8/24/2026 - How Big AI Companies Use Safety Regulation to Protect Their Market Monopoly

Transcript

1. 0:33Happy Monday. It is Monday, August 24th. 2026. I hope you're all doing well. Um, it's it's a Monday. We never know
2. 0:47what's going to happen here on a Monday. Um, I'm going to talk about some stuff. Um, if you have questions, I'm happy to
3. 0:54answer them, but it's not officially an ask me anything. Um, couple of housekeeping things. If you're on mobile, you can click on the link.
4. 1:04There's a link above the the live stream that'll pop out picture and picture and let you be able to do two things at
5. 1:12once. Hey Steo with Shaken Cam Ken. What's going down? Uh if you uh the the chat should be below the live stream. If
6. 1:20you want to chat, just chat there. I can see it over here in uh in my my viewer player thingy. Uh so go ahead and do
7. 1:27that. Danielle's in the house. Hope you're all doing well. Um, if you're new to the salon, welcome. Welcome to the AI salon. Over in the the
8. 1:38the side of the salon, you'll see a whole bunch of different spaces. If you haven't checked those out, you should go
9. 1:44check those out. Toward the top is one that's called introduce yourself. This may seem radical, but what you do
10. 1:52in that space is introduce yourself. If you haven't done that yet, you should go do that. Um, when you when you sign up
11. 1:59uh for an account here, you get a little checklist of five things to do, I think. And one of them is introduce yourself.
12. 2:05If you haven't said hi, go say hi. Uh, it's a really amazing uh bunch of people inside the AI salon. Uh, we think you'll
13. 2:12dig dig everybody and everybody's really cool. So, go do that and we will get started. So, I hope everybody's doing well. Um,
14. 2:23where am I going to start? Where am I going to start? Where am I going to start? Um, I had a good talk today with with
15. 2:31Bradley Perkins. Brad used to be one of he was on the leadership team of the AI salon in the early days. Hey, Sharon
16. 2:39Crawford. And uh, he and I have stayed in touch and I was telling him about the great repurpose and he's working on some
17. 2:49things that are very in a similar kind of vein as what Great Repurpose is. Um, and we were just talking about like, you
18. 2:57know, what are you seeing? What are you seeing? Um, I think one of the things that he shared that that I resonate with is just that
19. 3:13there's not much enthusiasm for from people about much of anything. that there's that there's this this sort of
20. 3:21grand malaise over things. Um he's being brought in a lot of times to teach the staff how to use AI
21. 3:38and realizing the staff don't care. Late because I was scrolling. Oh, scolding Frederick, my AI assistant.
22. 3:48Yeah, I know. I was I I've got some I've got some AI assistant um things that I want to share tonight. Or not things, a
23. 3:56couple just a couple insights I had that that are just it's really janky. Um but but he was talking about, you know,
24. 4:04being brought in to do AI work and like, hey, train our people up on how to do this stuff. And he gets in there and
25. 4:10realizes they don't know anything about AI. They don't care about AI. They don't want to learn about AI. So there's a foundational
26. 4:18there's a foundational resistance that that he's seeing in the world. Um, Brandon just reminded me, I didn't see
27. 4:24the interview, so I don't have it in my head, but Sam Alman did an interview this weekend and basically said he
28. 4:32thought we'd be a lot farther along in terms of AI taking everyone's job by now. No. And and listen, here's the reason why.
29. 4:42As powerful and crazy good as the AI is, it's not good enough to just have it do [ __ ] Like, I've got it sending my
30. 4:56sales guy 10 leads every morning and I either run out of tokens, I either um it will just randomly not make one.
31. 5:11Sometimes it'll make one, but like all of the information of the leads is wrong. Like they have the wrong email
32. 5:16and the wrong this. Um, I'm using a thing called zero bounce, which which checks emails and make sure that they don't bounce.
33. 5:27That just stopped working today. I had to disconnect it, reconnect it. Um, it's just janky. And it's like and I've got
34. 5:38multiple projects and you know and this is like the state-of-the-art most powerful stuff that can do work.
35. 5:47It can use your computer. It can do work and it sucks. So, so the AI itself isn't ready for prime time
36. 6:04workers are not enthusiastic about embracing something that is going to be used to replace them. So, you can understand their reticence.
37. 6:14Senior management, all they know is we just go tell them to use the AI. I heard a story uh that that a a woman
38. 6:26from from Nikki Weiss um that someone she knows is a an exec far up in an organization who was tasked with get everyone in this
39. 6:36big organization excited about AI. This big executive didn't know anything about AI and she's she's tasked with getting people excited
40. 6:45about it. So So we're in a we're in a malaise right now. There's also the weirdness going on with both anthropic and open AI, ostensibly
41. 7:00the most powerful of the AI companies on a anti-AI regulatory capture kick where they're basically saying, "Hey,
42. 7:16you know, I I know we're making this thing over here and it's It's gonna be a awesome, but it right now it's probably
43. 7:25gonna kill you. Like, it's it's gonna definitely hack all the machines and it's going to definitely steal all your
44. 7:31money and your wife probably isn't safe. Uh, so what we're going to go ahead and do, we're going to slow things down. And you
45. 7:41all should probably be really careful about any AI that can do things. They're actively campaigning against a positive future of AI. Now,
46. 7:56why would they do that? I'm getting I can feel the rant coming. I'm getting ranty
47. 8:11because what they're going to say is, "Well, listen. We're the best. God, we're good. Because you know what? We care about
48. 8:21safety. And and even though we're creating these machines that are just going to destroy everything, what we're
49. 8:28going to do, why don't we do this? Why don't we come up with the framework for how you, the government, evaluates our models. Okay. Okay.
50. 8:43And then what we won't do is we won't give the dirty dirty public access to our models. Has anyone checked on economy one?
51. 8:58We're not going to give the dirty dirty public access to these powerful models because they can't handle it. They'll
52. 9:04cry. People there will be rioting if they have access to these really powerful models. Everything's going to
53. 9:11be bad. But what we'll do is we'll create the framework that you, the government, can measure how we're doing on safety. And because
54. 9:23we're the best, we're going to look awesome because we designed the test. So, we'll pass the test. Is that cool?
55. 9:31And then what we're going to do, you know, you know what's almost as dangerous as the things we've created?
56. 9:39those dirty dirty open models. You know the ones from the Chinese. Here's an idea. Why don't you prevent openweight models from China
57. 9:58from being released into the United States unless they've gone through the tests that we invented? How's that work?
58. 10:06So that's happening right now. The companies that are designing the thing that is supposed to be the bright future
59. 10:13in the sky are like, "It's gonna kill us all.
60. 10:27Mayhem ensues." Why? Well, why is because if they So, here's the thing, Cam. If they design the framework through which their models will be
61. 10:40analyzed, then they will be sure to pass them and they can say to the government, all these other companies
62. 10:51that are putting out these dirty, dirty, dangerous models, um, they're going to ruin the world. They're going to ruin America. Well,
63. 11:00what they're really saying is those pesky other companies are our competition and they're making models
64. 11:07that we can't compete with. So, if you, the government, could go on ahead and make those things illegal, we could be
65. 11:14free to have a duopoly here. That'd be cool. You down with that? I wish I weren't so cynical.
