1 00:00:01,360 --> 00:00:03,359 Join us as we gather around the hedge, 2 00:00:03,359 --> 00:00:04,740 where we dig into technology, 3 00:00:05,040 --> 00:00:07,759 business, and culture with the finest minds in 4 00:00:07,759 --> 00:00:08,740 computer networking. 5 00:00:20,975 --> 00:00:22,195 Well, hello, Tom. 6 00:00:22,815 --> 00:00:23,795 Hello, Russ. 7 00:00:24,350 --> 00:00:26,829 You are there. Your plant is there. Your 8 00:00:26,829 --> 00:00:28,829 whiteboard looks far too neat. I think I 9 00:00:28,829 --> 00:00:30,510 need to call your wife and get your 10 00:00:30,510 --> 00:00:32,350 kids in there to do something about that 11 00:00:32,350 --> 00:00:32,850 whiteboard. 12 00:00:33,630 --> 00:00:34,850 I need to do some artwork. 13 00:00:35,469 --> 00:00:35,969 Yeah. 14 00:00:36,670 --> 00:00:37,329 Be creative. 15 00:00:37,965 --> 00:00:38,784 Yeah. Exactly. 16 00:00:39,324 --> 00:00:40,144 Be creative. 17 00:00:40,765 --> 00:00:42,144 What is the problem? 18 00:00:43,164 --> 00:00:45,104 Hello, Yvonne. You heard Yvonne there. 19 00:00:45,564 --> 00:00:46,384 I'm here. 20 00:00:46,765 --> 00:00:49,564 In the she shed, you moved the frog 21 00:00:49,564 --> 00:00:51,884 lamp. I wanna understand the story behind moving 22 00:00:51,884 --> 00:00:52,864 the frog lamp. 23 00:00:53,750 --> 00:00:54,969 Well, I was actually, 24 00:00:55,829 --> 00:00:57,449 working on some recordings 25 00:00:57,909 --> 00:01:00,309 for my day job, and I just thought 26 00:01:00,309 --> 00:01:01,909 it would be I would I'd I'd love 27 00:01:01,909 --> 00:01:03,769 the frog lamp to have a more prominent 28 00:01:03,909 --> 00:01:06,915 place. He was behind me. Mhmm. He's sort 29 00:01:06,915 --> 00:01:08,795 of a joke between me and a a 30 00:01:08,795 --> 00:01:11,474 a friend of mine. And so I figured 31 00:01:11,474 --> 00:01:13,394 I'd put him a little make him a 32 00:01:13,394 --> 00:01:16,515 little more visible because he's fun. Yeah. He's 33 00:01:16,515 --> 00:01:18,534 really fun. He's cool. That's cool. Great. 34 00:01:19,079 --> 00:01:20,920 Yeah. I always rearrange the stuff behind me 35 00:01:20,920 --> 00:01:23,239 from time to time. I take everything down, 36 00:01:23,239 --> 00:01:24,920 dust the shelves, and then I, like, wipe 37 00:01:24,920 --> 00:01:26,840 down the shelves, and then everything ends up 38 00:01:26,840 --> 00:01:28,359 going back in a different place for the 39 00:01:28,359 --> 00:01:29,099 most part. 40 00:01:29,479 --> 00:01:31,079 We don't we don't need to talk about 41 00:01:31,079 --> 00:01:31,819 how frequently 42 00:01:32,200 --> 00:01:35,094 I dust or vacuum the shed. It's not 43 00:01:35,094 --> 00:01:36,314 a topic for conversation. 44 00:01:39,495 --> 00:01:41,594 Oh, yes. So there you go. 45 00:01:42,375 --> 00:01:45,015 So it's time for a roundtable. Let's talk 46 00:01:45,015 --> 00:01:46,935 about a couple of articles that came off 47 00:01:46,935 --> 00:01:47,674 Mind Matters, 48 00:01:48,219 --> 00:01:49,579 which, by the way, if you're not following, 49 00:01:49,579 --> 00:01:50,879 it's kind of a cool blog. 50 00:01:51,899 --> 00:01:53,599 They do a lot of stuff on artificial 51 00:01:53,739 --> 00:01:57,200 intelligence and technology and stuff. It's not really 52 00:01:57,259 --> 00:01:59,500 a technology blog. It's more of AI and 53 00:01:59,500 --> 00:02:01,280 culture stuff, but it's pretty interesting. 54 00:02:01,739 --> 00:02:02,479 So, anyway, 55 00:02:03,575 --> 00:02:05,814 the one is the slow decline of a 56 00:02:05,814 --> 00:02:07,435 key aspect of creativity, 57 00:02:09,814 --> 00:02:13,275 and the author begins by talking about serendipity. 58 00:02:15,415 --> 00:02:17,495 Okay. Do we need to define this word 59 00:02:17,495 --> 00:02:17,995 for 60 00:02:18,800 --> 00:02:20,560 all of our listeners? Are we gonna let 61 00:02:20,560 --> 00:02:22,259 them go look it up in the dictionary 62 00:02:22,319 --> 00:02:23,460 because that would be 63 00:02:23,840 --> 00:02:25,840 they might discover the words around it and 64 00:02:25,840 --> 00:02:26,740 have some serendipity. 65 00:02:27,199 --> 00:02:29,300 I'd like to hear your definition of serendipity, 66 00:02:29,520 --> 00:02:30,020 Chris. 67 00:02:31,474 --> 00:02:33,314 When you're searching for one thing and you 68 00:02:33,314 --> 00:02:35,335 run across another that's really interesting. 69 00:02:37,875 --> 00:02:39,875 I've always thought but think of the rom 70 00:02:39,875 --> 00:02:42,355 com, Serendipity, that came out in the nineties, 71 00:02:42,355 --> 00:02:43,794 and I have a hard time hearing the 72 00:02:43,794 --> 00:02:45,415 word without thinking about 73 00:02:46,159 --> 00:02:48,099 Cusack and I don't remember. 74 00:02:48,800 --> 00:02:49,620 Kate Beckinsale? 75 00:02:50,159 --> 00:02:50,659 Yeah. 76 00:02:51,040 --> 00:02:53,280 I have no idea. Yeah. It's yeah. I 77 00:02:53,439 --> 00:02:56,240 yeah. Sorry. That that word is forever tied 78 00:02:56,240 --> 00:02:57,760 with that movie for me. So 79 00:02:58,935 --> 00:03:00,455 I I think I think of it more 80 00:03:00,534 --> 00:03:01,974 I don't know if I have the definition 81 00:03:01,974 --> 00:03:03,675 right, but I think of it more as, 82 00:03:04,855 --> 00:03:07,974 running into as happening to people, maybe ideas, 83 00:03:07,974 --> 00:03:11,034 but people, and running into somebody which 84 00:03:11,415 --> 00:03:13,435 in a way that turned out to be, 85 00:03:14,129 --> 00:03:16,229 the root of something really important. 86 00:03:16,530 --> 00:03:18,949 Alright. So just extend that to ideas. 87 00:03:20,689 --> 00:03:22,370 It's the same thing. Just extend it to 88 00:03:22,370 --> 00:03:22,870 ideas. 89 00:03:23,650 --> 00:03:26,389 And I also think it hints at surprise 90 00:03:26,449 --> 00:03:29,615 and delight. Like, it's it's it's you come 91 00:03:30,074 --> 00:03:31,615 up upon something unexpected 92 00:03:32,155 --> 00:03:32,974 that's delightful 93 00:03:33,514 --> 00:03:34,175 or interesting 94 00:03:34,794 --> 00:03:35,534 or sparks 95 00:03:35,835 --> 00:03:36,335 curiosity, 96 00:03:38,074 --> 00:03:38,574 imagination. 97 00:03:39,034 --> 00:03:40,155 Right? So that's, 98 00:03:40,634 --> 00:03:43,055 that's why the word's important for this conversation. 99 00:03:43,594 --> 00:03:44,094 Yeah. 100 00:03:45,969 --> 00:03:49,110 So this author talks about serendipity experiences 101 00:03:49,489 --> 00:03:49,989 online. 102 00:03:50,449 --> 00:03:52,689 You're clicking through a chain of links, and 103 00:03:52,689 --> 00:03:54,310 you suddenly discover something 104 00:03:54,930 --> 00:03:55,909 that you weren't 105 00:03:56,289 --> 00:03:58,870 looking for, but is really important or interesting 106 00:03:58,930 --> 00:04:01,245 or fascinating or sparks you to go do 107 00:04:01,245 --> 00:04:02,145 something else. 108 00:04:03,164 --> 00:04:04,944 I can give you an example, you know, 109 00:04:05,164 --> 00:04:08,044 I probably my grandfather kept bees. I probably 110 00:04:08,044 --> 00:04:10,125 never would have thought of keeping bees, but 111 00:04:10,125 --> 00:04:12,810 there's a couple of people that I follow 112 00:04:12,810 --> 00:04:15,290 on various social media sites who do, in 113 00:04:15,290 --> 00:04:16,270 fact, keep bees. 114 00:04:17,290 --> 00:04:18,830 And so there's, like, this serendipitous, 115 00:04:19,850 --> 00:04:20,350 oh, 116 00:04:21,209 --> 00:04:23,209 I could do that too. Like, that that 117 00:04:23,209 --> 00:04:24,350 gives me an idea 118 00:04:24,664 --> 00:04:26,024 of something I could do. I can't do 119 00:04:26,024 --> 00:04:27,064 it here, but I can do it at 120 00:04:27,064 --> 00:04:29,305 the new house whenever we build that. I 121 00:04:29,305 --> 00:04:30,985 could do that too. Like, that's that would 122 00:04:30,985 --> 00:04:32,904 be a cool thing to do is to 123 00:04:32,904 --> 00:04:34,664 is to keep bees, to get five or 124 00:04:34,664 --> 00:04:36,125 six hives and 125 00:04:36,504 --> 00:04:37,724 and keep some bees. 126 00:04:39,829 --> 00:04:41,610 So but I never would have thought about 127 00:04:42,069 --> 00:04:45,370 it without running into somebody who did it 128 00:04:45,669 --> 00:04:48,889 or does it and talks about their their 129 00:04:49,509 --> 00:04:50,009 stuff. 