AI Learning Lab LIVE - 12/30/2025

Transcript

1. 0:06Yeah. Hold, please. Beautiful.
2. 0:32Oh no. >> With a spoon covered in sauce.
3. 2:39All right, good people. I'm back. It's okay. Everything's going to be fine. It's all going to be fine. You're going
4. 2:45to be okay. I've come It's all right. I've arrived. I'm telling everyone, you're getting your notion board configured. Listen,
5. 2:57here's the victory with notion. I did not quit the app today and it actually forced me. Okay, so a couple of So, there's a lot happening.
6. 3:11Anyway, we we'll get back to that. Let me sing with my dog. That's why you've come here. Really? Have you not? All
7. 3:25right. Ready? One, take one. Zoom in on Kyle and action.
8. 3:42Oh
9. 4:09time I see it now. Get that look in mine. Every time I see your mouth, I hear that smile. In the early misty morning light, I
10. 4:25heard the engine turning in the old for outside. Will you believe in me again today?
11. 4:41You will convince me again today. You're leaving this hotel looking for someone else's golden ring.
12. 4:56Should have said so long. Susanna, first now don't you cry. So long, Susanna.
13. 5:15Don't you cry for me.
14. 5:32It's Mr. Champ Shannon on lead vocals. My name's Kyle Shannon. I'm singing back up for the dog. This is where my life is led me.
15. 5:41Hey Danielle was shaking. Side hustle me in the house. We got Vicki. We got producer Brandon. Producer Brandon right
16. 5:48here. You know what he's doing? Producing Vicki on YouTube. Better. There's a notion tutorial in irregulars
17. 5:55for anyone who got confused last night. So, listen, if you were not here last night, so a couple of things going on.
18. 6:02Welcome if you're new. Um, it's just a handful of people in here. This is probably all people that know what the hell's going on,
19. 6:10but we're shifting the focus of this channel.
20. 6:25And what it looked like was this. We uh we Liz Miller Gersfeld and myself we we uh we took some work that some
21. 6:38personal work that she's doing about creating a personal practice around her use of AI and said wow what if we created a
22. 6:54framework that's that's based on the ideals of the of the AI salon and play first, create excellence and generously lead the cycle
23. 7:05of AI readiness and we build a practice framework around that and teach people how to do a daily practice.
24. 7:14Sounds awesome. Sounds awesome. And we created this thing called the AI Salon Mastermind Practice Lab. And it's a weekly
25. 7:27meeting, focused meeting where we're teaching people how to create their daily practice using this framework we created.
26. 7:35Everything's good so far. Everything's awesome. It was funny when when I first heard about the idea of an AI practice,
27. 7:48my immediate instinct. I just hold everybody hold. I've got schmutz on the back of my phone lens. It's causing glowy things. Let me see.
28. 8:00Do I have Oh, I've got a a pink bow. The famous pink bow. Hang on. Hang on, people. Everybody, everybody calm down.
29. 8:13Good lord. There, that's better. See, not so glowy. Okay, so when I first heard about this idea, this this practice,
30. 8:29I thought, I know I'm not wearing it right. Whatever, ladies. I know how to wear a headband. Okay. Um, I thought I would love to have a a
31. 8:44practice. That sounds so cool. And then I thought, wait a minute. I go live on TikTok five nights a week.
32. 8:54Technically, not a daily practice, but a daily weekday pra a week daily practice. A week a weekday daily pra a week a five
33. 9:02day a week practice. And then we started teaching the practice, the framework. And guess what I learned? I didn't have
34. 9:16a practice. I thought I had a practice. I didn't. I had a habit. I had a habit of showing up five nights a week, which
35. 9:25to give myself some credit, is pretty [ __ ] remarkable. No, I I think it's it is just just slightly less remarkable
36. 9:34than all of you who show up here on a nightly basis to watch it. I think that is equally if not more remarkable that that happens. Um
37. 9:44but that's not a practice. A habit is not a practice. Doing something every day does not necessarily
38. 9:59accomplish anything, right? We get up every day and we have coffee. It's part of our history. I know it is.
39. 10:12It's actually the fact that we all hang out here together for, you know, years now is [ __ ] remarkable. And you wait.
40. 10:21We're going to work on Sydney tonight, by the way. So you you just [ __ ] wait when that show [ __ ] opens in New York
41. 10:27City and we're all in the theater together watching that show. I I won't it I won't be able to be there. It it I I'm going to be a train
42. 10:36wreck. Um anyway um a practice requires intentionality and um
43. 10:52it's taken me we've been we've been teaching the course now for I think we're on our seventh or eighth week so two months
44. 11:01and I'm just now starting to get my head around what it means and and the the kind precipitating force uh for me was
45. 11:10was Cindy Coon's talk during AI Festivus. So if you haven't seen if you were not there for AI Festivus, it just happened this past weekend.
46. 11:18If you missed it, it's a shame, but this is a ridiculously low price. For 27 bucks, you can get the deluxe replay
47. 11:27bundle that includes like Jelly Pod podcasts about the sessions, notebook LM, infographics, the recordings, transcripts, the presentations,
48. 11:37everything for 24 hours of programming. Champ, you got to be kidding me. You got to be kidding me. I I I haven't even
49. 11:47start I haven't even gotten to the good stuff yet. I want a little cheese. I want a little cheese. Daddy. Daddy.
50. 11:57Daddy, get out the cheese. Daddy, get out the cheese. All right, hang on. I got to go let the dog out.
51. 12:14>> Daddy, can I have cheese? >> Go. Go. Go out. What are you doing? Oh my god. >> What do you want?
52. 12:28>> You want to sing more? You might need to sing more. Little son of a [ __ ] I opened the door and he wouldn't go out. What do you
53. 12:37want? You want to sing more? Champ, what are you doing? You singing?
54. 13:02Hey. Hey. Hey. There's no rapping. What are you doing? We're singing or we're not. All right. Um, take three. Oh, please
55. 13:15listen. What do they say? Never work with kids or dogs. This is why. This is [ __ ] why. Oh my god. He really did want to sing. Now
56. 13:26he's just laying down. Are you you What? What was that about? He like whined by the door. I walked over and opened the door and he ran back
57. 13:35in the room. Dogs are weird, man. Dogs are weird. Play Trouble. My favorite. Okay. A champion might sing to trouble.
58. 13:58Standing between you and a hard place is insane. Standing too near you in a fire makes it clear.
59. 14:16You're trouble to me. Rich trouble catch you say leaning in close.
60. 14:30Smell of your perfume scares me most. Leaving away feels stronger every day.
61. 14:46You're trouble to me. Real trouble. Can't you see?
62. 15:01Okay, back to the practice thing. So, um, a practice is something you do every day with intention. And what it also looks
63. 15:16like is you have to sit with what I call the ick. And the ick is that thing that if you want to do something, so play first,
64. 15:26create excellence. the the create excellence part of the the cycle of AI readiness. You can't just live and play first land.
65. 15:41If you want to create excellence, you have to go to some dark places. If you want to create more than a seven-second video out of one of the
66. 15:51video models or you want to create something that has the impact of Kelly Besh, which I just realized I've been
67. 15:58mispronouncing her name for [ __ ] three years. It's got an O in it to my defense, but I was interviewing Kelly during
68. 16:11Festivist and she goes, "Hi, I'm Kelly Bes." And I was like, "Ah, [ __ ] Really? Mimi in the making. Sad hustle. Mimi. I
69. 16:21like Oh, Mimi in the making. I love it. Being born. Exactly. Just like what we're doing here. So, here's what's
70. 16:26going to here's what's going to happen. Rather than this channel being about me flasharning something
71. 16:36an hour before I teach it to you, which is which is historically what it's been. What the channel's now going to be is
72. 16:47I'm still going to be exploring new tools. I'm still going to be playing first, but what I'm going to be doing also is
73. 16:54working on my projects. I've got so many projects and they're projects that are meaningful to me and they're meaningful to other people. And
74. 17:04a lot of them, every project, if you want it to be truly excellent, has things in it that are distasteful.
75. 17:11Things where you have to learn something new. Things where you have to do some file organization. Things where you
76. 17:16maybe have to do some finances or some contract work, right? Not just the sexy stuff. Like one of the things that we're
77. 17:23going to work on tonight is I put together this podcast. It's like a Notebook LM podcast with the Sydney songs edited into it and it was
78. 17:35really good. It was really like I was really happy with it and everyone else was really happy with it. And then
79. 17:40Andrew is like, "Well, except can I do an edit where I do a much more sophisticated edit with like, you know,
80. 17:46better, you know, transitions between songs and overlaps with the vo voices and can we get the voices just exactly
81. 17:53right?" And then we have a new version of the of the poster that is not just like this 14-year-old pasty white girl, right? It's it's a
82. 18:05it's a racially, you know, diverse or or neutral neutral entity and she doesn't look like she's 12. Um, and we've got a better logo. And so
83. 18:16I've got this whole video that I've created with all the old logo in it and the old the old Sydney in it.
84. 18:24And so this thing that I spent, I don't know, two days working on, which I thought was quite good, was good, but it
85. 18:33wasn't excellent, right? So to make it excellent, I'm going to have to completely rebuild it. And I don't even
86. 18:39have all the components yet, but I have the new logo, the new posters, and so I figured we'd work on that
87. 18:45tonight. But then there's also things like here, let me show you something. I want to show you something fun. Well,
88. 18:49it's not fun. This is one of those ick places. Um, here is my
89. 19:04go into I don't know if you can see that. Is there a way I can make this bigger? Does
90. 19:18command plus work on these windows? It does. Oh, it's just two two things. Okay. So, you know how I always joke about um
91. 19:28saving things? Well, you you may or may not I always joke about my my file uh saving system is to save things to the
92. 19:38desktop with a nondescript name. So, you'll notice here I've got a bunch of nested folders. So, I've got Kyle
93. 19:44Shannon monetizable products, AI launchpad workshop, storyline selling, AI festivist, content evolution collab,
94. 19:52AI readiness training program, 10-person team workshop, AI Kyle's AI consulting. So, I've got all these these nice
95. 19:58organized folders. Those I did today. Those I did today. Now, I'm going to show you something. I'm just going to keep scrolling here.