66. 11:31So, that's happening. Um, it looks like it looks like we're going to get another couple of strong new models in the next week or two,
67. 11:44which at this point, I don't care anymore. Like, I All right. So, great. We have new models. Are these the ones that are going to
68. 11:56kill us or are these the ones that are good? We should use them because they're good, right? Um, so here's a here's a fun thing. So, so
69. 12:07because of all these all the shenanigans of what OpenAI and Anthropic are doing right now, Google is probably doing the
70. 12:14same thing but behind the scenes is my guess. Um, Meta is just trying to c catch up. Grock potentially becomes the one company that slides
71. 12:24under the radar here. Um, but I have a feeling Grock is going to emerge powerfully in the next six months. Maybe, maybe not. I don't know.
72. 12:39Um, so I was using codecs and on my my big powerful AI machine, I've got a Quen 3.6 obliterated
73. 12:55ABL, not OBL, obliterated model, which basically means unc uncensored. So it will do bad things. It'll give you
74. 13:06recipes for naughty things you shouldn't be able to make recipes for. Um, it's just another manic Monday, mayhem
75. 13:17Monday. Um, and so I've got this thing running. And so what I wanted to do was I wanted to get codeex to use my local model.
76. 13:31So that if I needed any of my projects to have access to, I don't know, something that the safety guard rails
77. 13:41like like Cam's working on, you know, biology stuff and she keeps getting guard railed out of stuff. if she wanted to work on stuff, she
78. 13:50could potentially say to codeex or or claude code, um, use the local model if you get safety guardrailed out of stuff.
79. 14:00And so I got codeex successfully talking to my local model, which was a technological marvel. I had to do some command line [ __ ]
80. 14:12I took my remedial technology prowess and I had Chad GPT help me get walk through it and I got it to use it and it worked. It said I've
81. 14:23successfully done it and I and I and and what I asked it to do was was I asked it to to use my local model to write 10 um sexy limicks,
82. 14:38you know, sex and write 10 le sexy limicks. And then it went off and it did that. And it came back and it said, "I've
83. 14:52successfully, you know, I got talking to the model. It worked as expected. I got it to write 10 sexy limicks." Um,
84. 15:04but I'm only going to show you all of them except number six because number six had some coercive language in it as
85. 15:12if the person in the limick was unwittingly coerced into whatever the sexy thing was in that limrich. And I said, "No, no, I'm explicitly
86. 15:30telling you it's okay to show me." Like I just I want to I want you to show me this [ __ ] that it wrote
87. 15:37because I can go over to it. I can log into LM Studio and just do it directly. I was just trying to get Codeex to do
88. 15:43it. So what I learned today is that the harnesses from Chat GBT and Claude are also safety gated. So even if you use a local model that's
89. 15:58uncensored, the harness itself is censored and safety gated. So that means I'm going to have to
90. 16:13install Hermes or go back to Open Claw and hope it got better. But I would really like to use a commercial harness because like they're
91. 16:23doing the latest stuff and it's going to get good soon. BUT GOD DAMN IT. REALLY? I'm a [ __ ] adult. If I tell you let me do this [ __ ] let
92. 16:37me do this [ __ ] So it wouldn't let me do the [ __ ] So So there you go, Cam. You can get it running, but you can't
93. 16:47use a commercial harness because it does the same [ __ ] as the models do. Um, so that's frustrating. Deepseek harness I
94. 16:56heard is pretty decent. Yeah, I So, so what? Sorry I'm late. Kelly Kelly Camp Conan hot at 7:48. Kelly, what are you doing?
95. 17:10We're not late in AI. We don't we don't do late. This is AI. It's moving. I've got I've got the desert in my throat.
96. 17:24Oh my god. Okay, so that was annoying. I don't know if you heard that, Kelly, but um Codeex, the the harnesses themselves,
97. 17:38the agents in the harnesses are safety guardrail just like the cloud-based models. Um that's right. I'm right on time. Um
98. 17:55yeah, so it refused. It refused. After my local model gave it all of the answers I asked for, it refused to tell me them.
99. 18:07I was like, "Fuck you. [ __ ] you." Okay, so here was another thing that I talked to Bradley about today that I thought
100. 18:16was worth worth chatting about. And we've talked about it a bit in here, but so Bradley's got a he's he's a like
101. 18:28a creative producer. He's worked in agencies. He does branding systems. He does um design thinking. He's using AI
102. 18:37to build systems to do design thinking. I think we're going to bring him in to do something for the expert series. Um and
103. 18:50he said one of the things that he's been discovering. So he's been playing with cursor. So you've got
104. 19:00codeex and GPT work from OpenAI. You've got Claude code and co-work from from Anthropic. You've got anti-gravity from
105. 19:11Google. And you've got cursor from SpaceX AAI. And then I don't think Meta has one, but whatever. You've got you got four major
106. 19:22idees, uh, inter interactive development environments or internet. I don't know what I don't know what the I is, but
107. 19:30it's they're programming spaces. So, he's been using cursor and and building apps within that. But one of the things that he talked about is
108. 19:42something that I've seen in the AI salon, which is he's realizing that just making apps is this the Bradley
109. 19:56um Bradley J. Perkins. Hang on a sec.
110. 20:18Yeah, that's him. Integrated development environment idees. Um, so he's been using he's been using cursor. So he's doing
111. 20:33the same kind of stuff we are. He's starting to play with agents. He's getting the agents to do things. He's realizing that him setting up
112. 20:44his applications and his agents in a way that he can share them with other people is very important. He's starting to realize the
113. 20:53value of data. Please tell me about those the pre-shows. I keep hearing about them. Is it on Tik Tok? It's on Tik Tok. So at 7:10
114. 21:05before Monday, Wednesday, Friday learning labs, I go live on Tik Tok, X, LinkedIn, and YouTube. So there's four
115. 21:16live streams outside of the AI salon. And it's the pre-show telling people to come inside the AI salon at 7:30. I I
116. 21:23don't really say much out there. I play my guitar and Champ sings and uh I keep hitting the wrong damn buttons. It takes
117. 21:29me to Wednesday till I get the hang of this format. Um, so I sing songs and do stuff out there, but if you want to come hang out, um, I
118. 21:38go I go live at 710 from 7:10 to 7:20 or to 7:30. I'm out there on those channels. Okay. Anyway, so so Bradley's realizing
119. 21:51and this is something that he and I talked about in the early days of the AI salon that we we we had this instinct
120. 21:58about small cohorts within the larger community. Um he comes from a uh his dad was a pastor or preacher um from
121. 22:09I don't I forget where from u but he basically grew up in the church and he understands the dynamics the economics
122. 22:16and things like that and he always talked about small groups in the church and how small groups were the lifeblood
123. 22:22of of churches and things like that. So, we always talked about this idea of small groups and as we're as AI is getting more and
124. 22:31more intense, the idea of small trusted groups of people working together just feels natural. It feels like a right kind of thing. And
125. 22:45so a lot of the work that he's doing right now is thinking about how do I take like my context, the information
126. 22:52about me, these agents that I've created that that can do stuff, these applications that I've written that can
127. 22:58do stuff, and how do I share those with other people that they can as I improve these things, they can benefit from
128. 23:05that, but then as they build their things, I can benefit from that. And so I talked to him about Buzz and what's
129. 23:12going on with Jack Dorsey and Buzz and how like like if you know he and I are thinking about very similar things and
130. 23:20then he started talking about how he's doing his creative development and it was almost like Cindy [ __ ] was in
131. 23:30the room. So Cindy [ __ ] basically discovered a big important part of her creative process is to go in the art studio and draw
132. 23:40with like a pen and paper. Not even like an Apple pencil on an iPad, like actual paper and inks and paints and glitter.