130 00:04:51,029 --> 00:04:51,849 Another one 131 00:04:52,324 --> 00:04:53,524 I'm trying to think of others, like, in 132 00:04:53,524 --> 00:04:56,324 the technology world, more closely related to the 133 00:04:56,324 --> 00:04:57,144 technology world, 134 00:04:58,004 --> 00:05:00,004 but there have been many times when I've 135 00:05:00,004 --> 00:05:01,704 been working on a problem. 136 00:05:02,644 --> 00:05:04,245 And in the middle of working on the 137 00:05:04,245 --> 00:05:06,485 problem, I realized I've solved another problem that 138 00:05:06,485 --> 00:05:07,785 I wasn't trying to fix. 139 00:05:08,279 --> 00:05:08,779 Right? 140 00:05:09,479 --> 00:05:11,479 That's Well, there there are all kinds of 141 00:05:11,479 --> 00:05:13,819 examples of that when it comes to, 142 00:05:14,839 --> 00:05:15,659 like, medication, 143 00:05:16,680 --> 00:05:18,919 you know, like or a a great one 144 00:05:18,919 --> 00:05:21,524 is the adhesive that's used for sticky notes. 145 00:05:22,004 --> 00:05:24,425 Like, the three m were making one thing 146 00:05:24,964 --> 00:05:27,044 and then that it didn't work out quite 147 00:05:27,044 --> 00:05:27,544 right. 148 00:05:27,845 --> 00:05:29,925 And and then they realized they could use 149 00:05:29,925 --> 00:05:30,664 that adhesive 150 00:05:31,524 --> 00:05:34,964 for sticky notes. Or Scotchgard was a similar 151 00:05:34,964 --> 00:05:37,040 kind of discovery. Somebody dripped it on their 152 00:05:37,040 --> 00:05:39,600 shoe and realized that that chemical and then 153 00:05:39,600 --> 00:05:41,839 realized that that spot on their shoe never 154 00:05:41,839 --> 00:05:44,160 got dirty and Scotchgard was born. So there 155 00:05:44,160 --> 00:05:45,620 are all these interesting, 156 00:05:46,240 --> 00:05:47,540 you know, serendipitous 157 00:05:48,314 --> 00:05:50,074 discoveries. You're trying to do one thing, and 158 00:05:50,074 --> 00:05:52,254 then you stumble upon something else. 159 00:05:52,795 --> 00:05:53,295 Yeah. 160 00:05:53,675 --> 00:05:54,175 So 161 00:05:54,955 --> 00:05:57,375 in this article that that the author, 162 00:05:57,915 --> 00:05:59,455 Eric J Larson, says 163 00:06:00,169 --> 00:06:02,250 that something else is a surprise because we 164 00:06:02,250 --> 00:06:03,930 weren't looking for it or at least didn't 165 00:06:03,930 --> 00:06:05,529 think we were, and yet it turns out 166 00:06:05,529 --> 00:06:07,769 to be exactly what we wanted or a 167 00:06:07,769 --> 00:06:10,269 more compelling or more compelling than we originally 168 00:06:10,329 --> 00:06:10,829 sought. 169 00:06:12,745 --> 00:06:13,245 So 170 00:06:14,105 --> 00:06:16,264 I think the thing is is that in 171 00:06:16,264 --> 00:06:18,685 the modern world, instead of serendipity, 172 00:06:18,985 --> 00:06:19,884 we have doomscrolling. 173 00:06:23,305 --> 00:06:25,324 I I think that's a that's a problem 174 00:06:25,689 --> 00:06:27,689 with our attention spans and the way we 175 00:06:27,689 --> 00:06:28,350 do things. 176 00:06:29,290 --> 00:06:30,810 That's not what this article is talking about, 177 00:06:30,810 --> 00:06:32,750 but I think it's absolutely true that 178 00:06:33,529 --> 00:06:35,790 we tend to think of doomscrolling 179 00:06:36,169 --> 00:06:36,910 as serendipity. 180 00:06:37,529 --> 00:06:39,930 Like, we're gonna discover something. I might find 181 00:06:39,930 --> 00:06:40,990 a new cat video. 182 00:06:41,495 --> 00:06:43,975 I might find a new duck video or 183 00:06:43,975 --> 00:06:45,035 whatever it is, 184 00:06:45,415 --> 00:06:48,134 or some person falling off the trampoline in 185 00:06:48,134 --> 00:06:50,475 a new and interesting way. I don't know. 186 00:06:51,415 --> 00:06:52,475 That's not really 187 00:06:53,335 --> 00:06:56,129 what this is talking about, because those might 188 00:06:56,129 --> 00:06:59,090 be fun to watch, but they aren't they 189 00:06:59,090 --> 00:07:01,189 don't lead to productive places. 190 00:07:01,569 --> 00:07:03,009 And I think one of the points of 191 00:07:03,009 --> 00:07:04,310 serendipity is 192 00:07:05,009 --> 00:07:07,350 that where you end up is something, 193 00:07:08,694 --> 00:07:10,475 I I hesitate to say useful, 194 00:07:10,935 --> 00:07:13,035 but maybe useful is the right word. 195 00:07:13,574 --> 00:07:15,814 I don't know. So it's something more than 196 00:07:15,814 --> 00:07:16,875 another laugh. 197 00:07:17,654 --> 00:07:19,995 I think that think go ahead, Tom. 198 00:07:20,375 --> 00:07:20,875 The 199 00:07:21,620 --> 00:07:23,319 the just thinking about doomscrolling, 200 00:07:24,099 --> 00:07:25,479 the money is a lot easier, 201 00:07:25,939 --> 00:07:28,439 if you feed engagement with rage 202 00:07:28,899 --> 00:07:31,539 and and, fighting with your neighbors and stuff. 203 00:07:31,539 --> 00:07:33,300 Like, it's just a lot easier to get 204 00:07:33,300 --> 00:07:36,360 people involved when they're mad. And so, like, 205 00:07:36,654 --> 00:07:37,615 I I don't know. I guess it's the 206 00:07:37,615 --> 00:07:38,915 cynical part of me, but 207 00:07:39,294 --> 00:07:40,115 the monetization 208 00:07:40,495 --> 00:07:42,035 of of our attention 209 00:07:42,415 --> 00:07:42,915 online, 210 00:07:43,294 --> 00:07:45,214 it's just easier to get people going when 211 00:07:45,214 --> 00:07:47,714 they're when they're upset about something. And so 212 00:07:48,175 --> 00:07:50,735 serendipity never felt to me like a negative 213 00:07:50,735 --> 00:07:52,115 thing. It never felt like, 214 00:07:53,020 --> 00:07:54,699 you know, I went from feeling bad to 215 00:07:54,699 --> 00:07:57,020 worse sort of experience. It always felt like 216 00:07:57,020 --> 00:07:58,860 the meeting of good ideas or the meeting 217 00:07:58,860 --> 00:08:00,540 of good people or whatever. Yes. And so 218 00:08:00,540 --> 00:08:02,860 it's kind of its antithesis is ruling everything 219 00:08:02,860 --> 00:08:04,720 because it's probably a lot more profitable, 220 00:08:05,415 --> 00:08:07,574 to to doom scroll than it is to 221 00:08:07,574 --> 00:08:08,134 have these, 222 00:08:09,014 --> 00:08:10,714 the the these positive connections. 223 00:08:11,334 --> 00:08:13,735 So go ahead. It's certainly more predictable. And, 224 00:08:13,894 --> 00:08:15,115 and I think too, 225 00:08:16,855 --> 00:08:18,454 a lot of a lot more of our 226 00:08:18,454 --> 00:08:22,170 online experience is curated than we want to 227 00:08:22,170 --> 00:08:24,269 believe. Right? The the algorithms 228 00:08:25,449 --> 00:08:25,949 really 229 00:08:26,730 --> 00:08:28,269 are finely tuned 230 00:08:28,569 --> 00:08:30,670 to detect our interests 231 00:08:31,370 --> 00:08:34,009 and then defeat us the things that they 232 00:08:34,009 --> 00:08:34,509 know 233 00:08:35,095 --> 00:08:37,674 we want, which is the opposite of serendipity. 234 00:08:38,375 --> 00:08:39,995 Right? If it's predetermined. 235 00:08:41,014 --> 00:08:42,475 And so often, 236 00:08:43,014 --> 00:08:43,674 you know, 237 00:08:44,054 --> 00:08:44,554 I've 238 00:08:45,815 --> 00:08:48,294 in my mind believed I was thinking about 239 00:08:48,294 --> 00:08:49,975 doing a thing, and then all of a 240 00:08:49,975 --> 00:08:52,570 sudden I start seeing those kinds of things 241 00:08:52,570 --> 00:08:53,789 on my feed without 242 00:08:55,049 --> 00:08:59,209 intentionally explicitly searching for it. Right? But clearly, 243 00:08:59,209 --> 00:09:00,829 there was something in my behavior 244 00:09:01,769 --> 00:09:03,470 that indicated to the algorithm. 245 00:09:04,024 --> 00:09:05,625 Whether it's like, you know, buying a new 246 00:09:05,625 --> 00:09:07,245 car, going on a vacation, 247 00:09:07,625 --> 00:09:10,024 or getting a new pair of shoes, like, 248 00:09:10,024 --> 00:09:12,924 things like that that I don't think I've 249 00:09:12,985 --> 00:09:14,764 told the algorithm about. 