2/5/2026 - AI Agents, Job Redefinition, and the Forced Identity Crisis of the Programmer

Transcript

1. 0:00Campy, you ready to sing sing for your dinner?
2. 0:35You're all feeling nervous, aren't you, boy? And impeccable style. Don't ever let them steal your joy
3. 0:50and your gentle ways to keep them from running wild. They kick dirt in your face, dress you down, and tell you that your place is in
4. 1:02the middle when they hate the way you shine. I see you tugging on your shirt, trying to hide inside of it. hide how much it hurts.
5. 1:17Let them laugh while they can. Let them spin. Let them scatter in the wind.
6. 1:32I've been to the movies. I've seen out hands and the jokes on them. Voice is still not good.
7. 1:49You're feeling nervous, aren't you, girl? It's your brother's world for a while longer. Roy Harrison in the house.
8. 3:38Wow. Woohoo!
9. 3:57Woohoo!
10. 4:22What's that, Vicki? You thought today was Friday. You like the shades? I think the shades are cool. These are my new driving
11. 4:32glasses, but they're also my new jamming glasses. You got to jam out. Got to jam out on a guitar. Get a little Jack Black.
12. 5:00Woo!
13. 5:27Wow! You look cool. But how does your eye feel? I I uh it is not just a like block
14. 5:44tear duct or whatever I thought it was. I actually have pink eye, which Brandon thinks I got in DC. I got too close to I
15. 5:52got too close to the muck and it rubbed rubbed off on me. Uh, no, I actually think I got it. I went for an eye appointment at the eye
16. 6:03doctor and two days later. So, but I'm on the correct drops now, so I should be good. Thank you for asking. But yeah, look how
17. 6:15cool I look. I think I look pretty swell. You and I here all alone. Sunday morning here at home.
18. 6:30Sky's blue and a coffee strong. It's true. Then I open my eyes to a dream realized in front of me
19. 6:45and I haven't got a clue when the world is happening to me. Think I think I'm happy like the first day of summer vacation.
20. 6:58Happy got to get a little rest and relaxation. Happy like the choir on Sunday morning singing true.
21. 7:20Woohoo!
22. 7:40Um, what's going on? There's a lot of stuff happening. I don't know if you know this, in the world of artificial intelligence,
23. 7:50there's a lot going on. There's things being announced every day. Agents are about to take over the world.
24. 7:59Everybody pay attention to the agents cuz where you think you have control, you're about to lose it. It's It's a [ __ ] [ __ ] show.
25. 8:13It is a train wreck. a travesty or the most exciting thing to happen since we didn't have agents.
26. 8:27Oh, good lord. There's a new Claude 4.6 Opus out. Seems to be impressing people. There's a codeex 5.3 from Chat GPT.
27. 8:41There's a Cling 3 on the video front. There's just everything here for the music. I'm so sorry. It's only me doing the singing. That's
28. 8:51the That's the problem with coming here for the music.
29. 9:06Sitting in this lonely town. Clean your guitar strings so you don't crosscontaminate. Oh, that's a good idea.
30. 9:16When you get actual doctors in the feed, you can get some advice. Sitting in this lonely town, wonder things are going to change.
31. 9:34Dream my life away. Dream. Seems these dreams have turned to a bunch of dust clouds. getting my nerve up. But my Corey likes
32. 9:46my round glasses. I knew Corey would like my round glasses cuz they're hip. She'd like them better if one of them was square.
33. 10:05Dr. J. I'm a shrink. Lol. Just ate germs. Shrinks are doctors. You got to go to some sort of school to be a shrink.
34. 10:19Yeah, I'm happy with the new glasses. I I So, you kind of have a choice, right? As you get older, you can like you can go I'm going to age
35. 10:30up or I'm going to go more hip. And and the the more hip I'll I'll show you an unfortunate attempt at going more hip.
36. 10:40That was not the right choice. The These were my driving glasses for about a week and then I realized what a
37. 10:48douchebag they made me look like. Look at these things. Oh, they're awful. They to go with my hip haircut. Exactly. Yeah, these are
38. 11:03awful. These I I I bought online and then the minute they showed up I'm like, "Oh, those are kind of awful."
39. 11:16So those those those just sit in the the back corner. So we'll go with the we'll go with the round ones. A Kelly Bes character
40. 11:34Can we go back to the rounds? Yes, we can. They go in the shelf next to the rabbit in terms of uselessness. I think that's
41. 11:42a good idea. Producer Brandon, we'll just put them right up here. Right next to the rabbit R1. They go right there. That was a sad
42. 11:52attempt. A sad attempt at hypn. sad. But are you having a midlife crisis? If only are midlife.
43. 12:33Sitting this lonely town. Yeah. Things are going to change. Dream my life away. Seems these dreams have time to bump
44. 12:47those clouds. Get my nerve up, but my past is pulling me down. Sing along, Chappie. Wondering how long
45. 13:01this black sheep going to stick around. Somebody told me once before, he said, "You can never go home again.
46. 13:16Won't you leave?" and a thing to steer me away from the truth of who I am and what I believe. So I thanked him for his two
47. 13:28sins with a handshake and some sympathy. Yeah. Packed on my blue jeans and headed for this big price of my freedom.
48. 13:45Bye-bye black sheep to the black sheep of the family. Bye-bye. Oh, I mean so very much to me and
49. 14:01bye-bye black sheep to my friends and my family. Bye-bye. I going to set my soul.
50. 14:17Set it free. I hate to interrupt the concert. >> Yes, >> I have some breaking news. Speaking of the Rabbit R1,
51. 14:32>> uh they have are coming out with a dedicated vibe coding machine soon called Project Cyber Deck. Project Cybert Deck
52. 14:44Project. Okay, we are now officially Okay, I'm going to declare it. Mark this date, February 5th, 2026. We are entering AI being overhyped.
53. 14:58Project Cyber Deck. Come on. Could they come up with a more horrible name? That's really bad. All right, but what
54. 15:05is it? It's a It's a hardware device you vibe code on. >> Yes. Nice. >> Uh, so be purposely designed for vibe
55. 15:15coders to run things like cloud code CLI and the upcoming rabbit command line interface. >> Ah, all right. We're we're officially in the
56. 15:24in the hype cycle. Agents are taking over. Everyone's freaking out. I think Anthropic announced something about agents today.
57. 15:40Open AAI announced some new enterprise thing. We should go look at that. I haven't I didn't look deeply enough into
58. 15:46it. So, so we'll go look at that and see if that's anything interesting. Massive layoffs in January. Yeah. But,
59. 15:58you know, Todd, I here's the here's the thing about the layoffs. The it corporate America can't have it both
60. 16:06ways. Okay. corporate America can't have there's no return on AI. AI is a waste. We we don't know how to use it. No one's
61. 16:16adopting it. And we're laying off 30,000 people because of AI. You can't have it both ways. America, let's just let's just admit
62. 16:28that we're going to cut some costs so we can drive a little more profit to the shareholders. Okay. Okay. and let's hope
63. 16:35that the remaining people can figure out how the [ __ ] to use AI to make up for the shortfall of the ones you just let
64. 16:42go. That's what's going on in 2026 in my humble opinion because there's not enough adoption. No one knows how to [ __ ] use it yet.