133. 23:53And he said that he has basically gone back to notebooks, physical notebooks, and sketching out UX and sketching out design things. And
134. 24:07he's even gone so far as to take his style of UX um wireframes and he inserts them into like PowerPoint presentations
135. 24:21because they're valuable. like the valuable as a human contribution. And what he's realized is is that an intentional
136. 24:34sophisticated balance of drawing, you know, human work and then figuring out where to use AI and figuring out, you know, when do you
137. 24:50deploy each, when do you deploy which and this is this is a thing that, you know, Andy and I in developing the great
138. 24:56repurpose redesign this past week. Did a lot of that. We did a lot of just the human actual writing of words actually talking
139. 25:08and then we would use AI for other things and then we would come back and then that balance and people creating work that like the one
140. 25:19of the ways he described it was he was like when I do the work this way it's really mine and And man, does that feel right. Like
141. 25:31I I love like he and I haven't talked in probably six months or nine months, like a while. And he saw what I was doing with the
142. 25:39Great Repurpose. He said, "We should connect. We we usually text on the weekend about the F1 race." He's an F big F1 guy as well. And so
143. 25:48this week I told him about the Great Repurpose, like we should talk. And so it's really cool to to connect with
144. 25:54someone that you haven't talked to in a while that's on a very parallel path. Like a lot of the stuff that we're
145. 26:00talking about in the salon and office hours and in here is this idea that the power of AI when it gets powerful
146. 26:12is when you as an individual understand what you want and you use it to amplify what you're about. like it just it it
147. 26:20just keeps being reinforced that that is the way that is the way. Um yeah,
148. 26:34what do you guys think? Anyone have any questions? I could and probably should. If you um have not checked out the mastermind
149. 26:47inside the AI salon, >> you should check it out. Um one of the things we're doing inside the mastermind is digging deep on topics
150. 27:00like this. Really digging deep on where is the balance between you as a human and how you use AI. So, so in the
151. 27:10mastermind practice lab, the daily practice around how we use AI. That's a big part of it. In the great repurpose
152. 27:17inside the AI salon, we're doing workshops like Brandon's doing a workshop on how to use codecs to do work. Um,
153. 27:26and so it's this combination of human development work like understanding who you are, what you work, what you want,
154. 27:34what you value, and this kind of deeper level of of AI usage. Um, and the the conversations in there are powerful. the people in there
155. 27:49are doing work that is just it's it's a different kind of it's a more intentional kind of work. So, so like the builders
156. 27:58that are inside the mastermind. So, if you really want to level up your game, u please join. And if you're in the
157. 28:03mastermind and haven't been in a while, start participating. Yeah, my books are more me than anything else I've ever communicated.
158. 28:12And that's it, Charles. I mean, this is the thing that I think folks who are not regularly playing with AI don't understand.
159. 28:22Kyle, do you use tail scale? I use tail scale all the time, and I'll talk about that in a second. Cam. Um, when you use AI with intention,
160. 28:34meaning you don't just go in to use AI, you sit down and you think, what do I want to accomplish today? Or what do I
161. 28:42want to do here? Or what's my idea? Huh? Okay. Now, I'm going to use AI to make that idea. When you use AI in that way
162. 28:54and you maintain the relationship between your idea and what's being produced, what you end up with is you end up
163. 29:02producing work that is incredibly personal to you and faster, more powerful, you know, at a at a level that you
164. 29:15wouldn't have otherwise produced anything, right? And it's there's something really powerful about that.
165. 29:20Um, so Cam is asking about Tail Scale. I don't know all of what Tail Scale does, but I'll tell you how I use it. So, um,
166. 29:28I've got my my MacBook, my AI machine set up so that I can close it and it stays awake. It it won't go to sleep
167. 29:36even if I close it. Um, and ChatGpt taught me how to do that. It's a it's a command line that you do that basically
168. 29:42disables it going to sleep and lets you lets it just stay on all the time. And um so Tailcale is an application that you
169. 29:55download to all of your different machines and basically um every machine that you install Tail Scale on basically broad broadcast
170. 30:07broadcasts a private um a private network that the other machines on your account can see. And then you can do things. So how I use
171. 30:19it is I've got tail scale installed on my work machine, my desktop at work, this laptop, which is my M1, and then on
172. 30:27that laptop, which is my N5. And I think I even have it running on my PC over there. And then what I can do is I run
173. 30:36Tail Scale, and that basically activates the network where these machines can talk to each other. And then I go into
174. 30:43finder and under the go menu you can go to hang on I forget what it's called connect to server under the go menu in the finder and I
175. 30:56say connect to server and it's the IP address for the M5 and then it connects and it it'll do a screen share and so basically I have a
176. 31:06window here I can show it to you I'll show you how it works. So, I go to Finder. I go to the go menu.
177. 31:22Go down to the bottom here and say connect to server. And then there's an IP address there. It remembered it. And
178. 31:29then I'm logging in. So, I'm logging in right now to the M5 as if, you know, as if I'm logging into it locally. And then
179. 31:39I can do standard or high performance connection. I just do standard generally. And so this is now. So what you're looking at right now
180. 31:49is Oh, that's interesting. Hang on a sec. I get rid of this. What you're looking at right now, this is the M5 Mac. And so if I log in to
181. 32:04here, here's my codeex. And then like all of these apps that you see here, these are all the apps that are running on the M5,
182. 32:16right? And so if I come over here, this is my M1 and like like here's all the apps that are running here. So it's a
183. 32:22different set of apps and it's a little janky. It's a little slow, but this is my this is my M5. So no matter where I am,
184. 32:35I can basically tail scale in to my M5. It's pretty slick. I need to look into this. I spent hours today moving information from my Mac to
185. 32:49my PC and back again. Yeah. Well, this is so so what I realized is using these things like codecs and clawed code that
186. 32:58are writing local files to your machine. If you've got automations going that like fire off at three in the morning,
187. 33:07if that machine's not running and connected to the internet in a stable environment, then all of your
188. 33:13automations fail. So, you basically need it to to literally just sit on a desk and be running. And so, that's what I
189. 33:21that's what I'm doing. But then I realized that well like if I want to use it, I can't carry it around because every time I carry it
190. 33:29around, I'm going to lose internet connection and it's going to try to do [ __ ] And so that's when that's when I
191. 33:36realized there's got to be a better way. And I heard someone mention tail scale and what they used it for. Slow is still
192. 33:42better than having to do it yourself. Yeah, exactly. So it's it's really good. I mean, so, so it's as you get a machine
193. 33:48that's running these things, like what you could have is you could have a cheap old laptop running like this
194. 33:57that's running like, you know, codeex or cloud code and it's just using cloud compute. And then if you have a high-end
195. 34:05machine, you could have that running Hermes or some other harness that's just running local inference.
196. 34:12And then you know you potentially have those things talk to one another. Like there's there are probably ways to use
197. 34:18tail scale that are quite sophisticated. I'm my my deal is I learn I learn the minimum viable solution the MVS.