250 00:09:15,544 --> 00:09:16,365 But suddenly, 251 00:09:16,745 --> 00:09:19,300 it's feeding me things and it's it's not 252 00:09:19,300 --> 00:09:21,460 because it's reading my mind. It's because it's 253 00:09:21,460 --> 00:09:24,660 doing behavioral analysis enough to determine what it 254 00:09:24,660 --> 00:09:26,600 believes I'm gonna do next. And it's 255 00:09:27,059 --> 00:09:29,540 shockingly right. It often knows when I'm getting 256 00:09:29,540 --> 00:09:31,059 into the car to go to the Taco 257 00:09:31,059 --> 00:09:32,684 Bell. Yeah. Right? 258 00:09:33,404 --> 00:09:34,304 And and 259 00:09:34,845 --> 00:09:35,985 those experiences, 260 00:09:39,245 --> 00:09:41,245 they they can be interesting. They can be 261 00:09:41,245 --> 00:09:42,384 troubling, but 262 00:09:43,004 --> 00:09:44,764 they they may have some of the same 263 00:09:44,764 --> 00:09:47,500 markers of serendipity, but they're really not. Right? 264 00:09:47,500 --> 00:09:49,339 They're they're they're more curated than we want 265 00:09:49,339 --> 00:09:51,819 to believe. Yep. So I was just gonna 266 00:09:51,819 --> 00:09:53,980 say, Tom, in relation to what you were 267 00:09:53,980 --> 00:09:55,740 saying, I actually don't think 268 00:09:56,379 --> 00:09:58,940 I think that we have two ideas of 269 00:09:58,940 --> 00:10:01,264 what sells in the world. The of them 270 00:10:01,264 --> 00:10:03,365 is is is sex sells. 271 00:10:03,824 --> 00:10:06,485 The other of which is is anger sells. 272 00:10:06,784 --> 00:10:09,184 I actually think that what really sells is 273 00:10:09,184 --> 00:10:09,684 fear. 274 00:10:10,865 --> 00:10:12,725 And if if you go back to 275 00:10:13,759 --> 00:10:15,679 a lot of the anger based stuff that 276 00:10:15,679 --> 00:10:18,179 you see on social media, it's actually fear. 277 00:10:18,960 --> 00:10:20,660 It's undergirded by fear. 278 00:10:21,040 --> 00:10:21,540 And, 279 00:10:22,160 --> 00:10:23,840 you know, I'll say the same thing about 280 00:10:23,840 --> 00:10:24,980 sex selling. Right? 281 00:10:25,514 --> 00:10:27,355 Okay. You see that you see the girl 282 00:10:27,355 --> 00:10:29,535 or the guy or whatever, but in reality, 283 00:10:29,915 --> 00:10:31,595 you don't wanna be with that person. You 284 00:10:31,595 --> 00:10:32,735 wanna be that person. 285 00:10:33,274 --> 00:10:35,055 You fear missing out. 286 00:10:35,595 --> 00:10:36,735 You fear not 287 00:10:37,529 --> 00:10:39,449 being with that being with that person or 288 00:10:39,449 --> 00:10:40,350 being that person. 289 00:10:40,970 --> 00:10:42,809 So I really I really think a lot 290 00:10:42,809 --> 00:10:44,190 of it comes down to fear. 291 00:10:44,730 --> 00:10:46,809 And if we just let go of our 292 00:10:47,209 --> 00:10:48,429 if we had a better 293 00:10:50,375 --> 00:10:50,875 focus 294 00:10:51,254 --> 00:10:52,394 on self forgetfulness 295 00:10:53,415 --> 00:10:55,575 and go get things done instead of worrying 296 00:10:55,575 --> 00:10:56,075 about 297 00:10:56,615 --> 00:10:58,934 what other people are doing or whatever, it 298 00:10:58,934 --> 00:11:01,035 probably eliminate a lot of that 299 00:11:01,589 --> 00:11:04,470 pathway in our brains. But anyway, that's outside 300 00:11:04,470 --> 00:11:06,389 the topic of serendipity. It's just something that 301 00:11:06,389 --> 00:11:08,230 occurs to me. I liked, 302 00:11:08,629 --> 00:11:10,709 where the article went with I'd never heard 303 00:11:10,709 --> 00:11:12,809 of these three phases of creativity before. 304 00:11:13,350 --> 00:11:16,549 Phase one generation, chaotic, undirected process of preparing 305 00:11:16,549 --> 00:11:17,714 for ideas to germinate 306 00:11:18,095 --> 00:11:20,815 permission where you step aside and let an 307 00:11:20,815 --> 00:11:23,695 idea surface, and then translation where you where 308 00:11:23,695 --> 00:11:25,695 you make it something usable. I I had 309 00:11:25,695 --> 00:11:27,955 never heard this particular model of creativity. 310 00:11:28,815 --> 00:11:30,654 You know, the author's thesis is that these 311 00:11:30,654 --> 00:11:32,115 things are disappearing. But, 312 00:11:33,029 --> 00:11:35,910 I don't know. My experience is that I 313 00:11:35,910 --> 00:11:38,149 I feel like I have experienced these things 314 00:11:38,149 --> 00:11:38,970 even recently. 315 00:11:39,670 --> 00:11:41,190 I don't know. What what about you all? 316 00:11:41,190 --> 00:11:43,670 Do you feel like you've had creative moments 317 00:11:43,670 --> 00:11:45,129 in the last week or two? 318 00:11:46,725 --> 00:11:48,965 Sometimes. Yes. I think one of the 319 00:11:49,605 --> 00:11:51,365 and it would be an interesting topic to 320 00:11:51,365 --> 00:11:52,665 explore further. The 321 00:11:53,045 --> 00:11:54,345 the the the permission 322 00:11:54,804 --> 00:11:55,304 component. 323 00:11:55,845 --> 00:11:56,985 I find, especially, 324 00:11:57,445 --> 00:11:59,285 you know, and and Jonathan Haidt, I've not 325 00:11:59,285 --> 00:12:01,879 read the book, but I've I've read some 326 00:12:01,879 --> 00:12:04,199 synopsis of it. His his book, The Anxious 327 00:12:04,199 --> 00:12:04,699 Generation, 328 00:12:05,000 --> 00:12:06,139 and talking about 329 00:12:06,519 --> 00:12:09,159 I I feel like a lot of and 330 00:12:09,159 --> 00:12:11,100 even the young people in my life, 331 00:12:11,959 --> 00:12:15,465 they want their they wanna know what's next. 332 00:12:15,865 --> 00:12:17,725 Right? And they get really uncomfortable, 333 00:12:18,425 --> 00:12:20,904 like, when the plans aren't fully set or 334 00:12:20,904 --> 00:12:21,465 when there's 335 00:12:22,024 --> 00:12:24,184 they they can't map out exactly what's gonna 336 00:12:24,184 --> 00:12:25,004 happen next. 337 00:12:25,625 --> 00:12:27,085 Part of creativity 338 00:12:27,545 --> 00:12:29,325 and this permission phase 339 00:12:30,105 --> 00:12:30,605 is 340 00:12:30,990 --> 00:12:33,870 being comfortable that you don't know what's gonna 341 00:12:33,870 --> 00:12:34,370 emerge. 342 00:12:34,830 --> 00:12:37,730 Right? And and, you know, leaning into, 343 00:12:40,029 --> 00:12:41,789 mystery. Even that word is here. It's a 344 00:12:41,789 --> 00:12:44,565 mystery even to those experiencing it That 345 00:12:45,185 --> 00:12:47,745 that the thing that's coming next, you don't 346 00:12:47,745 --> 00:12:48,805 know what it is, 347 00:12:49,904 --> 00:12:52,545 but you're open to experiencing it. And it 348 00:12:52,545 --> 00:12:54,865 could be good. It could be, you know, 349 00:12:54,865 --> 00:12:57,745 something that's surprising and delightful. And I I 350 00:12:57,745 --> 00:13:00,360 think that that's also something that's kind of 351 00:13:00,360 --> 00:13:02,779 been lost in our very safe, 352 00:13:03,399 --> 00:13:04,299 very curated, 353 00:13:05,399 --> 00:13:09,080 both physical and online experiences. Yeah. We we 354 00:13:09,080 --> 00:13:11,960 just want everything mapped out. And what that 355 00:13:11,960 --> 00:13:13,740 does is it eliminates 356 00:13:14,375 --> 00:13:15,754 the potential for 357 00:13:16,534 --> 00:13:19,495 curiosity, creativity, and all those all the things 358 00:13:19,495 --> 00:13:21,575 we've been talking about. And and we want 359 00:13:21,575 --> 00:13:22,075 permission. 360 00:13:23,575 --> 00:13:25,894 We want to know that somebody approves of 361 00:13:25,894 --> 00:13:28,610 us doing this before we do it. And 362 00:13:28,669 --> 00:13:29,950 the problem is you don't have to have 363 00:13:29,950 --> 00:13:30,450 permission. 364 00:13:31,230 --> 00:13:32,769 You can just go do things. 365 00:13:33,710 --> 00:13:35,230 I mean, really, you can just go do 366 00:13:35,230 --> 00:13:36,210 things. Like, 367 00:13:36,909 --> 00:13:37,649 you can 368 00:13:38,029 --> 00:13:39,809 you can go build a chicken coop. 369 00:13:40,350 --> 00:13:41,250 You can build 370 00:13:41,774 --> 00:13:42,995 beehives. You can 371 00:13:43,615 --> 00:13:45,615 start a newsletter. You can start a pot. 372 00:13:45,615 --> 00:13:47,074 You can just go do things. 