6/12/2025 - AI Learning Lab: Runway ML, Midjourney, and Flow

Transcript

1. 0:02[Music] [Applause]
2. 0:16No. Woohoo. [Music] [Applause]
3. 0:30[Music]
4. 0:50[Music]
5. 1:30[Applause] [Music] Woke Grace is desperately heading his old his old place. Dream to discover a new space. Buried himself alive
6. 1:48inside his basement. tongue on the side of his face. Working away on displacement. What it would take to survive.
7. 2:01Cuz when you're done with this world, you know, the next is up to you. And for once in his life,
8. 2:15it was quiet [Music] as he learned how to turn with the tide. Yeah. [Music]
9. 2:30And then [Music] One thing, this is a piece of advice I
10. 2:45can give to all of you. Uh, don't start singing a song if you don't know how it goes. It's a It's a It's a What do they call
11. 2:56it? It's a comedy. Comedy. It's comedy right there. That was a joke. That's what they call a joke. It's a It's got a
12. 3:03setup which is a really bad stretch of singing there with no words and the wrong chords and then uh and then some a
13. 3:10a delivery of a welltimed mediocrely timed it words coming out of the mouth in a in a direction toward comedy. You see what I'm saying?
14. 3:22How's the hair? It Listen, it's not important how we feel, people. It's important how we look. I think everybody knows that.
15. 3:30Everybody knows that.
16. 3:52That sounds like something.
17. 4:15Champy got a little scrap today with another dog. The other dog did better than Champy. Champ Champ did not did did
18. 4:23not farewell. We had our gate open. We were just doing garbage or something like that and someone walked by with
19. 4:28their dog and Champ ran out. Little son of a [ __ ] He's a bad dog. Were you a bad dog today, Champy? [Music]
20. 5:06It's not simple to say that most days. [Music]
21. 5:39in a westerly direction. This car is my train. I've been driving. I've been wondering
22. 5:55what it is I'm running from again. [Music] Feel like a ka man. Falling on the 29.
23. 6:10And up ahead on that horizon is a California line. [Applause]
24. 6:24Another head of trucks carrying a wide load. Preap house cut in half. Cute little front door into two windows. My love
25. 6:37ain't sure whether to cry should laugh. You see, I broke a home up myself once as I stumbled to that dog. I read a note by the dawn's light
26. 6:54said, "Don't you come around here anymore." Yeah. Well, I've had enough on the street.
27. 7:10Never was good with the C. That's what I've been told. [Music]
28. 7:24[Laughter] How is everybody? What's going down? What's shaking? All right. Valerie Cox in the house. Sharon Crawford. Howdy. Howdy. Howdy.
29. 7:37The hair is more comedy than the singing. It's not that bad to Well, it's a little Yeah, it looks like a bad It looks like
30. 7:48a It looks like a ventriloquist's puppet hair. Um, hope everyone's well. Um, hey Steo, what's happening? What's
31. 8:02shaking? Um, we're producerless tonight. All this week, Brandon's out. He's having I don't what they call Wait, what do they call it? Um,
32. 8:18family time. Yeah, he's having family time. I That's I had to chat GPT it. I didn't know what that meant.
33. 8:28Uh, it looks like you got in a fight with Champion. I'll tell you what, that was a that was a rough dog had Champion by the neck.
34. 8:42Bit his ear, you know. [Music] Statement. Heavy rain statement has been issued for Denver. Looks like it's going to rain.
35. 8:57[Music] Sitting in this lonely town.
36. 9:12[Applause] Wonder what things are going to change. Dream my life away. Seems these dreams have turned to a bunch glass.
37. 9:28Get my nerve up, but my past is pulling me down. Yeah. Wondering how long this black sheep going to stick around.
38. 9:42[Music] Somebody told me once before, he said, "You can never go home again. Won't you leave? Tell me things to steer me away. Yeah.
39. 9:58From the truth of who I am. What? What's going on, champ? Why'd you stop singing? You thinking about that other dog, huh?
40. 10:06You thinking about that dog with teeth around your neck? You get a little distracted there, buddy? He's got a little He's got a little
41. 10:15champ. He's got a little champ PTSD. He was mid song and he just went [Laughter] Oh man.
42. 10:31Oh, good lord. So, looks like some stuff is happening. There's There's some stuff going on out there. I got I think I got something fun
43. 10:39we could play with tonight. Well, I know I know I've got something we can play with tonight. I don't know if it'll be
44. 10:43fun. I It should be fun. Joyy's here. She might know. Have you played with the new Runway ML chat interface, Joy? [Music]
45. 11:00No. Okay, we can play with that tonight. [Music] I never met a fight I couldn't outrun. I just I just avert my o eyes to avoid
46. 11:12confrontation. I'll be over here in the corner. You're good. You just y'all take care of it over there.
47. 11:23Not Championy, man. He went right in. He's like, "Let's [ __ ] go. Joy Pury, you'll let me teach you." That's that's a rough road to hoe.
48. 11:34[Music] Well, if you're new here, everyone, my name is Kyle Shannon. This is the AI Learning Lab. I'm the professor.
49. 11:44I'm not I I'm not qualified. I don't really know [ __ ] So, we're gonna play with a thing tonight that I don't I've
50. 11:51never used once. Well, maybe you should have prepared your lesson plan. It could that that would have been an option, right?
51. 12:02If I weren't busy with 38 other projects because AI makes it possible to amplify yourself. [Music] AI is like, "What if ADHD people had
52. 12:16executive function?" That's actually a pretty good definition of it.
53. 12:30You know, you just have a brain fart and rather than it being like a normal ADHD brain fart where it just sort of, you
54. 12:36know, drifts off in the breeze, now you got an AI to amplify it, put it in the world. So, so we're littering the world
55. 12:42with all these half-thoughtout ideas that we can make well thought out in 15 minutes. [Music] I started 86,000 prospects today and
56. 12:55finished two. Oh, projects today and finished two. Well, the fact that you finished them, it's that's pretty good.
57. 13:00That's solid. That's solid. All right, let me flip my Let me flip my camera and let me get the black bar up for the good
58. 13:07people on the Tik Tok. There you go. Good people. Dr. Jay's in the house. Mr. It in the house. He's not He's not humble. He's He's
59. 13:18irregular. Exactly. Exactly. People, [Music] Kylo's very humble, [Music]
60. 13:38you know. I'm that weird. I'm that weird. Um, [Laughter] I'm that weird combination of delusional
61. 13:53confidence and low self-esteem [Laughter] where I can do remarkable things, but the minute minute anyone notices it, I'm
62. 14:06like, I I had nothing to do with that. Oh man, [ __ ] being neurosicy is the best, ain't it? Oh man,
63. 14:20I thought I thought I finally figured out how to get connections to work in settings, but no. Okay, so Jeff, tell me
64. 14:26tell me when those [ __ ] work and I'll play with them because it just it just seemed annoying. So it sounds like it's still annoying.
