759 --> 00:09:25.639 eating the Bible. 185 00:09:26.679 --> 00:09:28.159 One bit of feedback. 186 00:09:27.720 --> 00:09:31.519 Is Gary's dog is metal as f okay. The other 187 00:09:31.840 --> 00:09:35.720 is that you should knock out five hail mary's that 188 00:09:35.759 --> 00:09:37.960 you did not protect the Bible, that the dog was 189 00:09:38.120 --> 00:09:39.000 doing dog things. 190 00:09:39.039 --> 00:09:41.200 But it's up to you to protect the Bible. So 191 00:09:41.679 --> 00:09:42.559 protect the Bible. 192 00:09:42.639 --> 00:09:44.559 So you're good for five hail marys, which I think 193 00:09:44.639 --> 00:09:47.240 is low. I mean I would go with, yeah, protect 194 00:09:47.240 --> 00:09:50.519 the Bible. I don't do hail mary's. That's n what's 195 00:09:50.519 --> 00:09:53.759 a Christian prayer? Just like our father? You do that one, 196 00:09:54.120 --> 00:09:56.799 but you do the our Father right, everyone does, sire? 197 00:09:56.879 --> 00:09:58.879 Yeah, everybody kind of does that one. All right, how 198 00:09:58.879 --> 00:10:01.120 about give me five ur our fathers will call it again. 199 00:10:01.200 --> 00:10:03.759 That's not how it works. It is no, no, no, 200 00:10:03.799 --> 00:10:06.279 you don't pay a toll to get across the bridge. 201 00:10:06.320 --> 00:10:09.360 Well no, you've already crossed the bridge and then burned 202 00:10:09.399 --> 00:10:13.519 it down, so you now have to pay the damage 203 00:10:13.559 --> 00:10:14.639 fee for the Bible. 204 00:10:15.799 --> 00:10:20.039 Yeah it is good morning, Hey, Ary. 205 00:10:20.440 --> 00:10:22.879 I know you, you know your upbringing was a little 206 00:10:22.960 --> 00:10:25.799 unorthodoxed and you know, some weird stuff, But I think 207 00:10:25.799 --> 00:10:28.879 it's all cool. But the fact that you guys aren't 208 00:10:28.879 --> 00:10:30.919 going to bury your mom next to your dad, I 209 00:10:30.960 --> 00:10:35.200 find that kind of strange. I think that's a no brainer. 210 00:10:35.799 --> 00:10:37.159 You know if I'm wrong, aren't you? 211 00:10:37.679 --> 00:10:40.360 But I can't see any other place. I think that's 212 00:10:40.399 --> 00:10:42.799 the most appropriate thing to do. 213 00:10:42.879 --> 00:10:45.639 Feel very strongly about people. He doesn't know that. Well. 214 00:10:45.799 --> 00:10:49.080 First of all, f you, you don't get to tell 215 00:10:49.120 --> 00:10:51.240 me where they're gonna They get to decide where they're 216 00:10:51.240 --> 00:10:53.399 going to be buried. And yes, they're going to be 217 00:10:53.440 --> 00:10:56.639 buried together. That's such a weird. 218 00:10:57.840 --> 00:10:59.960 I'm pretty sure my husband has it somewhere in writing 219 00:11:00.159 --> 00:11:01.679 to not be buried next time. 220 00:11:04.200 --> 00:11:07.720 Don't care what you do in life separated and give 221 00:11:07.759 --> 00:11:13.000 me some f and pece around here. That's great. Good morning, Gary. 222 00:11:13.120 --> 00:11:15.240 Hey, did you ever get rid of that squatter in 223 00:11:15.320 --> 00:11:16.279 your parents' house? 224 00:11:16.360 --> 00:11:18.919 Yes, yes, that was That was one of the things 225 00:11:18.960 --> 00:11:20.759 you had to take care of before the burial. 226 00:11:21.759 --> 00:11:24.039 They would not have wanted us to have that around. 227 00:11:24.120 --> 00:11:26.360 Hi, Gary, I love you both. 228 00:11:26.519 --> 00:11:30.320 Hey. I am a nurse case manager for an insurance company. 