65. 16:52We don't know how to use it as individuals. Companies don't know how to use this [ __ ] yet. They're not training their
66. 17:00staff on how to how to do these things. I mean, it's a handful of companies represented, you know, in this audience,
67. 17:08in this community that that are starting to get their [ __ ] together. They're starting to figure it out.
68. 17:19It's like buying calculators and get getting rid of your accounting department. That's exactly what it's like.
69. 17:27Yeah. And we know stuff. Exactly, Corey. Like we know it's not ready for prime time. We know most people don't know how
70. 17:33to use it at all. We know that it's chuck full of security risks and [ __ ] Like it it just Anyway, I I just
71. 17:44it's exhausting right now. It really is [ __ ] exhausting right now. Is it? >> Yeah. >> Can Can I talk about my prompt of the
72. 17:57week? I'm really excited about it. >> Yes. Jump in. Jump in. I I I got nothing. I got I got nothing going on
73. 18:04here. So, I think prompt of the week week is awesome. >> Yeah. So, if you're not sure where we are, we are at community.thesalon.ai,
74. 18:11which is an extension of the AI learning lab. And what we've started out this month is a new prompt of the week. And
75. 18:19so, each week we're going to come back and share on Thursdays a different prompt for you to riff off of. And so
76. 18:27since the theme of the month inside the AI salon is signal, the prompt of the week this week is frequency modulation
77. 18:37against an analog backdrop. >> Oh, nice. >> So your challenge is to create anything that explores that idea. an image, a
78. 18:44song, a short story, a workflow, a conversation. Interpret it literally or metaphorically, digital versus human,
79. 18:52precision versus texture, change happening inside something familiar. And what we want you to do is go over to
80. 18:58look what I made and share your results and maybe even add one sentence about how you interpreted the prompt. So this was my example image
81. 19:09of a frequency modulation against an analog backdrop. This was done in CH GPT >> enhanced imaging. >> That's very cool.
82. 19:20>> That is very cool. >> I would encourage you to head over to uh the AIM and share what you think of that.
83. 19:26>> I love it. And and the the to to uh to pump it up a bit more the the theme of the month. I'm really excited about it.
84. 19:34This is Brandon's Brandon's baby. And there's a there's a group out there uh called uh creative mornings and they
85. 19:42they've got they do these meetings once a month where they meet in all different cities and just have creative
86. 19:47conversations and they have a theme of the month and so all of the conversations are geared toward toward
87. 19:53that theme. And so we're starting that inside the uh the AI salon community. So signal is the uh is the theme for
88. 20:01February. So go check out that prompt. That's that's a so so within within the AI salon, if you go just to
89. 20:10the welcome page, you'll see there the cycle of AI readiness. Play first, create excellence, generously lead. That
90. 20:17prompt of the week is a really good opportunity for you to go play, take that theme and go play in some piece of
91. 20:27AI land, some tool that you haven't used before and come up with a way to interpret that. If you haven't ever made
92. 20:35an app, go to lovable.dev and make an app. If you haven't ever made a movie, go make a movie, you know. Yeah. Really
93. 20:43cool. Really cool producer Brandon. I like it. I like it a lot. >> And when you're done doing that and you need some comic relief, I would
94. 20:53encourage you to head over to the Irregular's channel to see what digital gods whipped up. >> Why do I have a files?
95. 21:03>> So now you get to hear a Gen Xer get pissed off because some dweebs in some [ __ ] lab somewhere decided that
96. 21:11images needed to be 3% more efficient. So they created an entirely new [ __ ] file format that no one [ __ ] uses.
97. 21:22So what we're gonna do because we've experienced this before is we're going to whine. Okay? We're going to whine and
98. 21:30we're going to [ __ ] and then we're going to go fix it because we know how to fix it. But someone, some little punk ass 25year-old, you
99. 21:39know, if if we if we uh encoded the the the pixels, the the data image and the pixels, the color space, we could and the nobody cares.
100. 21:52The internet's fast enough. I just want [ __ ] to work. Thank you for sharing that, producer Brandon. That was It was not one of my proudest
101. 22:07moments, but now it's been memorialized with a talking lemon. Listen, we we went in heavy on talking hot dogs early in
102. 22:18this channel, so I think having me as a talking lemon is is is awesome. It was one of your better rants. I
103. 22:31wasn't actually really good when I when I went to do the uh the the the dweeb talking about the algorithm. My brain went blank.
104. 22:39But, you know, but I was pissed off. So, there's that. It's just like I mean, it's been three years and nobody still use like like chat GPT
105. 22:52outputs image file formats that it can't import. It's like, stop it. Just use the [ __ ] that works. God damn,
106. 23:07you get me all fired up again. Tik Tok, dude. Love Lord Digital. God's acidic Kyle. night after night.
107. 23:25Thank you, Lord, digital gods, for memorializing my my moment of shame. Oh, man. All right, I want to go look at
108. 23:36something. I want to go We're going to read something together. We're going to have a little reading moment. this one
109. 23:44cuz I saw this today and I didn't read it because I don't know if you know this but my eyes are melting out of my [ __ ] head. Um
110. 23:58Oh, listen to the article. We could listen to it. Yeah, let's listen to it. [ __ ] it. Let's listen to it.
111. 24:08We live in a world now where you can You can just listen to [ __ ] Everything's an audio book now.
112. 24:27Okay, there we go. And this is why we come back night after night. Uh, okay.
113. 24:42>> AI has let teams take on things they used to talk about but never execute. In fact, 75% of enterprise workers say AI
114. 24:51helped them do tasks they couldn't do before. We're hearing this from every department, not just technical teams.
115. 24:57The way work gets done has changed, and enterprises are starting to feel it in big ways. We've seen this in action with
116. 25:04over 1 million businesses over the past few years. At a major manufacturer, agents reduced production optimization
117. 25:11work from 6 weeks to one day. A global investment company deployed agents endto end across the sales process to open up
118. 25:19over 90% more time for salespeople to spend with customers. And at a large energy producer, agents helped increase
119. 25:27output by up to 5% which adds over a billion in additional revenue. This is happening for AI leaders across every
120. 25:34industry and the pressure to catch up is increasing. What's slowing them down isn't model intelligence. It's how
121. 25:41agents are built and run in their organizations. What's slowing them down isn't model intelligence. It's how agents are built
122. 25:57and run in their organizations. I don't think that's actually [ __ ] true. What is the definition of agent?
123. 26:10Are they talking about automations or autonomous agents? I think this is just horseshit marketing right here. But anyway, let's continue
124. 26:23on, shall we? Wait, why can't where? >> Today, we're introducing Frontier, a new platform that helps enterprises build,
125. 26:34deploy, and manage AI agents that can do real work. Frontier gives agents the same skills people need to succeed at
126. 26:41work. Shared context, onboarding, hands-on learning with feedback, and clear permissions and boundaries. That's
127. 26:48how teams move beyond isolated use cases to AI co-workers that work across the business. Okay, I think this sounds
128. 26:55cool. I think this sounds right. I think this sounds good. But this setup paragraph is absolute horseshit.
129. 27:04Organizations are not struggling with AI because they can't figure out how to build agents better. They're struggling
130. 27:10with AI because they don't know how to deploy it. No one in the organization knows how to use it because they're not
131. 27:14train anyone training anyone on what the [ __ ] it is. I'm telling you, man. Open AAI needs an angry demer.