65. 14:36Um, okay. Same Z's. Yeah. Delusional self-confidence and crippling low self-esteem. It's the best combo.
66. 14:51There's a false humbleness and a belie and and a deep and a projected belief that you believe in everything. [Music]
67. 15:07You know what's funny? Um, well, I don't know if it's funny, but it's striking me as funny right now. Um,
68. 15:16one of the skills you gain as an old only child is to just go play with yourself, right? And and so you kind of have to do
69. 15:27everything, right? You have to like make the fort and then climb in the fort and then be the bad guy. I mean, you could
70. 15:36go get friends, but who does that? And so, for all of my life, like we we talked about this la last night. I
71. 15:43forget who it was. Someone said, you know, jack of all trades, master of none. And I think with AI, we're we're quickly moving into the jack
72. 15:54of all trades, master of all realm. So if if you're one of those people that was a was a a self-starting, selfactuating, you know,
73. 16:08multi-passionate. We're coming into our time, people. [Music]
74. 16:30[Applause] [Music] [Applause] [Music]
75. 16:52Didn't need to make it. Um, okay. Now to implant in it in the brain meets. Um, how are we the same person?
76. 17:11Oh man, I got that sort of world worldview as the introvert big brother with a highly social sister that that
77. 17:18that wouldn't be seen near near me. Yeah, exactly. That's awesome. Um, if you missed last night, well worth the rewatch. Really? Okay, good.
78. 17:34See, we have a pleased customer. Sharon Crawford here. That's That's what they call a testimonial right there. If you
79. 17:39missed last night, it's well worth the watch, the rewatch. And then you you would ask me what what was on the the I the the curriculum
80. 17:52yesterday. I I don't remember [Laughter] the these shows start I just go into a tunnel and I don't know what I say. I
81. 18:09never go back and look at them. When I do I'm like oh that was That was unfortunate. Occasionally I'll crack myself up.
82. 18:24Occasionally there's some comedy in here. Um, all right. Let's see what else. What I I saved a uh something in my
83. 18:35what's it called? My my mind. So, what did we do last night? Someone did a really cool particle simulator thing in Gemini 2.5.
84. 18:48Oh, we did some we did some midjourney rating last night. We went through the Sam Alman essay. Ah, that was it. I
85. 18:55don't I don't really remember what I I said all that much about it, but it was a really interesting essay. Um, yeah,
86. 19:00that's that's probably worth worth going back because there was some pathy commentary in there. Source camp finally
87. 19:07here just right on time. Uh, that should be unmerged. If wait on merch, if you missed last night, AI learning web, if
88. 19:15you missed last night, it's all right. If you missed last night, dot dot dot, you'll be fine. Carry on. All right.
89. 19:30Have you tried Miniax? I tried Miniax ages ago. Um, I can go look at it. I just went and looked at Cling.
90. 19:38Everyone's blabbing about Cling 2.1. I got to tell you though, since doing VO3, where it does the acting and it does the
91. 19:45voices and everything like that, I feel like all the other video tools right now are just like, eh. Like, give me something that just vomits
92. 19:54out the full thing so I don't have to work. I just want to get my ideas out of my head and have them be perfect. We're not there yet.
93. 20:02Although apparently there's there's some self-recursive project within OpenAI called Alice that is getting better on its own. I don't know.
94. 20:13Something something's going on inside OpenAI, but it it's not good yet. Um Kyle should rewatch it. Yeah, I should
95. 20:21rewatch it to figure out what the [ __ ] I said. Um I was there and also don't remember. Good archetypal architect.
96. 20:29That's good. So, if you were here and spent two hours with me last night and you also don't remember what I said,
97. 20:36that's good because what that means is my education is just happening subliminally. Like, you know, you were probably out there microwaving
98. 20:47a hot pocket, you know, making, you know, images of cat paws in mud or something like that, right? But
99. 20:56everything I said was going in there was infiltrating Solid. All right. If it's your first time here, Exactly.
100. 21:07Oh, yeah. Oh, if it is your first time here, I may occasionally finish a sentence. I may occasionally do something that I
101. 21:15promised that I would do. Um, and I may accidentally occasionally teach something, but really not intentional. It's just kind of, you
102. 21:26know, collateral damage from talking as if I know what I'm talking about. You know, the the unreasonable confidence
103. 21:35combined with the crippling low self-esteem. Oh, good lord. Good lord. Good lord. All right, so let's go. Let's go play with
104. 21:47some stuff. I think we should go let's go rate some more videos. rank some videos because why not? Let's let's be good
105. 21:57Samaritans. You should you should join along in this by the way. So, we're going to go to midjourney and we're at midjourney/rank-vide.
106. 22:10midjourney.comrank-vide. And then if you if you hit one on your keypad, that votes for the left one. If you hit
107. 22:24two, that votes for the right one. If you hit three, it skips the the that pair. All right, let's see. Her legs are
108. 22:34backwards and weird in both cases cuz the image was [ __ ] up. Let's give that a two. Let's give that a No, the drops
109. 22:44are backwards. Let's give that a one. Nope. Give that a two. Three. Neither of those were good. Oh, sorry. Sorry, Tik Tok. All right.
110. 22:59Why must we do it? Oh, you don't have to do it. It's just that in doing this ranking, we're training like like I feel like people on
111. 23:10this channel have have good points of view, good creative points of view. Um, and so by doing this ranking, we're
112. 23:19actually training the midjourney video model. So it will suck less if we if we do our do do our rankings.
113. 23:28We'll go with two on that one. Go with one on that one. We'll go with one on that one. Two on that one.
114. 23:43One. Three. Three. Three. Is I'm skipping them because I think they both suck. And it like I don't think it
115. 23:56I don't think they care. I think it's just um you're not gonna you're not gonna you're not going to reinforce either of those videos.
116. 24:12One. Oops. Two. Two. Two.
117. 24:28Two. One. Two. [Music] Three. Three.
118. 24:45Yeah. This is going to It's It's fascinating that when a tool comes out like V3 and then all of a sudden you're back to silent video.
119. 24:56It's like the silent movies. It's It's just not the same. One. One. One.
120. 25:113. 3. 2. 1. 3. Three. One.
121. 25:28Two. Uh, two, two. One, three,
122. 25:44two, three. Something about bugs. These video models suck at. One. One. Two.