229 00:11:30.720 --> 00:11:33.720 I work in their workers' comp division, and we use 230 00:11:33.799 --> 00:11:38.960 AI every day to kind of as a collaborative partner 231 00:11:39.080 --> 00:11:45.159 for medical reports, looking at diagnostics and anticipating what type 232 00:11:45.159 --> 00:11:48.840 of care a patient or an injured employee might need. 233 00:11:49.320 --> 00:11:51.759 Okay, talk to you so well, that's good. I mean 234 00:11:51.759 --> 00:11:54.679 that makes sense. We're going to talk in a couple 235 00:11:54.679 --> 00:11:57.440 of minutes here about what Deep seek Ai is, why 236 00:11:57.519 --> 00:12:00.240 it caused such an earthquake on Wall Street yesterday. But 237 00:12:00.600 --> 00:12:03.000 this was a comment from President Trump just a few 238 00:12:03.039 --> 00:12:03.399 days ago. 239 00:12:03.559 --> 00:12:08.759 The release of Deepseek Ai from a Chinese company should 240 00:12:08.759 --> 00:12:11.200 be a wake up call for our industries that we 241 00:12:11.279 --> 00:12:14.519 need to be laser focused on competing to win. Because 242 00:12:14.559 --> 00:12:16.919 we have the greatest scientists in the world. We always 243 00:12:16.919 --> 00:12:18.399 have the ideas we're always first. 244 00:12:18.919 --> 00:12:22.320 We may always be first, But China did it cheaper. 245 00:12:22.440 --> 00:12:24.799 It's time for us to stop thinking that way, because 246 00:12:24.799 --> 00:12:26.399 it's just not the case anymore. 247 00:12:27.000 --> 00:12:28.519 I don't think it's the case anymore. 248 00:12:28.840 --> 00:12:32.039 If this is real, if China jumps out and says 249 00:12:32.039 --> 00:12:34.679 we can do it for cheaper, or we thought that 250 00:12:34.759 --> 00:12:38.720 we held the key to the expensive computer chips that 251 00:12:38.759 --> 00:12:42.559 make this all possible. I mean, as I mentioned, the 252 00:12:42.600 --> 00:12:46.879 Biden administration band companies from sending these powerful AI chips 253 00:12:46.879 --> 00:12:49.519 to China. So we must have thought that we were 254 00:12:49.639 --> 00:12:50.919 the big kid on the block. 255 00:12:51.000 --> 00:12:55.360 We're not. If they're able to do this, you raise 256 00:12:55.399 --> 00:12:59.360 a great question though, if this is real. Remember this 257 00:12:59.440 --> 00:13:02.279 is a China needs product, and with that comes a 258 00:13:02.320 --> 00:13:05.679 certain amount of mystery about its background. About its origins, 259 00:13:05.799 --> 00:13:09.799 about how it was actually created, who funds it, and 260 00:13:09.840 --> 00:13:12.120 what kind of information is it going to take when 261 00:13:12.240 --> 00:13:16.159 you use it? Where does that information go? Yes, echoes 262 00:13:16.200 --> 00:13:16.799 of TikTok. 263 00:13:16.919 --> 00:13:19.679 They have shown some of their math. They did publish 264 00:13:19.720 --> 00:13:23.879 a technical paper last week about exactly how they were 265 00:13:23.879 --> 00:13:27.320 able to do this PS. They're making it available for 266 00:13:27.480 --> 00:13:29.600 other people to get in there and tweak it or 267 00:13:29.720 --> 00:13:30.120 change it. 268 00:13:30.399 --> 00:13:32.639 And here's why. They want as much. 269 00:13:32.480 --> 00:13:35.399 Information about us or you or the smartest of the 270 00:13:35.440 --> 00:13:37.399 smart that they can get. And if you get into 271 00:13:37.440 --> 00:13:40.080 their system and you start tinkering around, guess what they're 272 00:13:40.120 --> 00:13:44.360 tinkering with all of your stuff. It looks like most Americans, 273 00:13:44.600 --> 00:13:47.879 more Americans, I