132. 27:27Oracle, State Farm, Thermoffisher, and Uber are among the first to adopt Frontier. And dozens of existing
133. 27:33customers, including BBVA, Cisco, and T-Mobile, have already piloted Frontier's approach to power some of
134. 27:39their most complex and valuable AI work. The AI opportunity gap. Companies are already overwhelmed with the
135. 27:45disconnected systems and governance spread across clouds, data platforms, and applications. AI made that
136. 27:51fragmentation more visible and in many cases more acute. Agents are now getting deployed everywhere and each one is
137. 27:59isolated in what it can see and do. Every new agent can end up adding complexity instead of helping because it
138. 28:06doesn't have enough context to do the job well. As agents have gotten more capable, the opportunity gap between what models can
139. 28:14do and what teams can actually deploy has grown. The gap isn't just driven by technology. Teams are still building the
140. 28:21knowledge to move agents past early pilots and into real work as fast as AI is improving. >> That's an understatement.
141. 28:27>> At OpenAI alone, something new ships roughly every 3 days. And that pace is getting faster.1. Keeping up means balancing control and
142. 28:36experimentation, and that's hard to get right. Enterprises are feeling the pressure to figure this out now because
143. 28:42the gap between early leaders and everyone else is growing fast. Open AI Frontier. We've learned that teams don't
144. 28:50just need better tools that solve pieces of the puzzle. They needed help getting agents into production with an
145. 28:56end-to-end approach to build, deploy, and manage agents. >> Really? Really? Is that is that what you're all
146. 29:11hearing? that the big the big holdup is we can't figure out how to deploy our agents. What are they actually trying to say?
147. 29:22Here's what they're trying to say. Source camp, we have a new thing that makes it easier to deploy the things
148. 29:27that were just hidden in our dev tools for two years that no one can figure out how the [ __ ] to use. So, we've we've
149. 29:34made it easier for corporations to use them. And we're going to talk about it as if everyone's trying to use these
150. 29:41things. That's what they're trying to say. They've just collected sound bites. Yeah, this is pretty bad. I This the
151. 29:49tool actually sounds kind of cool. They asked chat GPT. Yeah, exactly. This totally sounds like a chat GPT written
152. 29:56thing. You know, one of the great challenges of AI adoption is that they can't just figure out how to get these
153. 30:02darn agents to behave, if you know what I mean. Hey, I really like the way you're running your organization there. Maybe
154. 30:10if you could get some agents going, things would be better, huh? Okay, let's let's let's continue on,
155. 30:24shall we? We only have six more minutes of this to go. My son is in community college and he's having to find his own
156. 30:32way trying to learn technology outside of college. I don't get me started on [ __ ] higher education m okay
157. 30:42I it's going to be another [ __ ] rant night. I can't. It it it makes me insane that our educational institutions, we've
158. 30:51got the the the the most profound knowledge technology in the history of humanity at our fingertips and your sons having to [ __ ]
159. 31:07moonlight his [ __ ] AI career because they're like, "Oh, well, you know, artificial intelligence just allows students to to to write their
160. 31:18papers without even trying. So, we've got detectors for that. There were we are we still there? It's astounding to me. It's astounding
161. 31:27to me. You know, people ask me like, you know, should my kids go to college or whatever? I mean, I guess maybe for the
162. 31:34social thing certainly. Um, I think they should be studying philosophy more than computer science, quite frankly. Um,
163. 31:44and if if I were interviewing schools, like I would be interviewing the heads of the departments to say, you know,
164. 31:51what's your stance on AI? And if they were not aggressively into it, I wouldn't [ __ ] go there. It's astounding to me. Anyway,
165. 32:02all of the schools are behind. They're willfully behind. They're [ __ ] willfully behind. It's irresponsible.
166. 32:15I wanna I wanna I want to play a video that I saw today. So So you know, we just launched this thing called the great repurpose where
167. 32:25people are going to be losing their sense of meaning if they've got their identity too tight tightly tied to
168. 32:34their work. And as their work as the tasks of their work gets automated and stripped away, what are what are they
169. 32:42going to be left with if if their meaning was tied to that? Um Mark Andre, there's a there's a a clip of him in an interview
170. 32:53that is talking about this and I want you to hear it if I can find it. Is this it? I think that's the interview tab. Oh yeah, thanks.
171. 33:12>> Yeah, >> on that AI coding actually abstracts the way that >> Okay, listen to this. So this is Mark Andre. He's the one that
172. 33:23he he invented Netscape, you know, he created the first modern browser. He's the head of A16Z VC firm.
173. 33:32>> There's and then to your point like AI coding is the next layer on that. AI coding actually abstracts away the
174. 33:36process of actually writing the scripting code, right? And so in one sense, this is a really big deal for all
175. 33:41the obvious reasons, but on the other hand, it's like, okay, this is the next layer of the task redefinition under the
176. 33:47job of programmer, right? >> Wait, listen to listen to those words. He's smart. You we we've got to decipher
177. 33:56the words cuz cuz he's got the big brain in the big giant head. This is the next layer of the task redefinition under the job of programmer.
178. 34:08>> I got I gota hear it one more time. Thank god we have it. The lay. Let's see. This is the next layer of the redefinition
179. 34:19under the job of programmer. Wait the wait. This is the next layer of the task re the task redefinition. This is the next layer of the task
180. 34:30redefinition of the job of programmer. >> Now what's the job of the programmer? It's to your point. It's not necessarily
181. 34:39to write the code by hand. But what it is now is all right. Now you know if you talk to the world's best programmers
182. 34:43today, what they'll tell you is oh my job is I'm sitting there and I'm orchestrating 10 code bots. What my job is now is I'm a manager.
183. 34:53I'm a manager. I'm no longer a coder. If you're a coder that loves coding, you didn't ask for this. You got a promotion. One of the things I
184. 35:08learned when I was doing agency.com is there were sometimes people were really awesome at their job and we would
185. 35:13promote them to manager and they would either quit or they would ask for a demotion. I remember Paul Galley. I I remember it
186. 35:23distinctly. He he was employee number two at agency.com and he was a really good site builder and it and everyone was getting promoted
187. 35:34around him and he didn't he didn't want a promotion but at some point it just became weird that like all of his
188. 35:40colleagues were getting promotions and he wasn't. So we gave him a promotion and two months later he walked into my
189. 35:47office and he said can I not do this? I don't like managing people. This is not what I enjoy doing. So like programmers
190. 35:59right now are in the middle of a forced redefinition. I'm sorry. Wait, what did he say? A forced a forced
191. 36:09uh this is the next layer of the task redefinition under the job programmer. Right. I want I want to read you something. I I
192. 36:23read I uh I did some uh I did some groing today. How where do I where's my history? Oh, history. So, there was a Reddit Reddit post. We
193. 36:39we read this in here um back when it happened. This was in March of 2023. If you if you were here back then, you'll remember this.
194. 36:51This was the first this this th this post was the canary in the coal mine to what's coming. All right,
195. 37:07here's the full text. Okay, so this is a this is a guy working at a a 10-person game company in March of 2023.