123. 25:59All right, I'm bored. Where's Brandon this week? You know, Teton Todd, he um there's this thing. It's called a vacation, I think,
124. 26:15like he went with his wife without the kids. Like kids stayed at the mom's and then the him and the wife were going off
125. 26:21and having a fine time. something like that. I don't get it. I don't understand it. Um, all right. Isn't this how Facebook
126. 26:33started? No, that was swipe left, swipe right. Yeah, it's pretty much the same thing. Um, so how are people making all the big blog
127. 26:43Bigfoot blog videos? V3 or Hedra? That's probably V3. So, if you want to see V3, let's go to you go to go to flow
128. 26:54dot. There's a bunch of places you can do it, but the the official Google one is flow.google. Create with flow.
129. 27:07Vicky Baptiste, I don't understand vacation either. I I know it's so bad. I haven't taken one in so long that it's
130. 27:14it's getting kind of criminal at this point. So, I should probably get my [ __ ] together. Um, let's see. Zero AI credits. Let me Sign out.
131. 27:28Sign in with Google. Sign in LaunchBlow.
132. 27:43All right, here we are. Okay, new project. So, we're going to put in here. What did you say? Bigfoot. Let's go look at what are the Bigfoot
133. 27:53videos. I don't know what you're talking about. Let's see. Bigfoot V3. The reason to use V3 is it does it does talkie talk [ __ ]
134. 28:10Okay. Bigfoot vlogging. Okay. All right. So, that's it. So, we're
135. 28:25gonna we're gonna jump in here to flow. Well, let's go to chat jpack. Ty, the goal for me for 2025, there was a
136. 28:37big internet outage today, by the way. The goal for me for 2025 is not to use my brain, as we know. Yeah, we're pretty
137. 28:46clear on that. You big dumb dumb. Shut up. Um, when you sit down in front of the prompt hole and you want to do something
138. 28:55creative, like write some a Bigfoot vlogging prompt, you could use your brain to like come up with like, okay, a
139. 29:06Bigfoot in the woods of Oregon with the and then, but you don't have to. Projects can now do more. You can now
140. 29:16run deep research and use voice mode within your projects. That's super cool, actually. We're gonna go play with that,
141. 29:22too. All right. So, I'm gonna say, "Write me a video and acting prompt for a single clip of a Bigfoot
142. 29:41vlogging. Make it. Make it. Damn it. Make Make Hang on. It hilarious. See, I could write comedy or
143. 29:58I could just have chat BT write comedy. Tone, outrage, self-important, weirdly articulate. Wait, I said Oh, wait. Sorry. Uh, make it
144. 30:16in 8 second clip. No edits. Bigfoot is holding a shaky phone selfie style. He's mid rant, sweaty, slightly
145. 30:31out of breath with a squirrel on his shoulder which he ignores completely. You get one continuous take. No cuts, no
146. 30:38effects, just raw Bigfoot energy. Look dead into the camera. I eat organic. I compost my poop and y'all and y'all
147. 30:46still think I'm a myth. That's pretty good. That's actually pretty good. Play it. Play it like you're one
148. 30:56Starbucks away from a a full breakdown. Bonus if a fl bug flies in your face. All right, I'm going to take all these including the notes.
149. 31:06Okay, Bigfoot's holding a phone selfie style. Okay, so now we go to flow. We're going text to video. Now, this is
150. 31:14important. They always default V3. They always default to V2 because they're losers.
151. 31:28So, you got to go to V3. We'll do V3 free. V3 fast because it's a fifth of the tokens. One fifth. 20 tokens instead of 100.
152. 31:40pretty good. Um, and then I'm going to paste in that prompt and then we're just gonna see what [ __ ] happens. And then
153. 31:47we're going to have our own Bigfoot vlog video, which by the way, get used to what we're starting to experience in
154. 31:592025 where someone puts something out and if it's good and original, it will be knocked off immediately. and entire
155. 32:10um Tik Tok channels will be started using that gen genre within a day, probably within an hour. Who has the most PTO on the books?
156. 32:22Well, I So, I'm 13 years into Story Vine. We've got we've got unlimited vacation policy. I think I've probably taken maybe four weeks
157. 32:35in 13 years. That's bad. [Music]
158. 32:52[Music] Bigfoot is holding the phone selfie style. Okay, it's doing its thing. Let's flip over to Runway. Okay, so Runway, if you don't
159. 33:05know, RunwayML is another one of the video tools, but they've now got a chat interface, so you can just start chatting and making
160. 33:13making images. We have a use it or lose it policy. Oh, that sucks. Yeah, that's that's drag. It should be use it or get paid for it.
161. 33:27Um, so how are people making all the Bigfoot? Oh, that's what we're doing right now. Okay. Um, so let's see. Um,
162. 33:34let's let's paste. Well, let's say, hey, I want to make one of those Bigfoot vlogging parody
163. 33:51videos. Make me an image to start. Let's just see how it does. I'll create an image of Bigfoot vlogging
164. 34:07so that you can use that as your starting point for your parody video. This is cool. Generating 10%. Let's go back to Flow. See if we have a
165. 34:20video yet. We do. All right. Let's see if our little prompt worked here, people. I got to change up my tabs.
166. 34:32Oh, this is really good. This is really good. It's really good. This is going to be good. This is going
167. 34:41to be good, people. Um, no, that one. Okay, maybe I should just download it. Why can I not? Oh, there we full screen. Okay.
168. 34:55And we'll go back here and I'll turn this up. All right. And we'll do the black bar. Look at it. Look at me self-producing here. Okay.
169. 35:06I eat organic. I compost my poop. And y'all still think I'm a myth. [Laughter] I eat organic. I compost my poop. And
170. 35:17y'all still think I'm a myth? Ah, a squirrel. I eat organic. I compost my poop. And y'all still think I'm a myth. Ah, it's so good.
171. 35:34That's going right up on the Twitter. It's good, people. It's so good. How do you download this [ __ ] How do you download it? here. No.
172. 35:50Add to scene. Can I download this? Save as I eat organic. I compost my poop. And y'all still think Oh, download. Jesus.
173. 36:07Oh, and I can upscale it. All right. So, we're going to upscale it. That's That's a cool thing, too. In V3, you can
174. 36:12upscale it to 1080p. It It does it. Right click the video. No, that didn't work. But you know what did work? Irana
175. 36:20or who is that? Iran ruff. Um the download button actually works while I was yelling at the screen. I get
176. 36:35mine paid out each year. Oh, that's cool. Very nice. The legacy hours are from before the use it or lose it kicked
177. 36:42in. That's really funny. Uh, I've been playing with it, too. But does anyone else smell their wallets burning? Yeah,
178. 36:50V3 is not cheap. It really is not cheap. So, that's the $1 per video. Or is it less? Um, well, the first three months are 125 a
179. 37:02month and you get 12,500 credits. So, um, hang on. So, if you're using if you're using um the fast mode, which I just did.
