1 00:00:01,839 --> 00:00:09,040 You'll own no IP at all and you'll be 2 00:00:05,359 --> 00:00:10,719 happy. I will be sad. What? Nope. Not 3 00:00:09,040 --> 00:00:15,200 allowed. That's okay. It's the rules. The AI will 4 00:00:12,480 --> 00:00:18,960 stop you from being sad. How could you be sad when you live in a utopia where 5 00:00:17,039 --> 00:00:22,560 everything's controlled and oh, everything's generated? 6 00:00:21,199 --> 00:00:28,000 What are we talking about here, Luke? It's powered by the the IP grinding 7 00:00:25,039 --> 00:00:31,439 machine. We're talking about uh IP. We're talking about AI policy. We're 8 00:00:29,840 --> 00:00:35,840 talking about how this stuff is going to possibly work because there's a huge 9 00:00:33,440 --> 00:00:41,760 battle going on between North American and European AI companies and mostly 10 00:00:39,520 --> 00:00:45,360 Chinese AI companies. Deepseek came and took everyone's lunch and what are we 11 00:00:43,520 --> 00:00:49,680 going to do about it? You you uh it's like you host a stream or something. 12 00:00:46,879 --> 00:00:54,160 You've done this before. Maybe we haven't done a talk linked in a in a hot 13 00:00:51,680 --> 00:00:59,600 second but and you've never been on. No. I'm so excited. Yeah. And also get you 14 00:00:56,640 --> 00:01:04,360 get to talk about AI. Wait, I just when Lionus doesn't let you. No. On when he 15 00:01:01,840 --> 00:01:09,360 whips me. As soon as you get to a AI topic, and I know this because I prepare 16 00:01:07,040 --> 00:01:12,479 the topic sometimes. Yeah. I'm like, "Oh, they're going to like this topic." 17 00:01:10,799 --> 00:01:15,920 I stopped preparing them because he just gets there and he goes on his phone. 18 00:01:14,159 --> 00:01:18,880 Yeah. He has literally done that. And then he'll get at me for not paying 19 00:01:17,200 --> 00:01:22,880 attention during other topics. But whatever, man. Yeah. But now you get to 20 00:01:20,960 --> 00:01:30,000 talk about it. So the the the kind of impetus for all of this was Jack Dorsey, 21 00:01:26,080 --> 00:01:32,560 the founder, former CEO of Twitter. 22 00:01:30,000 --> 00:01:39,600 Yeah. Now X, which I still call Twitter. Yeah. Uh tweeting delete all IP law on 23 00:01:36,640 --> 00:01:43,840 April 11th. And it kind of sparked some discourse. Yeah. That's that's part of 24 00:01:41,360 --> 00:01:49,840 what in my opinion are like the three core topics around AI policy right now, 25 00:01:47,360 --> 00:01:54,000 which I laid out as open versus closed models, copyright and IP reform, and AI 26 00:01:52,320 --> 00:01:59,040 safety versus innovation speed. Oh, heck. And I think that's like one of the 27 00:01:56,720 --> 00:02:04,399 really core things is what do we do with IP law? Because when you're looking at 28 00:02:01,280 --> 00:02:06,439 the the race for AI stuff, it's becoming 29 00:02:04,399 --> 00:02:10,160 more and more like country or alliance-based instead of company based. 30 00:02:08,959 --> 00:02:15,760 So a lot of people are looking at it like okay the west versus the east I 31 00:02:13,120 --> 00:02:21,360 guess or right now basically deepse v everyone 32 00:02:18,040 --> 00:02:26,080 and there are differences in how the 33 00:02:21,360 --> 00:02:29,599 west and China deal with IP law. Do I 34 00:02:26,080 --> 00:02:31,599 understand Chinese IP law? Not even sort 35 00:02:29,599 --> 00:02:34,560 of. Yeah. Yeah. I kind of started to Google it in preparation for this and 36 00:02:33,120 --> 00:02:40,080 I'm kind of like oh this is going to be too too complicated. What my not enough 37 00:02:37,599 --> 00:02:44,640 Googling has told me uh is that they are much more lax on IP law especially when 38 00:02:42,879 --> 00:02:49,360 it comes to outside of their own country but also within their own country and 39 00:02:47,360 --> 00:02:53,280 especially especially when it has to do with anything that could benefit the 40 00:02:50,640 --> 00:02:58,000 state right because over here you know the the North American kind of 41 00:02:55,120 --> 00:03:01,840 capitalist mindset corpo focused right it's like let people go out there 42 00:02:59,840 --> 00:03:05,200 struggle whoever rises to the top is good and China also does that I know 43 00:03:03,440 --> 00:03:11,440 they do that with like but but it's more like it's they the government directs these little 44 00:03:08,400 --> 00:03:13,440 experiments to happen. Um so if 45 00:03:11,440 --> 00:03:18,519 something is like oh this will be beneficial to the nation then they can 46 00:03:15,599 --> 00:03:25,200 just kind of override yeah the ide which which when you're in a race 47 00:03:20,879 --> 00:03:27,519 like this could be very beneficial when 48 00:03:25,200 --> 00:03:32,239 the other side the west is like squabbbling over are we even allowing 49 00:03:30,159 --> 00:03:37,519 these companies access this data at all which I I think what a lot of the 50 00:03:34,159 --> 00:03:40,720 argument hinges on right is like the 51 00:03:37,519 --> 00:03:43,440 morality around letting these gigantic 52 00:03:40,720 --> 00:03:48,519 companies that don't care about you and want to loot you for everything you have 53 00:03:45,599 --> 00:03:52,959 take all of your IP as well or lose the whole race and have it not 54 00:03:51,200 --> 00:03:57,360 matter anyways because a company in another country that's just going to do 55 00:03:54,319 --> 00:03:59,760 that anyways ends up winning. Yeah. So, 56 00:03:57,360 --> 00:04:03,040 so for me tough argument for me the there's kind of two well there might be 57 00:04:01,280 --> 00:04:08,720 more than two but the two perspectives that are kind of popping into my head 58 00:04:04,640 --> 00:04:12,000 here are the like philosophical argument 59 00:04:08,720 --> 00:04:14,239 for AI copyright like as a society just 60 00:04:12,000 --> 00:04:21,199 regardless of like you know where China is at do we think that AI created stuff 61 00:04:18,320 --> 00:04:25,120 deserves copyright do we think that AI a model could count as the author of 62 00:04:23,360 --> 00:04:29,040 something there's all those questions and then there's the kind of more 63 00:04:26,759 --> 00:04:35,680 utilitarian like practical question of like is this what we have to do in order 64 00:04:32,080 --> 00:04:38,000 to keep up in the nuclear AI arms race 65 00:04:35,680 --> 00:04:40,919 with China. So which of those is more interesting to you? Well, I think right 66 00:04:39,759 --> 00:04:49,880 now the can an AI have ownership over 67 00:04:44,639 --> 00:04:52,560 something that it creates thing. Um is 68 00:04:49,880 --> 00:04:58,000 is technically somewhat decided in law right now and it's no. Right. Well, can 69 00:04:55,199 --> 00:05:02,479 can an AI model count as the author? Yeah, I think no. Right now, it's no. 70 00:05:00,160 --> 00:05:06,880 Yeah, the copyright office has ruled a couple times, I think, at least. Yeah. 71 00:05:05,040 --> 00:05:11,440 About uh as far as go, there's there's multiple presidents here. I think there's one guy in particular, we'll 72 00:05:09,759 --> 00:05:16,160 look it up later maybe. But like there's one guy in