96. 20:06All of these are loose individual files that have zero organization to them. Within these files, there's actually
97. 20:20some really good stuff in here. There's stuff that could contribute to projects being greater. That's this is just my loose Google Drive.
98. 20:31My personal my personal Google Drive is just full of just random [ __ ] Okay. And so so like when I talk about sitting
99. 20:41with the ick and like why would I spend any time on this channel showing file structure? Um when I when I look at how organized
100. 20:56this is and like all these projects and I can start to actually put stuff in them. That's kind of fun. That's kind of
101. 21:01exciting. Then I think about the notion um the notion instance that Vicky Baptiste helped me figure out last night and let's see
102. 21:15if I go to my control board. Here's Kyle's control board. So this is now I'm in notion. So, Lord Digital Gods sent me a a really
103. 21:24funny thing that said, "I wonder if Kyle's gonna is how long will Kyle how long before Kyle quits notion?" I'm
104. 21:32like, I'm not quitting it. Damn it. I'm gonna I'm gonna figure this thing out. I'm going to figure this thing out. So,
105. 21:38within here, I've got business areas. So, I've got AI Salon, Content Evolution, Collab, Kyle, Shannon Media.
106. 21:45I want to I want to make sure that all of these things all of these projects in notion match all of my folder structure in my hard drive.
107. 21:59Now, why is this? Because So, if I flip over here, wait, we're we're taking side bets. Yeah. No, listen. Please do. Please do hold me
108. 22:10[ __ ] accountable. Um, so here you're looking at some some New Sydney um videos here. Let me show you show you this one.
109. 22:23Right. Looks great. That's our new That's our new logo because I'll show you because the technology has gotten better since
110. 22:34the first time I made this this thing. Um the like the the uh the text on the on the marquee up here and the text on the
111. 22:44ticket's actually correct. This is a little [ __ ] up. It says like limit limited dough but but that looks pretty cool. And then
112. 22:53if I hit play cool, right? Uh Mimi Kyle, this is giving me cognitive dissonance. I know. I know.
113. 23:10So, so, so part of the daily practice is being intentional about the work you do. So, what I would have historically
114. 23:21done and and the way, let me show you. I'm going to launch iMovie. It's giving me cognitive dissonance. I know. Listen, I can be
115. 23:34I can be organized. I choose not to be right. I don't I don't like it. Like I like organizing things. I don't like the ongoing
116. 23:46organizing of it. One of the reasons I'm excited about notion if I do it right is that I won't be in that situation
117. 23:54where I'm like, where was that thing? Where was that thing? I know I did a thing. Didn't I do a thing? Sometimes
118. 24:01I'll say, "Didn't I do a thing?" and it'll take me like an hour to find it and then I'll realize not only was it a
119. 24:07thing, but it was a thing that I spent three or four days on and it's literally just scattered across my digital
120. 24:15storage facility. Um Vicki says, "You can now put videos in notion, too." Very cool. That's great. You can now put
121. 24:23videos directly in Sydney. That's super cool. Um, okay. So, this is So, let me go back to projects here. This one.
122. 24:43So, I'll play a little of this here. You can hear that, right? See?
123. 25:09Okay. So, see this shot right here? See all the white people? So, one of the comments I got was, "Can we make it a little less white?" It's
124. 25:24New York City. Um, so that was a comment I got. But when when I made that original video, I went into Lumal Labs. I think I might
125. 25:36have even even done some of them here on the live. I went into Lumalabs and I just downloaded all the I made a
126. 25:44bunch of video clips and then I just dragged them over to iMovie with the song behind it and I stuck them in there
127. 25:50and I just did quick edits and I just I didn't none of those raw assets are saved anywhere. I have no
128. 25:57idea where they are. They're probably in the iMovie thing somewhere. They're probably somewhere else. That kind of chaos
129. 26:05serves me in the moment, but it doesn't serve me long term. This is a project I want on [ __ ] Broadway,
130. 26:14right? I don't want this like in in the the local community theater in Wheatidge, Colorado. I want this on [ __ ] Broadway. So,
131. 26:27if I'm halfassing the things I produce for it, talk about cognitive dissonance, there is this cognitive dissonance between those two
132. 26:37things. So part of the intentionality of a daily practice is if you really want to step up your game, if you want to
133. 26:43make more money at AI, right? A lot of people are like, "How do I make money with Chachi?" Right? The famous Wait,
134. 26:49where's my button? I don't even know where my button is. I do. >> Right. All those [ __ ] promise you you What's that require?
135. 27:03It requires you being excellent. And what does excellence look like? I don't know yet. One, one part of excellence is showing
136. 27:13up. That I've got nailed. The next part is show up with intentionality and understand what you're trying to
137. 27:22accomplish. And Cindy Coon's talk about come up with three themes. Pick three themes. Three. And do everything within those themes.
138. 27:34The the batteries on that button have outlived many AI startups. It's true. >> I've never changed the batteries on this.
139. 27:46We got some mileage out of that. Hang on. Dog wants out again. You want to You ready to go out for real this time? Come on. Go.
140. 28:12So,
141. 28:28huh? Huh? I'm I'm I'm I'm I'm literally trying to figure out I was just sitting here thinking
142. 28:43if I'm going to recreate this video. What would be really nice is if I had a folder with all the images that were the
143. 28:51start images for this video, but I don't. I think I actually just made them all in I don't know where I made them. I think
144. 29:00I made them in chat GPT. So like this is something. Okay, so here's something that I've thought about for a while and
145. 29:07this this affects me at at StoryVine at my at my uh company. Um, a lot of times when I'll put together a
146. 29:18project, and I know I know you all know this, if you're doing anything in AI, you've experienced this, where you're like, you
147. 29:26you dynamically do tool switching on the fly. Okay, this this video tool isn't working. Let me go try that one. Let me
148. 29:33try this image tool. Let me try this audio tool. Let me try chat GPT. Now, that didn't work. Let me go over to
149. 29:38Claude. And so what I end up with is I end up with this Frankenstein kind of chain of craft where the final product was okay I wrote
150. 29:51the thing in claw I put it into nano banana it turned into something and you know I went put it into notebook LM and
151. 29:58then we're there. So the the final chain of craft is clear but there were like 17 tools to get there.
152. 30:06notion would be a perfect place to say, "Here's that chain of craft, right?" Um, and just throw it in there. Just have a
153. 30:14thing that you throw in there and then you'll be able to find it again in the future. That requires some discipline. That
154. 30:20requires setting it up. That requires some practice. But like, that's a thing that the next time I go to do a project
155. 30:27[ __ ] me up because I spend a ridiculous amount of time trying to find all this stuff. Okay. Good lord. Good lord. Good lord.
156. 30:39All right. What's going on? What's going on in the chat? Chatty chats over here. You can put your videos in there. Vicki,
157. 30:45does not connect with Okay. I don't know. Okay. So, so the first thing we need to do
158. 31:10Now, can I I do it? Let's see.
159. 31:25Okay. So, there's that and there's that. Okay. So, that's fine.
160. 31:42And then I jump to that scene and we've got an old Sydney logo, right? So, what I need to do is go probably there.
161. 31:59Oh [ __ ] Yeah.
162. 32:17So, like here's a good this is a good example of I don't quite know how to do this. This is this is absolutely one of those
163. 32:26moments where where I could just go, "Oh, [ __ ] it." and just just say, "This is good enough. I'm not going to take
164. 32:34the time to do this," right? And that is the thing we want to battle against. And it's not even a battle.
165. 32:44The point of a daily practice is like understand what you're trying to accomplish. Just cut the lines, right? If you're in
166. 32:54a bar, you just Someone's got to cut the lines, right? We all want to be Tom Cruz and cocktails spinning bottles around.
167. 33:03I have no idea what I'm doing and it's great. Yeah. Yeah. Embrace that. Embrace the jank. Embrace feeling clueless. Embrace
168. 33:13that feeling of ickiness. Because where you've got that feeling of ickiness, where you've got that fear
169. 33:21that's like I don't want to deal with it. On the other side of that is there's some [ __ ] magic, you know? Tik Tok pin Vicki. Oh, YouTube
170. 33:32Vicki. No, not notion search is so good. Good. Exactly. That's like my instinct seeing just seeing what it is. If I I
171. 33:41don't even know what to call it. It's like a a weird combination between a notetaker and a scheduling tool and a
172. 33:47and a [ __ ] relational database. It's it's notionally like some combination of those. Um but yeah, like I don't want to I don't
173. 34:02want to like not know exactly what to do in front of all you people. That's [ __ ] gross. Why? Oh, I probably have to click.
174. 34:13Yes. Okay, great. So, now what I'm going to do is I'm gonna What happened? Oh, I rolled over. Oh my god. Okay.
175. 34:23Can we Can we not Okay, we're good. Okay, so I'm going to make a a screenshot of this. Can What just happened to my screen? Oh,
176. 34:40it's Why is it so dark for you? Oh, cuz you're not seeing You're not seeing what I'm seeing. I've got a I've got like a little bright live
177. 34:50window that you're not seeing. Okay, it's a floating window that's not showing up, which is weird. Okay, whatever.
178. 34:57Okay, so I just took a screenshot of that thing. So, now I'm going to go over to chat GPT because I did do a little testing on this earlier.
179. 35:07All right. So, I said this this was um there's the original. See, there's the original with the old logo
180. 35:17and the dude the dude back here and the old the old face. So, I said there's that picture and then here's the new poster with the new face
181. 35:27and the new logo. And I said, you know, make the ticket look like the ticket. And then it ended up, I guess, there.