373 00:13:47,615 --> 00:13:49,634 And if they fail, well, they failed. 374 00:13:50,654 --> 00:13:51,634 And that's okay. 375 00:13:52,495 --> 00:13:54,975 But this permission phase is part of that 376 00:13:54,975 --> 00:13:56,355 too, I think, Yvonne. 377 00:13:57,139 --> 00:13:59,379 So the point the author is making with 378 00:13:59,379 --> 00:14:01,620 all of this talk about serendipity and permission 379 00:14:01,620 --> 00:14:02,120 and 380 00:14:02,820 --> 00:14:04,679 translation and everything else 381 00:14:05,139 --> 00:14:07,240 is that using Chat GPT, 382 00:14:09,460 --> 00:14:11,559 whatever the other ones are, Copilot, 383 00:14:12,245 --> 00:14:13,784 I don't know what they all are, 384 00:14:14,644 --> 00:14:15,945 is actually harmful 385 00:14:16,644 --> 00:14:18,264 to your creative process. 386 00:14:18,884 --> 00:14:21,065 And it's harmful at every stage 387 00:14:21,445 --> 00:14:22,985 that you mentioned there, Tom. 388 00:14:24,004 --> 00:14:25,304 It's badly harmful 389 00:14:25,830 --> 00:14:27,050 in all three stages. 390 00:14:27,590 --> 00:14:29,930 So for instance, if we think about 391 00:14:31,110 --> 00:14:31,610 generation, 392 00:14:33,830 --> 00:14:36,149 if we allow an AI to write a 393 00:14:36,149 --> 00:14:37,529 presentation for us, 394 00:14:38,149 --> 00:14:39,610 we're not doing the generation. 395 00:14:40,945 --> 00:14:42,325 We're not doing the work. 396 00:14:43,985 --> 00:14:44,644 The brainstorming 397 00:14:44,945 --> 00:14:45,445 disappears 398 00:14:46,225 --> 00:14:46,725 because 399 00:14:47,504 --> 00:14:48,565 why would I bother? 400 00:14:49,105 --> 00:14:51,125 The AI is doing it for me. 401 00:14:52,384 --> 00:14:52,884 So 402 00:14:53,379 --> 00:14:54,819 I don't know if you if either of 403 00:14:54,819 --> 00:14:56,120 you experienced that 404 00:14:56,500 --> 00:14:58,759 thing. I know I have where I've been 405 00:14:59,220 --> 00:15:01,720 working with an LOM for some reason, and, 406 00:15:04,339 --> 00:15:05,079 I don't 407 00:15:05,465 --> 00:15:07,544 I don't think that I need to spend 408 00:15:07,544 --> 00:15:08,764 as much time creating 409 00:15:10,105 --> 00:15:12,264 or thinking about what I'm creating because they 410 00:15:12,264 --> 00:15:14,585 all, is feeding me ideas. I become the 411 00:15:14,585 --> 00:15:15,085 filter. 412 00:15:16,665 --> 00:15:18,605 And the l l becomes the fill. 413 00:15:19,120 --> 00:15:21,200 Yeah. I never I guess I just never 414 00:15:21,200 --> 00:15:23,279 use it in the two phases. Like, I 415 00:15:23,279 --> 00:15:25,840 just never considered that. It's it's when it 416 00:15:25,840 --> 00:15:28,160 gets to, okay. Here's something that I know 417 00:15:28,160 --> 00:15:30,100 about what the shape it needs to be. 418 00:15:30,240 --> 00:15:30,740 Like, 419 00:15:31,865 --> 00:15:34,424 writing code for little little home automation projects 420 00:15:34,424 --> 00:15:36,825 is what my my latest, use of this. 421 00:15:36,825 --> 00:15:39,164 Like, I know what framework I wanna use. 422 00:15:39,225 --> 00:15:41,065 I know this is the thing I'm trying 423 00:15:41,065 --> 00:15:42,585 to accomplish, but I don't even tell the 424 00:15:42,585 --> 00:15:45,139 LOM that. I just say this is the 425 00:15:45,200 --> 00:15:47,139 this is the particular technical outcome. 426 00:15:47,600 --> 00:15:49,200 And, you know, there's lots of ways to 427 00:15:49,200 --> 00:15:51,120 use OOMs, and I'm not mine's not the 428 00:15:51,120 --> 00:15:53,120 only way, but I just never even considered 429 00:15:53,120 --> 00:15:54,559 that it would have anything to offer. I 430 00:15:54,559 --> 00:15:55,940 don't want a regurgitation 431 00:15:56,774 --> 00:15:59,014 of any fool on the Internet who had 432 00:15:59,014 --> 00:16:01,414 something to say. I'd rather just I figured 433 00:16:01,414 --> 00:16:03,174 my ideas are more valuable than that. Maybe 434 00:16:03,174 --> 00:16:06,054 that's being, you know, maybe that's hubris. But 435 00:16:06,054 --> 00:16:07,815 for me, the translation part of it, I 436 00:16:07,815 --> 00:16:09,894 think there is there is work for LMS 437 00:16:09,894 --> 00:16:10,634 to do there. 438 00:16:11,450 --> 00:16:13,129 But the two, I just never used it 439 00:16:13,129 --> 00:16:13,629 for. 440 00:16:14,490 --> 00:16:16,730 Well and and I'm reminded of an example 441 00:16:16,730 --> 00:16:18,649 in a book, and I cannot remember the 442 00:16:18,649 --> 00:16:20,250 the the book that I that I read 443 00:16:20,250 --> 00:16:22,350 this in. But it was talking about, 444 00:16:23,754 --> 00:16:25,294 it may have been the book range, 445 00:16:25,835 --> 00:16:27,215 but it was talking about 446 00:16:27,595 --> 00:16:28,975 Nobel Prize winners 447 00:16:29,355 --> 00:16:31,134 and how many of them 448 00:16:31,995 --> 00:16:36,235 had broad experience across multiple domains. And there's 449 00:16:36,235 --> 00:16:38,820 this school of thought that you take children 450 00:16:38,820 --> 00:16:40,899 when they're very young, you choose a discipline 451 00:16:40,899 --> 00:16:42,500 for them, and then you drill it. Like, 452 00:16:42,500 --> 00:16:45,720 whether it's violin or chess or piano. STEM. 453 00:16:45,940 --> 00:16:46,440 But 454 00:16:46,899 --> 00:16:49,480 but but what the data says is that 455 00:16:49,539 --> 00:16:51,000 often, it's best 456 00:16:51,315 --> 00:16:54,294 for children to have lots of multiple experiences 457 00:16:54,514 --> 00:16:57,154 across different disciplines and then let them specialize 458 00:16:57,154 --> 00:16:58,294 as they get older. 459 00:16:58,754 --> 00:17:00,835 The the other thing that that they found 460 00:17:00,835 --> 00:17:03,394 was that, like, with Nobel Prize winners, those 461 00:17:03,394 --> 00:17:04,054 that were, 462 00:17:05,090 --> 00:17:07,830 the the vast majority of them had expertise 463 00:17:07,970 --> 00:17:09,590 across multiple domains 464 00:17:10,049 --> 00:17:11,750 and had maybe a hobby 465 00:17:12,210 --> 00:17:15,350 that didn't align with their field of study. 466 00:17:15,809 --> 00:17:16,309 So, 467 00:17:17,015 --> 00:17:18,934 yeah. I mean, and it could be something 468 00:17:18,934 --> 00:17:22,634 like gardening or studying philosophy or reading literature, 469 00:17:22,855 --> 00:17:25,274 but but but their brains 470 00:17:25,815 --> 00:17:27,575 were able to take the data from all 471 00:17:27,575 --> 00:17:29,994 these different disciplines and combine them creatively. 472 00:17:30,420 --> 00:17:32,600 And they were able to see connections 473 00:17:33,220 --> 00:17:35,320 that other folks that are so narrowly 474 00:17:35,779 --> 00:17:38,519 focused in just one discipline 475 00:17:39,059 --> 00:17:42,420 can't see. And I think that that, like, 476 00:17:42,420 --> 00:17:44,820 is is a super important thing for us 477 00:17:44,820 --> 00:17:45,960 not to lose. 478 00:17:46,454 --> 00:17:49,414 And and it it may also give us 479 00:17:49,414 --> 00:17:52,134 permission to lean into a hobby or lean 480 00:17:52,134 --> 00:17:54,315 into a thing that we are interested in 481 00:17:54,694 --> 00:17:56,075 just for its own sake. 482 00:17:56,375 --> 00:17:59,015 And those things can spark creativity and actually 483 00:17:59,015 --> 00:18:01,255 make us more successful in our primary field 484 00:18:01,255 --> 00:18:01,835 of study 485 00:18:02,349 --> 00:18:05,389 because we have these broad interests. And I 486 00:18:05,389 --> 00:18:08,029 think all of that becomes part of the 487 00:18:08,029 --> 00:18:09,409 conversation about creativity. 488 00:18:09,950 --> 00:18:11,309 Yeah. Oh, yeah. Definitely. I 489 00:18:12,349 --> 00:18:14,509 you know, you talk about the breadth of 490 00:18:14,509 --> 00:18:16,369 knowledge. Another way to me 491 00:18:16,734 --> 00:18:18,994 that LLMs and relying on LLMs 492 00:18:19,694 --> 00:18:20,914 causes us problems 493 00:18:21,454 --> 00:18:22,674 is we now think 494 00:18:23,054 --> 00:18:24,734 we get to the point when we're using 495 00:18:24,734 --> 00:18:27,615 LLMs, like vibe coders. Right? Vibe coding just 496 00:18:27,615 --> 00:18:29,774 sounds really dangerous to me, the whole vibe 497 00:18:29,774 --> 00:18:30,755 coding concept. 