particular who has submitted 73 00:05:13,039 --> 00:05:18,000 multiple times to be like, okay, this is 74 00:05:16,160 --> 00:05:22,880 definitely like there's one for sure, guys. But I used an AI this time and 75 00:05:20,800 --> 00:05:26,039 this time this AI is sentient for sure, guys. and and the copyright office is 76 00:05:24,479 --> 00:05:31,360 like, "Stop coming back here and put on some pants." So, there's 77 00:05:28,720 --> 00:05:35,600 precedent against that dude. Yeah. Yeah, that makes sense. But, I mean, I think 78 00:05:32,960 --> 00:05:40,320 that the the like, do do we get to keep copyright in an era where all these 79 00:05:38,560 --> 00:05:44,560 companies are just kind of gobbling everything up anyway? Like that those 80 00:05:42,960 --> 00:05:49,160 are the interesting lawsuits that are moving through. The big one being OpenAI 81 00:05:46,479 --> 00:05:54,880 versus the New York Times. Yeah. Uh, it's so like CHAGBT has been out 82 00:05:52,720 --> 00:05:59,320 since 2022 and obviously transformer-based LLMs. They were out 83 00:05:57,520 --> 00:06:04,800 for a while. I mean, they were in research scenarios, but attention is all 84 00:06:02,160 --> 00:06:09,120 you need. It's been Yeah, it's 2017. Yeah, I've done research. I can't 85 00:06:07,360 --> 00:06:13,199 believe that this question is still kind of as open as it is. I would have 86 00:06:11,199 --> 00:06:18,560 expected there to be some precedent set and a ruling set by by now, but we're 87 00:06:16,160 --> 00:06:23,840 still just kind of like, you know, Meta just is also going through proceedings 88 00:06:20,720 --> 00:06:26,560 where they're like, you you literally 89 00:06:23,840 --> 00:06:30,720 torrented like every book ever. I mean, not not, you know, but like a massive 90 00:06:28,800 --> 00:06:34,160 archive of books. Yeah. Yeah. And like that's the that's the and this I'm not 91 00:06:32,880 --> 00:06:40,720 the first person to reference this at all, but u one of the co-founders I 92 00:06:37,280 --> 00:06:42,080 think of Reddit Mhm. Alex was Alexian. 93 00:06:40,720 --> 00:06:47,039 Yeah, I don't remember the exact details of this story, but I believe he was 94 00:06:43,840 --> 00:06:48,880 pirating um like research papers and 95 00:06:47,039 --> 00:06:53,440 then making them available to people for free, I think. Yeah. And then he was 96 00:06:51,440 --> 00:06:58,319 going to get thrown in jail forever and ended up 97 00:06:55,400 --> 00:07:01,759 unaliving himself. Oh. Over it because it like ruined his whole life. Wait, 98 00:07:00,160 --> 00:07:05,199 Aaron Schwarz? Are you talking about Aaron Schwarz? Oh, okay. Yeah. Yeah. 99 00:07:03,280 --> 00:07:08,960 Yeah. Yeah. Yeah. No, I'm familiar. Yeah. Yeah. And then now a company's 100 00:07:07,120 --> 00:07:13,840 doing it on a much grander scale and we're all like, "Ah, you know what? 101 00:07:11,840 --> 00:07:18,639 Seems fine." That totally reminds me. I saw an article recently like referencing 102 00:07:15,919 --> 00:07:24,800 Aaron Schwarz and like how bonkers it is that it's like actually kind of similar. 103 00:07:22,160 --> 00:07:29,520 Well, it it's it's crazy that that was the case just a few years ago. Yeah. And 104 00:07:27,840 --> 00:07:34,960 really not that long ago. He was his entire life was going to be ruined. Yep. 105 00:07:31,759 --> 00:07:38,720 And now we have every tech company ever 106 00:07:34,960 --> 00:07:40,160 is just gobbling up. I mean, I had as 107 00:07:38,720 --> 00:07:47,280 one of the points I wanted to talk about these like AI web crawlers that are 108 00:07:44,240 --> 00:07:48,880 taking up tons of bandwidth. Oh yeah. I 109 00:07:47,280 --> 00:07:53,520 we've referenced it a couple times on TechLink, but there are like a number of 110 00:07:51,280 --> 00:07:57,120 studies now and like reports from administrators being like, "Yeah, we 111 00:07:55,120 --> 00:08:01,599 looked into it and it's like 80% of our web traffic is just AI crawlers." So 112 00:07:59,759 --> 00:08:07,919 regardless, we can I don't want to bash on AI completely because I feel like 113 00:08:04,479 --> 00:08:12,080 there's an argument here and like for 114 00:08:07,919 --> 00:08:15,120 for AI for embedding it in society at 115 00:08:12,080 --> 00:08:19,759 every level for even maybe deleting all 116 00:08:15,120 --> 00:08:22,160 IP law and the argument would be is if 117 00:08:19,759 --> 00:08:28,000 it's actually if it if it actually leads to this scenario that Jack Dorsey and 118 00:08:25,919 --> 00:08:34,479 apparently Elon Musk and whoever because Elon Musk tweeted did as well. I agree 119 00:08:30,720 --> 00:08:36,719 to delete all IP law which is yeah uh if 120 00:08:34,479 --> 00:08:39,279 the argument by uh putting being put forward by them that like okay this is 121 00:08:38,159 --> 00:08:44,560 actually going to help us it's actually going to lead to a situation where we 122 00:08:41,560 --> 00:08:46,720 benefit do we like if you could know 123 00:08:44,560 --> 00:08:51,360 deleting all IP law will lead to a kind of a utopian society where you don't 124 00:08:48,640 --> 00:08:54,160 need money. It feels like, you know, there's a bunch of stories that are like 125 00:08:52,560 --> 00:08:59,360 this where you have utopian society and then like in the basement of city hall 126 00:08:56,160 --> 00:09:01,519 there's just like a a terrible demon 127 00:08:59,360 --> 00:09:05,920 machine that is like the reason why everything's okay. I feel like this is 128 00:09:03,279 --> 00:09:10,160 one of those scenarios. I I don't think um and I'm decently confident this is 129 00:09:08,320 --> 00:09:15,200 the route it's going to go to. I I I don't think deleting IP law is like the 130 00:09:13,120 --> 00:09:19,680 way to go. I I feel like what they're trying to do is like gambit an 131 00:09:17,640 --> 00:09:24,720 overcorrection. So, they're trying to be like, "Let's delete everything," hoping 132 00:09:22,000 --> 00:09:28,399 that they'll end up somewhere 50% of the way, if that makes sense. Because I I 133 00:09:26,560 --> 00:09:31,600 don't think they think that's going to happen either, to be completely honest. 