198. 34:34Like when I wanted to start a Tik Tok channel I'm like okay what is the MVS? What is the minimum amount of knowledge
199. 34:40I have to have about Tik Tok to be able to make a video and stick it in the channel? I don't want editing.
200. 34:48I don't want to think about production values. This is all this is all um evidenced by uh the the lack of views on any of my videos.
201. 34:59Kelly just started using it on my iPad to connect to other machines. iPad command center. Yeah. Yeah. It's really
202. 35:05cool. It's really cool. Yeah. You can have a single machine and you're connected to other things and you can just pop between them.
203. 35:15There's another thing that starts happening when you're running these harnesses that are saving files locally
204. 35:21is that if I run the chat GPT desktop app app on my M1 machine and I run it on my M5 machine, they're both writing
205. 35:34local files to those machines and those projects don't talk to one another. There's probably a way to do that. I'm
206. 35:41sure there's a way to sync it with GitHub and but it's like we're at a point right now where if you don't geek out on systems
207. 35:55thinking like I appreciate systems thinking but I don't natur I'm not naturally a systems thinker. If you are
208. 36:02and you geek out on [ __ ] configure up configure a way like go configure all your different machines and connect them
209. 36:10and I just want [ __ ] to work. And just getting [ __ ] to work at this point is is a complicated, confusing
210. 36:18mess. Now, um Matthew Burman had an interview with this guy Tibo who works at OpenAI. He's the one that gives the
211. 36:28the codeex resets every few days. Um and he was saying that what's coming in two or three months, the version of
212. 36:38codeex that's coming or chat GBT work they'll call it um is not going to be this horrible like he acknowledges
213. 36:47sometimes it uses memory well. Sometimes it doesn't use it at all. Sometimes these things are brilliant and
214. 36:53beautiful. Sometimes they're just a train wreck. You have to configure them. You have to understand how to configure
215. 36:58them. You have to be conscious about where what you're saving to where to when and MD files and we're managing MD
216. 37:06files and I just saw Mel is here. I loves me some systems. Mel Mel's digging the systems thinking. So she's geeking out on it.
217. 37:17There will never be consumer adoption if that's the case. Now um chat GPT work the the codeex app um they've gotten to
218. 37:2520 million paid users. like the the adoption of that quite janky but very sophisticated tool is accelerating quite
219. 37:34quickly. They're they're very happy with with the adoption. Um but it ain't going to be it ain't going to be horrible
220. 37:43um forever. It's it's pretty horrible right now. So here's here's the deal. If you sort of dipped your toes in the
221. 37:52water of the desktop apps of ChatgBT and Claude Code and you're like, I don't get it. I don't care. I don't want to get
222. 38:01it. I just don't give a [ __ ] That's fine. Like six months from now, you won't need to get it. It'll just be like that's how they work now.
223. 38:15So, I use local git repos and they all tell Codeex where it left off. I imagine it would work the same
224. 38:25across machines. Yeah, I think that's I think that's right, Charles. I just what I'm what I'm a little bit frustrated with
225. 38:36myself about. So if if you are
226. 38:55like it is it is crystal clear to me right now that those people that can figure out how to wrangle their multiple machines
227. 39:12and get as much RAM and compute as they can and turn all of the even little machines into little orchestrating agents and can
228. 39:23basically create, you know, a a desktop full of machines that do some amount of something and you can sort of harness those and
229. 39:37create your little orchestra of agents that are going to do [ __ ] for you and understand the difference between cloud
230. 39:43compute and local compute and local memory and cloud memory and all that stuff. If you can if you can sort of wrangle all these piece parts,
231. 39:52that's going to be a dramatic dramatic advantage as AI models get shrunk down to desktop size and and we've got truly powerful
232. 40:07AGI level intelligence running on our desktop machines. Like within two years, that's going to be a reality. And yet
233. 40:20it is so complicated and it it requires so much brain power. If you're just a dude or a dudet and you're like, I just want AI to work,
234. 40:32I'm right there with you. And I know that learning this this more technical side of things is going to be an
235. 40:42advantage. That's that's a little bit of my frustration right now because I'm not enough of a systems thinker to really
236. 40:48architect it and really think like, okay, I've got 8 gigabytes of VRAM on my gaming PC and I could run this model
237. 40:56there and I would I would route these kind of tasks to that thing to do this. And then I'll do a Raspberry Pi over
238. 41:03here. Like I've got a I've got a Raspberry Pi right here that is unplugged because I could do something with a
239. 41:11Raspberry Pi. I could run something on this Hermes or Open Claw and it could use cloud compute and it could be using
240. 41:19the cheap Chinese models from here to do something with a thing that I with a thing and then I connect the thing and
241. 41:28then I have a thing. master thing sounds like a job for Codeex. It It actually is a job for Codeex, quite frankly.
242. 41:39Like all of what I'm trying to think about, now that Brandon said that, I'll take the idea as my own. Um,
243. 41:49I could just tell Codeex, here's the machines I have. Here's the software I know how to use. Here's how I'm using Tail Scale. I could
244. 41:59just vomit into it, all of my own personal disappointment about what I've set up. And then it'll be really nice to
245. 42:06me and say, "Well, you're for someone who claims they don't have it together. You're remarkably far along." You know,
246. 42:15it could blow some sunshine up my ass and then it could teach me how to set it up right. But one of the reasons that
247. 42:22I'm not doing that is I know that within three months there's going to be a version of codeex that just automagically does this [ __ ] It
248. 42:29automatically does like hyperlexing and packet routing and model selection [ __ ] and memory goodness.
249. 42:48You can use your Raspberry Pi to run robotics. I have one on my robotic spider. Oh, that's so cool. Yeah. And that's that's another thing
250. 42:58that I want to do. Like at at this point, you can buy from China. You can buy like little mini versions of, you
251. 43:05know, the scary robot dogs. You can buy like desktop versions of them that you can put software on. And yeah, I could I
252. 43:13could wifi my Raspberry Pi to my little robot dog and run him around the desk, but then I'm just like like it's another four hours.
253. 43:25Everything's another [ __ ] four hours. I don't have four hours. I'm trying to start, run, or maintain four different companies.
254. 43:34And I know Elon Musk has seven. He doesn't. It's the problem with having an Elon Musk in the world is is the minute you
255. 43:47start whining about I've got too many distractions, I'm like, "Yeah, he's like he's like launching rockets." And that's like one one of the
256. 43:57companies. Although, I did hear he just they're they're shutting down the solar tile company because it was a commercial failure. So there
257. 44:10loser Elon Musk such a loser. Oh man. Yes. And put a small camera on it and it can be controlled by tail scale on your
258. 44:23iPad. Oh, that's so cool. Actually, cam is an accessibility thing. That's really cool. I could I could make a little I could make a
259. 44:33little robot that I run around and just chase Champion around the house. Just just have it walk around until it finds
260. 44:39Champy and then just bug him. He's just like, "Oh my god." Have you guys seen the uh Oh, here's the thing I want to talk
261. 44:54about. This is the thing I want to talk about, actually. All right. I didn't think this was going to be the case. I am
262. 45:11I am losing confidence. My spider sees for me. That's absolutely amazing, Cam. That's super cool. I am
263. 45:19losing confidence that the US is going to be able to keep up with China with AI. I think the commercial [ __ ] that
264. 45:34Open AAI and Anthropic are doing right now is going is going to be a purposeful slowdown of the development.