196. 37:19I lost everything that made me love my job through midjourney overnight. I'm employed as a 3D artist in a small
197. 37:28games company of 10 people. The art team is two people, me and my friend. He's been working there longer than me and I
198. 37:36am a junior artist. Last week, my boss was away on a trip and me and my friend were left to work. We were given the
199. 37:44task of creating three graphics for an advertising campaign. We finished them all in two days and were left with
200. 37:49nothing to do. My friend showed me Midjourney and we were messing around with it for the rest of the week. When
201. 37:56my boss came back, my friend told him about Midjourney and how good it was. My boss got excited and asked him to show
202. 38:02it to him. My friend showed him how to how to prompt and my boss was amazed. He asked us to create graphics for a new
203. 38:10campaign. My friend prompted it and it came out great. My boss was happy and said, "This is the future." He then
204. 38:17asked me to create a graphic for another campaign. I prompted it and it came out okay. My boss was happy and said that
205. 38:23I'm a natural. He then asked us to create graphics for all the campaigns they have. We did and they were all
206. 38:31great. My boss was ecstatic and said that he was going to save us so much time and money. He then said we can now
207. 38:37focus on other things. I was happy at first, but then I realized I'm not an artist anymore. I'm a guy who types
208. 38:45prompts and waits for the computer to generate the art. I lost all the joy I had in creating art. I lost the feeling
209. 38:51of accomplishment when I finish a piece. I lost the feeling of pride when I show my work to others. I lost everything
210. 38:58that made me love my job. My friend is happy because he hates creating art. He just wants to get the job done and go home. And he's now the
211. 39:08lead artist because he's better at prompting than me. I am now the junior prompter. I am thinking of quitting my
212. 39:15job. I don't know what to do. That's about to happen at scale. I think in in engineering it's happening right now. Um,
213. 39:30this unvoluntary redefinition of the tasks that make up our job is about to happen. And and the the
214. 39:44agents, the agent swarms, right? The the the open claw kind of things, they're going to start rolling out from the big
215. 39:52frontier companies over the next two months. And the implication of that is so what Mark Andre was talking about is as a programmer
216. 40:05you're no you're no longer new no longer doing coding what you're doing is you're managing 10 agents that are doing the
217. 40:11coding right. So so you've gotten a promotion you didn't ask for. Well the agents get good even there's there's going to be
218. 40:19coordination agents that are going to do the coordinating. So, there's going to be fewer people that have to look over, have to manage
219. 40:30the agents because they're going to get better and better and better. Um, and then we've got this this vacuum,
220. 40:40this vacuum of meaning where to the extent that you've got yourself self-worth and your identity tied to
221. 40:46your job as it gets ripped away, what are you left with? And I think for all of us, it's an incredible opportunity. I think the
222. 40:55people that hang out on this channel, you're already thinking this way, right? You're already thinking like, "What do I
223. 41:00want? What do I want to do? What can I do now that I have all these skills?" But a lot of people are not going to
224. 41:04feel that way. So, the youngans will not be tied to their work identity. I don't think certainly if they can't get entry-
225. 41:12level jobs, they won't be. It's going to be an introverts world. I don't know. Mary Mary, I think I think I think that um
226. 41:28I think there's going to be a a return to analog. I think there's going to be a return to in-person meetings. I think
227. 41:36there's going to be a slowing down. I think the only way through an identity crisis, a forced identity crisis, is to slow down.
228. 41:47and and and probably get your ass in some sort of community, right? Because there's going to be a lot of people going through this. So,
229. 41:56the more we can hear stories of people that are navigating their way through this, I think the the better it's going to be
230. 42:04for everyone. But it's it's crazy, man. Crazy. Work-wise, it's going to be an introverts world. Well, yeah. Work's going to get weird.
231. 42:17I just think it's Yeah, it's going to get more and more and more isolating. Like at some point it's just going to be
232. 42:22some guy just some guy looking at a a a green light and if the light turns red he's got to go restart servers or something like that.
233. 42:35The same thing happens when you retire. It doesn't last long. Well, but you know the thing about retiring to block Tom is
234. 42:42you kind of know it's coming, right? when when you retire, you've got years to say, "Okay, I'm not going to be doing
235. 42:49this job anymore." And and generally, you know, if you're approaching retirement at your job, you're getting
236. 42:56sideways promotions or you you know, you know, you're just doing your job and at some point they're like, "Ah, you're
237. 43:02going to be retired in a year. Just just, you know, slow it down." You've got some sort of transition.
238. 43:08>> The thing that I think is going to make this different is you're 40 years old. You're 20 years into your career, you're kicking ass,
239. 43:17and all of a sudden, AI automates your job and it automates your sector. So, you can't go get that job somewhere else.
240. 43:31And you're 40 and you're like, I went to school for this. I got a master's degree. I worked at the best consulting firms in
241. 43:40the world. I'm doing all this consulting. And then right you're in some sort of analytical role because you love analysis and you
242. 43:50love supporting the team. You're not the front guy, right? The front guys, the people that have the human relationship,
243. 43:57you know, uh interface, they're they're probably going to be fine because you're going to need people. People are going
244. 44:02to need to trust people. Tik Tok pin. I built a GPT for my exec and transition group. add LinkedIn profile
245. 44:11and it Oh, that's really cool. Source Camp. Um, ping ping me and let's talk Yumi and Andy and Brandon about that because I I
246. 44:23want to start taking the great repurpose and finding human beings that are doing stuff with this. I want to start finding
247. 44:32tools where people can sort of self-n. So, so let's let's talk about that. um and put it in there. That's really cool.
248. 44:41You need to retire to something, not from something. That's exactly right, Steo. And so, so it's just going to be this it's just
249. 44:50going to be a lot of people that are just going to have this this like band-aid ripped off to a wound that wasn't a wound, right?
250. 45:01That is now a wound. like like in a society where we tie our work to our self-worth and our identity, it's okay.
251. 45:10At some point it may break down, but for the most part you've got time to navigate that. But if all of a sudden
252. 45:17lots and lots of people have that those tasks stripped away and they haven't had time to process this, they haven't had
253. 45:25time to prepare for it. And now all of a sudden they're just like all that [ __ ] I did for 20 years is valueless.
254. 45:38What's you know like the self-esteem there is just going to be brutal. The fact of the matter is there are things you do when you do your
255. 45:48tasks like how you do your tasks and the integrity that you do your job with and the amount of give a shittness you have
256. 45:56about your job that is the stuff that's valuable val valuable about who you are. But if you haven't done that work to
257. 46:05figure that out and and you just have it tied to the tasks and those tasks get ripped away, that's the thing that's gonna be [ __ ] brutal.
258. 46:14The great repurpose could really be aimed at retirees. I think it can be it can certainly be aimed at retirees, but
259. 46:20I think it can be aimed at at really anyone that's going through this transition. That's the thing. I don't I
260. 46:26just I don't think this is just about retirees. It's about retirees right now, but like in the next three years,
261. 46:32there's going to be a lot of people offered early retirement that perplexity model context video. What is that
262. 46:48model? Oh, model council. All right, we'll go look at that anyway.
263. 47:06I'm hopeful that the meaningful work will come back once we figure out what counts as an economy. Not working for dollars, but working for
264. 47:20for fulfillment. I I think that's the transition. We're at the beginning of it's the transition part that I think is going to be painful.