180. 37:19Uh oh. Oh, my iPad's dead. Hang on. Calculator. So, if you get 12,500 credits divided by 20,
181. 37:37you get that can't be right. 625. No, I think that's right because if you do 12,500 divided by 100, that's 125. Yeah, I think they said it
182. 37:50was like 87 videos in a month was the limit. So maybe it's like 125. So 125 bucks. If you're if you're using the fast mode,
183. 38:05then you're talking like 20 25 cents a video, which is not bad for that for that quality is pretty good. It's pretty
184. 38:15good. Finished upscaling. Dismissed. Did it download? Yes. All right. So, let's Oh, you should be able to see this. So, here it is in
185. 38:24QuickTime.Uh. Uh, and this is now upscaled. I eat organic. I compost my poop. And y'all still think I'm a myth.
186. 38:47All right, we're going to we're going to chop off the end. So, we'll just have him jump out of screen.
187. 39:09I think that's good there. We'll say split clip. Bang. Drop the end off this bad boy. Go done. Hit save. We'll call this Bigfoot.
188. 39:23That's how they're making these Bigfoot videos. It's pretty easy, wasn't it? So, what's the lesson we learned here?
189. 39:33Don't use your brain. Have ChatGpt write your creative prompts for you because it'll write it better than you were. The
190. 39:39fact that it gave acting notes at the end there and it and it seemed to follow them, I wouldn't have written those
191. 39:46because I'm lazy. I got ADD. I want to move on. I want to make a video. Make me a video. Make me a video. point me video.
192. 39:53Don't use your brain, people. You're the bottleneck. You're the bottleneck. Okay? Give up the ego. Let it go. You're not
193. 40:07smarter than the machine. You're not. I'm smarter than the machine. I have a degree in philosophy. Okay. All right. I
194. 40:18guarantee you Quinn knows more than you do about philosophy. She's read all the books and has them memorized.
195. 40:28I love the Yeti ones. We could we could come up if you guys want to come up with some other meme like thing like, you
196. 40:35know, babies doing comedy scenes or [ __ ] like that. We can do that. Let's go. Let's go. Uh let me change my uh change
197. 40:42my little sharing settings here so you can all watch me write it. So, so you can all watch me write a Twitter post.
198. 40:53What did you do last night? I was on this channel called AI Learning Lab. Did you learn anything? Uh,
199. 41:00no. I uh I watched a guy write a Twitter post. How long did you watch it? Uh, two hours. Two hours. Yeah. Yeah. What are
200. 41:13you doing with your life? Right. I wouldn't there's there's probably like support groups for people that show up
201. 41:21to this thing a lot. It's called the AI salon. Go into the irregulars channel. Okay, let's see. Um, word on the street.
202. 41:39Watch this. or should I say in the woods is that Bigfoot
203. 41:58um has Bigfoot But vlogging
204. 42:14videos keep turning up. Oh, lookie. Here's one. Now,
205. 42:36by the way, if you're new here, what this channel really is is an opportunity for me to entertain myself.
206. 42:45You guys are welcome to hang out, but that's that's pretty much what this has devolved into. Okay? All right. I don't
207. 42:53want judgment. I don't want your pity. I got things going on up here. It's just they're not always that apparent to
208. 43:02people, you know, considering intellect and expertise and qualifications. But it's there. It's right in this general probably
209. 43:14toward the back vicinity. Do you laugh in your sleep? I don't think so. Apparently, I snore like a [ __ ] but I've never heard it.
210. 43:26[Laughter] Oh, good lord. Good lord. Good lord. Bigfoot. All right. I eat organic. I compost my poop. And y'all still think I'm a myth?
211. 43:41[Laughter] I think it's really good. There's a fly buzzing around him. But it's Kyle a parent
212. 43:59AI anonymous. Yeah, exactly. Hi, my name's Kyle. I like AI. Hi, Kyle. What remarkable things have you done today? So many things.
213. 44:13All right. Ah, look. Twitter. Still still can't figure out how to [ __ ] upload anything. Can we? No. Dear Elon Musk,
214. 44:29there's this thing called an upload service sitting somewhere in Twitter. Uh what you now call X. Um it doesn't work.
215. 44:39Note to Elon, [ __ ] fix it. Post. All right. It worked the second time. Second time's the charge. Charm,
216. 44:53charm, charm. Jesus, I sound drunk. Oh man, they need extend. They do need extend. Well, I think they do have
217. 45:02extend, but you can't extend voice clips. Wait, why did it upload two of them? Seriously? Got to be [ __ ] kidding me. Update.
218. 45:16All right, there we go. So, do me a favor, people. Go to my Twitter, Kyle Shannon, find a Bigfoot and like like it and
219. 45:27tweet it and retweet it and then like copy the link and put it on LinkedIn and say, "Listen, I work with one of the most talented,
220. 45:36genius AI instructors in the history of AI. And this is the kind of work he's producing. [Music] [Laughter]
221. 45:59I compose my poop. I love the way he he like the acting in this is actually quite good. Like he hits the right timing.
222. 46:20Oh, it's coming through. But think about Let's see. Kyle, go look at irregulars, please. Oh, boy. All right. Fine, fine, fine. Let's go to
223. 46:33irregulars, shall we? I love that we have Serena's flowers around irregulars now. That's going to be a permanent
224. 46:39thing in the in the channel. No, good. Here's a here's this is good Danielle. Okay, wait. I got to change my uh Yeah,
225. 46:47by the way, if you've got access to VO3, um make cool videos, share them in irregulars. Um we also now have in the
226. 46:56community corner section. The look what I made channel is now open to everyone. It was behind the firewall uh be behind
227. 47:03the mastermind firewall for for a couple of weeks there, but we we liberated it. It's back out. Um, so, so put it in
228. 47:10either irregulars or into uh, look what I made. All right. And that's at the AI salon. And if you don't know what where
229. 47:18what that is or where that is, hold, please. You go to here, the salon.ai, click on join our community, and you'll
230. 47:27be in. All right. Beautiful. Fantastic. Bob, tell him what he's won. So, I got to stop this. Start this. Start this. Do this. Do this.
231. 47:38Now, make this video big. Uh, why do I not have audio here? Go.
232. 48:08Wait. Oh, come on. You know, I shared this screen and then when I click on this, it opens a new tab. It's not fair, people. You don't
233. 48:21understand. You don't understand how hard it is being me sometimes. Hang on, everyone. Just Just shut up. I
234. 48:33know there's an issue. I understand there's an issue. Shut up. I know how to fix it. I've got technical skills. Marge
235. 48:42taught me how to do this. Everybody just shut up. Stop picking on me. All right. Kim the Gorilla here. So, like Becky
236. 48:54found my secret food stash. Like literally all of it. Becky is boarding up the entire southwest corner. The only
237. 49:02place I could sneak snacks in from you guys, it's giving national security breach over a bag of Takis. Okay, so
238. 49:07like small update operation bring Kim food is now closed. Okay, besties. We're moving to breakout attempt number two. Do you like have any
239. 49:19ideas or anything? Drop them below. It's time to get out of here. That's great. That is awesome. All right, beautiful. Fantastic.