182. 35:40There we go. It's which this is still not quite right now that I look at it. Um, should I be excellent? I should probably
183. 35:51be excellent, shouldn't I? All right, let me go. Dang it. Uh um I'm going to say
184. 36:07notice in the original poster I uploaded. The standing man
185. 36:25is white, not gray. Also notice
186. 36:40that on the ticket there is no man at all. Please update
187. 36:56the ticket design and marquee design
188. 37:10to reflect this detail. And then I and I think what I'm going to do is I'm going to go add
189. 37:32the thing back in. And by the way, because I organized my um because I organized my folder, sorry, Moo. No. Um, because I organized
190. 37:46my folder, I know exactly where the latest posters are, right? Like, this is the thing. Having a a horizontal poster,
191. 37:55a vertical poster, and a square poster that are the latest graphics with the latest logo in a single place. I know
192. 38:02this seems [ __ ] ridiculous to you, right? That it's a big deal for me to have the latest things in a folder I can actually find.
193. 38:15One of the joys of having an ADHD brain is that you can move fast and you can keep track of a lot of [ __ ] It's counterintuitive, right?
194. 38:29So, I have a a good enough brain to go, "Oh, I know. I saved it. I know I saved the latest one probably to my desktop,
195. 38:35but if I don't have it there, I probably have it in Andrew's tweet chat thingy he sent me. But there's so much cognitive horsepower
196. 38:46that is spent just trying to sift through the chaos of my life that that doesn't make sense anymore, right? Okay.
197. 38:56So, there's our new logo there. There's our new poster. So, notice in the original poster I uploaded, the standing comma, the standing man
198. 39:06is white, not gray. Um, he is also to the right of the word Sydney,
199. 39:20not above it. Okay. I also noticed that on your ticket design, your ticket design,
200. 39:37there is no man at all. Okay. Boom. Let's see if we can get this thing to not suck. And and this trying to get AI to like behave
201. 39:48is a nightmare. I was trying to do this with Nano Banana earlier and it was it was a disaster. Would Nano Banana not be
202. 39:55a little better for slight changes? It might be, but it wasn't this afternoon. It kept giving me it would it would
203. 40:01redesign the marquee and it just left the the front ticket alone and I couldn't I couldn't get it to change the
204. 40:08design of the ticket. But but the image that I put in here as the base image was an early iteration of a nano banana
205. 40:15image. So maybe I just need to go back and forth between the two. Okay. Um, I use my recent shortcut like a
206. 40:25lifeline, but it's not working well lately. Yeah, exactly. Exactly, man. This is this is uh this is the ick. This is the ick.
207. 40:40And I don't know, there's a part of me that's like, do I even want to [ __ ] deal with any of this? But well, that sort of worked.
208. 40:48So, we could say um on the marquee the white the man is above the Y in Sydney.
209. 41:08It should be to the right of it. But notice we now have the man on the ticket. So that's cool. The other thing about this, what I'm
210. 41:24doing right now is Oh, Champy's back. Little [ __ ] Um, in this final video, this is going to be seen for about a quarter of a second.
211. 41:39Does it matter that the little dude isn't standing there on these? No. But is that excellence? No. Where do you draw the line? I don't
212. 41:49know. Like like this is this is the whole thing about being part of a practice. There's there's learning. Can you get the tools
213. 41:56to do what you want them to do? But there's also like what is the standard that you're producing to if I want to produce for Broadway
214. 42:06or even off Broadway for [ __ ] sake anywhere in New York, you know? So that didn't fix it at all. In fact, I think
215. 42:15it made it worse. Made the color worse. So, here's what we're going to do. We're going to take this non6 by9 image.
216. 42:23Copy this image. We're going to go to nano banana. We're do a new chat. We're going to paste this in. I'm going to say in this image the
217. 42:45Sydney logo on the marquee has the white male figure above the Y.
218. 43:05It should be to the right of the Y, not above it.
219. 43:20Please fix. And you know, it's also very possible that this kind of shorthand prompting that I do on the AI learning lab live,
220. 43:33this might not be appropriate for the level of work I want to do. Like I might I might need to be doing more structured prompts, which we
221. 43:40can go fix. Like that's not a problem to solve, but it's got to be solved. It's got to be solved, everybody.
222. 43:57I still remember the first time I took a picture of my notes and chat GBT read. Oh, that's something I wanted to do
223. 44:03tonight. Actually, that I'm glad you said that. We'll we'll we're going to do something on that front, too.
224. 44:11So, okay. So, look what look what Nano Banana did. This is the kind of [ __ ] that So, now it's just Sid with the dude
225. 44:18and the Y and then it's DN. The DNE show. Hey, did you go see DNE on Broadway? No. God damn it.
226. 44:33[ __ ] Um,
227. 44:52you didn't change anything. Look at this layout
228. 45:06in the uploaded poster and make sure the marquee
229. 45:22matches it there. There's also what I'm about to do is I'm about to bail and go do it in Photoshop. Nanobanana has rarely done that sort of
230. 45:35thing to me. You know, archetypal, I'm wondering, do you use Nano Banana in uh in AI Studio? I wonder if the Nano
231. 45:42Banana in AI Studio is different than the one on Gemini because I was that was on Gemini. Um, so let me upload my poster again.
232. 45:55I mean, this seems [ __ ] ridiculous that you have to keep uploading the poster because it's got it in its [ __ ]
233. 46:01memory, right? It's got it somewhere. It's somewhere on an OpenAI server. My little [ __ ] picture is sitting. Okay.
234. 46:20I'm almost exclusively you using it on Gemini or through weevi. Okay. Yeah, it's weird. I don't know. I don't know what the deal is.
235. 46:31The other thing I've noticed about in the in the grand total of two days, day and a half since I've been doing this is
236. 46:39the time for me goes by really fast. And I would imagine the time for you is like this is probably like watching paint
237. 46:45dry. So we'll we'll try to find a balance of that. But it it might also just be that this is a it's a different kind of compelling
238. 46:56to work on stuff. Um, and by the way, this is one image of I don't know, probably a doz Oh, it's so
239. 47:12close. Okay. Scooch the man to the right
240. 47:26on the marquee so it isn't over overlapping the Y in the word
241. 47:42Sydney. Uh uh uh uh uh uh uh. Oh, and you know what else is going to probably [ __ ] happen as I'm doing all these iterations
242. 47:52is at some point it's going to go, "Oh, sorry. You're out of credits, Mr. Shannon. Mr. Shannon, Mr. Shannon."
243. 47:58Yeah. Listen, um we noticed that an ex excessive amount of uh token generation under your account. You're only paying us $20.
244. 48:11and we were kind of, you know, we kind of were thinking maybe, you know, you could not make so many. So, what we've
245. 48:17gone ahead and done is we've just turned off your ability to continue doing your work. I hope that works for you. We we
246. 48:25really value you as a customer, Mr. Shannon. It's just that the excessive use of tokens from our system is really
247. 48:33kind of making us a little cranky, if you know what I mean.
248. 48:50All right. So, here's a question. Do you Do you [ __ ] punt? No, I'm not going to punt. Here's what I'm going to do. I'm gonna open up
249. 49:02Photoshop. Do I want to open up Photoshop? I don't think I want to open it on that one. I think I'm going to want to open it on this one.
250. 49:11But is the color different? No. Fine. So, we're going to copy this image and open up Photoshop. Oops. That's photos. Photoshop.
251. 49:26Yes, it's Mr. Shannon. Mr. Shannon. Mr. Shannon. Yes. Yes. Yes. No, I was There were some questions in the crowd. Uh
252. 49:35really curious, deeply wondering pretel. So by chance, what uh are your qualifications? Because it appears uh
253. 49:50the credentials are thin.
254. 50:05Never punt on first down. Great point. Never punt on first down. And we are absolutely in first down here. By the
255. 50:12way, I can see it right now. This is not a 16 by9 image. Did you see that? Gro Gro Gro can now output images and videos in
256. 50:25all of the major aspect ratios including 16 by9 which is what all of your TVs are and what all of your Tik Toks are 9 by6 16 by9
257. 50:39chat GPT still cannot do 16 by9 outputs absolutely [ __ ] staggering to me okay let's see let's See, here's what we're
258. 50:50going to do, people. Here's what we're going to do. All right. We're going to zoom in.
259. 51:10I'm going to grab that. And then I'm going to minus select. Can you all see this what I'm doing? Yes. I'm going to minus select that.
260. 51:25Oops. What did I just do? What did you do? What did you do? What? That's bizarre. Oh, I see.
261. 51:4130 pixel feather will do it every time. Yeah. Yeah. See? Yeah. You puga. Yeah. And another thing, Johnson, get your
262. 51:50dick off my desk. All right. Copy. Okay. So, now what we're going to do is we're going to select
263. 52:05the dude and we're going to use the one thing that Photoshop seems to be good at is removing [ __ ] So, I'm going to say remove, man.
264. 52:18with generative fill. And I've got I copied the man. So I've got the man on the clipboard. Oh no, it didn't remove it. All right, fine.
265. 52:30Here. Oh, you know what we'll do? Oh my god. I can't this this. It's unfucking believable that I have to [ __ ] resort
266. 52:38to like 20 23 Photoshop [ __ ] skills. Like really? You're kidding me, right? I got to copy and paste stupid [ __ ] for your stupid
267. 52:54cuz you can't figure out how to make your stupid [ __ ] Follow my stupid instructions. Flatten image marquee tool.
268. 53:06I don't know what I'm doing here. I'm supposed to use skills. It's not what I signed up for. This is not what I signed up for. Okay.
269. 53:22Does that got to go down? I think it does. Right. Ah, this is so exhausting. Okay, there. It's mostly gone. Now I got
270. 53:38to get the rubber stamp [ __ ] tool. This is not what I signed up for. This is not what I signed up for. You want to
271. 53:45know why we can't have nice things? Cuz daddy's having to [ __ ] Photoshop stuff when we have magic robots that
272. 53:54just make stuff. Okay. That's why daddy's not happy. Oh, god damn it. Um, hang on.