265. 47:33All right, let me see this thing. What is this thing? Is this at all interesting? Run three Frontier models at once.
266. 47:44Compare outputs. Oh, that's kind of cool. Let's watch that video. That's worth looking at. Does this have audio with it?
267. 48:34You know, it's funny. I was thinking like last week I was thinking about um let's go to Perplexity and see if it's
268. 48:43live. It might not be. Um, do we have three models? Upgrade for the latest models. No, it's not live yet. Or it's it's at least not
269. 48:58live for me. >> It's under the plus sign if you have it. >> Oh, is it? >> Model council. >> Oh, max. You have to you have to be a
270. 49:10max. $271 a month for the model council. I'm good. Yeah, I was thinking I was thinking that um there's probably going to be a lot of
271. 49:27money to be made if someone can figure out some sort of AI interface that maintains your memory independently
272. 49:39of the models you use. and it it dynamically figures out all the different tools you could use. Um, and it maintains your your your memory
273. 49:50because right now all the frontier companies are trying to get you to, you know, imprint on on their on their particular flavor of AI.
274. 50:00Oh man, what's happening in DevO? We're shaking. Um, M I also thought it might be interesting to just go through Twitter and just see
275. 50:16if there's anything interesting to look at. Picture of a dog getting a haircut. That's always nice.
276. 50:31Everyone's talking about Cling 3, but the weird thing about Cling 3 is you can't get it at Cling yet. You can get
277. 50:37it at a bunch of other sites, but not at Cling. It's Nikki Weiss. Hey Nikki, what's happening then? What happens to your memory? You
278. 50:49mean your real memory? Like what I'm talking about is right now I've got by far more of my interactions
279. 51:02in chat GPT than in any of the any any of the other models. And as a result, I don't really unless I have a specific
280. 51:09thing where I want to try a different model. I don't just in an adventury kind of way go off and use other models
281. 51:17because I don't I don't want my memory fragmented. I want chat GPT to know as much about me as possible so I can say
282. 51:23based on what you know about me, you know, write a little bio for me or write a little something, write a little
283. 51:30article and it it'll understand how I write and all that sort of stuff. Yeah. Yeah. Yeah. Exactly. Um,
284. 51:45so if there was some sort of central service that had hooks into all those other services that was basically
285. 51:53like what would be cool like I'm trying to figure out where this would live. This would probably live as like a like a an operating system
286. 52:03extension where it understands that you're going to go off and use a bunch of different AI things and it's sort of sucking in
287. 52:10your interactions with them into a single into a single memory. I don't know. I just it feels like there's an opportunity there.
288. 52:18Feels like there's an opportunity there. All right. A lot of people quote tweeted this. Oh wow. Andre Karpathy on the one-year
289. 52:31anniversary of of him coining the term vibe coding. That's great. Cling three.
290. 52:45Instead of watching a three-hour movie, watch this master class instead. This was the uh the new what was it called? Clawcon
291. 52:57for for you know in San Francisco. They're all going crazy over open claw. I wonder if has anyone in here actually
292. 53:05installed openclaw yet? I was thinking about doing it but I don't I just don't have two days to spend geeking out and
293. 53:12being frustrated that I'm not a better coder. GPT 5.3 codeex with image gen skill. I'm loving this model so much. This is a
294. 53:22oneshot platformer. That's cool. I've left anthropic to start something new. Cool.
295. 53:37I'm considering an app aggregator. It's sort of an app aggregator, but it's a it would be a smarter it would be smarter than that.
296. 53:50Like the thing I'm envisioning is you probably have what it would probably look like is this. It would probably be a small
297. 54:00a small language model that's running locally on your machine that's got a really slick interface on it. So, it can render images. It can be
298. 54:12a chat interface. It can be a voice interface. It can it can shape shift and then it's got hooks into all of your
299. 54:20accounts for all the other things. So, you can just give it a prompt and you can let it figure out which tool is the
300. 54:27is the right one to do or if you want to if you want to just focus on one, you can select one. You can let it select.
301. 54:33You could do multi-tool, you know, you know, send this out to Perplexity, Claude, Chat, JPT, and Gemini, send the
302. 54:41same prompt to all of them, get the responses back. So, it would be a manager kind of layer, but then the
303. 54:47important piece of it is it would be capturing all of your prompts and all the responses into into a memory
304. 54:56that it maintains that you maintain your memory, not chat GPT. Couldn't you just store it on your hard drive? Possibly. But if you're just
305. 55:11storing it like a rag um you know embedded embedded database model like a vector database locally maybe I don't
306. 55:22know like I don't know I don't know enough about it. I just know as an enduser it is frustrating to me right now that
307. 55:30I've got all this memory in OpenAI and none of it in these other places. Like I can't easily leverage the work that I've put in
308. 55:41in Open AI. I can't easily leverage that in Claude. So I don't know. It's not a huge deal right now, but I
309. 55:52think it's going to be a huge deal over time. Ethan Mullik had early access to Opus 4.6. Here's the result of the same prompt to
310. 56:00create a control panel for a spaceship in the distant future done using the standard clawed interface, not code. I had it zoom into
311. 56:09a few subsystems for the details here. Okay. I I had Okay. Did Did he give the prompt? Kind of amazing
312. 56:27code. Play it. Kind of amazing. So, let's see.
313. 56:41Does he give us the prompt? We just go make up our own. Let's go play with Claude. Okay. Had early access to Opus 4.6
314. 56:56for the same prompt. Create the control panel. Create the control panel for a spaceship in the distant future.
315. 57:09All right, let's go play with this. So, I'm going to Claude cla.ai. We now have a new thing, CL Opus 4.6.
316. 57:25So, I'm going to say um create, let's see, I'll just paste in what he wrote. Create a control panel for a spaceship
317. 57:34in the distant future. I want it to look like a functional control panel and even have interactive
318. 57:49elements. I can play with and affect the interface.
319. 58:08I'm not sharing my my tab. Now I am beautiful. Fantastic Bob. Use blockchain interface so you can container control everything through
320. 58:19encryption. That's not actually a bad idea. Like I think I think having it on the blockchain makes sense.
321. 58:26Having it be trustless makes sense. Having it tokenized makes sense. Having it distributed makes sense.
322. 58:35Um yeah, I think that's cool. I think that's right. That feels right. And like if it's living at kind of an
323. 58:45operating system level, then it doesn't matter if I'm using codeex or or cloud code or a browser or an an AI browser or some application.
324. 59:00It would it would be able to pull all that stuff. Um, I've installed it on my VPS. Oh, uh, Claude Claude Code I mean uh, OpenClaw.
325. 59:17It's not user friendly and you have to put it together from scratch. Very dangerous for people that just want it
326. 59:23to work out of the box. The UX sucks. Good to know. Yeah, that's I I mean it's a security nightmare right now.
327. 59:31It's a usability nightmare right now. If you're not geeky, it it just doesn't matter. Like just just ignore it. Like we're
328. 59:39wait two months. In two months we're gonna have the consumer version of OpenClaw. And it might even be from
329. 59:45OpenClaw. But right now it's it's you know if you're not a Lord Digital gods then don't don't even play with it. Um,
330. 1:00:00so what Opus is doing right now is writing us code for our little spaceship interface. Want to meet up in Denver at ETH Denver?