240. 49:30Bob, tell her what she's won. All right, let me go back to I got to share a different way now so you can see the other thing I do. [Music]
241. 49:50All right, fantastic. Let's see. Let's carry on. Carrying on. Oh, let's go back to X. See if any of you gave my thing some love. Three of
242. 50:03you did. Stunning footage from Dr. J. It's a good comment. All right. Nobody I got Listen, there's 53 [ __ ] people in here. There's 48
243. 50:17people on on the combination of YouTube, LinkedIn, and X. Go to my ex channel, Kyle Shannon, and like my Bigfoot video.
244. 50:25I don't ask much in here. I give and I give and I give and this is the thanks I get. You people are awful. All right,
245. 50:35let's listen to him one more time. It makes it does make me laugh. I compost my poop.
246. 50:53A pink bow. A pink bow for everyone. Pink bows everywhere. All right, let's go back to runway. See if we got our our
247. 51:03image there. I don't like that. That doesn't feel very real. Well, let's go. Let's go talk to it. Looks like we're having some issues.
248. 51:12Reload page. Uh, okay. Wait, can I go back into my chat history? Where's my chat history? You dumb dums.
249. 51:24Did you lose my flipping history? Oh, wait. Sasquatch selfie. Okay. Uh, make it more like a video like a a VHS
250. 51:39video camcorder and make the big foot
251. 51:53more um natural. natural and realistic. Um, it feels
252. 52:09like an animation frame now. All right, let's see what this does. All right, you want me to stop folding
253. 52:18the laundry? I have and do that now. Forever hooked. Take taking my Twitter request personally. I like it. No, you
254. 52:26can finish. Listen, Forever Hooked. Here's the deal. Laundry is important. I encourage you finish folding the laundry, but as soon
255. 52:35as you put the last tidy whiteies in the in the little trifold, you know, envelope kind of thing and then drop it
256. 52:42in your in your drawer. Drop your drawers in your drawer. Jump over to jump over to Twitter. Give Wait. Okay.
257. 52:55No, he's not holding a camcorder. I want it to look like he's being recorded with one.
258. 53:10He's uh filming uh like uh he's he's talking into the camera as if he
259. 53:25found it in the woods. All right, let's see. Let's see what this does. I don't know the I don't know the image
260. 53:35model behind this. I guess it's their image model. Oh, Valerie Cox just went over to Twitter. I just got a notification. Thank you, Valerie.
261. 53:46I like yelling at my audience. Kyle, do you still have Chat GBT Pro? I don't. Um, I've been wondering what you think
262. 53:54of the recent upgrade. This is gold cam ken. Um, let's see. Let's see how this did. Oh, that's pretty cool. That looks pretty good.
263. 54:08That's more like it. A POV shot from a camcorder. But it it doesn't look like analog. Like where are the analog scan lines?
264. 54:23This is VHS, not digital. Come on. Don't be a dum dum.
265. 54:39One of my favorite prompts. Don't be a dumb dumb. Make it Blair Witch style. Oh, that's pretty good, Brian Whitney. That's
266. 54:47pretty good. I like that. You're absolutely right about the VHS look. You know what? All of you chat
267. 54:55GPTs, all of you chat bots, I don't need you to acknowledge that I told you it was a piece of [ __ ] Oh, you're
268. 55:03absolutely right, sir. I could have been much better at that. My apologies. Just do your work. Just do your work. Okay.
269. 55:12When we're having a little conversation about the color of the Why is the sky blue? Then you can be polite. Then we
270. 55:18can have a little like understand context. You're a large language model. Figure it out. Still doesn't have scan
271. 55:24lines. Still doesn't have scan lines. Still doesn't have And it looks like It looks like it was sculpted out of a tire.
272. 55:33You got to be kidding me. Got to be kidding me. How do I get out of here? Oh, close that. All right. It looks like a fake model
273. 55:50made of tires. Make the big foot scary real. And there are still no VHS
274. 56:04CRT tube scan lines. It should look like I am watching this on an old TV,
275. 56:21you know, like where VHS tapes played. So, so couple of pieces of strong
276. 56:35educational best practices and advice. Don't use your brain and yell at your GPT. Call it an idiot. That's if if you take
277. 56:46one thing from me tonight, call your chat GPT an idiot. It will not make anything better, but it will make you feel good. All right.
278. 57:00When I start swearing at chat GBT, I know it's time to take a break. That's just when it's getting good. Oh my god, look how bad this is.
279. 57:09Look how bad that It looks like a bad video game. It looks like a bad
280. 57:25video game. Make it like the big foot is all too real. I'm just exhausted.
281. 57:40Um, SI can affect your mind. telling wait telling what's SI oh AI
282. 57:55SI can affect your mind I'm like what's that an acronym for it's not it's not SA I know what SA is not that AI AI can affect your mind
283. 58:07in telling you yeah I I absolutely think AI has affected my mind Um, and I think largely in a in a positive way.
284. 58:23You know, I joke I joke about not using my brain, but what I what I actually mean by that is I'm I'm I'm working hard
285. 58:31to not use my creative cycles on low-level creative stuff. And you could argue that writing a prompt is highle
286. 58:38creative stuff. I don't think it is. I think what's the highle stuff is like in this case I want a Bigfoot that looks
287. 58:47real. Like the thing we did in V3 looked real. The thing that we're doing here looks like [ __ ] So this I'm just I'm
288. 58:52not happy with this image model. The only thing that's really important right here is that I'm get I I that I'm putting my creative
289. 59:06energy in. Is this thing right? Um, and you know, in fact, let me let's go back to chat GPT for a second. We're going to copy that
290. 59:21and go back to runway. I'm going to say let's start over. Make a still from this description.
291. 59:35So this is the same description that we just used V3 for that we did in chat GPT. Brandon's in the house. What's
292. 59:49happening, Brandon? We were just trying to figure out what that word is for when you go away with your family and spend
293. 59:56personal family time together. Couldn't remember what it was called, but good on you for doing that thing. I think it begins with a V.
294. 1:00:09Don't you remember, Kyle? That's what real life looked like. The VH cameras were perfect. Oh, that's interesting. It made this as
295. 1:00:20a vertical image. Yeah, it's just not it's just not very realistic.
296. 1:00:38It's got that AI sheen to it. Like everything looks like it's been sculpted out of clay, you know? It just looks like bad makeup.