134 00:09:30,399 --> 00:09:36,720 But I think they're trying to do the like Trumpian move of like I'm going to 135 00:09:34,800 --> 00:09:40,720 ask for this and then I'll send end up somewhere here. You know, we'll delete 136 00:09:38,640 --> 00:09:45,279 half of IP law. Yeah. Well, I think it's going to be IP reform. Uh because like a 137 00:09:43,200 --> 00:09:51,680 lot of the laws that were made weren't even really made uh in the internet age, 138 00:09:48,399 --> 00:09:54,560 let alone let alone the the like AI 139 00:09:51,680 --> 00:09:58,959 webcrawwler crazy situation we are now. So it's likely a lot of modernization 140 00:09:56,880 --> 00:10:03,440 and reform that needs to happen. Uh I mean we even have this problem with like 141 00:10:00,720 --> 00:10:09,760 YouTube videos. Um like there's this huge question mark of like can I have a 142 00:10:06,160 --> 00:10:12,880 clip from a movie in my video? Yeah. 143 00:10:09,760 --> 00:10:15,279 Like maybe. But also maybe not. Yeah. 144 00:10:12,880 --> 00:10:18,800 I've been scared to put anything. Yeah. That could go over the line. But then 145 00:10:16,720 --> 00:10:22,720 like all short form content seems to just be able to use whatever music they 146 00:10:20,240 --> 00:10:27,360 want. Yeah. And like that seems fine. That's completely different, Luke. 147 00:10:24,560 --> 00:10:31,839 That's totally different. One is watched by young people. And the other they 148 00:10:30,399 --> 00:10:36,560 don't have money. Why do they matter? They don't. They're too dumb. They're young and dumb. They don't know about 149 00:10:34,399 --> 00:10:41,200 copyright. Yeah. Elijah, one of the parallels that I would draw is to 150 00:10:38,800 --> 00:10:44,720 patents. And it's not perfect, but it's it's an interesting argument because 151 00:10:42,959 --> 00:10:48,560 there's a huge debate on whether or not patents have driven or stifled 152 00:10:46,480 --> 00:10:54,320 innovation in the past. And I think it really depends on 153 00:10:50,720 --> 00:10:55,920 the time, the country, uh the the field 154 00:10:54,320 --> 00:10:58,880 of innovation. Like it it depends on tons of stuff. A lot of people, this is 155 00:10:57,680 --> 00:11:04,800 the Wright brothers conversation I warned you about. A lot of people look at the Wright brothers and are like, "Oh 156 00:11:02,959 --> 00:11:09,000 my god, they brought in a this is amazing, right? We we didn't think we 157 00:11:06,720 --> 00:11:14,800 were going to fly for a long time." And then the 158 00:11:11,160 --> 00:11:17,600 air sick immediately turned into patent 159 00:11:14,800 --> 00:11:21,279 trolls as far as I could be wrong about this. A little bit of online research 160 00:11:19,279 --> 00:11:26,240 led me to find that they got their patent and then just 161 00:11:24,480 --> 00:11:30,959 just wrecked everyone else as much as they could and like very close to 162 00:11:28,480 --> 00:11:34,320 stopped innovating. Oh jeez. They they like got aviation off the ground and 163 00:11:32,640 --> 00:11:40,240 then just tried to make money off of what they did. Did you get this info 164 00:11:36,399 --> 00:11:43,680 from Grock? No. Okay. But there's also 165 00:11:40,240 --> 00:11:46,000 like in my opinion the which I don't 166 00:11:43,680 --> 00:11:53,360 think is as common anymore, but being an inventor as your trade as like what you 167 00:11:49,480 --> 00:11:55,440 do probably became a thing because of 168 00:11:53,360 --> 00:11:59,360 things like patents and IP law. Oh yeah, he's an inventory. Like no one has 169 00:11:57,120 --> 00:12:05,600 thought of, you know, an automatic egg making machine or something. But like 170 00:12:02,320 --> 00:12:08,320 it's a chicken. At one point IP law, 171 00:12:05,600 --> 00:12:12,480 copyright law, patent law, whatever. uh spurred innovation and if we get rid of 172 00:12:10,240 --> 00:12:15,480 it, you're worried that we would lose. Well, but there's but there's sto 173 00:12:13,600 --> 00:12:21,279 there's stories in both sides because like let's come back to more IP specific 174 00:12:18,480 --> 00:12:26,160 law, right? Um Disney kind of owned children's entertainment for a long 175 00:12:23,120 --> 00:12:29,120 time, most of our lives and a lot of 176 00:12:26,160 --> 00:12:33,839 before it. Um and now they're losing but they're not losing because you can use 177 00:12:31,279 --> 00:12:37,519 Mickey Mouse in other content now because that is a thing that happened. 178 00:12:35,519 --> 00:12:41,120 That was like a tiny splash. A bunch of games all came out all at once right 179 00:12:39,279 --> 00:12:44,800 when that got lifted that all featured Mickey Mouse and none of them were good. 180 00:12:43,120 --> 00:12:48,399 Everyone stopped caring almost immediately. No one talks about it 181 00:12:46,160 --> 00:12:52,240 anymore at all. And Bluey is just completely taking their life. Heck yeah. 182 00:12:50,480 --> 00:12:56,800 Have you watched Blueie? I have seen some Bluey. Blue is freaking It makes me 183 00:12:54,880 --> 00:13:02,480 cry. They like lost their ability to protect their IP on Mickey, right? And 184 00:12:59,360 --> 00:13:05,519 it didn't matter. And that's why the MCU 185 00:13:02,480 --> 00:13:07,519 didn't is isn't as good anymore cuz they 186 00:13:05,519 --> 00:13:13,200 they're losing their because they stop they didn't stop uh Steamboat Willie 187 00:13:10,320 --> 00:13:17,200 from escaping. He was going to join the Avengers. 188 00:13:15,360 --> 00:13:22,880 Steamboat Willie was Yeah, honestly I would go watch that. I would I haven't 189 00:13:19,440 --> 00:13:24,560 watched a Marvel movie in many years. If 190 00:13:22,880 --> 00:13:29,120 Steamboat Willie joined the Avengers Avengers, I would go. I mean, I I will 191 00:13:26,800 --> 00:13:35,440 say that like one argument for the deleting of IP law is the the the few 192 00:13:33,680 --> 00:13:38,800 times there's been a couple times where Disney Point puts out a project and it 193 00:13:37,360 --> 00:13:43,040 seems like the entire point of the project is just to brag about how much 194 00:13:40,399 --> 00:13:48,160 IP they have. Like did you see the Ralph Rex the internet movie like quite a few 195 00:13:46,160 --> 00:13:52,480 years ago? Did you you saw it? I think so. Okay. Yeah. It was just the whole 196 00:13:50,480 --> 00:13:56,320 movie just seemed like they were like wait Ralph. So there's Wreckit Ralph and 197 00:13:54,639 --> 00:13:59,680 then the sequel was Ralph Breaks the Internet. Wreck-It Ralph, they had a 198 00:13:58,240 --> 00:14:03,040 couple recognizable people. There was Sonic, it was there, and like Bowser and 199 00:14:01,440 --> 00:14:07,519 stuff, but then they go to the internet in the second one and they they go to 200 00:14:05,279 --> 00:14:10,959 like Disney the Disney website and it's just like, "Wow, what a magical place. 201 00:14:09,519 --> 00:14:14,800 Look, there's Star Wars and there's these and Lilo and Stitch and like 202 00:14:12,959 --> 00:14:18,560 they're just showing all the princesses are there." So it was basically like 203 00:14:17,279 --> 00:14:24,160 look at all these things that used to be cool. That watching that movie made me 204 00:14:21,360 --> 00:14:27,920 think like this is kind of like if I if if I watched that movie and then Jack 205 00:14:25,920 --> 00:14:32,639 Dorsey called me up and like we should delete all IP law and I'd be like Jackie 206 00:14:30,079 --> 00:14:37,760 you're so smart why don't we talk more often. No I please it kind of makes me 207 00:14:35,760 --> 00:14:44,480 feel that way because it's like we're heading into this AI 208 00:14:40,519 --> 00:14:46,480 future where if AI gets power like 209 00:14:44,480 --> 00:14:52,360 there's kind of two paths that I see. Yeah. 210 00:14:47,639 --> 00:14:54,880 Um, one is we lock this down. And we're 211 00:14:52,360 --> 00:14:59,199 like, AI companies, you're going all over the place. You're you're slurping 212 00:14:56,639 --> 00:15:03,360 up everyone's stuff and making slop. We need to stop this. You if you want to 213 00:15:01,600 --> 00:15:06,800 train on material, you got to license it. You got to pay all the people you've 214 00:15:05,120 --> 00:15:13,279 already trained on their material. Like, like either that that's not going to happen. But like that's either we kind 215 00:15:10,480 --> 00:15:19,199 of move in that direction or things just kind of keep progressing as they have 216 00:15:15,040 --> 00:15:22,240 been and copyright and IP just kind of 217 00:15:19,199 --> 00:15:25,040 become this kind of nebulous like like 218 00:15:22,240 --> 00:15:31,440 the the walls get torn down a lot more. Yeah. Uh and it's a lot more 219 00:15:28,519 --> 00:15:34,720 sketchy as to what's infringing and what's not. You think we're going there? 