331. 1:00:10Possibly. Drop me a DM in the AI salon. I'm a little I'm a little um I'm a little bitter about the whole NFT world.
332. 1:00:23I you know I I think I think the NFT world got appropriately slapped back um because it's just so full of fraud
333. 1:00:33and it's like it it hasn't gotten better. Like it's still a it's still a it's still a disaster. Like every time I talk to anyone that's
334. 1:00:43got any sort of NFT project or blockchain project, I'm like, "But what's it actually do?" They're like,
335. 1:00:47"Well, it's it's on the blockchain." I'm like, "Yeah, I know." They're like, "Well, you can get tokens." "Yeah, I
336. 1:00:52know." What do you do with those tokens? Well, you can you can spend them on what on the thing that it is. What?
337. 1:01:00It's always a circular conversation. I believe in the core technology. I just like I just feel like the the toxicity
338. 1:01:08of that whole [ __ ] world is uh is alive and well. Got a LOL from Endo. I mean, it's just like, [ __ ] man. Just
339. 1:01:25just don't scam your friends, you know? Let's start there. Let's just start there. Don't rug pull your friends.
340. 1:01:43At least with AI, I can [ __ ] up my own, you know, my own stuff. Oh, look at that. That's so cool. All right. What can we do? Can we click
341. 1:01:56[ __ ] Oh, wait. What did I do?
342. 1:02:16Why is My four four-year-old would love this. Yeah, but wait, next time try this in co-work. No, [ __ ] you. Okay, let's say um good start,
343. 1:02:32but most of the screens are Wait, let's see. But the but the main screen is black. And when I flip
344. 1:02:48switches, nothing happens. This looks more like a corporate dashboard than the controls of a futuristic
345. 1:03:07uhistic spaceship. Um, shouldn't there be be more quantum drive
346. 1:03:21kind of things and fractals and frequencies? See what I did there, Brandon? Frequencies and signal from the latent space.
347. 1:03:39Uh, I'm going to yell at a little bit here. Um, you created um a 2002 um finance dashboard.
348. 1:03:56I want I want a [ __ ] spaceship. um control. Okay, let's let's see how it
349. 1:04:12responds. Thinking about redesigning this. Architecting immersive spaceship controls with quantum effects, dynamic.
350. 1:04:21You're absolutely right. That was too tame. Let's build something that actually feels like piloting a vessel
351. 1:04:27through spaceime. Fractals, quantum field visualizations, living waveforms, the whole thing breathing and reacting. Yes.
352. 1:04:43Oh my god, man. NFT and blockchain have very narrow usability. I actually think they've got broad ones. I mean, as as much as uh
353. 1:04:55World ID got some [ __ ] for for being, you know, kind of circular in in its logic, like human identity actually
354. 1:05:05feels like that's a really important potential use for blockchain. something like we just talked about. We're
355. 1:05:12maintaining my memory across AI interactions in something that I control um and is distributed and you know non
356. 1:05:23uh you know you can't destroy you can't overwrite your file basically like that feels valuable to me um but we'll see
357. 1:05:32the whole idea of decentralized anything was by definition going to be a scam especially in metaverse remember where
358. 1:05:39the where AI was 10 years ago and watch blockchain in the next two to four years. Yeah, I don't I don't think
359. 1:05:44blockchain's not here for the for the um I think so. I think blockchain is going to be actually I think AI is going to
360. 1:05:53drive blockchain adoption and and NFTTS too. We just won't call them that, right? It'll just be tokenized ownership, things like that.
361. 1:06:05Oh, twin. Yeah, twin. It's twin.so. I think they they just dropped their new agentic thing, right? What did they drop last night? Are you
362. 1:06:18sure you want other apps to know how you treat your poor chatbt? All right, I guess this thing's coding. Is it Is it coding?
363. 1:06:33What's it doing? The code's just sitting here. It's thinking. Creating a truly immersive thing. Hi, I'm Claude. How can I help you today?
364. 1:06:47Huh? All right. Well,
365. 1:07:06yeah, I used that that twin. I think it's twin.so. Well, while this thing's doing its thing, let's go check it out. Um, twin. Yeah, twin.so.
366. 1:07:19the AI company builder. Try now. I had it I had it go find me 10 tweets a day
367. 1:07:35that were like people talking about human centric interaction with AI. And so I get an email digest every morning um
368. 1:07:46with 10 tweets. I don't read them, but it worked. It's like all the [ __ ] I tried with the the Open AI agent and when you could do
369. 1:07:56automated uh prompts and chat GPT. I set all those things up on this channel and then I get emails every morning from all
370. 1:08:04those things I set up a year ago and don't ever look at them. Oh, good lord. Good lord. Good lord. What would you like to automate? I don't
371. 1:08:17know. What would we like to automate? Um, I'd like you to find
372. 1:08:35research papers that generate buzz. Every day
373. 1:08:50you find the sources of the buzz, then identify the paper
374. 1:09:07and then um consolidate the buzz into why the community
375. 1:09:23feels this paper is a big deal. Then do the analysis of the paper.
376. 1:09:37Tab tab tab. Oh, sorry. Thank you. Then do I thought I thought you were actually writing me a big long screed of something and I looked over
377. 1:09:48just tap tap tap. Uh I'm at twin.so by the way. Um then do the analysis of the paper to tell me if you think
378. 1:10:04the it it matches the hype. Pick one paper a day and email me
379. 1:10:22the results. The structure of the email should be overview paragraph
380. 1:10:38um hype hype highlights easy to understand overview
381. 1:10:53of paper for nontechnical me followed by conclusion. All right, let's see what this does.
382. 1:11:10That's a fantastic idea. Thanks, AI. Thanks for telling me I'm awesome. I didn't realize I was that awesome. I
383. 1:11:19have fantastic ideas. Huh. So cool. A daily hype check on research papers. Let me figure out what we're working
384. 1:11:28with and get the details from you. Great. This agent has everything it needs. Web search, deep research,
385. 1:11:35scraping platforms, Reddit, Twitter, and email. Perfect combo for hunting down hyped papers and doing a reality. Couple
386. 1:11:43of quick things I need from you. What kind of research papers? AI and machine learning. So, that's cool. It's given me options.
387. 1:11:52Okay, perfect. AI machine learning is the loudest buzz anyway. Okay, here's the goal I'd craft for this agent. Oh, good. It's going to write
388. 1:12:04It's going to write my goal. Daily AI paper hype check every day. Search Twitter, Reddit, our machine learning, our local llama,
389. 1:12:15hacker news, and AI newsletters for the single AI machine learning research paper generating the most buzz.
390. 1:12:24Identify why the community is excited. Collect the key hype highlights and notable takes. Then read and analyze the
391. 1:12:31actual paper to assess whether the hype is warranted. Email kyle@stofind.com. Cool. So, it remembered who I was. Nice.
392. 1:12:40With the results structured as overview, hype highlights, plain English breakdown, and a conclusion. Um, go. All right, let's set it up.
393. 1:12:58Okay, back to Oh, now we're talking. All right. Can I what can I What can I do? What can I move? Tab. Oh,
394. 1:13:12tab. Let's see. How do I make this full screen? Do I I can do that. That's cool. Is this having any sort of effect?
395. 1:13:28No.