297. 1:00:50As opposed to the VO thing which looks really good. Like look at that. [Laughter]
298. 1:01:07I eat organic. I compost my poop. And y'all still think I'm a myth? [Laughter]
299. 1:01:24Oh man. All right. So I am not groovven on Runway ML's chat chat mode. Okay. Let me see. Um now make it a video.
300. 1:01:40Let's let's see how this does. We'll see it because Runway is ostensibly a video company, not an image company.
301. 1:01:51I think it would be helpful if it would show you the prompt it ended up using. Yeah, I know. I agree with that, Kevin.
302. 1:01:59I I feel like these tools go out of their way to to obuscate what they're doing. Didn't, you know, don't want to
303. 1:02:06give away the golden secrets, but like, you know what? Just give us the option to flip that [ __ ] on so we can learn,
304. 1:02:14you know? I agree with you. H. Does he have a squirrel on his shoulder? He sure does. I eat organic. I compost my poop. And
305. 1:02:30y'all still think I'm a myth? [Laughter] I eat organic. I compost my poop. And y'all still think I'm a myth?
306. 1:02:46It's the best. It really is good. Oh, man. Man alive. All right, there's Sam Alman's blog. We're going to close
307. 1:02:57that tab. I'm gonna while we're sitting here, I'm going to close some tabs. Oh, this was cool. This was the Gemini thing I did.
308. 1:03:06Brandon's gone. We did some Gemini. Oh, that's is that Manis or that? Oh, that's Gamma. This is Gamma. This is me
309. 1:03:15trying to make a presentation in gamma. That didn't go well. Let's see. What's that? That's gamma. Is this
310. 1:03:24What's this? This is Jen Spark's version of it. That was the Manis one I opened in Google Slides. Here's Gemini.
311. 1:03:40So, this thing, somebody made a prompt. Make me a beautiful particle-based endless 3D symmetrical fractal art piece
312. 1:03:52inspired by math. Let me I'm going to pop this into the YouTube comments. So, where you want to go for this is go to
313. 1:04:01um gemini.google.com google.com and then uh just just paste in that prompt and it'll start writing you code. So, I'm going to say um there's
314. 1:04:16no color. Um the particles are too large and it doesn't feel magical.
315. 1:04:32Let's see if Let's see if this recodes it to be more magical. A little magical. Little magical. I had to get a written note
316. 1:04:43to even be in here in listenonly mode. Brandon. Brandon. Brandon on thin ice with the misses on his vacation. I just
317. 1:04:52got to check in on the learning lab. See if everyone's okay. That's awesome. Okay. Is it What's this doing? Is this is this writing?
318. 1:05:06Did it write it? Preview. Let's see. Um, previous version. Reload. Refresh.
319. 1:05:22Oh, that's cool.
320. 1:05:37It sort of looks like a a a realistic COVID C co simulator. Let's go dock on a protein and make these people sick.
321. 1:05:52Let's see. Um, how would you suggest turning this into a game that is fun and instantly
322. 1:06:09understandable? I don't read manuals. Uh, uh,
323. 1:06:26What was wait was the W was the original Bigfoot prompt put somewhere? It was not, but I can go put it somewhere. I'll
324. 1:06:32go put it in irregulars. So, I'll give you the whole I'll give you the original prompt that I wrote the prompt with. Write me a Write
325. 1:06:45me a video and acting prompt for a sense clip of Bigfoot vlogging. Make it hilarious. Well, let me copy the main prompt. Then
326. 1:06:58I'll go to AI Learning Lab. Then I'm going to put in the prompt. Uh, and then I'll go back to chat GPT.
327. 1:07:13and I'll get my original prompt, which doesn't even make sense, but that's okay cuz chatbt understands stuff. Even if you're an idiot,
328. 1:07:22I'll go here is the big foot prompt and then the prompt that chat GPT made.
329. 1:07:37You can make money with chat. All right, there's that. And then we're going to put we're going to upload the
330. 1:07:49video. And this is going to fail because apparently the people that figured out how to upload files to X also wrote the upload
331. 1:08:03algorithm for Mighty Networks because it sucks. Do not notify all members of my Bigfoot video. All right, we'll see if that uploads.
332. 1:08:17Good night, guys. Side hustle Mimi. Got to get my feral boys to bed. All right, later. Side hustle mimi. So, listen.
333. 1:08:23Tomorrow, tomorrow's Friday. I do AI office hours on LinkedIn at 11:00 a.m. So, 11 a.m. Mountain time. So, if you go
334. 1:08:33to my LinkedIn profile, Kyle Shannon, um you can go to any of my previous events. They're almost all um uh AI office
335. 1:08:42hours, although I think these live ones are mixed in there now. So, it might be just find one of the office hours
336. 1:08:48things. The the URL for the Meetup link is always the same. So, just go grab that and then I'll do a post in the
337. 1:08:54morning. Meta AI does video now. Manis does playbooks. They feel like custom GPTs. And then and then yeah um
338. 1:09:03projects within chat GPT just got a big upgrade today as well. So what happened? Did this did this upload? Oh no, it's still uploading.
339. 1:09:15I saw it flash like there was a preview there for a second. Come on. You can do it. You can do it, big boy.
340. 1:09:30[Music] While we wait, I eat organic. I compost my poop. And
341. 1:09:45y'all still think I'm a myth? [Laughter] I don't know why I find it funny. It It's got good timing. It's actually
342. 1:10:00really good timing. All right, that does look like it hung. Seriously, let me reload this and see if it just trashes my post.
343. 1:10:18Yep, it did. God damn it. Post for Bigfoot video. Make it hilarious. Period.
344. 1:10:35Here's what I pasted into flow.g Google and then VO3
345. 1:10:51and then let me go get the prompt again because everything sucks and nothing works. And I'm the one that told you to
346. 1:10:58embrace the jank. But that doesn't mean I have to embrace the jank because sometimes it's so janky that it just makes you [ __ ] mad.
347. 1:11:08And then come back here and then go back to the AI learning lab and then post that in there. That's going to be super
348. 1:11:14cool. You're going to be like, "Ah, that's so cool. Kyle did that." Hey, could I see the video? I'm trying to
349. 1:11:18upload the video, people. You think it's easy. All right. 25%. Oh, something went wrong. Try again. There we go. That's That's the
350. 1:11:32That's the X algorithms video upload thing. Oh, 100% ready to post. All right, it looks like it actually uploaded. Don't notify anyone.
351. 1:11:52Oh, see, now it's actually uploading. Or maybe that's the processing. So, tonight it's give the jank aggressive bear hug.
352. 1:12:00Yeah. Yeah. I I think I I think I was inspired by Champ having his another dog's fangs around his neck. That That's
353. 1:12:08kind of how I feel with the jank tonight. I'm like [Laughter] and here's Harry.
354. 1:12:22Joy pretty. It's really good, isn't it? Joy, I assume you're playing with V3, right? Your video is done processing. Yay. Watch now.