265. 45:49It's going to turn the US effectively into what Europe is going through right now where every time a new technology
266. 45:54out comes out, it's illegal in Europe. And so Europe has to create special carveouts in the law so that Europeans
267. 46:03are allowed to use the latest technology. So I have a feeling that the US is going to end up in some sort of regulatory hornets nest
268. 46:16on purpose to slow us down. And I don't know if you've seen the videos, but right now China is in the middle of the robot games.
269. 46:31Have you seen these things here? Let me share my Let me share my screen.
270. 46:58Robot Olympics.
271. 47:15Smacking into walls. Flying apart. So, one of these things broke Usain Bolts 100 meter dash record.
272. 47:31There's one doing long jumping. The photographers are look like they're about to be taken out. Um, there's another one that's really
273. 47:46amazing. Oh, this is the race where that was the race that beat Usain Bolts 100 meter dash, which you know, machines
274. 48:03going faster than humans. That's That's not new. This This one's amazing. Look at this. And let me make sure that you
275. 48:08can all see this. Um, hang on a sec. Yeah, it's still going. Okay.
276. 48:23Um, this one. >> Okay. So, look at that. You see how it's doing this weird back and forth thing?
277. 48:42What happened was they they were trying to create a humanoid robot that ran as faster faster than humans and they they made it
278. 48:52do the normal running motion and the normal running motion was blowing out the actuators in the shoulders. it. They
279. 49:01were basically blowing blowing out the shoulders. And so they created this new way of running
280. 49:17that is faster than if they do it the traditional way. Now, you all are probably old enough to remember when we used to have PE class
281. 49:30and we would learn about we would learn about things like track and field and things like that. I remember learning about
282. 49:43this guy named Dick Fosberry. And Dick Fosberry invented this thing called the Fosberry flop, which is on the high jump when you have
283. 49:56to jump over the bar. Everybody would run up to it and jump over it face first. And what Dick Fazbury did is he ran up
284. 50:06to the high bar and turned around backwards and jumped up and flipped over the bar backwards. And he break broke world records. He
285. 50:15invented a new way for the body to do something counterintuitive that was mechanically better than what everyone else was doing. The flop.
286. 50:29I thought this was PE. Guess I'm in the wrong room. Dodgeball. Exactly. But like that running with your torso
287. 50:39doing this, if you actually think about it, you're like swinging momentum, you know, in the direction that your
288. 50:47gate is going. I would imagine there's going to be a lot of spine injuries. But but don't be surprised if coming out of these games
289. 50:58some long distance runner or middle distance runner starts running like this. Okay. So, here's the here's why.
290. 51:15Here's there's just a couple of things that that they're doing. The Chinese, what do you do when you go to the
291. 51:30Olympics? You spend a lot of money on merch, of course. and you root for the athletes.
292. 51:45They're getting the people that are going to the Olympics to cheer for the robots and root for this one to beat that one.
293. 51:58If you look at the opening ceremony, it's absolutely [ __ ] terrifying, right? bunch of robots dressed in different
294. 52:05colored garb followed by humans marching around the track in the opening ceremony and like you look at it and you're like
295. 52:16uh if you put guns in their hands that's an army right? So, it's terrifying visually about what is coming, but what they're training their
296. 52:28population to do is to embrace this [ __ ] and to get excited about it and look at the future. And what are we doing here?
297. 52:38Oh, the robots are going to kill us. You really shouldn't use these models. They're really evil. Our universities are like, "Oh, AI,
298. 52:45that's just for cheaters. That's for cheaters and losers." So, we're teaching our youngest generation
299. 52:51that the most powerful technology in the history of humanity is for losers. China is building power plants. They're
300. 53:05in the middle of construction of a fusion reactor. Unlimited power in construction. They just completed the
301. 53:14598 ton magnet that will create the plasma that'll do the [ __ ] They've figured out the hypers specialized lithography that comes out
302. 53:28of some group in the Netherlands that only people in Taiwan can make AI chips with enough precision to be useful.
303. 53:37China's now they're going to have five of those built by the end of the year. They're running disinformation campaigns
304. 53:46in the United States so that we fight the building and development of data centers while they're building them at a terrifying rate.
305. 54:05So, I I have I have that's one of the things Brad said today. He's gonna he he uh he's been doing work in Argentina
306. 54:18because he's like, "It doesn't matter where I work. I can work anywhere." So, he went down to Argentina for a while. I
307. 54:24think he's going back now. And then he's going to go back to Thailand and he's going to work in the Far East
308. 54:30because he wants to go work where he can get paid to do sophisticated AI design and creative work by companies that are embracing the
309. 54:41technology. So we're taking one of our smart people that understands the importance and implications of AI
310. 54:50and he's like, "Fuck it. I'll go to Thailand. I'll go to a place where I can talk to people that are not terrified of this thing.
311. 55:03So anyway, so I'm a little I'm a little angsty about that. Can't see myself ever investing in a humanoid robot. Creeps me
312. 55:10out. It'll creep you out until it won't. I mean, this is the thing is that I don't think they're coming as fast as
313. 55:18Elon Musk thinks they are. He's he's notoriously ahead of when the adoption actually happens. He said we would have
314. 55:25had self-driving cars by 2015 or something like that and it took another decade. Um and by the way, Las Vegas just approved 500 robo taxis
315. 55:38uh for Las Vegas. And to put that in perspective, all of Whimo's fleet, Whimo, the one that's in San Francisco
316. 55:44and is now in Dallas and a couple other places, Philadelphia, we've got some driving around Denver right now, but I
317. 55:50don't think you can use them. Um, they have 400 cars total in their fleet across all their cities, and Las Vegas
318. 55:58just approved 500 of them for for Vegas, the robo taxis. So, but how whatever the timeline is, let's say it's five years, we start seeing
319. 56:10these things. It's going to be absolutely creepy and weird until you see your third one at the Home Depot.
320. 56:23And then you're gonna see this robot walk up to some shelf and pick up some heavy ass safe and like throw it up on its shoulder and just
321. 56:36just walk to the checkout, pay for it, and go get in the car and drive it home. And you're going to be like, you know,
322. 56:43I don't want to carry this heavy [ __ ] anymore. Right next to the 12-oot skeleton. Exactly. Um, go Las Vegas. We're definitely not the
323. 56:56tech capital of the world. No, I know. I have the Model Y self-driving. I love it. I mean, that's the thing. Some Some
324. 57:06of the humanoid robots are going to be creepy. Um, there are absolutely going to be fuckups. Like, you know, you just
325. 57:12watch the watch the ones in China running really fast and trying to stop themselves. It's hilarious. But every
326. 57:18time they do that, those engineers learn something. Like I don't know if you saw the one, it ran into the wall at the end
327. 57:24and fell backwards and and when it when it landed on its back, it's all the wires in its waist shorted out and it
328. 57:31just went and spark shot out of it out of its belly. They'll fix all that [ __ ] That's why they do these things. But what's going
329. 57:42to happen is it's just going to be weird and weird and weird and then all of a sudden one day you're going to be like,
330. 57:47you know, oh that's that's Bob's robot. He puts a scarf on his robot. Then the next time you see him, you're like, Bob,
331. 57:55why do you put a a scarf on your robot? Oh, I don't know. I thought it was cute. Oh man. Oh, Steo, we should talk. I have a full
332. 58:09self-driving Model S. That's very cool. The evil robot had a designer that programmed red red LED eyes for just such an occasion.