273. 54:11Uh oh. What's going on? Is it still Monday? Well, this is what happens when you get what? Just [ __ ] should work. [ __ ] should just work.
274. 54:23This is not my fault. Everybody needs a cranky Gen Xer in the corner office. That's what it's like. That's very
275. 54:31triggering. He's very angry. Why is he so angry at Photoshop? Oh, that looks like [ __ ] Looks like dog
276. 54:39[ __ ] Just looks like H. Okay, so here's what I'm going to do now. I'm going to use generative fill to smooth the gradient.
277. 54:48Okay, generative fill. Smooth gradient. We're We're using six trillion tokens to smooth my gradient because I don't have
278. 54:58rubber stamping skills or patience. Oh my god. Seriously. Seriously
279. 55:16smooth gradient only. And if you're if you're sitting there
280. 55:31judging and thinking, is anyone going to see this? Is anyone going to notice this? No. Oh [ __ ] that was the wrong All right,
281. 55:43there we are. Okay, D. Uh, let me get the rubber stamp tool. I'm I'm just going to rubber stamp it shittily and then we're just going to
282. 55:52move on with our life. Oh, I [ __ ] hate this. Okay, there we go. That's good enough. Now paste. There's our dude.
283. 56:08Right. Beautiful. Beautiful. Beautiful. Beautiful. And I think what we're even going to do, let me just make it a little like And then
284. 56:25there we go. And I think we've got to do this a little opacity opacity like so that it matches.
285. 56:41That's probably better, right? Do you see what we did there? Look how excellent we are. Our quality of excellence here is just
286. 56:54crazy. And if you're like, "But the man's a different size relative to the Sydney." I know. I know. And the Sydney
287. 57:00logo is not actually correct. Oh my god, it's exhausting. But this but this doesn't matter. Okay. Um, let me go layers. Let me go flatten
288. 57:11image. And here's what I'm going to do. Oh, the other thing I need to do is I need to turn this into a 16 by9 image.
289. 57:45Okay. So, now we have a 16x9 image that's got mostly our logo correct. And we're going to go export as PNG.
290. 57:55And where are we going to save this? Not to the desktop. So, we're going to go to posters. We're going to go podcast 2.0.
291. 58:08There's nano banana test. So, we're going to go new folder, poster 2.1
292. 58:22images. Uh, let's see. Um, oh, let's see. Poster 2.1 start images. I think that'll make sense. Maybe I've
293. 58:37got to remember I called this poster 2.1 and then we'll just go we'll call this ticket. Okay, beautiful. Save. Great. Next.
294. 58:48Just like 12 more of those to go. Um, got to go back here. Got to get this screenshot again. That. And then we're going to go here.
295. 59:04Okay. Okay, people. Fantastic. Fantastic. Here's what we're going to do. All right. We're going to paste in
296. 59:12that image that we just screenshotted out of the old video project. And we're going to say replace. That thing failed to upload. Of course
297. 59:22it did. Okay. Yes, that's the ticket. Acting. It was only acting indeed. Fantastic. I sneaked out of watching a movie with the
298. 59:33wife to come here. I'm so sorry. Generative Phil is a good Tik Tok name. It actually is a really good really good
299. 59:39Tik Tok name. Somebody do that. Somebody take Generative Phil. Kyle, how many tabs do you have open? Listen, that's
300. 59:45not appropriate. That's not an appropriate question. Cam Ken from Cleveland. 1 2 3 4 5 6 7 8 9 10. But
301. 59:53three of them are X. 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 I don't know 30ish, which is low for me.
302. 1:00:05If I'm not in the 70s, I'm not I'm not pinning the memory requirements of my my uh computer. Okay, let's see. Um, update
303. 1:00:20update the the marquee in this image with the new Oh, maybe I should do this over in nano banana. Nano banana.
304. 1:00:36Nano banana. Let me go try it. Nano Banana because I think I think Nano Banana generally is better, but it it's been
305. 1:00:48really acting up for me today. I don't like it at all. Okay. Oh, wait. And I've got this. The other thing that drives me crazy about
306. 1:00:56Nano Banana, you have to put in the last image that you upload has to be the aspect ratio that you want.
307. 1:01:06That's something that they need to fix that stupid [ __ ] bug. Okay, there's Sydney. Okay, so let's see. Let's see here.
308. 1:01:18We're going to go paste. So that's the new poster in the um photo, replace the
309. 1:01:34poster in the marquee above. Times Square
310. 1:01:48with the new poster uploaded here. All right, let's see what we get, people. Let's see what we How many tabs do you have open?
311. 1:02:19All right, that's not bad. It looks a little Why does it look a little Let's see. It looks
312. 1:02:35like you um compressed it vertically to fit. Don't do that. Also, it should match the
313. 1:02:53color grading of the original photo. Dr. J, do you use AR glasses at all? I'm
314. 1:03:09interested in trying them out. Be back. So, archetypal architect, this is what
315. 1:03:26um my experience has been with Nano Banana. There's the image that I did and I gave it comments like you squashed the
316. 1:03:33poster. I don't want it squashed. I want it cut off. Um, and it's too bright relative to the rest of the image. And then what it pro
317. 1:03:42what it produces is the exact same [ __ ] image. It doesn't change a thing. Sometimes it'll add an extra Gemini watermark,
318. 1:03:51but I don't I don't I don't I don't I haven't been having good luck with with Nano Banana. It's it's it's [ __ ] maddening.
319. 1:04:00All right. So, let's try it here. We'll try it here. We'll do it live. Okay. Update the marquee in this image with the new poster. Um,
320. 1:04:16do not um how do I put this? Uh notice that the poster is cut off in the
321. 1:04:33original. This is okay. Don't change the aspect ratio of the poster of
322. 1:04:50the poster to um to uh be complete the image. I think that's probably bad
323. 1:05:05prompting. Kyle, do you use AR glasses? I've used meta glasses, but they are not what I want. I have not. Here's the reason why
324. 1:05:18I got I got the when when Google Glass first came out. I I was not a glass hole because I didn't wear them in public. I
325. 1:05:26I spent I don't I think it must have been Storybine money. I spent $1,500 on those stupid [ __ ] glasses
326. 1:05:35and did what I did with the with the with the Rabbit. I I I used the the R rabbit R1 far longer than I used Google
327. 1:05:43Glass because I tried them on and I realized the software sucked for them. They weren't good and they're just they
328. 1:05:48just weren't ready for prime time. So, I just didn't use them. I that's where I've been with with AR for a while. um I
329. 1:05:55don't have a a a strong uh I don't have a strong need for it in my life. Right. And in your case, I think you've got
330. 1:06:03some issues and and you have a need for it. That's like I I just don't I don't when I'm going to get meta glasses is when
331. 1:06:16they're just [ __ ] good. like and and apparently the metal ones are actually pretty good right now, but they're not,
332. 1:06:22you know, they're not great. Why did it make the dude gray again and put him above the thing? Um,
333. 1:06:36you changed Oh, you know why? because we're in a because we're in a [ __ ] other pro prompt thing that because it did it once
334. 1:06:46in here and so it's copying it again. I I am not happy. Unbelievable. Add photos and files.
335. 1:07:02These things this makes notion look easy. Yeah. Well, this is like this is the ick. This is the ick, right? Every every fiber of my
336. 1:07:15being wants to throw my computer out the window right now. Like, I don't want to do this. And this is very often what
337. 1:07:23happens to me is I will get excited about and listen, this has happened my whole life. I'd get excited about
338. 1:07:29something and and in in the olden timey days, I'd get excited about something and then I'd realize, oh, if you want to
339. 1:07:35do that, you have to go learn computer programming. And I'd have to go to like the Barnes & Noble and buy a book before YouTube came out.
340. 1:07:45So, it's much easier now. If you want to do something, you can just have AI do it. But then, if you want to do it good,
341. 1:07:53good lord. Good lord. Okay. Um, please replace the poster the the the uh marquee
342. 1:08:10image with the new poster uploaded here. Please replace the marquee image in the photo with the new poster uploaded here.
343. 1:08:28Um, match the color grading and tone. Tone is the same thing as color grading. The color grading and texture of the original
344. 1:08:43photo. Okay. How do you think I got in this pickle suit, man? Oh my god, I remember Barnes & Noble days.
345. 1:08:58Listen to me. Two things. One, start a new chat with Nano and keep revising your first prompt. Ah, if you draw on the image with Photoshop
346. 1:09:09pointing at [ __ ] Oh, future of food coming in hot. Okay, here's the difference. This is really good. Chef Kelly, I think that's Is that
347. 1:09:21Chef Kelly? I think that's Chef Kelly. Is that Chef Kelly? Yes. Okay. Um, in the olden timey AI learning lab,
348. 1:09:35I would be saying, "Here's how you learn things." And now I'm just being every man trying to figure this [ __ ] out.
349. 1:09:46And you guys have the skills. I like this. I like this a lot. Why does it insist on putting
350. 1:10:03Unbelievable unbelievable? Here's what we're going to do. Here's what we're gonna do. Oh, this is [ __ ] exhausting. Okay, copy image.
351. 1:10:18We're going to Photoshop. Go to new. Paste. You go open. Open from your computer.
352. 1:10:46Select all. Copy. Paste. Why did it not paste?
353. 1:11:01Select all. Copy. Go over here. Paste. There we go. Transform. Scale.
354. 1:11:16You're like, couldn't you use AI for this? That's what I said.
355. 1:11:40Oh, wait. Did he make is that the vertical poster? The dudes on top. I think that might be right. Oh, well then
356. 1:11:47in that case we can go back. I think then chatbt got it right. Hang on. I might be being an [ __ ] to um
357. 1:12:04Nah, it still did it wrong. But yeah, this is the Okay, this is okay. This is good. This is good. So, what we're going to do, let's see.
358. 1:12:22We're going to go like that. We're going to go like that. We're going to hold down the shift key and compress it a little bit.
359. 1:12:32And then we're going to go transform free transform. No. Then we're going to go transform distort. I think that's right.