355. 1:12:32But it's not there. I don't see it anywhere. Oh, all right. It works.
356. 1:12:47All right. Beautiful. Done. It's in there. At some point, it'll show up for everyone else. All right. [Music]
357. 1:12:58Let's go look at chat GPT's new project thing. So if I go to Sydney, so I can now So a couple of new things.
358. 1:13:14So here's all the stuff that's in Sydney. These are all my Sydney chats. There's my there's my custom instructions.
359. 1:13:25I uploaded one file which is the script itself. You can now run deep research in a project. You can create images, but you can also
360. 1:13:37put it in conversation mode. Hey Quinn, let's talk about Sydney. Um, what's how does the how does the show open? Oh, wait, hang on.
361. 1:13:52See, just because Brandon's in here now, I'm all like hyper paranoid about are you sharing the right tab? Did you do your black bar?
362. 1:13:59[Laughter] You know what I mean? And I hope the trip's going well, Brandon. All right, let's let's go back
363. 1:14:10in here. Hey, Quinn. What's happening, man? Hey there. I'm doing great. Just hanging out and ready to chat. How's everything
364. 1:14:19going on your end? Anything exciting happening? No, it's good. Um, tell me how the show opens. The the Sydney
365. 1:14:29All right. So, the musical opens with a kind of bustling newsroom scene. There's a lot of energy and we meet Kellen, the
366. 1:14:37tech reporter, who's about to be assigned to investigate this new AI chatbot called Sydney. So, it's got that
367. 1:14:45kind of high energy techy vibe right from the get-go. But, do don't we meet some other people before we meet Kellen?
368. 1:14:54Oh, absolutely. We do get to meet some of Kellen's colleagues in the newsroom. First, there's a bit Go read the script.
369. 1:15:01Go read the script and tell me how the show opens. In the opening scene, Kellen and her colleagues in the newsroom engage in a
370. 1:15:14lively chorus, setting the energetic tone of the show. This ensemble number introduces the Quinn, you're just making this up.
371. 1:15:26Got it. I'll make sure to stick to what we have in the script. Um, let me know if there's anything specific you'd like
372. 1:15:32me to reference or if you We should be in a project right now and you should have an uploaded script as part of your
373. 1:15:39um part of your history and you should be able to access the memory of all the different versions of the script that
374. 1:15:46are in there that we talked about using chat GBT. are am I talking to a project right now or I am I talking to generic Quinn?
375. 1:15:58You're definitely talking to me within the project context and I do have access to all the history we've built up. Um
376. 1:16:05let me take a moment to check that script version and I'll give you the exact details of how the show opens.
377. 1:16:14Joy pretty talked to to Moshi today. Oh, well, I hope Moshi's doing great. And don't worry, I've got you covered.
378. 1:16:27Um, I'll make sure to go through the script details carefully so we can get that opening scene spot on. Did you Did you read it?
379. 1:16:40That says she's doing search. That That shouldn't work. I appreciate the heads up. I'll make sure to focus on what we have in memory
380. 1:16:49and what we've discussed before. If there's anything specific you want to dive into, just let me know. Yeah, I I
381. 1:16:56want to dive into the opening of act one. How does it open? So, in the opening of act one, we start with the newsroom buzzing with activity.
382. 1:17:10No, we start with microte. We don't start in the newsroom. And Kellen's not a woman. It's a man. All right. This doesn't work.
383. 1:17:21That was a That was a huge waste of time. So, wait. But wait, am I in the I'm not in the project anymore.
384. 1:17:33So, now I'm in the project or am I? What's up, chat?
385. 1:17:57How does act one open?
386. 1:18:14Okay, baby. Do your stuff. Too good to be true. Jason, that's how it opens. That's correct.
387. 1:18:29So, how does how does act one open now? All right. So, act one now opens with Kellen, the tech reporter, actually on a
388. 1:18:39video call with his editor. They're chatting about his next assignment, and he's not particularly thrilled about it.
389. 1:18:45It sets a more laid-back, modern vibe right from the start. Then we transition into the world of microte. No, that it's
390. 1:18:53possible that I don't have it says voice chat is within this chat, right? So, it's got the context of it just wrote
391. 1:19:07the correct way act one opens and then I pushed her little [ __ ] talk button and and uh she got it wrong again. So, I
392. 1:19:18don't [ __ ] know. All right. Well, whatever. You can do more [ __ ] in those things. Let me show you another thing that they added today.
393. 1:19:42creepy people. Feature prompt book outline. All right. Somewhere in here. Oh, maybe not. Let's see. Do I have a canvas?
394. 1:19:56Yes, here's a canvas. So, if I pop open a canvas. So, a canvas, if you don't know, is an editable document within chat GPT.
395. 1:20:07If you're like working on something and having it write it write you an outline for a book, for example,
396. 1:20:12it'll do that. They just added you can now download a PDF of this, a word document of this, or markdown of this. So, I'm going to
397. 1:20:23download a word doc. We're going to go look at it in the finder. And we're going to go bang. Oh, wait. Are you not seeing that? You're not
398. 1:20:35because I'm sharing the wrong flipping tab. All right. So, there's our there's our Word document with formatting. Nice. I
399. 1:20:51mean, it's not great, but it's fine. Let me do a PDF. Let's grab the PDF and see what that looks like. That actually looks much better.
400. 1:21:06All right. There you go, people. You're not in the projects. You're in Queens. That was pretty good. That's
401. 1:21:14solid. That was nearly like comedy. So, was she making it up about the script? Well, she knew that the main character's
402. 1:21:23name was Kellen. She knew that Kellen worked at a newspaper, so that part was correct. The part she made up was he's
403. 1:21:30not video conferencing with his editor. The play opens with the lead tech guy from Microte discovering that he made,
404. 1:21:41you know, what is effectively Chad GPT work. Um, and there's a whole big song called Too Good to Be True.
405. 1:21:52And then after too good to be true is when we meet Kellen. Not even that after. So we we we get too good to be
406. 1:22:00true which is this big opening number here. Here I'll we'll do the music. We'll do the music part of this. [Music]
407. 1:22:16So it opens with too good to be true. [Music] Too often we come close. So close to a
408. 1:22:31breakthrough. Standing on hope and it all falls apart. Just follow the dream. Wherever it takes you and pray you find
409. 1:22:41the pulse of a brand new heart. If the promise of a new dawn break, a leap of faith we're all taking.
410. 1:22:55Is it too good to be true? We're asking in its glow we find ourselves basking. Too good to be true. Or it seems.