498 00:18:31,134 --> 00:18:31,634 Right? 499 00:18:32,179 --> 00:18:34,819 So I don't need to know about this 500 00:18:34,819 --> 00:18:35,319 field. 501 00:18:36,339 --> 00:18:38,759 I can treat the LLM as an expert 502 00:18:38,819 --> 00:18:40,359 in the field for me, 503 00:18:41,460 --> 00:18:44,039 and therefore, I never really learned it myself. 504 00:18:45,865 --> 00:18:47,724 Like, I think there's a real danger 505 00:18:48,345 --> 00:18:51,144 in something like vibe coding where the LLM 506 00:18:51,144 --> 00:18:52,765 is writing code that you can't 507 00:18:54,424 --> 00:18:57,884 read. That, like, you're missing an essential component 508 00:18:59,410 --> 00:19:01,490 of the breadth of knowledge that you need 509 00:19:01,490 --> 00:19:02,470 to be creative 510 00:19:03,330 --> 00:19:04,390 when you do that. 511 00:19:04,769 --> 00:19:05,269 So 512 00:19:05,809 --> 00:19:07,750 that again but that's more in the translation 513 00:19:07,890 --> 00:19:08,869 phase, but still, 514 00:19:09,330 --> 00:19:11,190 that is, I think, part of the problem 515 00:19:12,194 --> 00:19:13,794 to me with this. It's interesting that you 516 00:19:13,794 --> 00:19:15,654 say that. I had a had a conversation 517 00:19:15,794 --> 00:19:16,835 with one of our, 518 00:19:17,234 --> 00:19:19,654 one a developer I know at work who's 519 00:19:19,794 --> 00:19:22,615 been doing a lot of development with LLMs 520 00:19:22,835 --> 00:19:26,115 and code assistive technologies. And in talking to 521 00:19:26,115 --> 00:19:26,615 him, 522 00:19:28,009 --> 00:19:30,269 like and, of course, he is 523 00:19:31,130 --> 00:19:33,609 a long term software architect and has been 524 00:19:33,609 --> 00:19:35,849 writing code for years. But but what I've 525 00:19:35,849 --> 00:19:37,069 kinda surmised 526 00:19:37,529 --> 00:19:38,509 is that ultimately, 527 00:19:39,404 --> 00:19:41,164 you still have to have a certain degree 528 00:19:41,164 --> 00:19:44,065 of expertise to effectively use the tools. 529 00:19:44,525 --> 00:19:46,684 And I think the further we go, the 530 00:19:46,684 --> 00:19:47,184 more 531 00:19:47,565 --> 00:19:51,005 we're going to find that. That the tools 532 00:19:51,005 --> 00:19:52,545 can only get you so far 533 00:19:53,005 --> 00:19:53,505 because 534 00:19:55,859 --> 00:19:56,599 they need 535 00:19:57,700 --> 00:19:59,559 human intelligence to guide them, 536 00:20:00,740 --> 00:20:02,579 which which might be a good segue for 537 00:20:02,579 --> 00:20:04,200 us. But at the same time, 538 00:20:04,579 --> 00:20:08,099 you know, like, that requires human experience and 539 00:20:08,099 --> 00:20:08,599 creativity 540 00:20:08,900 --> 00:20:11,355 and all the all the things that, you 541 00:20:11,355 --> 00:20:13,375 know, I think make us innately 542 00:20:13,755 --> 00:20:15,674 human. I think the bottom line of this 543 00:20:15,674 --> 00:20:16,575 article is 544 00:20:18,315 --> 00:20:18,634 we 545 00:20:19,434 --> 00:20:21,674 it is possible for us to lose our 546 00:20:21,674 --> 00:20:22,174 creativity 547 00:20:23,970 --> 00:20:24,950 by outsourcing 548 00:20:25,569 --> 00:20:28,049 the whole creative process to something like an 549 00:20:28,049 --> 00:20:28,549 LLM. 550 00:20:29,410 --> 00:20:31,829 Yes. And it actually reduces us 551 00:20:32,529 --> 00:20:34,130 as humans. Like, you can't get rid of 552 00:20:34,130 --> 00:20:35,809 us being humans. Like, that's a that's a 553 00:20:35,809 --> 00:20:37,410 fundamental part of what we are and what 554 00:20:37,410 --> 00:20:39,934 we do. But you can narrow the scope 555 00:20:39,934 --> 00:20:40,595 of humanness 556 00:20:42,335 --> 00:20:44,414 by doing this to some degree, and I 557 00:20:44,414 --> 00:20:45,154 think that's 558 00:20:45,455 --> 00:20:46,674 a dangerous outcome. 559 00:20:47,134 --> 00:20:48,654 So the other thing we're gonna talk about, 560 00:20:48,654 --> 00:20:50,515 like you said, segue, is 561 00:20:51,809 --> 00:20:53,490 the same place Mind Matters. They have an 562 00:20:53,490 --> 00:20:55,349 article by Gary Smith talked about 563 00:20:55,890 --> 00:20:58,690 a large language models, a lack of progress 564 00:20:58,690 --> 00:20:59,190 report. 565 00:21:00,049 --> 00:21:02,390 It's very interesting because there's a guy, 566 00:21:02,849 --> 00:21:04,529 and I can't seem like I can't I 567 00:21:04,529 --> 00:21:06,134 can't I don't know if he's in mind 568 00:21:06,134 --> 00:21:07,755 matters or if he's over in ACM, 569 00:21:08,375 --> 00:21:10,215 who every six months or so, he just 570 00:21:10,215 --> 00:21:12,475 takes German words, German sentences, 571 00:21:12,934 --> 00:21:15,414 and he has all the different LLMs translate 572 00:21:15,414 --> 00:21:15,914 them. 573 00:21:17,095 --> 00:21:19,355 And he says it's actually getting worse. 574 00:21:20,099 --> 00:21:23,000 Like, their translations are getting worse, not better. 575 00:21:23,779 --> 00:21:26,039 And he's very disturbed by this. 576 00:21:28,019 --> 00:21:28,519 So 577 00:21:29,140 --> 00:21:30,679 I don't know. There's this whole 578 00:21:33,045 --> 00:21:33,545 thing 579 00:21:34,884 --> 00:21:35,865 about LLMs 580 00:21:36,565 --> 00:21:37,065 and 581 00:21:37,845 --> 00:21:39,785 are they actually making progress? 582 00:21:41,365 --> 00:21:42,105 Are they 583 00:21:43,045 --> 00:21:43,545 becoming 584 00:21:43,924 --> 00:21:45,465 better at what they do? 585 00:21:46,289 --> 00:21:47,809 So, Tom, I don't know. You may use 586 00:21:47,809 --> 00:21:49,250 it more than I do. You use it 587 00:21:49,250 --> 00:21:50,930 more for code than I do. Do you 588 00:21:50,930 --> 00:21:52,529 think it's getting better? Do you think they're 589 00:21:52,529 --> 00:21:53,910 learning or do you think it's 590 00:21:54,210 --> 00:21:55,190 fairly stable? 591 00:21:56,690 --> 00:21:58,609 Like it doesn't like it's not really giving 592 00:21:58,609 --> 00:22:01,544 you more clever solutions or it's not really 593 00:22:01,544 --> 00:22:03,085 doing more. It's just 594 00:22:03,384 --> 00:22:05,164 the same thing you've used it for. 595 00:22:05,865 --> 00:22:07,544 Yeah. And I haven't used it a ton 596 00:22:07,544 --> 00:22:08,764 for code, but I think 597 00:22:09,065 --> 00:22:10,825 whenever I have seen it, I I didn't 598 00:22:10,904 --> 00:22:11,964 I did not detect 599 00:22:12,424 --> 00:22:14,284 that it was getting worse. I 600 00:22:14,929 --> 00:22:16,849 I I guess I detected that, 601 00:22:17,329 --> 00:22:18,390 I didn't understand 602 00:22:18,849 --> 00:22:20,710 how to tell it anything useful. 603 00:22:22,769 --> 00:22:25,009 But but that said, even even with my 604 00:22:25,009 --> 00:22:25,990 limited vocabulary, 605 00:22:27,250 --> 00:22:29,535 I was I've been able to, you know, 606 00:22:29,535 --> 00:22:31,695 give it tasks that it could make me 607 00:22:31,695 --> 00:22:33,455 make make me go faster. I have I 608 00:22:33,455 --> 00:22:35,075 have yet to find something where, 609 00:22:35,775 --> 00:22:38,575 I gave it instructions, and it revolutionized the 610 00:22:38,575 --> 00:22:40,515 way I thought about a problem. It accelerated 611 00:22:40,734 --> 00:22:41,234 mostly. 612 00:22:42,539 --> 00:22:44,220 But do you think it's getting better at 613 00:22:44,220 --> 00:22:45,740 accelerating or do you think it's I think 614 00:22:45,740 --> 00:22:47,740 it's getting better. You think it's just staying 615 00:22:47,740 --> 00:22:48,240 steady? 