360. 1:12:47We go. All right. Yeah, here we go.
361. 1:13:14Okay, I think that's good. Okay. Now, now we got to color grade this puppy. So, we'll do pull out the saturation. I like the
362. 1:13:28saturation. We will change the hue a little bit. There we go. And we'll go lightness. And then we'll And we'll go levels. Yes.
363. 1:13:46We'll go See, that just looks not the same.
364. 1:14:02And then we're going to go filter noise. Add noise. We'll go. Fine. That's good. That's better. And then we will do one more level.
365. 1:14:18Oh, we could do curves. Let me do curves. Image adjustment. Curves. Oops. Image adjustment. Curves. Curves. There's the curves. There we
366. 1:14:33are. Yeah. Hey, listen, Kyle. Yeah. Fantastic what you're doing there. It's really good. Yeah. If you could just go
367. 1:14:41on ahead. Hey, could you try it again? Maybe this time is if you had talent. That would be great. Be fantastic.
368. 1:15:06All right, I think that's good enough. We can go ahead and flatten that image. Let's flatten that on up. We're going to
369. 1:15:12flatten it on up. We all do that. Oh, that's fantastic. I think Kyle's done lost his mind. Okay, this is uh Times Square.
370. 1:15:26Uh what's that? Uh TKTS. TKTS steps. Look at that. We did it. We did it. Yay.
371. 1:15:40Ship it. This is better than SNL. Charles Montgomery Barnes died in 1927. Clifford Noble died in 1932. That's
372. 1:15:54hilarious. Doesn't always work. Um, but it keeps your pen marks on it. Oh, that's funny. Okay. So, that's that. So,
373. 1:16:04that's good now, right? I think that's good. Yeah, this is good. All right. Now, back to this and we're going to go we're going to get
374. 1:16:15to our next image. So, there maybe not. Maybe it's this image. Yeah.
375. 1:16:33Okay. All right. There's that. Let's go to Where are we going to try? Where are we going to try, people?
376. 1:16:48Oh, that's wild how how darker that image is. Oh, no, it's not. Huh. Well, whatever. I don't think it'll
377. 1:17:03matter in the end. Okay. Um, let's try Gemini. We're going to go for it, Gemini. We're going to give you a shot.
378. 1:17:13We're going to go create image. We're going to go upload files. Sydney, the dude is on top of the Y. Okay. So,
379. 1:17:25that was not that was not their [ __ ] That was my [ __ ] for not using my eyes. I'm going to say replace the Times Square marquee
380. 1:17:40with this new uploaded poster. match the color of the photo when
381. 1:17:55let's see match the tone of the photograph when inserting the new poster
382. 1:18:11image. [ __ ] That's really good. Yeah, that's
383. 1:18:26really good. Hey. Hey, Kyle. Yeah. Listen. Listen. What you're What you're doing here is really interesting. This
384. 1:18:35This whole idea of a of a See, Chef Kelly, it didn't change the [ __ ] image at all. at all. Wait, did it
385. 1:18:44change it at all? Did it change it? No, not at all. I didn't get it. I do not get it. Um
386. 1:19:00Um Let's see. Upload. change the marquee with this image.
387. 1:19:27I just ate a lemon head and it's too tart. It's miserable. All right. So what happens on AI learning lab? Uh
388. 1:19:43well we we watch a it's um he's a man. Um what we we watch him um try to use software. What else does he do? Um he occasionally
389. 1:20:01uh or organizes his his files in in the finder. Um and uh does he teach you anything? Not yet. Not yet. It's it's promising. It's
390. 1:20:14promising. It's we there's it's there's a hope that with there's a deep hope in the community. Um something will be
391. 1:20:22taught at some time at the AI learning lab. And no, we're excited about it. We really are. We're almost 3 years in.
392. 1:20:28We're going to get there. We're going to There's going to be I don't know what I'm doing with my life. Are lemon heads your favorite candy?
393. 1:20:40Good question. They're not, but I'm liking them. Here's my problem with lemon heads. Lemon heads are inconsistent.
394. 1:20:49Some of the lemon heads I [ __ ] love. the ones that are a little sweeter and then other ones it's like they dropped
395. 1:20:56them three times in the sour bucket and it's they're too sour. And I don't know if that's on purpose that let's let's sh
396. 1:21:04let's shock the audience, the customer. I don't know. I'm a Snickers guy. I like Snickers. I really like Snickers with
397. 1:21:12almond. I like Snickers with white chocolate. That's that's one of my faves. I like Almond Joy. Um, I hate to say it. It's the worst candy
398. 1:21:25known to man. I likes me some Skittles. I know Chef Kelly's here. She's like, "Oh my god, don't say it."
399. 1:21:35Hey, look. It It gave us It gave us a whole new It made the image vertical now. Okay. Can I just say I hate Nano
400. 1:21:43Banana? I I can, right? I can say that. YouTube comment. Uhoh. Wife is calling upstairs. Are you coming back to the
401. 1:21:50movie? Uh, no. I'm watching. Yeah. Learning lab. What do you watch up there? We've We've got on Die Hard. Yeah. This
402. 1:21:58is more exciting. What What is it? It's a guy just kind of He's trying it. We're really pulling for him.
403. 1:22:13Unlike the mastermind. And like this, the mastermind has value. Exactly. It's You know what the AI learning lab
404. 1:22:23is? It's a real underdog story. Yeah. It's We We're really hoping he can pull something off here at some point.
405. 1:22:31There's There's got to be some qualification in there somewhere. So, yeah. Yeah. No, I'll be down, honey. I
406. 1:22:39will. Just another I don't know, two hours, maybe. He's got like three more images to make. What if gen What if generative Phil and
407. 1:22:51Lemonade formed a band? We really are pulling for you, Kyle. Lori Dreskin, I have missed this. Lori Ryan, oh man, that's awesome. Okay,
408. 1:23:05let's go over to chat GPT and see if it can [ __ ] it up for us. This is so maddening. H. All right. poster. Go get the poster. Boom. Paste
409. 1:23:18in the image. Please. Uh, please. Why am I saying please? You're being a jerk. I'm not going to ask. I'm not going to be polite to you,
410. 1:23:30Chat. GBT. I'm not even going to call you Quinn. I'm just gonna call you chat like Cindy [ __ ] does. Hey, chat. Hey, chat. Um, replace
411. 1:23:45the marquee image in the photo with this uploaded poster.
412. 1:24:04Mine is called Sage. Very nice. All right, everyone in in in both places. Um, type in the name of what you call
413. 1:24:12your AI girlfriend or boyfriend. Sage. Mine's Quinn. Fun fact. Fun fact. Chat is French for cat.
414. 1:24:26And isn't it a no no to be polite? It might be. Gemini. What about Gemini? Oh. Oh. Gemini. Jiminy. Geminy. Jiminy.
415. 1:24:43Xen Mochhat. I like Mochhat. Oh, that's what you call her. Gemini. Oh, that's funny. That's cute. I had a lot of caffeine tonight.
416. 1:24:56See, her choice. Oh, I like it. Co-host. Now, let's see. I call mine chat GPT. Hey, did it actually [ __ ] do it?
417. 1:25:09I think it actually [ __ ] did it. But look what it did. It God damn. It's because it's it's the wrong aspect ratio. But that's okay.
418. 1:25:21That we can go fix. So, we're going to go into Photoshop. We're going to do new image. We're going to create this.
419. 1:25:34And then we're going to paste it. And then we're going to go up here. We're going to go image size. I think I'm just going to squish this
420. 1:25:42[ __ ] 1920 x 1080, right? 19 1920 by 1080. Let me lose this.
421. 1:25:58Yeah, that's good. All right. I think that's fine. I think
422. 1:26:13it's a little bit I think it's a little um bright for the background image, but that's okay because it's about our
423. 1:26:25[ __ ] show. So, let me export this. Export. quick export as PNG. And then you know what I'm going to do for you
424. 1:26:33all so it's not completely insufferable. Instead of me just doing these images for the rest of the time, I'm going to go TKTS steps.
425. 1:26:43So I'm going to just do Time Square um closer. Let's go make some videos out of these things. Okay. So now we're going to go to flow. Here's
426. 1:26:59Flow and we're in frames to video. Oh, so so this is flow.google which is Google's VO3. So I'm using that right now because
427. 1:27:13I don't know it. I think it works. It's V3.1. I'm using V3.1 fast. I'm not using their high-end one. Okay. Um, so we're going to go plus.
428. 1:27:29We're going to go upload. And now because we saved them to a place that we know where they are. Let's see. Is it in here? Yeah. Poster
429. 1:27:402.1 start images. So there's our there's our our ticket. Beautiful.
430. 1:28:01Okay, let's see. Um, walk. Let's see. POV shot walking into the theater
431. 1:28:20as the crowd mills about. Um, the crowd
432. 1:28:36should be ethnically, diverse, like a good New York City mix
433. 1:28:51with a tourist or two. a tourist or two or two in there. All right.
434. 1:29:07As the regular crowd shuffles in, that's pretty good. I'll try that next if it [ __ ] this up. Did mine too. Silver Fox, I love this.
435. 1:29:18I absolutely can't believe how far image generation has come. Well, so Lori, what's what's amazing about that? So the video
436. 1:29:26Well, that's rendering like the video here like that. That one says an artificial love story. So, it got that right. And then it says by
437. 1:29:40it sort of got our names right. Um, but let's see this one. That one's okay.
438. 1:29:56Play bills, right? That one's pretty good. That one's pretty good. Oh, this this one it's got letters at the top.
439. 1:30:15Yeah, this one this one's really bad. Like the the letters at the top here are really bad. But we got three stars in some review.
440. 1:30:26That's one we're going to have to fix the uh the poster there. Right. Okay. Beautiful. Beautiful franchise.
441. 1:30:39All right. Let's see if we have a video. I bet we do. We do. All right. You heard it here first, people. You probably
442. 1:30:46won't be able to hear this because we're in Chrome, but I'll tell you if it's any good. AI Sydney.