should say, are working two jobs. Rising 274 00:13:48.000 --> 00:13:51.440 caused a strong side market. Do you have a side gig? 275 00:13:51.480 --> 00:13:54.200 A side market? Are you selling feet picks? What's going on? 276 00:13:54.840 --> 00:13:56.120 There's no judgment here. 277 00:13:56.200 --> 00:13:58.240 Yeah, they would be total judgment. No, there wouldn't. 278 00:13:58.639 --> 00:14:00.559 Hey, if your feed are nice and have to send 279 00:14:00.600 --> 00:14:03.639 feet picks, I mean I would have to pay people 280 00:14:03.720 --> 00:14:04.799 to look at my feet. 281 00:14:04.840 --> 00:14:10.279 They're so gross. Be careful what you'd say. People are 282 00:14:10.279 --> 00:14:15.960 into gross feet people lid for every pot. Wow, that's 283 00:14:16.039 --> 00:14:18.480 really bad. I never thought about that. An eyeball for 284 00:14:18.519 --> 00:14:19.120 every corn. 285 00:14:19.879 --> 00:14:21.600 Also, at the top of the hour, we're going to 286 00:14:21.639 --> 00:14:23.120 tak an eyeball for every corn. 287 00:14:23.200 --> 00:14:24.399 Yeah, the corns on your feet. 288 00:14:24.440 --> 00:14:29.320 Oh oh, I don't have corns bunions, No, but look 289 00:14:29.519 --> 00:14:34.919 co fungus. Look, oh god, they're not great. 290 00:14:36.159 --> 00:14:43.200 Ah here, wipe off the You're gonna throw a bare 291 00:14:43.240 --> 00:14:47.879 foot up there. Also at the top of the hour, 292 00:14:49.159 --> 00:14:51.519 the creek fire. If you remember that from twenty seventeen, 293 00:14:51.600 --> 00:14:54.159 so cal Edison said, hey, it wasn't our power lines. 294 00:14:54.759 --> 00:14:57.919 The FEDS think so cal Edison may be lying about that. 295 00:14:58.000 --> 00:14:59.840 Of course you think of that in the context of 296 00:15:00.039 --> 00:15:02.960 the eating fire, and now people pointing their fingers towards 297 00:15:03.000 --> 00:15:04.960 the utility for that one as well. So we'll talk 298 00:15:04.960 --> 00:15:05.679 about that coming up. 299 00:15:05.879 --> 00:15:09.320 A game changing move that does not come from Open Ai, 300 00:15:09.559 --> 00:15:12.960 Google or Meta. There is a new model that has 301 00:15:13.159 --> 00:15:14.919 all of the valley buzzing, but. 302 00:15:14.960 --> 00:15:17.799 From a Chinese lab called deep Sea. 303 00:15:17.919 --> 00:15:20.440 It's opened a lot of eyes of like what is 304 00:15:20.480 --> 00:15:22.519 actually happening in ai in China. 305 00:15:23.639 --> 00:15:27.039 Deep Seek is a Chinese AI startup. It was founded 306 00:15:27.080 --> 00:15:30.879 in twenty twenty three by Langwin Fang. He's the co 307 00:15:30.919 --> 00:15:34.559 founder of a hedge fund high flyer. High Flyer uses 308 00:15:34.600 --> 00:15:37.840 AI algorithms to predict market swings, so you can see 309 00:15:37.840 --> 00:15:40.200 why he's got stock in the game, right. If you're 310 00:15:40.200 --> 00:15:43.240 going to use AI to protict market swings and make 311 00:15:43.279 --> 00:15:46.080 an s ton of money, not for yourself alone, but 312 00:15:46.440 --> 00:15:50.559 for everybody else, you want your AI to be tip top. 313 00:15:50.879 --> 00:15:52.919 You want it to be accessible and cheap. So that's 314 00:15:52.960 --> 00:15:54.039 what he set out to do. 315 00:15:55.200 --> 00:15:59.440 They claim that it's AI software is competitive with many 316 00:15:59.559 --> 00:16:02.720 main made by US tech giants like open Ai and 317 00:16:02.759 --> 00:16:09.080 Google at a fraction of the cost. Last week, Deep 318 00:16:09.120 --> 00:16:13.960 Seek did