616 00:22:48,940 --> 00:22:50,940 I would in my limited experience, I would 617 00:22:50,940 --> 00:22:53,099 say staying steady, but I don't know. There's 618 00:22:53,099 --> 00:22:54,700 probably people use it every day and they 619 00:22:54,700 --> 00:22:56,474 would they would know. Well, I know it's 620 00:22:56,554 --> 00:22:58,234 getting better at graphics. I know it's getting 621 00:22:58,234 --> 00:22:59,214 better at 622 00:23:00,075 --> 00:23:02,714 moving like picture, like videos and stuff like 623 00:23:02,714 --> 00:23:04,075 that. I know it's getting better at those 624 00:23:04,075 --> 00:23:04,575 things, 625 00:23:05,035 --> 00:23:06,875 but it's solving what I would consider to 626 00:23:06,875 --> 00:23:08,095 be real world problems. 627 00:23:09,380 --> 00:23:11,700 Not not just the I'm gonna make a 628 00:23:11,700 --> 00:23:14,019 dog cry type of videos, you know, or 629 00:23:14,019 --> 00:23:15,860 I'm gonna do whatever. I'm gonna have this 630 00:23:15,860 --> 00:23:18,340 put this person in this movie that wasn't 631 00:23:18,340 --> 00:23:20,500 there before in this picture or whatever it 632 00:23:20,500 --> 00:23:21,000 is. 633 00:23:21,694 --> 00:23:23,555 I know it's getting better at that stuff, 634 00:23:23,855 --> 00:23:25,694 but is it getting better at real world 635 00:23:25,694 --> 00:23:26,194 problems? 636 00:23:26,734 --> 00:23:29,295 I haven't it's not not apparent to me. 637 00:23:29,295 --> 00:23:30,194 It's not obvious. 638 00:23:31,454 --> 00:23:32,914 I think, you know, 639 00:23:33,454 --> 00:23:35,875 for the eighteen months of AI, 640 00:23:36,335 --> 00:23:36,994 the models 641 00:23:37,509 --> 00:23:39,289 were improving so rapidly 642 00:23:40,390 --> 00:23:41,210 that I think 643 00:23:41,669 --> 00:23:44,009 I I think people got a false impression 644 00:23:44,069 --> 00:23:45,130 of what the 645 00:23:45,990 --> 00:23:46,490 trajectory 646 00:23:46,869 --> 00:23:49,605 for improvement was gonna be over time. I 647 00:23:49,605 --> 00:23:52,005 think they're improving, but they're not improving at 648 00:23:52,005 --> 00:23:53,625 the same hockey stick rate 649 00:23:53,924 --> 00:23:55,605 as they were as far as the pure 650 00:23:55,605 --> 00:23:56,105 models. 651 00:23:56,565 --> 00:23:58,505 I think the thing that's happening now 652 00:23:59,045 --> 00:24:01,605 is the entire marketplace is figuring out where 653 00:24:01,605 --> 00:24:03,444 do they fit and how do we use 654 00:24:03,444 --> 00:24:06,659 them. So now instead of doing code complete, 655 00:24:07,279 --> 00:24:09,380 you have tools that are that, 656 00:24:10,159 --> 00:24:11,220 for software development 657 00:24:11,839 --> 00:24:12,339 that 658 00:24:13,279 --> 00:24:13,859 are feeding 659 00:24:14,240 --> 00:24:15,539 the, you know, 660 00:24:15,839 --> 00:24:16,819 they're more interactive. 661 00:24:17,279 --> 00:24:19,700 And so a developer writes code, 662 00:24:20,055 --> 00:24:20,555 the, 663 00:24:21,414 --> 00:24:24,134 the model generates some code, and then there's 664 00:24:24,134 --> 00:24:25,275 this inter iterative 665 00:24:25,815 --> 00:24:27,835 and interactive process with the developer 666 00:24:28,215 --> 00:24:29,755 that's much less, 667 00:24:31,255 --> 00:24:33,029 it's more automated than it was, 668 00:24:33,670 --> 00:24:34,970 before. And so that's 669 00:24:35,590 --> 00:24:37,990 increasing the ability to use them. I also 670 00:24:37,990 --> 00:24:40,230 think, like, we're still figuring out how to 671 00:24:40,230 --> 00:24:40,730 integrate 672 00:24:41,670 --> 00:24:42,170 LLMs 673 00:24:43,190 --> 00:24:43,690 into 674 00:24:44,710 --> 00:24:45,849 larger systems. 675 00:24:46,150 --> 00:24:47,369 I think, you know, 676 00:24:47,775 --> 00:24:48,275 and 677 00:24:48,575 --> 00:24:50,414 and that's where we're gonna see, I think, 678 00:24:50,414 --> 00:24:51,955 the next set of advancements 679 00:24:52,335 --> 00:24:54,494 is how are we gonna practically plug these 680 00:24:54,494 --> 00:24:55,474 tools in 681 00:24:55,775 --> 00:24:57,855 to something to make a bigger thing. I 682 00:24:57,855 --> 00:24:58,355 mean, 683 00:24:59,799 --> 00:25:02,460 the the models are truly fascinating, 684 00:25:03,240 --> 00:25:05,240 the when when when you see them and 685 00:25:05,240 --> 00:25:06,920 experience them. But the more you play with 686 00:25:06,920 --> 00:25:07,420 them, 687 00:25:08,039 --> 00:25:08,940 the the more 688 00:25:11,815 --> 00:25:14,615 you can detect patterns and and and maybe 689 00:25:14,615 --> 00:25:16,055 I've just played with them enough that I'm 690 00:25:16,055 --> 00:25:17,035 not as astounded. 691 00:25:17,654 --> 00:25:19,734 But I think there's a lot of work 692 00:25:19,734 --> 00:25:21,115 still yet to happen 693 00:25:21,575 --> 00:25:24,934 to integrate them in ways that have really 694 00:25:24,934 --> 00:25:25,434 meaningful 695 00:25:26,589 --> 00:25:29,069 impact. And that is gonna take a lot 696 00:25:29,069 --> 00:25:31,409 more work and frankly, a lot more skill 697 00:25:32,029 --> 00:25:32,529 than 698 00:25:33,629 --> 00:25:35,470 I think the the it's it's not a 699 00:25:35,470 --> 00:25:37,309 magic bullet. Like, the market has been like, 700 00:25:37,309 --> 00:25:39,255 oh, these things are, and they are gonna 701 00:25:39,255 --> 00:25:40,075 change everything. 702 00:25:40,535 --> 00:25:42,215 I firmly believe that at the end of 703 00:25:42,215 --> 00:25:42,795 the day, 704 00:25:45,174 --> 00:25:47,255 AI is gonna be integrated into our world 705 00:25:47,255 --> 00:25:48,234 like light switches. 706 00:25:48,695 --> 00:25:50,535 Like, something that was at one point in 707 00:25:50,535 --> 00:25:51,269 time this 708 00:25:51,670 --> 00:25:54,390 phenomenal amazing technology. You flip a switch and 709 00:25:54,390 --> 00:25:55,830 in the middle of the night, an entire 710 00:25:55,830 --> 00:25:57,690 room is illuminated and that's amazing. 711 00:25:58,789 --> 00:25:59,450 I think 712 00:26:00,070 --> 00:26:02,150 LLMs are gonna be more integrated into the 713 00:26:02,150 --> 00:26:04,355 fabric of our lives, but it's going to 714 00:26:04,355 --> 00:26:06,215 be that sort of thing. I think, ultimately, 715 00:26:06,275 --> 00:26:08,295 it's gonna be the more boring stuff, 716 00:26:09,555 --> 00:26:11,095 that that they're gonna do. 717 00:26:11,875 --> 00:26:14,595 So there's this great quote by Andreessen, Mark 718 00:26:14,595 --> 00:26:15,555 Andreessen, that, 719 00:26:16,299 --> 00:26:17,839 Andreessen, I guess it is, 720 00:26:18,299 --> 00:26:19,920 that this article points 721 00:26:20,460 --> 00:26:22,079 to where Mark is saying, 722 00:26:22,380 --> 00:26:24,870 we've really slowed down. This is in November 723 00:26:24,870 --> 00:26:25,680 2024. 724 00:26:25,820 --> 00:26:27,660 We've really slowed down in terms of the 725 00:26:27,660 --> 00:26:28,720 amount of improvement. 726 00:26:29,265 --> 00:26:30,884 We are increasing GPUs, 727 00:26:31,345 --> 00:26:33,125 but we're not getting the intelligence 728 00:26:33,424 --> 00:26:34,805 improvement you might expect 729 00:26:35,184 --> 00:26:36,644 from increasing the GPUs. 730 00:26:37,105 --> 00:26:38,164 And I think that's 731 00:26:38,545 --> 00:26:40,865 I think that's really the thrust of the 732 00:26:40,865 --> 00:26:43,259 focus of this particular article is yeah. But 733 00:26:43,259 --> 00:26:44,859 the other thing he says is very interesting 734 00:26:44,859 --> 00:26:45,599 in this article 735 00:26:46,299 --> 00:26:46,799 is 736 00:26:48,779 --> 00:26:51,519 LLMs are getting better at fact giving factually 737 00:26:51,660 --> 00:26:52,640 correct answers 738 00:26:53,259 --> 00:26:53,759 because 739 00:26:54,059 --> 00:26:56,720 thousands of experts with specialized knowledge 740 00:26:57,085 --> 00:26:59,184 are doing extensive post training 741 00:26:59,965 --> 00:27:01,505 to correct the errors. 742 00:27:02,605 --> 00:27:03,105 Right? 743 00:27:04,285 --> 00:27:06,065 That is absolutely true. 744 00:27:06,605 --> 00:27:09,265 And what that means is is that 745 00:27:10,130 --> 00:27:12,210 if you ask the LOM to do something 746 00:27:12,210 --> 00:27:14,470 that it's never been asked to do before, 747 00:27:16,210 --> 00:27:18,210 it has never been trained to find the 748 00:27:18,210 --> 00:27:19,109 correct answer, 749 00:27:19,889 --> 00:27:21,909 and the answer is gonna be unreliable. 