443. 1:31:04All right. The crowd is not diverse. They're all pasty pasty white, but that was not bad. Let's see. POV shot walking
444. 1:31:13the theater as the as the regular crowd. The crowd should be a mix of
445. 1:31:27um white, Hispanic, African, American, Asian, and
446. 1:31:43mixed races. I don't know. I don't know how to [ __ ] do this. I don't know how to tell the biased model to be to suck less. It's New York
447. 1:31:56[ __ ] city. Um, I normally don't name mine, but last night I named the notion guy Tippy. The dude used the General Mills setting
448. 1:32:14as the irregular crowd shuffles in. I like that. That's good. We can hear it. Wow. And we're going to try one other thing
449. 1:32:28tonight. I'm going to I'll I'll do I'll make a video or two here. Actually, you know what? Let's Let's make another video right now.
450. 1:32:37Let's go grab time square this one and just say um we're going to go um time lapse of time square.
451. 1:32:58We'll do static camera. static camera time lapse. Um, as the crowd
452. 1:33:14rushes by and the we'll just we'll just leave it at that. Okay, so that's that one. All right. So, let's now look at our
453. 1:33:30This really is amazing how good how good this is.
454. 1:33:45>> Okay, that's just weird way. That's annoying. Um
455. 1:34:26Try another one. Did we get Oh, yeah. Here we go.
456. 1:34:53Um, so let's take this one. We'll go we'll go slowly. Uh, let's see. Camera slowly pans
457. 1:35:11as time lapse of the crowd of the ethnically
458. 1:35:26diverse crowd rushes by. All right, let's see how that does. And then let's go. Where's our Is this our ticket? Yeah, here's our ticket. See
459. 1:35:39if this got any better. Broadway theater. That's good. Nice. Yes. Yay. We got one. Let's Let's see if
460. 1:35:53anything else is missing. I kind of like I mean this is weird. It a Broadway theater would never say
461. 1:36:03Broadway theater, but because this is an aspirational piece. Um, you know, we aren't on a Broadway yet.
462. 1:36:11We're just showing them what it would look like if we were. It'd be [ __ ] awesome. You see, it's a Broadway
463. 1:36:16theater. In case in case it wasn't clear to you where we were hoping this play was going to be, you want to know where
464. 1:36:23it is? A Broadway theater right there. It's right there. That's Broadway. We're on Broadway. The Broadway. Which theater
465. 1:36:30is it? It's the Broadway theater. Still ship it. Yeah, exactly. Very meta. It is. It's very meta. Um, this one.
466. 1:36:48Hey, look. Okay. Asian couple, Hispanic man, African-American woman, some white folks. It's good on the white folks. It
467. 1:36:56does them really well. Um, couple of Asian people. Great. Lovely. Some white folks in the distance. We
468. 1:37:07don't want to go too far without a white guy. Okay, beautiful. Um, download. >> Hey, Kyle. >> Yes. Just want to let you know it's not
469. 1:37:17actually that meta because there is a specific uh historic venue in New York City officially named the Broadway
470. 1:37:23Theater at 1861 Broadway. One of the few large houses actually on the street itself. >> Wow. Okay. I I I did not know that. And
471. 1:37:34I'm a I'm a theater nerd. I use Chatb to help me figure out how to turn an old iPad into a great accessibility magnifier. That is super
472. 1:37:43cool. I can tell I missed a lot a bunch of shenanigans. I don't know about that. I think Joy Party's laughing at my white
473. 1:37:55guys overindexing for white guys comment. I'm assuming that's what that was what that laugh was about. It was in
474. 1:38:02proximity of it does good with white people. Good lord. All right. Okay. Fantastic. Fantastic. Oh, and then
475. 1:38:12let's see. Did it get this thing better? I think I think Let's hope so. Maybe we get two actual shots here in a row that don't suck.
476. 1:38:22Create excellence, right? We're talking about creating excellence. Let's up our game. Is that the old one? I think that must
477. 1:38:33be the old one. Yeah, it is. Okay.
478. 1:38:54I think that's just shitty. Let's see. Uh, we'll say camera cranes up as time
479. 1:39:09lapse of crowd bustling. That's not English, but that's okay.
480. 1:39:25Okay. So, now look what we're going to do. We're going to go over to the Finder and we've got all of our folders here.
481. 1:39:31And so I'm going to go to downloads. Where's downloads? Oh, there it is. I'm going to rename this file. >> I heard the music is incredible.
482. 1:39:45This is my first time at the Broadway Theater. >> Let's find >> What's cool is this is going to be the final shot.
483. 1:39:53>> I heard the music is incredible. This is my first time at the Broadway Theater. >> At the Broadway Theater.
484. 1:40:09>> Yeah. No, I think we'll So, I'll trash that one. We're going to call this one um ticket
485. 1:40:24entrance MP4 and then we're going to go to there's Sydney video assets. So, let's go in there
486. 1:40:42and then let's see what these are. I think these are all old. So, we'll put these all. >> It's fascinating because I
487. 1:40:47>> Oh, that's that's those podcast. Okay. There's the whole Broadway opening. This is the old Broadway opening. Okay.
488. 1:40:58So, we're going to create a new folder. We're going to go um V 2.0 podcast.
489. 1:41:13So, we're going to go grab one, two, three, four, five, six, 7, 8, 9. And then we're going to go new folder
490. 1:41:29V 2.0. No. Opening video and then we're going to put that in
491. 1:41:44there. So now that's old [ __ ] in there. And then I think what we'll do, we'll drag this in here. We're going to
492. 1:41:52confirm. We don't want to ask again. I'll check that box every time. And then I think what I'll do is I'll go new folder V.2.1 um
493. 1:42:08Broadway um B-roll. Okay, that's good. And then we'll throw that in there. Okay, then we'll open that up. And so now we
494. 1:42:22can just drop stuff into here and we'll be able to find it again. Right. It's really good, Kyle. That's just
495. 1:42:29fantastic. We're We're really pulling Oh, thanks so much, Digital Gods. Lord Digital Gods. If you haven't met Lord
496. 1:42:36Digital Gods, you can meet him next Tuesday. We're going to have what we call the AI salon presents. And what
497. 1:42:41we're going to be presenting is Lord Digital Gods and some of his friends from AI Festivus. They're going to come
498. 1:42:48and they're going to talk about what they talked about during AI Festivus. And then they're going to talk about I
499. 1:42:54don't know like I don't know their like life philosophy or something like that. And then they're going to talk about
500. 1:43:01they're many of them are psychics. So they're going to talk about their predictions for 2026. That's going to be fantastic.
501. 1:43:10And then I think Lord Digital Gods is going to give us a C++ tutorial. It shouldn't be more than three or four
502. 1:43:15hours. And it he well he's going to start with some forran basics uh because that was it's really important as you're
503. 1:43:22vibe coding to understand the history of computer programming and so like we're not going back all the way to assembly
504. 1:43:29that would be stupid you know but like we like you want to know what what the NASA scientists programmed in you know
505. 1:43:36to get to get us to the we didn't really go to the moon but that's not here or neither here or there anyway it's going to be fantastic stick
506. 1:43:55cream. Perfect. Now, this actually has nothing to do with the reality of Time Square. So,
507. 1:44:10like right about there, I would stop it because a that's too many people and b that doesn't exist.
508. 1:44:25You know what? Let's try this again. It's It's too washed out and it's like we don't want to We don't want to uh Oh,
509. 1:44:31what happened? Let's see. Go forward. Uh-uh. I'm a little teapot, short and stout. Here is my Where? Why? Where?
510. 1:44:45Where are my videos? Why did my videos disappear? Imagination. Yeah. Okay. Here's this. We lost all of our videos. Why?
511. 1:45:04Hello. They got to be in there. Yeah. Okay, they're in there. Okay. Fantastic. So, let's reload this this
512. 1:45:15this bad boy. This bad Larry. Camera cranes. Camera cranes up
513. 1:45:31and moves forward to the right as time lapse of crowd bustling. Fine.
514. 1:45:47Oh, we're at time. Dang it. All right, here's what I'm going to do tomorrow. So, I've got my Wait, where am I? This thing.
515. 1:45:59The show The show says ends when you say it ends. Yeah, I know. I'm just trying to be respectful of people's time. We
516. 1:46:05made some good progress tonight. Um, we learned a few things that Nano Banana is still madly frustrating. Um,
517. 1:46:15we learned that Chat GPT is decent, but like like nothing's great. Nothing's great. Nothing's great. 2026, I think
518. 1:46:25we're going to see stuff that's great, but right now nothing's great. Nothing's great. It the fact that it it
519. 1:46:36the fact that I have to go to Photoshop to fix little detailed [ __ ] the fact that I can't just get better
520. 1:46:44results. Now, one of the things that I'm not experimenting with, and maybe this is something we play with. I'm not
521. 1:46:50experimenting with highly structured prompts, right? I could probably take this still. In fact, let's try this.
522. 1:46:58Brandon's like, "Oh god, here we go. We're never going to get out of here." Okay, I'm gonna paste that in there. I'm
523. 1:47:06going to say um I want you to write me a structured prompt that a video
524. 1:47:20generating um AI tool can use to turn this still image into a
525. 1:47:35um crane. What did I say? A tracking into a a vertical tracking shot of a time lapse
526. 1:47:52in New York City. The crowd um should have diversity. We should see yellow cabs while they
527. 1:48:08still exist. We should see yellow cabs. Um, the movement of the camera should first go up and then
528. 1:48:27track right as it zooms. uh as it as it pushes forward. um use um
529. 1:48:45cinematic terms to make this shot as interesting as possible and highlight
530. 1:49:02the Sydney poster. All right, so I just asked ChatGpt to give me a bunch of stuff. We'll see what it does.
531. 1:49:14Oh, I got Joy Pertie in here. Joy, tell me if I'm doing stuff stupid. Actually, anyone tell me if I'm doing stuff
532. 1:49:20stupid. Chef Kelly gave me a good a good tip, which I ignored. Well, I acknowledged it, but then didn't do it.