220 00:15:33,360 --> 00:15:38,480 I think we're going there. So Jack Dorsey is going to mostly maybe get his 221 00:15:36,800 --> 00:15:41,600 way. I think he's going to partially win. I don't think, like I said earlier, 222 00:15:40,240 --> 00:15:51,680 I don't think we're going to be deleting IP law, but I think there will be significant reform because I I don't see 223 00:15:47,279 --> 00:15:54,800 America effectively just kind of being 224 00:15:51,680 --> 00:15:57,279 willing to lose the AI race because I 225 00:15:54,800 --> 00:16:02,399 see this as their in in their opinion, this is this is computers. This is uh 226 00:16:00,320 --> 00:16:07,440 they have to fight again. So, I I think I agree with you if we're talking about 227 00:16:04,399 --> 00:16:10,079 like the powers that be. Yeah. the uh 228 00:16:07,440 --> 00:16:14,720 you know tech corporations, government, they definitely don't want to lose to 229 00:16:11,279 --> 00:16:16,639 China. Yeah. The question then is what 230 00:16:14,720 --> 00:16:24,480 what do how do the people feel about it? Oh yeah. Because every time I encounter 231 00:16:21,040 --> 00:16:25,920 sentiment about AI and copyright online, 232 00:16:24,480 --> 00:16:31,600 this is why this tweet was so interesting because the vast majority of 233 00:16:29,360 --> 00:16:36,000 the discourse around AI and copyright online is like, yeah, there these AI 234 00:16:34,560 --> 00:16:40,160 companies are stealing from people. Well, they are. Yeah. Yeah. And you have 235 00:16:38,560 --> 00:16:45,199 the tech bros, you know, whatever coming out every once in a while being like, we 236 00:16:42,639 --> 00:16:49,079 need to chill, guys. we like how are we going to get exponential growth if we 237 00:16:46,880 --> 00:16:53,519 don't steal all this copyrighted stuff. Um and but they don't usually like 238 00:16:52,000 --> 00:16:57,680 that's not getting a ton of engagement. It's like this but this is saying the 239 00:16:55,199 --> 00:17:01,120 quiet part out loud pretty much. It's cuz it's because it doesn't feel good. 240 00:16:59,360 --> 00:17:04,640 They are stealing stuff. They are taking other people's work and they are 241 00:17:02,399 --> 00:17:09,520 directly aggressively profiting off of it. The 242 00:17:07,199 --> 00:17:12,480 well profiting off of investment I guess most of these companies are really not 243 00:17:10,959 --> 00:17:16,319 making a profit. Yeah. They're not they're not taking the copyright machine 244 00:17:14,400 --> 00:17:19,839 and putting it into a different machine and it's like wow it's chewing up the 245 00:17:18,000 --> 00:17:23,039 copyright and spitting out profit. They're tricking people into thinking 246 00:17:21,520 --> 00:17:26,799 the copyright machine putting a cool little bow on it and being like you want 247 00:17:24,559 --> 00:17:31,440 to give me money for nothing. Yeah. Sick. Um 248 00:17:29,280 --> 00:17:34,960 the bubble man. Yeah, the bubble's pretty sorry you you 249 00:17:33,520 --> 00:17:40,000 were saying something though. No, it's just it doesn't feel good. Yeah, it's 250 00:17:36,720 --> 00:17:41,600 morally I think it's like bad. you're 251 00:17:40,000 --> 00:17:46,320 you're taking significant work from people especially h and we already see 252 00:17:44,400 --> 00:17:50,720 this from Google right a lot of small brands are getting screwed over by these 253 00:17:47,760 --> 00:17:55,360 search engines and now also by AI systems because everything's 254 00:17:52,960 --> 00:17:59,360 intercepting like even right now if you're using Google search not even if 255 00:17:57,039 --> 00:18:03,600 you're in a chat program if you're using Google search it will give you an AI 256 00:18:01,280 --> 00:18:07,840 summary at the top and the AI summary will often make it so that you do not 257 00:18:06,160 --> 00:18:12,960 have to go through and click onto one of those websites and that is making it so 258 00:18:10,080 --> 00:18:18,480 that independent small time research, writing, whatever, is all effectively 259 00:18:15,760 --> 00:18:23,760 being defunded by these platforms that are using them in order to succeed, 260 00:18:20,720 --> 00:18:25,760 right? And that is a crazy concept. 261 00:18:23,760 --> 00:18:31,520 Yeah, because like, okay, we might win the AI race, but at what cost? The cost 262 00:18:28,640 --> 00:18:34,880 of like literally everything else. Um, I was going to say like, okay, yeah, 263 00:18:32,799 --> 00:18:39,200 definitely at the cost of like, you know, smaller creators, smaller like 264 00:18:37,120 --> 00:18:43,160 copyright holders or whatever. uh smaller companies that are trying to 265 00:18:40,880 --> 00:18:48,320 get up in the SEO rankings and they just can't. But also at the cost of like I 266 00:18:45,520 --> 00:18:53,280 feel like our minds there have been some uh studies looking at uh people who use 267 00:18:51,360 --> 00:18:58,160 AI a lot versus people who don't use AI a lot and it's like looking at their 268 00:18:55,120 --> 00:19:00,880 like problem solving ability novel task. 269 00:18:58,160 --> 00:19:05,280 Yeah. They're worse at it. And I I know that when I go to Google something it 270 00:19:03,039 --> 00:19:08,960 gives me an AR overview. I know how to Google stuff. My whole job is like 271 00:19:07,120 --> 00:19:12,799 researching stuff and like reporting on it. So like I've learned how to be 272 00:19:10,640 --> 00:19:17,360 really good good at like Google foo. Yeah. But most people are not and 273 00:19:15,600 --> 00:19:21,600 they're going to be even worse now if they don't even have to click into like 274 00:19:18,960 --> 00:19:27,039 one site to see what the to see the answer or what the quote unquote answer 275 00:19:24,880 --> 00:19:31,280 because it might be it might be fake. And it's it's crazy that we we all know 276 00:19:29,600 --> 00:19:35,440 that AI hallucinates and just makes stuff up all the time, but a huge 277 00:19:33,760 --> 00:19:40,640 portion of the internet, like genuinely a lot of people seem to be going under 278 00:19:37,679 --> 00:19:46,320 the the route of blindlessly verifying things with AI. Mhm. So like you see 279 00:19:44,320 --> 00:19:51,520 this on the tweeters, which I'm still going to call it, um where someone will 280 00:19:48,400 --> 00:19:53,760 be like Grock, is this true? Don't call 281 00:19:51,520 --> 00:20:00,000 it the tweeters. The tweeters. Just call it Twitter. Twitter. All right. 