publish the technical infrastructure, which talks about the 319 00:16:14.000 --> 00:16:18.399 AI model behind the app. It claims cost savings from 320 00:16:18.480 --> 00:16:22.320 using far fewer and less advanced computer chips than advanced 321 00:16:22.360 --> 00:16:25.200 AI projects usually do, like the ones we do. It's 322 00:16:25.279 --> 00:16:28.279 punitively expensive here in the United States because we've thought 323 00:16:28.759 --> 00:16:32.000 all along that we need to use these advanced computer 324 00:16:32.120 --> 00:16:34.240 chips that are more expensive. 325 00:16:34.679 --> 00:16:38.639 Turns out you don't. That is the biggest issue. 326 00:16:38.960 --> 00:16:42.080 That's the biggest gift though to us as well. If 327 00:16:42.120 --> 00:16:46.720 we can harness that and create our own AI infrastructure cheaper, 328 00:16:47.159 --> 00:16:48.720 it's a win for everybody, isn't it. 329 00:16:49.080 --> 00:16:52.840 Chaitan Puta Gunta is a general partner at Benchmark, and 330 00:16:52.879 --> 00:16:55.639 I heard him on CNN explaining, we still don't know 331 00:16:55.679 --> 00:16:57.399 a lot. I mean, as much as we do know 332 00:16:57.440 --> 00:17:00.000 about deep Seek, there's plenty other questions. 333 00:17:00.080 --> 00:17:03.559 So did they actually assemble this talent, how did they 334 00:17:03.840 --> 00:17:06.200 assemble all the hardware? How did they assemble the data 335 00:17:06.279 --> 00:17:08.279 to do all this? We don't know, and it's never 336 00:17:08.319 --> 00:17:10.559 been publicized. And hopefully we can learn that. 337 00:17:10.640 --> 00:17:13.640 Yeah, hopefully we can learn that. So here's a terrifying thing. 338 00:17:13.759 --> 00:17:17.160 Alexander Wong is a founder of Scale Ai, and he 339 00:17:17.240 --> 00:17:26.000 actually that company is helps AI systems learn. It kind 340 00:17:26.000 --> 00:17:29.920 of puts together the information that AI then dives into 341 00:17:29.960 --> 00:17:34.039 to learn from. And this is a terrifying aspect of 342 00:17:34.680 --> 00:17:36.880 There are two parts of his answer here that are terrifying. 343 00:17:36.960 --> 00:17:41.440 Today we released Humanity's Last Exam, which is a new 344 00:17:41.519 --> 00:17:44.799 evaluation or benchmark of AI models that we produced by 345 00:17:45.160 --> 00:17:50.839 getting math, physics, biology, chemistry professors to provide the hardest 346 00:17:50.880 --> 00:17:53.559 questions they could possibly imagine. Deep Seek, which is the 347 00:17:53.640 --> 00:17:58.240 leading Chinese AI lab. Their model is actually the top 348 00:17:58.279 --> 00:18:01.519 performing were roughly on par with the best American males. 349 00:18:02.000 --> 00:18:04.480 And again I mean that point of it's using it 350 00:18:04.519 --> 00:18:08.839 with just a tiny percentage of the money that we've 351 00:18:08.839 --> 00:18:12.960 put into AI with chips that we I mean, the 352 00:18:13.519 --> 00:18:17.839 lower quality, slower chips that we have been thinking, Will 353 00:18:17.839 --> 00:18:19.839 they never be able to put together an AI system 354 00:18:19.880 --> 00:18:22.599 if they use those low quality chips? Can I can? 355 00:18:22.640 --> 00:18:25.400 I just want to point something out here, he referred to. 356 00:18:26.079 --> 00:18:32.319 He referred to something called Humanity's Last Exam. This is 357 00:18:32.599 --> 00:18:38.079 a this is a test that they put forth for 358 00:18:38.160 --> 00:18:43.480 these different AI models and the Center for AI Safety. 