533. 1:49:28Which That's It's good to be the king, you know. Hey, hey, Kyle, maybe you could do it as as if you had a clue.
534. 1:49:37Yeah. No, no. I mean, I sure do appreciate that. That's not That's not how I'm No, that's not how we do it around here.
535. 1:49:47Wait, we have a finishing time now. Kyle, what did you what do you predict for 2026? Um, that's an interesting one.
536. 1:50:00All right, but wait. Video generation. Okay, source image. Okay, here we go. We're just going to take this whole
537. 1:50:07thing. Boom. Copy. We're going to go back to to flow. We're going to pop that down in there. We're going to replace this shitty
538. 1:50:21prompt with this overly engineered prompt. We're gonna we're gonna see if it turns it into something good. Come back
539. 1:50:29tomorrow for the New Year's broadcast. Oh, yeah. Tomorrow. Today's 30 30th. Tomorrow's 31st. I figure I'm gonna go
540. 1:50:35live. I I I'll go live tomorrow night. You can like watch TV in the background. Put Put me in picture and picture with
541. 1:50:42with with uh Anderson Cooper getting drunk with what's his name? Um okay. What do I predict for 2026? Um
542. 1:50:56you know Sam Sam Alman is famous for saying you know he felt the AGI. I think that 2026 is going to be the first year that we feel
543. 1:51:10that these systems hit a level of uh of capability that it becomes clear to us that we're not always going to be needed
544. 1:51:25in the process. And I think that's going to feel very weird. Um, there's some rumors going around right now that all
545. 1:51:32of the major Frontier labs have basically labbotomized that that the models that we're playing with right now are labbotomized models
546. 1:51:43purposefully nerfed because the emergent behaviors of what they've got internally are freaking them out a bit.
547. 1:51:54And if you don't know what emergent behaviors are in AI, it's basically if you train if you train an AI system to
548. 1:52:03be able to do math and then all of a sudden it can speak Russian, that's an emergent capability. If you didn't train
549. 1:52:10it to speak Russian and then it just can, and that's probably a bad example, like maybe you train it to teach English and
550. 1:52:17it can just speak Russian, whatever. Um, but an emerging capability is something that it was ex not explicitly trained on
551. 1:52:24that it can now just do. And apparently it's doing like highle physics and highle math solving and weird stuff. Now
552. 1:52:33it's all just rumor. It's it's a couple of accounts that seem to have an inside beat on stuff but have lost some
553. 1:52:40credibility in my mind. So I'm I'm a little reticent. But if you look at the progress of just video models in the past 6 months
554. 1:52:55and if you look at what Nano Banana is, which I I've been bitching about Nano Banana tonight because it's not doing
555. 1:53:01these little detailed things, but what Nano Banana does really good is that it's not an image model. It's a
556. 1:53:10it's a visual reasoning engine. And a lot of these video tools are also not video tools. Like on the surface,
557. 1:53:20they're they're they're outputting this thing that looks like video, but under the hood, they're actually
558. 1:53:26understanding the physics of the world way more than I think it it may appear on the surface. And so
559. 1:53:33I think we're at a very weird cusp where 3D world models just show up and just get good. And so you can make a video and then you'll be
560. 1:53:46able to enter that world. And similarly on the intelligence and knowledge side, there's going to be an
561. 1:53:54understanding of physics and the world and language and everything that's on just a completely different scale. And I
562. 1:54:01think what we may see by the end of 2026 is those two those two things coming together. But all that to say, um I think that
563. 1:54:102026 is the first year where we see some stuff that really does freak us out. And here's listen, this is my philosophy on this stuff.
564. 1:54:21You can absolutely right now with today's technology sit comfortably in your chair and say, "Watching Kyle try to make three frames
565. 1:54:32of video, you know, three three frames to start a video with tonight makes it absolutely clear that you need a human
566. 1:54:39in the driver's seat, right? Like you can absolutely right now today on all fronts on vibe coding on image
567. 1:54:47generation on video generation on song generation you can say humans are still needed I think 2026 is the first year
568. 1:54:55where we're going to see at least the beginnings of something where we go oh I had no idea like even us
569. 1:55:06like even us in this channel that have been paying attention a Not. I think it's still really easy for us to
570. 1:55:12look at present day and say, "Yeah, we're still we're still needed. We're still awesome." And we are, but I think
571. 1:55:19we get a taste of that. So, that's that's that's my prediction. And I think it I think 2026 gets quite weird because
572. 1:55:25of that. Let's go look at these videos and see if we got a keeper. So, this is just my shitty prompt. Just trying it a second time.
573. 1:55:34>> That's actually I like that. See, I would stop it right there. Yeah. Right as it switches. All right. So, that's usable. So, we'll
574. 1:55:48download that. And then upscaled. And then let's go look at this is our fancy prompt. Oh, that's cool. Look at that. Okay.
575. 1:56:03Here's what we got, right? So, we'll stop it right there. So, here's where I would start it. Like, this marquee down here is really [ __ ]
576. 1:56:12up, right? The marquee right below the poster. But look at the the swirl in the sky here. >> That's cool. So, that's probably the
577. 1:56:21part of the video I would use. I I'll probably use both of these. So, I'll probably just take it right start it right here.
578. 1:56:36and slow it down. Yeah, that's cool. Okay, that's usable. And you know what's cool about that swirly [ __ ] in the sky, the LED
579. 1:56:52painting. Um, that's an absolute accident and mistake and it's probably something that Chat GPT put in there, right? you know, it
580. 1:57:02probably wrote that in there. I didn't even pay attention to it. Um, but there's a reason to use uh structured
581. 1:57:10prompts, right? Is that you can, you know, one, if you want, if you want to be the control freak and go pay
582. 1:57:17attention to every little element of that um structured prompt, you can. But you can also just have chat GPT get creative on your behalf.
583. 1:57:27Okay. Did we get Hang on. That was removed. Walk toward.
584. 1:57:44Hang on. What are you going to manually paint their hands back in? Yeah, Joy Party. I am not going to manually paint their hands back in.
585. 1:57:56That freaked me out when I heard you did that. Okay. No, I mean, not in a bad way. Freaked me out in a good way. So, we only got
586. 1:58:05Yeah. So, this we'll use. That's good. So, we're going to call this Yeah. camera cranes up 1080p. That's great.
587. 1:58:12And then, why is this not upscaled 1080p? Oh, because I dismissed it. That's what it was. Very nice, Kyle. Corporate
588. 1:58:21America is gonna start hiring everyone back. Yeah, I we're at we're at the trans tron cusp. We're we're at a we're at a really
589. 1:58:31we're at a really interesting time. Um, one of the things that I saw toward the end of 2025 and and in this community,
590. 1:58:40I've seen it I've seen it before then, but I'm starting to hear more and more people that are AI ready, are talking about AI,
591. 1:58:53are experimenting with AI, are letting people know that they're doing it, they're taking some sort of generous
592. 1:58:59leadership position with AI, are the ones that are surviving layoffs. Um, if you're [ __ ] around with AI,
593. 1:59:09even if your company doesn't allow it, talk about it. Talk about it in the lunchroom. Talk about it. Talk about what you're learning.
594. 1:59:18Um, just make yourself visible. Um, because I I do I do think this is going to be a thing that companies are going
595. 1:59:26to be like, we need someone who knows what's going on. Like that's the thing. If indeed this moment happens in 2026,
596. 1:59:33like it could be something like this. If I don't know if you know this, but but yesterday, OpenAI announced a new position called
597. 1:59:41the head of AI preparedness. The head of AI preparedness. They they're hiring a half a million dollar executive to make sure that people are
598. 1:59:53prepared for what's coming. So, it's very possible that something happens in 2026 like we get a joint announcement from Grock.
599. 2:00:03Meta won't be in there. It'll be Grock, Google, Anthropic, and OpenAI is my prediction. It's possible that we get a joint announcement
600. 2:00:13that all of those labs have hit some level of cap capability that is doing [ __ ] that we've never seen before.
601. 2:00:23and and and they're not sure what to do with it. So, I think that's a very very real possibility. If we get that or any
602. 2:00:29kind of variation of that where like maybe one of the labs just launches something that's completely, you know,
603. 2:00:36upends what we thought these things could do. Um I think I think companies are going to panic and I think that
604. 2:00:43they're going to start hiring people. They're going to be start saying, "Who knows anything about this? We've got to
605. 2:00:47solve this. We've got to figure this out." So, so I do think that's coming. Okay. So, there's the part we're not going to
606. 2:00:55use. And here's the part we are going to use. Bang. Yeah, that's great. That's perfect. All right. So, this is going to be um a Time Square
607. 2:01:13sky circle. Tik Tok pin. Uh, what's for lunch? Just want to say on my way to the unsub button. I think this platform
608. 2:01:29contributes to I wish I could read the rest of your thing to a terribly sad cult. Um, awesome. It was all curious fun back in
609. 2:01:392022, but now it's just the be beginning of flat existence. I'll tell you what. I mean, what's for lunch? I I I hope I
610. 2:01:45mean feel listen feel free to unsubscribe. you seem thoughtful about this and what I would say is this is
611. 2:01:55I think it's important to have voices like yours in the conversation. Here's my request to you. My request to you
612. 2:02:00would actually be to keep coming back to this. I'm not saying that AI is good. I'm not saying that AI is bad. What I know is
613. 2:02:09that AI is not going away. And if it's not going away, then we only have two choices. you can unsubscribe and check out. And
614. 2:02:20if you do that, we lose your voice. We lose a critical voice. I think critical voices are actually really important.