282 00:19:57,600 --> 00:20:03,360 But they'll ask like Grock or Perplexity or whatever other ones will respond to 283 00:20:01,600 --> 00:20:08,080 tweets if something is true and then they will just take that fully at face 284 00:20:05,440 --> 00:20:11,600 value. I which is brutal. I made a joke about this in tech the other day cuz 285 00:20:09,600 --> 00:20:16,720 yeah I've seen this where someone makes a claim and someone's like I don't think 286 00:20:14,000 --> 00:20:20,720 that's true and they're like well it is cuz look at this website and then they 287 00:20:18,320 --> 00:20:25,840 say hey Grock is this true and it's like just click on the website just Google 288 00:20:23,039 --> 00:20:30,480 it. Why are you asking an AI that may or may not lie to your face? Like they've 289 00:20:28,400 --> 00:20:33,760 gotten a lot better, but they still hallucinate enough that it's like, why 290 00:20:32,240 --> 00:20:38,240 are you trusting? They hallucinate a lot. Yeah. And then it was even worse is 291 00:20:36,080 --> 00:20:43,760 the people who are like, I collaborated with Chat GBT on on this and then they 292 00:20:40,960 --> 00:20:47,600 like post some giant freaking thing. It's like what was your participation in 293 00:20:45,919 --> 00:20:51,679 this? I was wondering about whether this is the case. So I asked Chat GBT and it 294 00:20:49,679 --> 00:20:55,919 told me that. So, I guess ChatGpt likes this and it's like, "Guys, stop." Yeah. 295 00:20:54,159 --> 00:20:59,520 The confirmation bias is crazy, too, because a lot of the chat programs will 296 00:20:57,919 --> 00:21:04,080 try to detect what you're leaning towards and just affirm it. Yeah, cuz 297 00:21:01,760 --> 00:21:09,840 people like those responses. We've done a lot of bashing on AI so 298 00:21:06,240 --> 00:21:14,000 far. Yeah. But I deserves it. Well, it 299 00:21:09,840 --> 00:21:15,600 does, but I know that you and I Well, I 300 00:21:14,000 --> 00:21:19,840 don't know, but I have a suspicion that you and I share an appreciation for some 301 00:21:17,840 --> 00:21:25,280 of its upsides as well. And I want to talk about that after this segue to our 302 00:21:22,440 --> 00:21:31,280 sponsor, Vessie. And Luke, let me tell you, I've been wearing Vessie for years 303 00:21:27,919 --> 00:21:33,679 now. Nice. And uh you know, Vessie says 304 00:21:31,280 --> 00:21:38,960 they're waterproof. They're comfortable. Oh, are your feet dry? They are. And we 305 00:21:36,480 --> 00:21:43,919 live in Vancouver here. It's tough. It's a wet place. Rains all the time. It's 306 00:21:40,799 --> 00:21:45,440 wet and and moist. And sometimes the 307 00:21:43,919 --> 00:21:51,760 water just gets in places you didn't even know it could critically. What? 308 00:21:48,080 --> 00:21:53,280 Don't worry about it. Critical hit. 309 00:21:51,760 --> 00:21:58,600 You said moist. Don't worry about it. It's fine. Oh, moist. Critical. 310 00:21:58,799 --> 00:22:04,960 Critical hit. By the way, this is completely real. So, like we had an 311 00:22:02,559 --> 00:22:09,919 atmospheric river here in Vancouver. Yeah. I don't know if it affected you. 312 00:22:07,200 --> 00:22:14,159 Seems like a new in your ivory castle. Oh, yeah. Up on the hill. You're safe 313 00:22:12,240 --> 00:22:17,919 and sound. I have a groundf flooror apartment. 314 00:22:16,080 --> 00:22:22,640 Okay. And all of the pipes for my building go to this corner where my unit 315 00:22:20,720 --> 00:22:27,760 is. You know, you know what got me? What? The snow drift on the roof. Oh, 316 00:22:25,120 --> 00:22:32,400 really? Separated the the like shingles on the roof and it rained in my 317 00:22:30,159 --> 00:22:36,080 apartment. Oh my gosh. So, you know, the ivory tower ain't that great either. Oh 318 00:22:34,159 --> 00:22:41,679 jeez. Yeah, you need to find a different ivory tower. Yeah, maybe an ebony tower. 319 00:22:40,159 --> 00:22:46,720 So, the atmospheric river hit. It overloaded the drain system. My patio 320 00:22:43,919 --> 00:22:50,559 flooded. It's an enclosed patio and it flooded. I've I've woken up in the 321 00:22:48,799 --> 00:22:56,640 morning multiple times over the past few years to find water streaming into my to 322 00:22:53,600 --> 00:22:59,520 my uh house. I've used Vessie shoes 323 00:22:56,640 --> 00:23:03,679 multiple times waiting through bailing out my patio, throwing water over the 324 00:23:02,000 --> 00:23:08,240 side of the fence. Now, all of these issues are fixed, but being out in the 325 00:23:06,320 --> 00:23:12,559 rain, in the cold, cuz they're they're warm, too. These Vessie shoes, these 326 00:23:10,480 --> 00:23:17,360 crazy Vessie shoes, they're warm. They keep your feet dry and you don't have to 327 00:23:15,679 --> 00:23:21,039 do up your laces every time. They're slip on, slip off. Yeah. Actually, my my 328 00:23:19,120 --> 00:23:23,440 girlfriend Emma went out and bought a pair just cuz she thought they sounded 329 00:23:22,400 --> 00:23:28,080 cool. She didn't even know they were sponsors of ours. Oh. Which bit of a 330 00:23:26,000 --> 00:23:32,440 failing on my part. What did she Yeah. What are you doing? I don't know. But 331 00:23:30,240 --> 00:23:38,240 she loves them too. Just up in that castle just imagining daydreaming. 332 00:23:36,640 --> 00:23:42,240 Yeah. Now, for me, my vesties have been very helpful. They kept my feet dry 333 00:23:40,400 --> 00:23:46,880 during these horrible, horrible nights I had to endure. Pumping water out of the 334 00:23:44,320 --> 00:23:51,760 sump, my whole body was wet, but not my shoes. Not your Not your Not my feet. 335 00:23:49,120 --> 00:23:56,000 Feet dry. Everything else dry and I looked great. Now you need a Vessie 336 00:23:53,360 --> 00:24:00,400 suit. Vessie, make it happen. These shoes work with any outfit. They're very 337 00:23:57,919 --> 00:24:04,480 comfortable. They keep your feet dry. So check out Vessie in the description 338 00:24:01,760 --> 00:24:08,480 below if you're tired of squishy socks. Okay, now AI. Essentially, the argument 339 00:24:07,039 --> 00:24:12,080 we're talking about is AI safety versus innovation speed because it's like, is 340 00:24:10,400 --> 00:24:16,000 is it more worth going for the moral argument of this sucks, we're stealing 341 00:24:13,679 --> 00:24:19,600 things from people, or is it more worth noting to China, right? And this is why 342 00:24:18,159 --> 00:24:26,840 I think the reason why there hasn't been precedent set is these are such 343 00:24:21,799 --> 00:24:30,480 different weights that it's it's hard 344 00:24:26,840 --> 00:24:33,679 to and like the same person can want 345 00:24:30,480 --> 00:24:37,320 both sides of it very deeply. Yeah. 346 00:24:33,679 --> 00:24:40,320 Yeah, the same person can want to not 347 00:24:37,320 --> 00:24:44,799 steal all thought creation from humanity 348 00:24:40,320 --> 00:24:47,039 ever and also not want to lose this war 349 00:24:44,799 --> 00:24:53,679 because if you lose this then this side didn't matter anyways. So it's like it's 350 00:24:49,919 --> 00:24:55,440 especially hard because you can like as 351 00:24:53,679 --> 00:24:59,400 you were saying that my first kind of like gut reaction is