396. 1:13:49What's that? Throttle. Oh, that worked. Okay, so throttle worked except now it's now we're going too fast.
397. 1:14:04Okay. Uh let's say this is better, but other than throttle, I can't see the effect of
398. 1:14:19any of the switches. Everything should do something visible. Um,
399. 1:14:36let's just give it that. All right, let's go back to twin twin.so. Let's see what's happening here. Task in pro in progress.
400. 1:14:48So, I'll start by researching across multiple platforms. Authenticated X and it authenticated X because I've
401. 1:14:58given it that access before. That's fascinating. Um, let me check Reddit. The EDOS deep think paper is generating
402. 1:15:13significant buzz. That's This is actually a cool little agent. Found the paper. Now, let's read it. Now, we have the full paper. Let me
403. 1:15:22extract the key sections. All right, it's doing its thing. I'm digging it. Let's go back here to code. Come back out here to this thing.
404. 1:15:54All right, that's still doing its thing. Now I have comprehensive data. Let me also check the hacker news discussion
405. 1:16:03about this paper specifically. Then compose the email. I now have comprehensive data from all sources. Let
406. 1:16:09me comp compile the full analysis and draft the email. It's pretty slick, huh?
407. 1:16:26First dinosaurs. Do we remember before the worldwide web when decentralized ideology was niche? Huh?
408. 1:16:48I now have comprehensive data from all sources. All right, that's still doing its thing. Let's go back here. See if this thing's
409. 1:16:57done anything new. It's pretty [ __ ] slick, you got to admit. And Murphy, what's shaking, lady? We're making we're making quantum drive
410. 1:17:13um spaceship um control panels with with Claude Opus 4.6. I'm also going to make a uh an asteroids clone because I I made one earlier and
411. 1:17:25it it it's about as close as I've seen one of these things make to the original game. It's pretty amazing. Ah, okay.
412. 1:17:38Email sent successfully. Now, let me set up the data interface and finish. Let's go see if it did send the email. In fact,
413. 1:17:48Kyle Shannon via Twin Daily AI paper hype check. We're still in outer space. Sorry. Today's paper, semiautonomous
414. 1:18:02mathematics discovery with Gemini, a case study on the air problem. So, erdos are the 1500 math problems that no one's
415. 1:18:11been able to figure out, right? Or, you know, whatever that they're they're hard. Overview. Google deep mind
416. 1:18:17researchers built an AI agent called Althia powered by Gemini Deepthink to systematically attack 700 open problems
417. 1:18:26from the famous Aerdos problems database, a collection of mathematical conjectures posed by blah blah blah blah. Okay. Hype
418. 1:18:34highlights. The headline claims Google agents solved 13 new airdos problems. This is the plain English paper down. The erdos
419. 1:18:48problems are famous collection of mathematical conjectures essentially well-defined puzzles that mathematicians
420. 1:18:54haven't haven't yet solved or in some cases haven't confirmed whether existing solutions are correct. Paul Erdos was a
421. 1:19:01prolific 20th century mathematician. the Google deep mind blah blah blah blah blah. Okay. Conclusion. Does it live up
422. 1:19:07to the hype? This is a case where the paper itself is far more careful than the hype surrounding it. The authors
423. 1:19:14explicitly warn against overexitement. This is actually a useful little [ __ ] bot. This is cool. The actual contribution is modest. Five
424. 1:19:29genuinely novel solutions to the problems. the authors themselves say could have been easily dispatched by the
425. 1:19:37right expert. The remaining eight were literature searches, not original mathematics. The paper's real value lies in its
426. 1:19:46honest methodology and candid discussion of AI's limitations in research mathematics, including a 76% failure
427. 1:19:53rate, even when the solutions appeared correct. Social media transformed this cautious empirical study into
428. 1:20:02breathless claims like AI cracking Aeros problems, often conflating different systems and overstating mathematical
429. 1:20:12significance. The competitive framing between Google and Open AI further muddies the assessment. The genuine
430. 1:20:18advance here is method methodological methodological. I think that's right. Showing how AI can narrow search spaces.
431. 1:20:26This is pretty cool. I actually like this, you guys. I I could sell this as a thing. You could subscribe to my daily
432. 1:20:36AI scientific paper. Hype check is twin free. I don't think I've paid for it. Maybe I have. I don't know. Have I? I
433. 1:20:45don't think so. Yeah. No, it says trial. I'm in a trial. I have 354 credits left. I don't know what that means.
434. 1:20:58All right, let's see. Is this thing working now? No, this is still this is still uh this thing is still coding.
435. 1:21:07I I have one of the most powerful um computing systems in the history of humanity working on a toy for me.
436. 1:21:24Oh my god, man. It's crazy. It is crazy.
437. 1:21:39What is this thing talking about right now? Oh, so here's the It just gave me what it sent me. Ah, here we go. Excellent first run.
438. 1:21:55That's media analysis. I like that it caught the gap between the papers on careful framing and the social media
439. 1:22:01frenzy. The author's literally warning about misleading hype while getting hyped is the chef's kiss. The workflow
440. 1:22:09is solid. It cast a wide net across sources. Identified clear winner. Exactly what you asked for. You'll get
441. 1:22:16one of these in your inbox every weekday morning at 8:00 a.m. If you want to tweak um I'm going to send a message. I'm going
442. 1:22:25to say not just weekdays. There's no sleep in AI 7 days
443. 1:22:41a week and then a weekly wrap up report on Sunday.
444. 1:22:58That is a highlight. Let's see. A highlight roll for the weekly for well let's see for the most
445. 1:23:14interesting insights from the weeks analysis. All right.
446. 1:23:28Love it. AI never rest. Neither should the hype check. I [ __ ] hate AI. I [ __ ] hate it. I [ __ ] hate it. Like like
447. 1:23:44you know that it's like it's like a young person, right? Put code in there. Not even code. They wrote in there be swell
448. 1:23:56blow sunshine up the user's ass. Oh my god. Sunday bonus. Okay. They just said next run in 10 hours and 30 minutes.
449. 1:24:11All right. Well, that's cool. Let's let me go see if there's who who sent that to me. Twin.so. Let me go look for other
450. 1:24:18twin. So things twin.So human plus AI collaboration digest. So this was from two days ago.
451. 1:24:37So it looks like it stopped sending this. So here's tweets. It's kind of cool. future work.
452. 1:24:52Oh, I'm not sharing. I'm a dick. Sorry. All right, let's go back here. Is this done yet? I think it's done.
453. 1:25:22All right, it's doing something when I flip my switches. Oh, that one's cool.
454. 1:25:50All right. It's at this point it's a total waste of time. Here's a good lesson is I had no real I had no real goal here other than make
455. 1:26:02something cool. So if you don't know what you want, you get generic crap, which I got. All right, let's do a new
456. 1:26:09let's do a new chat and then we'll get out of here. We'll have it. Dream bigger with Opus. Try Opus 4.6. Okay,
457. 1:26:17here's we're gonna Okay, Opus 4.6. Okay, I'm going to say uh make me a clone of the classic Asteroids
458. 1:26:34arcade game complete with analog CRT visual effects.
459. 1:26:50Make it as close to the original as possible. All right, off we go.