750 00:27:24,164 --> 00:27:25,144 That's an interesting 751 00:27:27,845 --> 00:27:28,904 the author says, 752 00:27:29,845 --> 00:27:32,325 elements cannot be trusted to answer prompts that 753 00:27:32,325 --> 00:27:33,865 experts have not anticipated 754 00:27:34,725 --> 00:27:37,525 nor to generate answers that require specific timely 755 00:27:37,525 --> 00:27:38,025 details 756 00:27:38,490 --> 00:27:39,869 and involve uncertainty. 757 00:27:41,369 --> 00:27:42,509 And that's probably 758 00:27:42,809 --> 00:27:44,190 always going to be true. 759 00:27:44,970 --> 00:27:46,730 I think that's just that's gonna be a 760 00:27:46,730 --> 00:27:47,230 limit. 761 00:27:48,329 --> 00:27:48,809 I saw, 762 00:27:49,529 --> 00:27:51,069 an article the other day, 763 00:27:51,755 --> 00:27:55,034 called generative I a hey. Sorry. Generative AI 764 00:27:55,034 --> 00:27:57,595 runs on gambling addiction. Just one more prompt, 765 00:27:57,595 --> 00:28:00,234 bro. And it talks about how Oh my 766 00:28:00,234 --> 00:28:00,734 goodness. 767 00:28:01,115 --> 00:28:03,274 It is actually really interest it's pretty entertaining. 768 00:28:03,274 --> 00:28:05,434 It's really interesting. It talks it's mostly around 769 00:28:05,434 --> 00:28:08,089 software developers use of, LLMs, but, 770 00:28:08,549 --> 00:28:10,710 it talks about how it it works the 771 00:28:10,710 --> 00:28:12,869 same way with the same reward system that 772 00:28:12,869 --> 00:28:14,869 the gambling addiction kind of feeds into and 773 00:28:14,869 --> 00:28:16,329 the devil knew path and everything. 774 00:28:16,710 --> 00:28:18,410 And basically talks about uncertainty 775 00:28:18,924 --> 00:28:21,325 and how, you know, as a developer using 776 00:28:21,325 --> 00:28:23,404 it, at it's, like, so awesome. It's like 777 00:28:23,404 --> 00:28:25,325 winning when you give it an instruction and 778 00:28:25,325 --> 00:28:27,164 it comes back and tells you what you 779 00:28:27,164 --> 00:28:29,085 need to do. But then a little while 780 00:28:29,085 --> 00:28:31,085 in, you get you you get information that's 781 00:28:31,085 --> 00:28:33,920 not useful from it anymore. And then that's 782 00:28:33,920 --> 00:28:35,680 kind of like kind of like losing, you 783 00:28:35,680 --> 00:28:37,440 know, in the whole in the whole addiction 784 00:28:37,440 --> 00:28:39,680 cycle. And so you're constantly looking for a 785 00:28:39,680 --> 00:28:41,839 win. And so, really, what it comes down 786 00:28:41,839 --> 00:28:43,200 to is can I prompt it the right 787 00:28:43,200 --> 00:28:44,480 way? And, 788 00:28:45,075 --> 00:28:46,514 so if I just if I just try 789 00:28:46,514 --> 00:28:48,755 another time, just one more prompt, maybe I'll 790 00:28:48,755 --> 00:28:49,875 maybe I'll be able to say it in 791 00:28:49,875 --> 00:28:51,335 a way that the machine understands. 792 00:28:51,954 --> 00:28:52,454 And, 793 00:28:53,394 --> 00:28:55,815 it it it's really compelling, actually. I 794 00:28:56,835 --> 00:28:58,619 I thought I was quite and and the 795 00:28:58,619 --> 00:29:00,220 the thing that he focused on was it's 796 00:29:00,220 --> 00:29:02,779 the uncertainty of the outcome that drives the 797 00:29:02,779 --> 00:29:05,900 cycle. And so, you know, I I don't 798 00:29:05,900 --> 00:29:07,420 know if this is that they're trying to, 799 00:29:07,500 --> 00:29:09,660 you know, ring a warning bell or anything, 800 00:29:09,660 --> 00:29:13,234 but, absolutely, our our brains are wired in 801 00:29:13,234 --> 00:29:14,994 a way that if we're not careful on 802 00:29:14,994 --> 00:29:16,054 how we use the tools, 803 00:29:17,154 --> 00:29:18,674 I think we could come out with some 804 00:29:18,674 --> 00:29:21,875 outcomes we were definitely not wanting here. There's 805 00:29:21,875 --> 00:29:25,174 a great older book called Addicted by Design. 806 00:29:26,880 --> 00:29:28,960 It's all about gambling, and it talks about 807 00:29:28,960 --> 00:29:31,680 this exact thing that it's not it's not 808 00:29:31,680 --> 00:29:33,940 that the reward cycle is random. 809 00:29:34,400 --> 00:29:36,640 That's what drives you that's what addicts you 810 00:29:36,640 --> 00:29:39,059 to the process. The reward pro the reward 811 00:29:39,519 --> 00:29:40,580 cycle is random. 812 00:29:41,194 --> 00:29:42,555 You would think if you wanted to get 813 00:29:42,555 --> 00:29:43,055 somebody 814 00:29:43,434 --> 00:29:45,355 addicted to something, you would give them positive 815 00:29:45,355 --> 00:29:47,855 resort resort rewards and negative rewards. 816 00:29:48,394 --> 00:29:50,795 Yeah. But you don't give them positive rewards 817 00:29:50,795 --> 00:29:51,615 all the time. 818 00:29:52,315 --> 00:29:53,694 You actually randomize 819 00:29:54,075 --> 00:29:54,815 the rewards, 820 00:29:55,440 --> 00:29:57,299 and that is what makes it addictive. 821 00:29:59,119 --> 00:30:01,039 I I think the reason this belongs in 822 00:30:01,039 --> 00:30:03,440 our current discussion is that the human factor 823 00:30:03,519 --> 00:30:05,119 because we talk about what can LMs do, 824 00:30:05,119 --> 00:30:06,399 what can they do. Well, they can only 825 00:30:06,399 --> 00:30:08,554 do what you prompt them to do. And 826 00:30:08,694 --> 00:30:10,934 it the human running it and and prompting 827 00:30:10,934 --> 00:30:12,615 it, if the human is is a is 828 00:30:12,615 --> 00:30:14,714 a is an AI addict, 829 00:30:15,335 --> 00:30:17,494 like, the human's interaction with the machine is 830 00:30:17,494 --> 00:30:18,714 gonna be highly variable. 831 00:30:19,654 --> 00:30:20,154 And, 832 00:30:21,119 --> 00:30:22,480 I don't know, but I think that plays 833 00:30:22,480 --> 00:30:24,240 into how effective these things are. And it 834 00:30:24,240 --> 00:30:25,539 also goes back to creativity. 835 00:30:26,880 --> 00:30:29,779 Addiction does not breed creativity, man. 836 00:30:30,480 --> 00:30:33,039 If you're addicted to gambling or computer game 837 00:30:33,039 --> 00:30:34,339 or whatever it is, 838 00:30:34,644 --> 00:30:37,285 you're not being creative. And, actually, this author 839 00:30:37,285 --> 00:30:39,464 makes this point. Right? Is that, 840 00:30:42,565 --> 00:30:45,444 LLMs are enormously expensive in terms of the 841 00:30:45,444 --> 00:30:47,944 resources devoted to chips, coolers, and other machinery. 842 00:30:48,539 --> 00:30:50,640 The energy needed to train and run them, 843 00:30:50,700 --> 00:30:53,600 and the human talent diverted from more useful 844 00:30:53,740 --> 00:30:54,240 pursuits. 845 00:30:56,779 --> 00:30:57,920 Sir. Right? 846 00:30:58,460 --> 00:31:00,860 Sure. That's When the early days of of 847 00:31:00,860 --> 00:31:01,360 LLMs, 848 00:31:02,115 --> 00:31:04,914 people were having these gotcha moments where they're 849 00:31:04,914 --> 00:31:07,255 trying to ask the LLM to do math. 850 00:31:07,634 --> 00:31:08,134 Right? 851 00:31:08,515 --> 00:31:10,755 And it's it's just the wrong tool to 852 00:31:10,755 --> 00:31:11,575 do math. 853 00:31:12,035 --> 00:31:13,954 If you wanna do math, you need to 854 00:31:13,954 --> 00:31:16,214 throw the math problem at a CPU 855 00:31:16,940 --> 00:31:19,420 in a very clear and deterministic way and 856 00:31:19,420 --> 00:31:20,960 let it just do the math. 857 00:31:21,740 --> 00:31:22,559 And I think, 858 00:31:22,940 --> 00:31:23,440 like, 859 00:31:24,619 --> 00:31:27,819 what are the right questions and tools or 860 00:31:27,819 --> 00:31:28,319 or, 861 00:31:29,259 --> 00:31:30,285 projects or 862 00:31:30,684 --> 00:31:31,184 experiments 863 00:31:31,565 --> 00:31:32,785 for the LLM 864 00:31:33,484 --> 00:31:35,965 versus other mechanisms. One of the things that 865 00:31:35,965 --> 00:31:36,625 I found 866 00:31:37,005 --> 00:31:39,565 is that people start talking about AI and 867 00:31:39,565 --> 00:31:41,805 they have all these ideas about, well, it 868 00:31:41,805 --> 00:31:43,345 could do this or it could do that. 869 00:31:43,509 --> 00:31:46,230 And half the time, the scenarios that they 870 00:31:46,230 --> 00:31:46,730 envision 871 00:31:47,990 --> 00:31:48,970 are just automation. 