like but what about 352 00:24:57,360 --> 00:25:03,360 principles you know like there's kind of like you know the utilitarian thing of 353 00:25:02,080 --> 00:25:08,000 like this is what's necessary in order to win but I'm like oh but at what cost? 354 00:25:05,360 --> 00:25:14,080 What if we lose our soul as as a as a society? But at the same time, it's 355 00:25:10,320 --> 00:25:16,080 like, well, if China develops a super 356 00:25:14,080 --> 00:25:19,279 intelligence that kind of hacks the entire world and makes everything, you 357 00:25:17,600 --> 00:25:23,919 know, whatever they want it to be, then we won't have a society. That's like 358 00:25:21,840 --> 00:25:29,520 very far on the end of the possibility. Yeah, but I mean there's also the the 359 00:25:26,240 --> 00:25:32,240 problem of how consumers act about 360 00:25:29,520 --> 00:25:35,520 things, which is everyone will say online, not everyone, I'm exaggerating, 361 00:25:33,840 --> 00:25:40,559 a lot of people will say online that they want the uh, you know, don't steal 362 00:25:38,480 --> 00:25:47,039 everything from everyone route, but then if a company that does do that releases 363 00:25:43,200 --> 00:25:50,159 a model that's better, they will use it. 364 00:25:47,039 --> 00:25:52,960 Well, and yeah, that's and like that the 365 00:25:50,159 --> 00:25:58,159 precedent of that is like carved into every stone on the planet. Like they 366 00:25:55,120 --> 00:26:01,840 when you say they will use it users 367 00:25:58,159 --> 00:26:04,320 if if if everyone knows that many uh 368 00:26:01,840 --> 00:26:09,600 things are made in sweat shops by slaves and they keep buying them. Um people 369 00:26:06,640 --> 00:26:15,120 won't like uh gambling but they'll watch kick. People won't like a certain thing 370 00:26:12,480 --> 00:26:19,279 but then if the thing is better they'll buy it anyways. And I guess this kind of 371 00:26:17,120 --> 00:26:25,679 comes into a part of the conversation that I wanted to address, which was the 372 00:26:21,919 --> 00:26:28,880 usefulness of AI, the the pros of AI, 373 00:26:25,679 --> 00:26:31,679 because I remember I had friends that 374 00:26:28,880 --> 00:26:38,159 are not techy people. Um, one is a teacher. And in like 375 00:26:35,240 --> 00:26:42,960 2023, like shortly after CHACPD is released, I go and talk to him and he's 376 00:26:40,400 --> 00:26:46,320 like, "Yeah, Chachi, I use chatbt all the I'm like, first of all, I was like, 377 00:26:44,480 --> 00:26:50,159 you know what chatbt is because these are like not techy people. Yeah. And 378 00:26:48,400 --> 00:26:53,520 he's like, yeah, use it to help me like make lessons. User growth in history. It 379 00:26:52,240 --> 00:26:56,720 has to have hit a lot of people, right? Yeah. And he and he was using it to make 380 00:26:55,279 --> 00:27:02,720 lesson plans for his kids and like coming up with ideas for stuff. And I was like, wow. Like cuz that that was at 381 00:27:00,159 --> 00:27:07,480 a point where even I wasn't really I I haven't found it to be super super 382 00:27:04,799 --> 00:27:12,880 useful for my specific uh work. Yeah. Um my my dad's a plumbing 383 00:27:10,960 --> 00:27:17,679 instructor and they use it for Yeah. lesson plans. Yeah. And once it gets 384 00:27:15,360 --> 00:27:22,799 like once these actual like multimodal updates get out like it's Gemini live is 385 00:27:20,080 --> 00:27:27,279 now available with the live camera feed. Yeah. I think chat GPT has I know you 386 00:27:25,600 --> 00:27:31,919 can feed images but I think they have a camera mode for the Gemini for the live 387 00:27:29,200 --> 00:27:37,279 mode too. There's um I think a lot of people dramatically misuse AI. Um and 388 00:27:35,679 --> 00:27:40,400 like for instance like I I say that about my dad's lessons plan lesson plans 389 00:27:38,960 --> 00:27:46,559 for plumbing and people are like oh hopefully it doesn't like teach people to go outside of code and it's like well 390 00:27:44,000 --> 00:27:52,159 no he's controlling those aspects. He's using it to build the structure right to 391 00:27:49,760 --> 00:27:56,320 build the pacing to come up with the document that he can put the information 392 00:27:53,840 --> 00:28:00,799 on. He's using it to help organizational skills those types of things. The way 393 00:27:57,760 --> 00:28:03,760 that I'll often use it is to review my 394 00:28:00,799 --> 00:28:07,760 output and give me feedback on my output. I don't use it for output, but 395 00:28:06,159 --> 00:28:14,559 like what kind of output are you talking about? My most common use case is I will 396 00:28:10,960 --> 00:28:17,600 write an email that is like I sure hope 397 00:28:14,559 --> 00:28:19,279 this goes well. Uh so but I you know 398 00:28:17,600 --> 00:28:22,240 maybe I have to be rather serious or whatever in the tone of the email, but I 399 00:28:21,120 --> 00:28:27,840 want to make sure that I get my point across but not come across as aggressive or whatever else. So, I'll ask it for a 400 00:28:25,679 --> 00:28:31,279 sentiment analysis of the email. And then I might ask it like if I'm like, 401 00:28:29,600 --> 00:28:35,520 okay, that's not what I wanted someone to get from this. Yeah. And it got that 402 00:28:33,600 --> 00:28:38,720 from this, then I'll be like, okay, how would you change it to get it there? And 403 00:28:37,039 --> 00:28:43,120 then I won't use its output, but I'll use it to help guide me in that 404 00:28:40,880 --> 00:28:46,399 direction. Did I mention lizards too many times? Well, like sometimes your 405 00:28:45,120 --> 00:28:51,279 email is pretty good, but you seem weirdly focused on lizards. 406 00:28:49,919 --> 00:28:56,880 Maybe you should draw back on the lizards. I think they're not like a reptile person. 407 00:28:54,480 --> 00:29:02,480 maybe cats. We we've had um I I think it's also a good um Kickstarter of 408 00:28:59,919 --> 00:29:07,919 sorts, like for for writer's block, for instance. If if you have to write um a 409 00:29:05,520 --> 00:29:11,200 video on something, you could use it to give you your skeleton. I found I've 410 00:29:09,679 --> 00:29:15,279 tried to get it to write YouTube videos. Man, it sucks. I don't know if you've 411 00:29:13,279 --> 00:29:20,000 tried. I've used it a couple times when I like can't think of a word for 412 00:29:17,440 --> 00:29:23,840 something or like I'm like writing a joke and I'm like, "Oh man, this is like 413 00:29:22,000 --> 00:29:27,760 this." And I'm just like drawing a blank and I'm like, "What's this thing? What's 414 00:29:26,320 --> 00:29:31,039 something like when this guy says this and then another guy comes at it like 415 00:29:29,360 --> 00:29:37,840 that and they're like, "Maybe this." And I'm like, "Okay, thank you." Yeah, that makes sense. But but writing whole 416 00:29:34,320 --> 00:29:41,520 scripts, no. No good. But you can use it 417 00:29:37,840 --> 00:29:43,919 as like a research assistant. Um, again, 418 00:29:41,520 --> 00:29:47,679 there's huge pitfalls there because you shouldn't just believe what it outputs, 419 00:29:45,919 --> 00:29:52,240 but you can use it as like a