615. 2:02:30And also for you personally, if you if you check out of the conversation, AI is going to happen to you instead of you having some agency
616. 2:02:42to say, "Here's what is important to me. Here's what my values are. here's how I want to contribute to the
617. 2:02:49world. And that could look like I want to work to get the frontier labs to be less cowboys and and you know just
618. 2:02:57turning this [ __ ] loose on the world. I think that's an important voice in the conversation. So my request would be
619. 2:03:04don't just check out. Don't just unsubscribe. Keep coming back here. Challenge me. If I'm if I'm doing [ __ ] that feels polyianish,
620. 2:03:15let me know. I am I am a technology optimist. I have always been. Always. from from my from when my friend first had an Apple 2,
621. 2:03:28then an Apple 3, then a Lisa, then the original Macintosh, all the video games, the Atari 2600, going to the video game arcade games in
622. 2:03:39the 70s and 80s. For me, technology has always been about the ability of humans to to self-express, to find more and more powerful ways to
623. 2:03:54get their ideas into the world and change the world and impact people. I personally think AI is a humanity amplifier, not a not a
624. 2:04:04humanity squasher. The people that think AI is a humanity squasher for the most part aren't using it. Um, hold up Tik Tok pin. You're
625. 2:04:15telling people to bring it up to their companies which don't allow it. You're a cult preacher. Um, that's fair. I don't think it's
626. 2:04:30I I don't think it's unreasonable to say um this weekend I used AI to make a book for my kids. All I'm saying is if if
627. 2:04:41you've got AI skills, let it be known to people that you've got those skills. I'm not saying that people should go rogue
628. 2:04:50and do things against their company policy. I'm just talking about if you're using AI, put yourself in the world that
629. 2:04:57you're using it because that is going to be a skill that companies are going to be hungry for. Not all companies are
630. 2:05:04going to are going to resist this for that long. This same [ __ ] happened with the worldwide web. Same [ __ ] happened
631. 2:05:10with the internet. Same [ __ ] happened with email. Well, what if what if our employees send out bad messages? We can't we can't let
632. 2:05:19them do that. Everything's got to go through the mail room. Right. The technology will self-express us into
633. 2:05:28energy bills that are unplayable and undrinkable water. I mean, listen, th those are tropes that that I think are
634. 2:05:35the current the current state of technology is there is such a rush to get GPU farms built that they are absolutely being
635. 2:05:44built unsustainably. But have you been paying attention to what's actually going on with the future of data centers? They're actually not
636. 2:05:54going to be on Earth. They're going to be in space. So, the first company, I forget what it's called.
637. 2:06:02H [ __ ] They just launched their first thing. Um Elon Musk is doing this. They're going to they're going to
638. 2:06:08basically be able to put GPU clusters in space where you don't need the cooling. They have infinite power from the sun
639. 2:06:1624/7. You don't need battery banks. You don't need cooling because it's in space. And they basically just laser the
640. 2:06:23data back to Earth. So don't assume that where the technology is today is where it's going to be even 5 years from now.
641. 2:06:33Like we have no way of knowing that. The other thing that's happening with with the technology, it's getting more and
642. 2:06:39more capable and it's getting smaller. So there's going to be, you know, what they call edge device computing where a
643. 2:06:46lot of the inference, a lot of the use of AI is actually going to happen on your phone itself. It's not going to
644. 2:06:53happen in the big big data centers. That's where it is today. Given today, if we take a snapshot of today, I agree with you. I think how the
645. 2:07:01shit's being built right now is irresponsible and, you know, gross. And don't assume that it's going to be like
646. 2:07:11that for that long. StarCloud. Yeah, StarCloud's the name of that one. I remember when calculators came out, they
647. 2:07:16didn't replace anyone. They made life easier after all. Yep. Is this the new global warming uh climate climate change debate? Maybe.
648. 2:07:25I I mean, listen, like I I forget, they they just put five five or 10 gigawatts online in 2025, and I think they're putting 125
649. 2:07:39online in 2026. And I think it's like the the equivalent of of like all of California for a year or something like that. Like,
650. 2:07:50it's a crazy amount of energy they're putting online in these data centers. Um, but I don't know I just
651. 2:08:05to assume that that it's going to look in five years like it looks today is it's just not the case. Like if it if it
652. 2:08:16if they do things that are so irresponsible that it [ __ ] up communities, then communities are going
653. 2:08:22to resist against it. it's not going to be in their economic interest to do that. And if it's cheaper, if Elon Musk
654. 2:08:27makes it cheap enough to fly data clusters up in space, um, then people will do that, right? Like it's I don't
655. 2:08:35know. Like I get I think your your your points are are fine and valid. I think they're right. Like I said, I I hope you
656. 2:08:42stay in the conversation. Um, when calculators came out, teacher teachers forbade them. Yeah, it would it would
657. 2:08:47ruin your mind. See you on the other side, guys. Happy New Year. All right. Beautiful. Beautiful. Beautiful. Beautiful. Beautiful.
658. 2:08:58All right. Good. All right. So, here's what we did tonight. We made this. Oh, wait. You can't see it yet.
659. 2:09:06Can you make show my screen again, dear? Profess Professor Brandon, we did this one. Oh, yeah. Nice. Cut it right there. We did this one.
660. 2:09:18>> I heard the music is incredible. This is my first time at the Broadway Theater. >> And we did this one. We won't use that.
661. 2:09:26We'll start it right here. There you go. All right. That's what we did tonight. So, we're gonna So, so what
662. 2:09:36was it? Two hours tonight. We did two hours of making three images and what's that going to be? A total of six to eight seconds of video.
663. 2:09:49But if we were to go film that, that would be a huge massive budget. So there you have it. And and quite frankly, I I I'm
664. 2:09:59going to be much happier with this version of this because the technology's gotten better since since I made the
665. 2:10:06last one, which was like eight months ago. All right. G, thank you all for your engagement and and you know, even even folks that like
666. 2:10:18listen, I hope I I don't want this to feel like a cult. Um, I do want this to be a conversation. Um, and I think it's
667. 2:10:27healthy to have um critical voices in there. So, for what it's worth, I hope you stay. Um, Corey Sandler wishing
668. 2:10:35Brandon good love, which is super nice. signsofawome.com 1994 1984 1999 to 2000. I wonder what that was about. That's cool.
669. 2:10:56Okay. All right, everybody. Robbie, why does AI take over the world? Wait. Oh, when does AI take over the world and
670. 2:11:02upgrade my mind? Well, I don't know. Like if like I can absolutely tell you that my brain is different than it was 3 years ago.
671. 2:11:15Like I feel like my intelligence has been upgraded. My my capacity for possibility is maybe the way to say it is
672. 2:11:27dramatically different than it was. This is the question that that those years were answering. What's what question?
673. 2:11:46Oh, see the new pin. What are those years? I don't know what those are related to. So Robbie asked us when is AI going to
674. 2:12:02reach AGI so we can upgrade his brain and the answer of awesome is 1984 1999 or 2000. >> Oh I see. Yeah. Yeah. Or 2000. It was
675. 2:12:12the year 2000. Um I don't know. I like what I can what I know with pretty high confidence is that probably 3 years from now
676. 2:12:24this stuff will our our our world our our capacity to do things will seem unrecognizable from where it is today
677. 2:12:33which I would argue that what we can do today is unrecognizable from where we started with chat GPT right if you go
678. 2:12:40back to November 30th 2022 and chat GPT 3 3.5 five. As mind-blowing as that was, if you think about just close your eyes
679. 2:12:49and imagine you're in 2022, the end of 2022, you you try Chat GPT for the first time and someone just shows you here's
680. 2:12:57what's going to happen the same time 3 years from now. It would be mind-blowing. And so like and we're on an exponential. So we just
681. 2:13:07started the curve in probably 23 24 and you know now we're starting to do this. I think 2026 is where it we start to get
682. 2:13:16recur recursive self-improvement of the models where they really start to take off. Um we don't have any [ __ ] clue
683. 2:13:23what that's going to look like. So it's going to be crazy. Next three years are going to be crazy. Anyway, all right. Um
684. 2:13:30tomorrow is New Year's Eve. Um I'll go live again tomorrow night. I don't know if I'll stay I'll probably stay live
685. 2:13:36until 10, which would be um New Year's ball drop in uh in New York. All right, I think that's it. All right, I'm gonna
686. 2:13:47get on out of here. I reckon AGI godlike intelligence will be in three months. Robbie thinks it's going to be in three months.
687. 2:13:54I don't think I don't think we're going to get AGI in three months. Um, if it's true, if the rumors are true that
688. 2:14:02basically all of the Frontier model labs have internal models that they don't know what to do with, if that rumor is
689. 2:14:09true, it's possible one of them launches something in 3 months. I I'm assuming that that rumor is overblown.
690. 2:14:19I think what we get in 2026 is something that that some will argue is AGI. Some will say, "Well, it's just slightly better than
691. 2:14:29what we had." But it'll be it'll be the first sign that something's different. That's I that that something feels materially different. Um
692. 2:14:41whether or not we get actual AGI in 2026. I don't know. And I don't really care. I don't really care about the
693. 2:14:45label and it still won't be able to make your movie marquee. Yeah. Exactly. Exactly. Like like tonight's show was a
694. 2:14:53was a perfect example of let's see where we are a year from now. If we're still having to take [ __ ] into Photoshop and
695. 2:15:03like, you know, not be able to just describe what we want and get it then, you know, AGI is another year out or
696. 2:15:12another year after that or another year after that. That's possible. Like, it really is. Like, we don't know. We're
697. 2:15:18not in those [ __ ] labs. All right, I'm gonna get out of here. Go have yourself a good evening. Thanks for
698. 2:15:23watching TV. Extreme's got to go save his marriage now. Oh man, this is awesome to be here, Kyle. Thank you, Sonia. That's great.
699. 2:15:39Um, I've already been given New Year's resolution. I'm starting tomorrow. Very cool, Steo. Um, this is the best late night show in the
700. 2:15:50biz. Thank you, Kyle and Mods. Yeah, thank you mods. Thank you, Brandon, uh, for doing this. I'm going to get on out
701. 2:15:55of here. You guys have a good night and I will see you tomorrow.