359 00:18:43.559 --> 00:18:45.960 It's a test that they say as a groundbreaking AR 360 00:18:46.119 --> 00:18:49.200 benchmark designed to test the limits of AI knowledge at 361 00:18:49.200 --> 00:18:52.440 the frontiers of human expertise. And as he mentioned, they 362 00:18:52.519 --> 00:18:54.599 just ask some of the smartest people in the world 363 00:18:54.920 --> 00:18:58.200 come up with the absolute hardest questions that you can, 364 00:18:58.839 --> 00:19:01.200 some of which may not act actually have answers that 365 00:19:01.240 --> 00:19:03.920 we know, and see if AI can do anything with it. 366 00:19:03.960 --> 00:19:06.839 I'm going to give you one example of the kind 367 00:19:06.880 --> 00:19:10.480 of question that they ask AI to come up with 368 00:19:10.599 --> 00:19:17.279 an answer for question one. Hummingbirds within a podiformis uniquely 369 00:19:17.319 --> 00:19:22.599 have a bilaterally paired oval bone a sesamoid embedded in 370 00:19:22.680 --> 00:19:27.880 the quadrilateral portion of the expanded curetiate upeneurosis of insertion 371 00:19:28.119 --> 00:19:33.240 of M depressor kaudia. How many paired tendons are supported 372 00:19:33.279 --> 00:19:38.680 by this sesamoid bone answer with a number twenty three. 373 00:19:41.440 --> 00:19:42.799 I don't know if that's true. I don't know if 374 00:19:42.839 --> 00:19:43.519 that I. 375 00:19:43.519 --> 00:19:47.799 Do deep Seek in terms of if it's safe, by 376 00:19:47.839 --> 00:19:50.680 the way, that's very convoluted that question, like yay for 377 00:19:50.799 --> 00:19:53.119 their AI if they can. They also ask about you 378 00:19:53.119 --> 00:19:54.240 should know the answer. 379 00:19:53.920 --> 00:19:58.680 To biblical translations from a Hebrew in Greek mythology, who 380 00:19:58.720 --> 00:20:02.640 was Jason's maternals, Jason's maternal great grand. 381 00:20:02.440 --> 00:20:05.640 Thoughts, So it's got a wide breadth of knowledge there. 382 00:20:06.200 --> 00:20:08.640 Deep Seek, by the way, in terms of safety, can 383 00:20:08.759 --> 00:20:12.400 collect and store a massive amount of personal information, including 384 00:20:12.720 --> 00:20:17.240 any conversations you have, technical information about things like your device, 385 00:20:17.440 --> 00:20:21.200 your Internet connection. This data can be accessed by the 386 00:20:21.279 --> 00:20:22.279 Chinese government. 387 00:20:22.720 --> 00:20:23.079 Okay. 388 00:20:24.440 --> 00:20:27.880 The app also appears to be censoring certain information like 389 00:20:28.000 --> 00:20:31.680 I don't know Tienneman Square and the protests there. Yeah, 390 00:20:31.799 --> 00:20:35.720 right in the line of what China does now. Deep 391 00:20:35.759 --> 00:20:38.519 Seek has released a version of its AI model for 392 00:20:38.599 --> 00:20:42.799 others to use and modify. Outside developers can make it 393 00:20:42.839 --> 00:20:45.920 more secure. Both more secure and less restricted by running 394 00:20:45.920 --> 00:20:50.440 it locally on a powerful enough device. But yeah, the 395 00:20:50.480 --> 00:20:52.920 fact that they've made it freely available for others to 396 00:20:53.000 --> 00:20:56.319 download and modify means that they just want more people's 397 00:20:56.400 --> 00:20:58.279 information in its clutches. 398 00:20:58.960 --> 00:21:02.440 Actually just USEPT in my car this morning. 399 00:21:02.519 --> 00:21:04.720 I'm on the way to go do some sales calls, and. 400 00:21:04.680 --> 00:21:07.200 I bounced some ideas off of chat GPT as far 401 00:21:07.240 --> 00:21:09.480