kicking off point. Mhm. So like something that I've 420 00:29:49,440 --> 00:29:56,880 done is like okay here's a topic what are the like key fields of interest in 421 00:29:54,640 --> 00:30:01,520 this topic and it'll give me that list and then I'll go off and learn things 422 00:29:58,799 --> 00:30:05,760 about those. All of the AI chatbots now have like deep research modes. It does a 423 00:30:03,679 --> 00:30:09,760 good job of like giving you kind of a beginner summary of stuff. But if you 424 00:30:07,840 --> 00:30:13,760 know a lot about a topic and you're like what was the finding of this research 425 00:30:11,279 --> 00:30:17,760 like in the 19 whatever and it's like they found up this they found this stuff 426 00:30:15,600 --> 00:30:21,840 that I just made up because I'm an AI and you're like well I know that's not 427 00:30:19,440 --> 00:30:24,880 true. So it's like but that's why like when you mentioned earlier in the show 428 00:30:23,360 --> 00:30:29,600 that it's getting better with like hallucinations and stuff a little bit. 429 00:30:27,840 --> 00:30:35,080 It's only a little bit. I think it's getting better at convincing people that 430 00:30:31,919 --> 00:30:38,279 it's not doing it. Yeah. Um I I don't 431 00:30:35,080 --> 00:30:42,640 know. I don't personally think it's 432 00:30:38,279 --> 00:30:44,320 getting dramatically better overall. Um, 433 00:30:42,640 --> 00:30:47,840 there are measures that it's getting a bit better, but I really think it's just 434 00:30:46,159 --> 00:30:51,760 getting better at convincing people and it's getting better at using slightly 435 00:30:49,919 --> 00:30:57,120 more vague language and stuff like that to get away from um actually being 436 00:30:55,279 --> 00:31:03,200 pointed as doing things wrong because I catch it hallucinating stuff like Oh, 437 00:30:59,360 --> 00:31:05,200 yeah. all the time. I will say that I've 438 00:31:03,200 --> 00:31:10,240 I've I've tr I've I've just started trying to use it a little bit to help me 439 00:31:07,279 --> 00:31:14,720 research techlength stories. Um it's kind of like one of the steps in the 440 00:31:12,000 --> 00:31:19,120 process and it has not been it has not been very uh useful. But most of the 441 00:31:16,720 --> 00:31:25,039 time it's not hallucinating fake stuff as much as it's like I'm asking for uh 442 00:31:22,880 --> 00:31:28,240 stories that are recent and it like brings me something from a month ago. 443 00:31:26,399 --> 00:31:32,039 Here's something from last quarter. Yeah. Yeah. On balance, if you look at 444 00:31:30,880 --> 00:31:38,720 the whole field, do you feel like the majority of 445 00:31:35,600 --> 00:31:40,559 AI tools that are available are like 446 00:31:38,720 --> 00:31:46,320 helpful or just kind of taking up space and taking up resources? Whoa. Okay. 447 00:31:44,799 --> 00:31:51,840 Majority of AI tools. So, if you include every AI tool in that pool, then 448 00:31:48,880 --> 00:31:57,720 honestly, no. There's a lot of compute, a lot of resources being taken up. Yeah. 449 00:31:54,720 --> 00:32:00,960 And that they're able to do that 450 00:31:57,720 --> 00:32:02,640 because these companies vacuumed up IP, 451 00:32:00,960 --> 00:32:08,000 vacuumed up copyright, and massive investment because that's the the hotness right now. I think most of the 452 00:32:05,760 --> 00:32:12,399 tools are are garbage and we'll lose and we'll die. And that is how these like 453 00:32:10,320 --> 00:32:15,919 very startup focused fields work. So that's pretty natural. I don't actually 454 00:32:14,240 --> 00:32:20,399 think that's weird. I don't think that's AI specific. I think that's just that's 455 00:32:18,159 --> 00:32:23,440 true big new fancy field with lots of investment dollars and everyone trying 456 00:32:21,600 --> 00:32:28,559 to dive in all at the same time. A lot of those people are going to make garbage for for the handful of like 457 00:32:26,559 --> 00:32:33,360 crypto projects that were actually kind of cool and maybe like we could they did 458 00:32:30,640 --> 00:32:38,000 exist. They did like the idea of having some like records and stuff on a 459 00:32:35,679 --> 00:32:43,799 blockchain system that was cool. Y but for every one of those there were 460 00:32:40,799 --> 00:32:43,799 250,000 461 00:32:43,840 --> 00:32:53,679 uh rugpull yeah garbage things. So it's the same for AI. But maybe that all 462 00:32:49,760 --> 00:32:56,880 won't matter if you we we only need one. 463 00:32:53,679 --> 00:32:58,559 We only need one ASI. Artificial super 464 00:32:56,880 --> 00:33:05,519 intelligence and then it'll all be worth it. Ask. You 465 00:33:02,640 --> 00:33:09,200 can do it. Yeah. What? Yeah. ASKI. Artificial super general intelligence. 466 00:33:07,919 --> 00:33:12,960 I've never heard that together. I've heard AGI and ASI. You can't just put 467 00:33:11,600 --> 00:33:16,799 them together like that. We were talking earlier about innovators. That's true. 468 00:33:14,799 --> 00:33:21,519 I'm an inventor. Wow. Yeah. Here's a patent, sir. Thank you. That was a You 469 00:33:19,679 --> 00:33:25,840 just witnessed innovation. Jack Dorsey, this is what you're trying to kill. 470 00:33:22,880 --> 00:33:29,840 Yeah. Stop it. Think of the children. So, maybe Jack Dorsey's right and we'll 471 00:33:27,840 --> 00:33:36,279 have a utopia and none of us will own any IP and everything will be fine. And 472 00:33:32,559 --> 00:33:39,320 maybe Jack Dorsey and Elon Musk and Sam 473 00:33:36,279 --> 00:33:41,919 Alman, especially 474 00:33:39,320 --> 00:33:46,640 Samman, are going to enslave all of us. Yeah, definitely. Especially Sam Alman. 475 00:33:44,799 --> 00:33:51,200 And those are the two options. Elon just wants to impregnate all of us. He can be 476 00:33:49,519 --> 00:33:55,760 persuasive. I think this has been a huge wakeup call 477 00:33:52,799 --> 00:33:59,600 for people about uh Chinese position on value chain. America has traditionally 478 00:33:58,000 --> 00:34:05,840 been positioned very high on the value chain and then I think in a lot of North 479 00:34:02,640 --> 00:34:08,159 American views u you you go down the 480 00:34:05,840 --> 00:34:11,440 value chain from America and end up in China and and a few other countries at 481 00:34:09,839 --> 00:34:15,679 the bottom of the value chain. But I think this and uh the tariff war going 482 00:34:13,679 --> 00:34:19,200 on and a few other things have revealed that China is much higher on the value 483 00:34:17,280 --> 00:34:24,879 chain than we previously thought and has the active ability now to take rungs on 484 00:34:22,879 --> 00:34:30,240 the value chain. Yes. And I think one of those Yeah. 485 00:34:28,800 --> 00:34:36,480 No, I was just saying that because I was thinking about a recent story that just came up with their uh they have uh chips 486 00:34:34,159 --> 00:34:39,679 that they say are more performant than NVIDIA's equivalent. But anyways, go 487 00:34:38,320 --> 00:34:44,000 ahead. Yeah. I'm like, I don't know about that being true, but they're 488 00:34:41,919 --> 00:34:49,440 they're progressing very quickly in a lot of these fields, CPUs, GPUs, AI 