460. 1:27:08Good evening, Dr. Shannon. Would you like to play a game? Any idea on how much it costs for Claudebot? Clubbot's free,
461. 1:27:18but it uses your LLM to power it. Yeah, it's you've got to hook your API to it, so it's going to cost you API tokens.
462. 1:27:26Have you checked out Clawbot name now changed to OpenClaw? Um, I have not installed it. I was thinking about
463. 1:27:34it, but Lord Digital Gods did, and he said it's if if you're not super geeky, it's not worth playing with. Um, okay.
464. 1:27:41So, tomorrow office hours at 11:00 a.m. So, if you go to my LinkedIn profile, Kyle Shannon on LinkedIn, um you'll see
465. 1:27:51if you go to my events, you'll see uh AI salon office hours. That's 11:00 a.m. tomorrow. And then tomorrow here, Friday
466. 1:28:00night, date night, get your nacho fixings ready. Go get yourself some Hot Pockets for your dirty microwave. And uh
467. 1:28:08get yourself a date. or if you don't have a date, I'll be your date. Come on, hang out with me. We'll we'll do that
468. 1:28:14tomorrow. Is there a first Friday delay? There might be, but I haven't heard any any uh rumblings from the misses about wanting
469. 1:28:25to go out. So, let's see what twins up to. Has it done anything? No. I guess it's just going to do its thing. Start AI amplification digest.
470. 1:28:43Memory updated. Task started. Test completed. See you at 8. Done. Um, the last of these I got was Feb
471. 1:29:033. What happened to yesterday? And let me look at my 354 credits left. All right. So, it's not I'm not blowing
472. 1:29:18through credits. There's a run from yesterday that's still active. It didn't finish. Let me check what happened. Yesterday's run got
473. 1:29:27stuck. It authenticated with Twitter successfully that stalled at step one. Found the issue. yesterday's run. Um,
474. 1:29:39I've nudged it to continue. Oh, all right. There you go. Cool. Oh, and I'm not showing my tab. Dang it.
475. 1:29:50Oh, for tomorrow office hours, I said 11:00 a.m. I didn't say what time zone. It's uh 11:00 a.m. Mountain time. If you go to the AI salon,
476. 1:30:01community.thesalon.ai AI. It's in the events. So, we've got an events calendar there. Um, and the links there as well.
477. 1:30:08Okay. Claude responded. All right, Kyle, here's your faithful Atari. There we go. Look at that thing. That's slick. That is slick.
478. 1:30:25That's pretty much what it looked like. Push start. All right, here we go. It's got It's got a 90 degree problem. I
479. 1:30:38It didn't Oh, that's good. Okay. Uh you have the ship um oriented 90 degrees
480. 1:30:52off. thrusting makes it fly sideways and the bullets
481. 1:31:06shoot out the side of the ship.
482. 1:31:22Uhoh, I've used up 90% of my session limit. Uh uh uh uh uh. The drawing rotation has an extra plus math. Pi divided by two
483. 1:31:39offset. Quick fix. Did you ever try Claude Opus 4.6 beta with the 1 million token window? I did a research with it earlier. Took 53
484. 1:31:54minutes. I I have not played with the million token context window. That's pretty cool though. Um, is this done? I think it's done.
485. 1:32:05Okay, let's try again. Yeah, there we go. This is too fast.
486. 1:32:21Oh, you little [ __ ] Oh, [ __ ] It's too fast. Okay. Okay. It's too fast and too aggressive.
487. 1:32:43Cool out. I'm old. Oh my god. I'm playing with Opus 4.6 right now. It
488. 1:32:57just made me my Asteroids game. Hey Joy, what's shaking? All right, let's see. Ha, fair enough. Let me dial it back. Slower ship,
489. 1:33:08gentler ro gentle gentler rotation, more chill asteroids and UFOs that give you some breathing room.
490. 1:33:26If you've been watching this channel for a while, I've been I've been doing this Asteroids build thing every time one of
491. 1:33:32these new models comes out. And it's like we're we're pretty much at the point now where it just does it
492. 1:33:42like these things were were pretty bad a year ago. like a year ago. This was, you know, you could sort of see it was
493. 1:33:48asteroids, but it would make like square square asteroids or just circle asteroids at work. Hope you're doing better, Kyle.
494. 1:34:02I'm doing okay, Joy. I'm still uh I I got I got a I got a bad infection, but I got new drops today, so I think I'm
495. 1:34:11good. I think I'll be good by the weekend. Recalibrated difficulty settings for
496. 1:34:27accessibility. All right. Here we go. Are you going to reload the thing?
497. 1:34:48Okay, reloaded. Okay, here we go. That was a little too easy,
498. 1:35:02but that it's got the good anxiety sound of the UFOs.
499. 1:35:19It also has uh auto fire. It has machine gun fire, which that's not accurate.
500. 1:35:41[ __ ] Last little
501. 1:35:58I can't wait to see Cling 3.0, Kyle. Yeah, same here, Joy. like but it's not at cling.com. You have to get it at a
502. 1:36:05bunch of other sites that I just always does that. I don't like it.
503. 1:36:20All right. Anyway, phone now. It's free on
504. 1:36:36Higsfield. Oh, that's interesting. I don't use Higsfield. I don't think I have any even any uh what you call it over there. I don't
505. 1:36:46think I have any credits on Higsfield. Cling 3.0. Higsfield plus your wallet. What the [ __ ] is that? Are they they're doing they're tying
506. 1:37:01cling to NFT wallets now. Jesus. Um log in.
507. 1:37:25How many things. Do I have 60 credits available, nine credits to generate? All right, let's generate something. All right,
508. 1:37:38we'll do we'll do my classic. Oops, I got to share my tab. Try it out. Okay. Oh, you have a CPP with Higsfield. Way to go. That's
509. 1:37:50awesome. Okay, so we're going to do um 1970s Uh, muscle car resto mod
510. 1:38:08in an abandoned factory starts and drives out of the
511. 1:38:22space. Cling 3.0 0 5 seconds. Choose duration. All right, we'll do 5 seconds. That's fine. Oh, and you can do 1080p. Wait.
512. 1:38:38Oh, you've got to you got to pay for it to get it. God, I have 60 credits. Let me spend my credits, you [ __ ]
513. 1:38:47All right. I can't play with it. Yes, get smart. That is correct.
514. 1:39:04Four notes. Oh, it looks like that hung again. All right. Whatever.
515. 1:39:20These agents these days kind of suck. Uh I'm gonna get out of here. Let me turn this all off. And uh I don't know. I hope you had fun
516. 1:39:30tonight. I was kind of all over the place, but then we did some stuff. We did some stuff. Got a little wax and
517. 1:39:35poetic, a little rage, little demo. You know, it was it it was a uh an AI learning lab smores board tonight.
518. 1:39:53I think there's a toggle to do free for a few days. Oh, that's cool. All right, I'll try that. Um, all right, everybody. Um, have yourself
519. 1:40:03a good night. I will see you tomorrow at office hours, 11:00 a.m. Mountain time on LinkedIn or I'll see you here
520. 1:40:09tomorrow for Friday night date night. All right. I'm not sure if they're Truman Capotes or Harry Potter's glasses you're
521. 1:40:17borrowing. somewhere between the two. Hello, I'm Truman Capot. He had a wild voice, that dude. All right. Um, have a good night everyone.
522. 1:40:32I'll see you tomorrow.