872 00:31:49,509 --> 00:31:50,009 Yeah. 873 00:31:50,390 --> 00:31:53,929 Right? Yeah. They're not actually large language models, 874 00:31:54,470 --> 00:31:56,730 or you don't even need one. It's like, 875 00:31:56,789 --> 00:31:58,835 oh, yeah. That's a process that that can 876 00:31:58,835 --> 00:32:00,755 be automated with a set of if then 877 00:32:00,755 --> 00:32:02,914 statements. And so I think we're still trying 878 00:32:02,914 --> 00:32:03,894 to figure out 879 00:32:04,355 --> 00:32:05,954 where they're fit for purpose and where they're 880 00:32:05,954 --> 00:32:08,914 not because as the author points out, it's 881 00:32:08,914 --> 00:32:09,734 very expensive. 882 00:32:10,275 --> 00:32:10,775 And 883 00:32:11,234 --> 00:32:13,734 and having an LLM do math, 884 00:32:14,240 --> 00:32:16,000 that that's a layer of abstraction you don't 885 00:32:16,000 --> 00:32:18,000 need because it's gonna be less reliable, and 886 00:32:18,000 --> 00:32:21,599 it's gonna require way more compute than just 887 00:32:21,599 --> 00:32:24,000 doing the math problem. And so we gotta 888 00:32:24,000 --> 00:32:26,180 figure out where the where the right places 889 00:32:26,240 --> 00:32:26,740 are. 890 00:32:27,825 --> 00:32:29,825 And and we're gonna do a little fumbling 891 00:32:29,825 --> 00:32:32,244 and bumbling in the process too, I think. 892 00:32:32,464 --> 00:32:33,684 Yeah. I think so. 893 00:32:34,704 --> 00:32:35,904 Well, I think that's a good place to 894 00:32:35,904 --> 00:32:37,744 wrap up. I don't have any other articles 895 00:32:37,744 --> 00:32:38,884 to cover, and 896 00:32:39,664 --> 00:32:41,299 I I think we've covered these two pretty 897 00:32:41,299 --> 00:32:43,960 good. Any last comments, Tom, before we 898 00:32:45,299 --> 00:32:46,279 go away and 899 00:32:46,900 --> 00:32:48,680 you can say LinkedIn really fast? 900 00:32:49,700 --> 00:32:51,700 Yeah. But that's for later. That's for later. 901 00:32:51,700 --> 00:32:54,440 Okay. Yeah. Yeah. What what about you, Yvonne? 902 00:32:55,494 --> 00:32:58,134 Did you use an LLM to reposition your 903 00:32:58,134 --> 00:32:58,954 frog lamp? 904 00:32:59,734 --> 00:33:01,974 I did not. I did that with my 905 00:33:01,974 --> 00:33:02,474 own 906 00:33:03,015 --> 00:33:04,234 sense of creativity 907 00:33:04,775 --> 00:33:06,575 and and spatial Serendipity. 908 00:33:07,015 --> 00:33:10,054 Serendipity. Yeah. Yeah. I just was like, man, 909 00:33:10,054 --> 00:33:12,420 nobody I should no longer put the frog 910 00:33:12,420 --> 00:33:13,400 lamp in the corner. 911 00:33:13,779 --> 00:33:15,000 Like, it should be visible. 912 00:33:15,539 --> 00:33:17,700 And I have an outlet right behind the 913 00:33:17,700 --> 00:33:19,460 frog lamp that was visible that I didn't 914 00:33:19,460 --> 00:33:20,420 like the way it looks. I was like, 915 00:33:20,420 --> 00:33:22,740 how can I cover that up? Frog lamp 916 00:33:22,740 --> 00:33:23,535 is the answer. 917 00:33:24,575 --> 00:33:26,575 It was serendipity. That's right. You solved two 918 00:33:26,575 --> 00:33:29,295 problems with one with one frog lamp. With 919 00:33:29,295 --> 00:33:30,434 one frog lamp. 920 00:33:33,695 --> 00:33:36,894 Oh, boy. We're getting deep now. Now I'm 921 00:33:36,894 --> 00:33:39,509 gonna have to go ask Midjourney to create 922 00:33:39,509 --> 00:33:41,589 an image of a frog lamp and see 923 00:33:41,589 --> 00:33:42,410 what it makes. 924 00:33:42,789 --> 00:33:44,250 Yeah. Okay. Good. 925 00:33:45,029 --> 00:33:46,890 Alright. Yvonne, if people 926 00:33:48,070 --> 00:33:50,150 if people wanna follow you or keep up 927 00:33:50,150 --> 00:33:51,509 with what you're doing, I know you have 928 00:33:51,509 --> 00:33:53,690 a different podcast, your own podcast now. 929 00:33:54,115 --> 00:33:56,274 Yeah. Not that the hedge isn't yours too. 930 00:33:56,274 --> 00:33:57,974 I mean, you're a ghost, so whatever. 931 00:33:58,434 --> 00:33:59,954 I mean, to put it that way, but 932 00:33:59,954 --> 00:34:01,634 talk to us. Where where can people find 933 00:34:01,634 --> 00:34:02,134 you? 934 00:34:02,755 --> 00:34:04,674 Yeah. So, you know, I've been a little 935 00:34:04,674 --> 00:34:06,994 quieter on social media lately because the day 936 00:34:06,994 --> 00:34:09,349 job is keeping me busy. But you can 937 00:34:09,349 --> 00:34:10,250 still find me 938 00:34:10,630 --> 00:34:13,030 on Twitter x at Sharp Network, and I'm 939 00:34:13,030 --> 00:34:13,690 on LinkedIn. 940 00:34:14,390 --> 00:34:16,309 The other place you'll find me is with 941 00:34:16,309 --> 00:34:18,230 the Cloud Gambit. And one of the big 942 00:34:18,230 --> 00:34:20,150 things that we have coming up in the 943 00:34:20,150 --> 00:34:21,050 July, 944 00:34:21,405 --> 00:34:23,244 Our podcast will be moving and joining the 945 00:34:23,244 --> 00:34:25,724 Packet Pushers Network. So Oh. You'll be able 946 00:34:25,724 --> 00:34:28,204 to find the Cloud Gambit over at Packet 947 00:34:28,204 --> 00:34:28,704 Pushers, 948 00:34:29,485 --> 00:34:32,125 mid July. That's really cool. I thought about 949 00:34:32,125 --> 00:34:33,965 bringing the hedge over to Packet Pushers, but 950 00:34:33,965 --> 00:34:35,744 every time I think about it, I get 951 00:34:36,440 --> 00:34:38,380 dissuaded by various people. 952 00:34:40,679 --> 00:34:42,779 Because Russ likes to be independent. 953 00:34:43,319 --> 00:34:45,559 Oh, I don't know. I get wrong with 954 00:34:45,559 --> 00:34:47,400 that. I get in I get all nervous 955 00:34:47,400 --> 00:34:48,299 about the whole 956 00:34:49,224 --> 00:34:50,284 advertisers thing 957 00:34:51,464 --> 00:34:53,324 and all that stuff. So, anyway 958 00:34:53,784 --> 00:34:54,284 alright. 959 00:34:54,744 --> 00:34:56,525 Tom, where can people find you? 960 00:34:57,784 --> 00:34:59,244 We can. See. 961 00:34:59,545 --> 00:35:01,224 That's, like, the shortest part of the whole 962 00:35:01,224 --> 00:35:02,250 episode. Alright. 963 00:35:03,289 --> 00:35:05,050 You didn't need an LLM to tell you 964 00:35:05,050 --> 00:35:06,650 he was gonna say that. That's right. No. 965 00:35:06,650 --> 00:35:07,309 I didn't. 966 00:35:08,170 --> 00:35:09,690 Alright. I'm Russ White. You can find me 967 00:35:09,690 --> 00:35:10,730 here at the hedge. You can find me 968 00:35:10,730 --> 00:35:12,889 at rule eleven at tech on I do 969 00:35:12,889 --> 00:35:14,730 log in to x Twitter every now and 970 00:35:14,730 --> 00:35:17,264 again. I'm on LinkedIn a lot more. So 971 00:35:17,264 --> 00:35:18,944 if you wanna DM me or give me 972 00:35:18,944 --> 00:35:20,545 ideas or if you wanna become a guest 973 00:35:20,545 --> 00:35:21,364 on the hedge, 974 00:35:21,744 --> 00:35:24,164 let one of us know. We would be 975 00:35:24,304 --> 00:35:25,824 glad to listen to you as long as 976 00:35:25,824 --> 00:35:28,724 you're not talking about worms eating plastic. 977 00:35:32,239 --> 00:35:34,579 It's a real thing. Don't even start. 978 00:35:35,119 --> 00:35:37,679 Anyway Hey. And maybe plastic eating worms will 979 00:35:37,679 --> 00:35:39,679 save the save the world, but it's just 980 00:35:39,679 --> 00:35:41,760 not the right Yeah. It's not it's not 981 00:35:41,760 --> 00:35:44,579 a networking problem. For us. Yeah. That's right. 982 00:35:44,844 --> 00:35:47,425 It's not a networking podcast thing, so cool. 983 00:35:47,644 --> 00:35:48,704 Alright. Awesome. 984 00:35:49,244 --> 00:35:50,844 We thank you for listening all the way 985 00:35:50,844 --> 00:35:53,244 to the bitter end of this episode of 986 00:35:53,244 --> 00:35:55,965 The Hedge. We know that attention is what 987 00:35:55,965 --> 00:35:58,445 drives our economy right now, and we thank 988 00:35:58,445 --> 00:36:00,305 you for giving us your attention 989 00:36:00,680 --> 00:36:03,739 as we talk about silly LLMs and creativity. 990 00:36:04,360 --> 00:36:05,960 Thanks for listening, and we will catch you 991 00:36:05,960 --> 00:36:06,700 next time.