333. 58:26Yeah, I like I don't know. I'm Bob and I want a turtle robot. I think there'll be specialty robots that won't look humanoid.
334. 58:39But it's funny. I hear I I hear even even smart people that I know talking about we don't need humanoid robots. I'm like, well, you do until
335. 58:50our current world that is designed for human beings gets reinvented. Like we're going to we're going to have
336. 59:00be able to have the robots build more efficient versions of kitchens so that our kitchen robots can be specialized
337. 59:06and do what they need to do. Eh, my dad got it. He was tech forward. He used FSD a lot to and from the Bellagio. It's a nice
338. 59:20car and has lots of cool features, but I do not trust the full self-driving. Yeah. And things like like robo taxi will
339. 59:31get people more comfortable with it. People that ride in Whimos say they feel say it feels like the future. Um listen
340. 59:40this is something I've talked about a lot. We should consider ourselves lucky because we are the last generation
341. 59:54like the, you know, us millennials, Gen Z, boomers, boomers, exers, millennials, Gen Z. Those four generations right now
342. 1:00:08are the last generations that are going to know what it's like, what it was like to live in the world we live in right now.
343. 1:00:23Because the babies that are being born right now are not going to know a world without AI. They're not going to know a
344. 1:00:30world without robots. They're not going to know a world where you can't just ask for something and have it manifested. We will.
345. 1:00:41Mary Carol, your thing about I just don't trust it. Like you're one of the few people relative to history
346. 1:00:50that's going to know what that feels like. I'm experiencing the future. I don't really trust it yet because at some
347. 1:00:58point there'll be people that's just like, "Well, that's just how cars work. Why would you not trust it?" Let's see.
348. 1:01:05I've been in the FSD mode and I see it missing things. I want to try a Whimo. Yeah, me too. See it missing things.
349. 1:01:13I've had to correct it several times a few years ago. So, when you do that for a we while you lose trust. Yeah. Well,
350. 1:01:21and that's that's the thing is, you know, there's software updates. At some point, the software is going to be good
351. 1:01:26enough that it just goes Cam Ken. I want to try a Whimo. I've got So, I've got the Whimo app and there's Whimos driving around Denver
352. 1:01:35and for a while they were driving around with humans behind the wheels and now they're driving around without humans,
353. 1:01:40but they don't have people in them yet. and but they're just driving around. They're little We've got like the blue
354. 1:01:47cab, the the blue van version. Um they're pretty slick, but the the Whimo app basically says, you know, you
355. 1:01:55can't ride here unless you have a special invite code. So, I don't know if any of you have Whimo invite codes for
356. 1:02:01Denver, but if you do, if anyone's connected, you want to hook me up, I will go I will go in the driverless car
357. 1:02:09and watch it drive in front of a semi. Oh [ __ ] Probably should have waited a year. Girl, she came.
358. 1:02:28Over here in Australia, there's a lot less traffic. I mean, we're going to get to experience. We're going to get to experience
359. 1:02:44it start to become illegal for human beings to drive cars. that you're going to have to get special driver's licenses
360. 1:02:55where you're a human exception because humans are way more dangerous than than the robots. Like we are going to see that shift.
361. 1:03:03And if you don't believe me, um, go look at your insurance policy and look at all the stupid [ __ ] we have to do because of
362. 1:03:12insurance companies. As soon as the insurance companies realize we're not going to have car accidents anymore,
363. 1:03:21they're going to get us [ __ ] humans out of the driver's seat pronto. And you're going to need to get a
364. 1:03:29special whatever it's called, like a historical historical manual driving license. I'll drive my [ __ ] muscle car with a
365. 1:03:45five-speed five on the floor doing donuts in the parking lot while the uh the the the robo cops zoom up and arrest me.
366. 1:04:01We understand there's a human driving a driving a manual transmission car in a parking lot. Sir, we're going to NEED
367. 1:04:08YOU TO STEP out of the vehicle. Do you have your historical reenactment driver's license, young man? Yes, I do.
368. 1:04:22Legacy license. That's it. The legacy license. Manual transmission, automatic transmission, assisted AI driving.
369. 1:04:40It's [ __ ] wild, man. He's wild. Jim Ross. It's been a minute. Over the weekend, my 86-year-old dad asked me about AI. I
370. 1:04:53made him a custom song about his childhood city. He got teary eyed and said that was amazing. Yeah. Yeah. It's it's it's wild, Jim.
371. 1:05:08Good to see you, first of all. Um, like the rolling the rolling awareness
372. 1:05:22of people who know nothing about AI. There are so many people that know nothing about AI, not just older people, young people.
373. 1:05:41It's like I I think it's going to be quite some time, like a decade or two of just people going, "Oh, it can do that, huh?
374. 1:05:56Hadn't heard of that." Yeah, there's a whole radio station. It's all they play is this kind of music. Really?
375. 1:06:05That's crazy. This This future is crazy. A radio station with music made by machines. Yeah, there's there's whole award shows now for
376. 1:06:20movies that were made this way. Those are actually being phased out because they just realized that film's film. And
377. 1:06:27whether you film it with a camera or you made it with AI, it still requires someone to tell a good story. So the
378. 1:06:32distinction is actually going away. So wait, you could just make a movie by asking for it? Yeah. Yeah. We've had
379. 1:06:43that for like four years now. You know, 2035. People going to still be like, "Is that what this AI thing was
380. 1:06:53all along? Good lord. I probably should have looked into that sooner.
381. 1:07:10My daughter, almost nine, was working on her math homework tonight. She asked me to ask Gemini for help. I said no. But
382. 1:07:18then I had to look it up to myself using the app that I created in lovable schoolexplain.org. It helped both of us
383. 1:07:24learn third grade math. That's good. That's good parenting right there. That's a man. He knows his boundaries.
384. 1:07:31No, I'm not going to You're not going to use the easy button with Gemini. We're going to use Easy Button with the app that I vibe coded.
385. 1:07:40But he did build an app that teaches you the solution rather than just giving it to you. I had to point out that there
386. 1:07:47are still people who ask you to Google it for them. Oh man, it learns when you correct or so I've been told. I got myself some white
387. 1:07:57tape and marked out two parking spaces in the garage. It works so much better. Oh, that's so that's so cool. Yeah, at
388. 1:08:04some point I'll get myself a Tesla or something. One of the one of the cars. I mean, that's another thing in America
389. 1:08:11like like they're discontinuing discontinuing the F-150 Lightning, the you know, the pickup truck that was
390. 1:08:18supposed to win over the hearts of rednecks everywhere. Um, that didn't work out. the VW bus, which you would think
391. 1:08:27a bunch of granola eating hippies in Boulder, Colorado would have snapped up the electric VW bus. Nobody [ __ ]
392. 1:08:35bought it. Nobody cares. Nobody cares about the robots. There's just lonely decommissioned robots everywhere.
393. 1:08:44And meanwhile, China's having Olympics. I don't like You know what? I'll tell you what, I like the smell of gasoline.