489 00:34:46,399 --> 00:34:50,560 technology, movies, random one. Um, lots 490 00:34:49,440 --> 00:34:56,320 and lots and lots of fields where they're they're really taking major steps forward. One of those places, 491 00:34:53,760 --> 00:35:00,480 insane story that I I don't understand how it doesn't get talked about more, 492 00:34:57,760 --> 00:35:06,320 but uh the quote that I have here is Ford CEO Jim Farley admitted he's been 493 00:35:02,800 --> 00:35:08,160 driving a Xiai SU7 for 6 months and 494 00:35:06,320 --> 00:35:13,280 doesn't want to give it up. Wow. The Ford CEO drives a $30,000 Chinese 495 00:35:11,119 --> 00:35:17,200 electricade vehicle, sorry, Chinese-made electric vehicle. Um, unavailable in the 496 00:35:15,440 --> 00:35:21,839 US due to tariffs and safety regulation issues. Is this something that he 497 00:35:18,720 --> 00:35:23,119 revealed uh on a podcast like like of 498 00:35:21,839 --> 00:35:27,520 his own valition? We are now. We're basically the Ford CEO. Yes, he did. It 499 00:35:25,760 --> 00:35:31,200 was of his own valition. Why would he do that? I feel like that's of a motor 500 00:35:29,440 --> 00:35:35,520 company mentioned that he didn't want to give up a $30,000 vehicle from a 501 00:35:33,839 --> 00:35:39,680 competing company because it was so good. That should be Is he about to cash 502 00:35:38,240 --> 00:35:43,920 out or something? Like what? Stunning wakeup call to people about the 503 00:35:41,520 --> 00:35:48,480 automotive industry outside of America. Yeah. I think in EVs in particular, oh 504 00:35:46,160 --> 00:35:53,359 yeah, China is like leapfrogging the auto falling apart and have panel gaps 505 00:35:50,640 --> 00:35:56,560 that I can shove my fist through. Um and are like, 506 00:35:55,119 --> 00:36:01,440 you know, they're they're just partying on Slack sharing people's videos and 507 00:35:59,119 --> 00:36:04,000 stuff like that. So like if you're And I think that's a big part of the problem 508 00:36:02,640 --> 00:36:08,480 too where like when you when you think about Chinese products, people traditionally go like, "Oh, but like 509 00:36:06,720 --> 00:36:15,119 they're going to steal your information, right?" It's like, "Yeah, but we we know 510 00:36:12,320 --> 00:36:18,800 that our companies are Yeah, for sure. Well, yeah. Tons of documented evidence. 511 00:36:17,200 --> 00:36:22,400 That's a whole another that's a whole another conversation I feel like which 512 00:36:20,400 --> 00:36:27,040 I'm still I don't know what to think about cuz that's been the main reason 513 00:36:25,040 --> 00:36:32,839 why I'm like ah don't I don't want to use that like Chinese phone or whatever. 514 00:36:29,520 --> 00:36:35,760 Who knows what data they're sending. 515 00:36:32,839 --> 00:36:39,920 But I know my data is going out to all these other companies. So like so at a 516 00:36:37,359 --> 00:36:43,680 certain point it's like this is the future. This is what they're 517 00:36:41,839 --> 00:36:47,440 saying. They're like IP doesn't matter. Your personal information doesn't 518 00:36:44,880 --> 00:36:52,560 matter. You'll own nothing, including your own personal information and your 519 00:36:49,920 --> 00:36:56,720 own thoughts, if you can't sell them anymore, your own link and you'll be 520 00:36:54,240 --> 00:37:01,280 happy. Yeah, but we'll have to see. I don't know if this tweet that Jack 521 00:36:58,960 --> 00:37:07,200 Dorsey reposted is going to come to pass in which so he tweeted he he retweeted a 522 00:37:04,160 --> 00:37:10,240 interview with Microsoft AI CEO Mustafa 523 00:37:07,200 --> 00:37:12,800 Sullean who says the future isn't UBI 524 00:37:10,240 --> 00:37:16,880 universal basic income it's UBP universal basic 525 00:37:14,359 --> 00:37:20,599 provision abundant intelligence as the new currency you won't need more money 526 00:37:18,800 --> 00:37:26,079 because knowledge won't be something you buy not cash capability we won't need 527 00:37:24,079 --> 00:37:31,920 cash because we'll have AI assistants who can do everything for us. You don't 528 00:37:28,800 --> 00:37:34,640 need to go to the store and buy a pop, a 529 00:37:31,920 --> 00:37:39,760 soda. Your AI will just synthesize it for you. Okay, this feels like that 530 00:37:37,520 --> 00:37:46,240 South Park episode where Randy can't fix his oven because he can't fix the oven 531 00:37:41,440 --> 00:37:47,920 door. And then the the like handyman 532 00:37:46,240 --> 00:37:51,520 people end up being the wealthiest people in society because no one can fix 533 00:37:50,079 --> 00:37:55,359 anything because no one knows how to do anything. M um and they they I don't 534 00:37:54,000 --> 00:38:00,160 remember if this happens in the episode, but I'm going to make it up, I guess. Um 535 00:37:57,599 --> 00:38:02,880 I'm hallucinating. Um but they they like ask an AI to fix the door, and it just 536 00:38:01,760 --> 00:38:07,040 tries to teach him how to do it, but he's like, "No, I want you to fix the door." He keeps giving them steps, but 537 00:38:05,760 --> 00:38:12,240 he's like, "No, I want you to fix the door." Like, "We don't have things that can do our laundry right now. We don't 538 00:38:10,240 --> 00:38:14,560 have things that can build houses. We don't have things that can fix plumbing. 539 00:38:13,599 --> 00:38:21,119 We don't have things that can do any of that kind of stuff." It's like he's talking about provisions of like this 540 00:38:18,160 --> 00:38:25,599 thing that can draw Studio Gibli for me. um or like write a terrible email or a 541 00:38:23,839 --> 00:38:29,119 really really really bad YouTube video that would get one view. Like is that 542 00:38:27,440 --> 00:38:33,119 really I'm really excited about that being a universal provision. Thanks, 543 00:38:31,520 --> 00:38:36,880 bro. I think that's what makes the whole conversation just so so hard to come 544 00:38:34,640 --> 00:38:41,839 down on one side of because it's so uncertain. You need your apartment to 545 00:38:38,960 --> 00:38:47,680 not leak. Is is is chat GPT going to help you with that? Yeah. Like it's hard 546 00:38:45,040 --> 00:38:51,520 and there's there's droughts and and uh like there's peaks and valleys, too. 547 00:38:49,440 --> 00:38:54,560 Like sometimes it seems like the AI whole thing is just hype and sometimes 548 00:38:53,119 --> 00:38:57,680 it's like oh this is okay it's happening. Stunning. Yeah. Yeah. Well, 549 00:38:56,560 --> 00:39:04,240 we're not going to solve it in this podcast. No. But maybe in the next one. 550 00:39:01,440 --> 00:39:10,720 So subscribe possibly. Uh I I wouldn't want to miss it. It might happen. Maybe 551 00:39:06,560 --> 00:39:14,400 we'll invent AGI on right here. A SGI. 552 00:39:10,720 --> 00:39:17,040 Watch out. Open AAI. 553 00:39:14,400 --> 00:39:23,440 We're going to delete your IP. Hey, thanks for watching. Uh, that was a talk 554 00:39:19,040 --> 00:39:24,800 linked. Uh, subscribe to Techl. Uh, 555 00:39:23,440 --> 00:39:30,240 see you later. Maybe I'll do another one of these in half a year. Nope. Never 556 00:39:26,720 --> 00:39:30,240 happening. Never again. Bye.