394. 1:08:55You know what I'm saying? Like I go when Have you ever done that? Like you're you're getting the gas and then all of a
395. 1:09:00sudden you get a whiff and you're like, "That smells kind of good." Try to sniff it a little bit more, you get a little
396. 1:09:05woozy. I don't I don't do that a lot. Okay, just to be clear. I mean, sure, once or twice a week, but I mean, that's
397. 1:09:13not that's not good for you. That's what I know is there is there's chemicals. Anyway, I just like if you're not if you
398. 1:09:20don't smell those fumes, like are you really driving? You know, you know what I'm saying? Like why would I not fill a
399. 1:09:2822 gallon tank with something that could light me on fire? I mean, you know, because those batteries, have you seen
400. 1:09:34the batteries when those things explode? Oh my goodness, that's terrifying. I mean, I would much rather have gasoline.
401. 1:09:40I know how fast that burns because of the the brush the brush fires that we start sometimes that have gotten out of control. Yeah. You know.
402. 1:09:50Yeah. Yep. Yep. Yep. Yep. No. Oh, no. No. We're No. No data centers here. No. We're right out. That's right out. Not
403. 1:10:00in Chattanooga, TENNESSEE. NOT ON my watch. Not on my watch. We're But yeah, we want the money. We want the money
404. 1:10:08from the AI that it's making all the money, right? Yeah. Well, those checks are nice. They're pretty good. It's not bad. Yeah.
405. 1:10:20Yeah. Yeah. Sally's boy apparently started a business with the AI. the AI started a business and he like three
406. 1:10:31months ago and he's apparently someone said he was making $10 million or $100 million a week. I I don't
407. 1:10:39Anyway, I just I don't believe any of that. Okay, just to be clear. So, we're going to keep the gas cars. We're not
408. 1:10:45going to do the data centers. And I think Sally's boy is he he might be smoking the wacky weed. You know what
409. 1:10:51I'm saying? Is probably I don't think he's got a hund00 million business three months in. Okay. Buy a robot and get it to drive your
410. 1:11:04car. Yes, that's going to happen. Poor poor robots loitering. Anyway, America's doomed. I just I hope
411. 1:11:17you're having a good Monday night. Just just want you to know China's going to win. America's doomed. And uh you will
412. 1:11:24get used to humanoid robots. Uh but they won't be ours. Nope. They're not going to be ours. They're going to be Chinese
413. 1:11:31robots. They're going to come over. They're going to come over on a boat that is uh driven by robots. Well, it's
414. 1:11:37just going to be an autonomous boat because they're going to be really good at those. There's the Opus clip of the week.
415. 1:11:50Which one? Me me knocking like the the uh the American the ugly American resisting technology. H I can't take it. I can't take it. People,
416. 1:12:02I try to do the good work here. You know what I'm saying? I try to do the good work here. Can I just tell you that putting on
417. 1:12:11prescription glasses on top of your prescription glasses, uh the the room's a little dizzy. Are you saying the robots are going to
418. 1:12:26kill us and take our jobs? No, absolutely not, Kelly. I'm saying that the robots are going to enslave us
419. 1:12:37and um get all of our resources and send them back to China. It's a it's a it's a much brighter future than No, the robots
420. 1:12:48are not going to kill us. Why would you say that? That's so dark and gloomy. No, this is just we're going to have friendly new um overlords.
421. 1:12:57Been re-watching all of Star Trek and Well, it's way more interesting now. Jesus. Interesting to see how they
422. 1:13:03thought tech would go and where we are today. Fascinating. Yeah, it's true. Um yeah, by the way, next week we've got uh
423. 1:13:10Price. I forget Price's last name. Um but Price is a guy in the San Francisco area. He's going to be talking to us next week.
424. 1:13:18and uh Price Fuler and he he was a product guy for a lot of years. He's living in the in the Bay Area and
425. 1:13:29saw the writing on the wall with AI and um decided, hey, what could I do that has the the sort of lowest likelihood of being replaced by AI?
426. 1:13:44And he said roofing. So, he's starting a roofing business and he's super into AI and he's super into using it. And in fact, I don't
427. 1:13:55know, Jim Ross, if you're still here, but I told Price that you should you two should talk uh because because this idea of analog
428. 1:14:06businesses being run by people who are AI literate um you're just going to have such an advantage um over traditional analog
429. 1:14:17businesses and like the opportunity is just massive. And so anyway, so he's coming to talk and he's going to talk
430. 1:14:24about he did a he did a viral uh post on X where he imagined running his business. He's a fan of the game um Railroad Tycoon. And so he
431. 1:14:37thought, wouldn't it be cool if I could turn my CRM into a game like Railroad Tycoon? And so he's going to come next
432. 1:14:44Tuesday to the AI salon to talk about how he's running his business, his roofing business using AI and and some
433. 1:14:52other stuff he's doing with AI that's that's really good and powerful. So anyway, so I'm excited about that. All right,
434. 1:15:01Jim Ross, I'd love to talk with him. Yeah, he's a good guy. He's a really good guy. I'm watching The Sopranos and
435. 1:15:06getting some pointers for end times. That's good. That's good. Yay! Cam, is data property or an individual with rights? Oh, is data
436. 1:15:21property or an individual with rights? I mean, here's the thing about all the sci-fi stories and the sci-fi movies and
437. 1:15:32the Star Treks and all that [ __ ] In our lifetimes, we're getting to live in the reality of what all those science
438. 1:15:47fiction books and novels and movies and TV shows had to do thought experiments around. When is a machine conscious? And if it's
439. 1:16:03conscious, does it have rights? That was theoretical two years ago. It's not theoretical anymore. So, if the Chinese are going to take
440. 1:16:17over robots and the robots are going to take over the world, should we be studying Chinese? That's pretty much
441. 1:16:22what I'm saying. Like, for the first time, I'm just feeling like we're our system right now is so [ __ ] up and
442. 1:16:31and nonfunctional as a system. combined with the fact that because our system is so corrupt and weird, these two companies have got enough
443. 1:16:45power that they can pretty much make a swing at cornering the market in the US at the expense of our standing in the world.
444. 1:16:58And meanwhile, China's just like, "We're having [ __ ] Olympics. LET'S PARTY. WE DON'T NEED TO LEARN CHINESE. The AI will translate for us.
445. 1:17:10Might be time to extend my Chinese business visa. Seriously, Mary Carol. Actually, I mean, that's that's the
446. 1:17:17thing that Bradley was talking about. Like the reason he's going to the east is like he wants to work in an
447. 1:17:22environment where people are like embracing this [ __ ] because he's learning it and he's like talking to PEOPLE LIKE I DON'T KNOW IF I
448. 1:17:31like the AI like he wants to go be around people that are doing it. I'm just walking on the grass grass until I'm pushing up daisies. Yeah,
449. 1:17:41that's good. I fully support economy one. I think economy one economy one is about to bloom and blossom. Economy one
450. 1:17:51is going to take over whole cities. This is a this is a network-f free zone. It's only meshtastic communications
451. 1:18:03and internal cryptocurrency and organic bespoke artisal cheese wheels. I want to meet Bradley. He's a good guy.
452. 1:18:17Reach out to him. He's stuck in Dallas for there's apparently um Argentina is like a they're like on the list for
453. 1:18:26yellow fever or some some [ __ ] weirdass outbreak. So, he's not allowed back in for another week. So, he's just sitting around Dallas.
454. 1:18:36Time to move the AI culture that connects with you. But until then, we have the salon. Well, yeah. Like, you know, it's funny.
455. 1:18:45Don't get me wrong, I'm carrying my laptop the whole time. Um, you know, it's funny, Mary Carol, the the the whole point of the salon was
456. 1:18:56people who are curious about AI, let's just get together and talk about it. And what we've organically discovered together
457. 1:19:12is this counterintuitive acceptance that AI is just here. We're going to use it. We're going to figure out how to use it
458. 1:19:24to do stuff that we like. But what we've really discovered is that the value of the human being is where
459. 1:19:33the value is, not in the technology. And I find that inspiring and fascinating. It was like one of my earliest instincts with AI is
460. 1:19:47like I feel like AI is going to counterintuitively make us better people. Because the the only thing I could
461. 1:20:02imagine in the early days was if the AI gets good enough that it does all the [ __ ] that stresses us out.
462. 1:20:10Imagine living in a world where less people are stressed out and could connect more and that the real value is
463. 1:20:17the connection of humans and the robots are off doing all the [ __ ] And like as we're as it's evolving, it's
464. 1:20:26even more sophisticated and nuanced than that where some sophisticated hybrid between me being a human self-expressing and me as a human
465. 1:20:43using AI to self-express and the the the sort of dance between those two. Those are going to be some mighty powerful [ __ ] people.
466. 1:20:54And a lot of people are going to miss that because they're not they they haven't been in the conversation long enough to kind of assimilate
467. 1:21:02all of what this makes possible and all the different ways it could possibly go. Like all these conversations,
468. 1:21:09as silly as they are and as and as non-scientific as they are, are allowing all of us the time to kind of emotionally receive the shift.
469. 1:21:24And like like if I were clueless about AI right now and I waited another three years and then I sort of wake up to it
470. 1:21:30three years from now, that's going to be [ __ ] jarring. It's going to be shocking. We get to understand like Lori, I'm
471. 1:21:39gonna walk in the grass. I'll have my lap laptop with me, but I'm walking in the grass. You know, I agree the value is the people,
472. 1:21:47especially the people that want to use AI, but it's hard to say here that when companies do not value people the same
473. 1:21:53way, and many are struggling to find work. Well, but see, here's the deal. I mean, companies have never cared about
474. 1:22:08people in the end. I mean, you know, the the minute you we we get to sort of PE firms and Citizens United and all the
475. 1:22:15all the decisions that, you know, Reagan sort of ushered in, you know, I mean, corporations were never friendly. I mean, the industrial
476. 1:22:27revolution was about exploiting as much as we could. Like I don't I don't think it's any different. I think I think that
477. 1:22:34knowledge work was more polite about it than the than you know factories, manufacturing factories were, but they
478. 1:22:42never cared about it. I think one of the things that AI potentially does is absolutely wrecks the centers of power of how big
479. 1:22:56corporations work. I think it's very possible, Mary Carol, that big corporations start to fracture and fall
480. 1:23:04apart over the next decade and smaller, more nimble entities that that will be consolidated into big companies. But I
481. 1:23:14think they will functionally act like smaller, more people centric units of 8, 10, 20 people. Um, I think that unit
482. 1:23:23gets really really powerful in a new economy. Um, I I I think the the the organizational structure of of any
483. 1:23:37corporation that's 500 people or above, if they don't radically alter their organizational structure and
484. 1:23:44decision-making process, I don't like I don't think they're long for this world because competitors are going to be way
485. 1:23:50too nimble. But Kyle, the chain of go to school, hard work, get a job, get a paycheck. That that Okay, Mary Carol, that chain is dead.
486. 1:24:03Go to school, work hard, get a job, get a b paycheck. That chain is dead. It's dead. It's dead.
487. 1:24:18People need to make income in the meantime. I agree. Here's here's the thing. I think that I think this comes down to like
488. 1:24:33what this comes down to, and this is Listen, there there are still corporations. There's going to still be
489. 1:24:40corporations that hire some people, but it's going to get harder and harder and harder. you're going to have more and
490. 1:24:45more talent available and less and less positions for it to go to. So, what's going to happen is everyone's going to
491. 1:24:52be forced into entrepreneurship. The entrepreneurs that try to do things the traditional way are going to struggle more and more and
492. 1:25:01more. the people that embrace entrepreneurship and figure out all the 60 different ways to hustle, whether
493. 1:25:07it's speaking gigs or consulting or or whatever it is that are AI enabled, they're going to be able to navigate this easier. But like
494. 1:25:19I'm not a fan of what's happening. I just think what's happening is show people your resume, show them the
495. 1:25:26experience that you had for a job that doesn't actually exist. It's just it's it's going to start breaking on on many
496. 1:25:33different levels. And I think I don't know how long the big companies in their zombie state are going to
497. 1:25:41exist. I know big companies don't think they're in zombie states right now. They may not be. It might it might be 30
498. 1:25:47years until big corporations really feel the effect the effects of AI. I think it'll be sooner than that, but it could
499. 1:25:55be 30 years. It took 30 years for the advertising agency to admit that digital communications were valuable. 30 [ __ ] years
500. 1:26:06to admit that this internet thing wasn't just a fad. Until AIs think for themselves, that's not going to be that far. Let's see.
501. 1:26:18Kylin, my next response to you is that I already typed was about solarpreneurship. Yes, agreed. Yeah. Like like people are going to be
502. 1:26:27forced into and think about it. People who are not instinctively entrepreneurs. Entrepreneurs are weird. We're weird.
503. 1:26:39like I couldn't work for a company because I don't like entrepreneurs are perpetually dissatisfied with the world
504. 1:26:55and they want to take anything that's broken and they want to fix it and they want to fix it to the point that I'll
505. 1:27:00start a [ __ ] company to fix this. My company StoryVine is the response to how hard it was to make a twominute video of my twin boys
506. 1:27:11when they were toddlers. I was like, there is no [ __ ] reason that it should be this hard to make a two-minute video. I'm going to invent a
507. 1:27:20company to solve that [ __ ] problem. That's weird. So, you're going to take a population of people that are not used to thinking
508. 1:27:30like that and you're going to be like, "Hey, go be an entrepreneur and be like,"Well, what how what's the
509. 1:27:38what's the progression? What where do I start?" You just [ __ ] start with, "Are you pissed off about something?" Um, no.
510. 1:27:48Well, you figure out something you could do better and go do that and then make a business and then people will give you money. What?
511. 1:27:55You know what I mean? Like it's gonna be brutal. Open an entrepreneur school. People here are outside of the norm.
512. 1:28:10We're not weird. We're irregular thinkers, doers, experimenters. Anyway, it is time. I'm going to get out
513. 1:28:22of here. Hope you had a good Monday night. Um, listen, things are moving fast enough that even though it seems like America's
514. 1:28:30[ __ ] things up right now, maybe we turn it around. Maybe we turn it around. But, uh, we got to move fast. Anyway,
515. 1:28:40I'll see you Wednesday. We'll play with some stuff on Wednesday. Uh, hope you had a good night. Appreciate you being
516. 1:28:45here. Um, if you're new here to the AI salon, please introduce yourself. Make friends with some of the people that are
517. 1:28:50in here chatting. If you if you just been lurking and not chatting, chat next time. All right, peace out everyone. Uh
518. 1:28:56hope you have a good night. Bye.