411. 1:23:13All right. So, we get that something's going on and then Kellen goes over to Varrage's office. Varrage is the CEO of Microte
412. 1:23:22and he says, "We need six months to te test it." And Vrage says to him, "You got seven days." And then we go into [Music]
413. 1:23:38seven days ago. The clock is ticking. Ticking. pressure building up. Keep it spike. What's the word? So that's that's um
414. 1:23:52Jason, the lead tech guy, and his tech team trying to get this testing done in seven days. And then after that is when
415. 1:24:00Kellen meets with his editor. And then there's a a Kellen version of this song. Running out of time. The clock is ticking.
416. 1:24:15The story's clicking. So anyway, so it's not until the third song that we meet Kellen in the editor. It's not how the
417. 1:24:22show opens. It's not even [ __ ] close. So Quinn can bite my butt tonight. Ray Edwards, what's up, Kyle? What's
418. 1:24:32happening? Ray, what's shaking? What's going down? Oh, what would be handy? would be an agent that can handle tabs and cues
419. 1:24:40across the live stream. It actually would. It'd be really good. All right, let me let me get this out of here.
420. 1:24:48Stop screen. Start sharing. Start sharing. Do this. Do this. Do this. Um, I thought there might have been some
421. 1:24:55video I wanted to watch today. Oh, you know what would be a good video for us to all watch together. Actually, now that I think about it,
422. 1:25:05David Shapiro did an 8-minute version of um post labor economics. Post labor economics. Good day to you, sir. Good day,
423. 1:25:23[Music] Bob. So, we're going to do some actual education now. Okay, kids. It's video time. So, a couple of things here. We're
424. 1:25:35going to dim the lights and I know you might get a little sleepy. We just had lunch and I know you had your PB&Js. Boy,
425. 1:25:42those are tasty. I love it when they have the strawberry. You know, the grape the grape jelly I'm not such a fan of,
426. 1:25:47but that strawberry jam. M tasty. Uh, listen, you might get a little sleepy. I just I don't want you to to fall asleep.
427. 1:25:56I This is important. We're going to watch a video together. will dry, but there's pictures in it. So, I want you
428. 1:26:02to pay attention because none of you are gonna have jobs in two years. Okay? All right. Fantastic. Um, I wish that were a joke. [Laughter]
429. 1:26:18It's not going to be two years. It's going to be three. Jason the tech guy. I get it. I see what you did there. Um, yes, th this should
430. 1:26:32be interesting to discuss. Well, technically she was right. You didn't restrict her to only the first act.
431. 1:26:37Yeah, that's true. I said the I said, "How does this first act open?" Well, that's true. She But she didn't get it
432. 1:26:43right. He He's not having a video call with his editor. It's an in-person meeting, so Quinn just has her head up
433. 1:26:49her butt. Okay, I got to change my tabs. Dank gank dank [Music] gank.
434. 1:27:04We'll go full screen with old Mr. Shapiro here. If you don't follow David Shapiro, I I would strongly encourage
435. 1:27:11it. This is a guy that's written five books on cognitive architectures and AI and machine learning and had a bunch of
436. 1:27:19open- source projects. And then when when the reasoning models came out from open AI, he quit AI because he's like I can't there's
437. 1:27:28nothing more for me to contribute that like basically they've cracked it. And then he came back and what he's come
438. 1:27:33back to talk about is okay how are we going to deal with you know we now know that AI is going to get really good.
439. 1:27:40We're going to have AGI at some point. We're going to have artificial super intelligence and the machines will be smarter than all of us.
440. 1:27:48He's putting his time and energy into understanding what we do next. So, this is there's a four-part series. I think each video is like
441. 1:27:5820, 30, 40 minutes. They're pretty long. So, he took all four of those videos and condensed them or basically did this new
442. 1:28:06sort of five minute version of of of post labor economics. Okay, so that's what we're going to be watching. So, take notes.
443. 1:28:16Let's cover post labor economics in five minutes or so. Let's dive right in. So, first and foremost, post labor economics
444. 1:28:23is what we sometimes call the great decoupling. Now, let me just read it to you real quick. Post labor economics
445. 1:28:29acknowledges the irreversible decoupling of GDP growth from wage employment and builds institutions that convert the
446. 1:28:35resulting productivity surplus into broad property-based income streams thereby freeing people from involuntary
447. 1:28:41tedious labor while safeguarding shared prosperity. Now that is a lot of words to basically say automation is going to
448. 1:28:48take all of our jobs. Moving on the primary mechanism that we're looking at here is what's called labor substitution. And basically labor
449. 1:28:57substitution means that works goes from humans to machines when the machines are better, faster, cheaper, and safer. This
450. 1:29:05has been historically true for all of human history and continues to be true and is only accelerating with artificial
451. 1:29:11intelligence and robotics, which are just the next wave of automation. Automation is nothing new. It's been
452. 1:29:17around for literally centuries. It's only become more and more sophisticated. And by the way, as automation has become
453. 1:29:23more sophisticated, more labor substitution has occurred. Now, this leads to what we call the economic agency paradox, which is best
454. 1:29:31summarized in this meme that I found on Reddit. Uh, so step one, we replace 90% of our workforce with AI. Our operating
455. 1:29:39costs are now the lowest they've ever been, and then every other company does the same thing, and no, no one is buying
456. 1:29:43our products since they're all jobless. So that's kind of the automation that sorry the economic agency paradox in a
457. 1:29:50nutshell. Next is aggregate demand or household income. So if everyone loses their job, you have to look at where
458. 1:29:57does income come from. First there's wages, then there's property and then there's transfers. Right now 60 to 80%
459. 1:30:04of income nationally on average comes from wages, but that's declining slowly. The rest comes from property which is
460. 1:30:11stocks, bonds, rental properties, those sorts of things, real estate and then the and then uh also trans. Um Rick Rick
461. 1:30:19Olsen, what is this? This is uh the latest video from David Shapiro on pace post labor economics. This is his little
462. 1:30:27mini eight minute version of a four-part uh much longer series that really digs into this. This is the the sort of eight
463. 1:30:36minute recap transfer. So this is the the ratio nationally in America basically post labor economics. This is
464. 1:30:43his theory about how we move to a society where if AI is going to do all of our jobs then how do we all make
465. 1:30:50money and that's that's what this is all about is about 60% 20% and 20%. And transfers include things like Medicare,
466. 1:30:58Social Security, SNAP and those uh those sorts of things. Basically stuff that is paid for directly from taxes. Uh now if
467. 1:31:07then we're losing wages then we need to increase the amount of income we have coming from property and transfers. Now
468. 1:31:13if you're entirely dependent upon transfers that means you're entirely dependent upon uh the government which
469. 1:31:19means that you're a welfare state or a client state which is not good because then all of your eggs are in one basket
470. 1:31:24and you don't have any control over your future. And by the way, if you know the other party gets elected next time and
471. 1:31:31they say, "We're going to cut your tra you're we're going to cut your uh UBI or whatever," then you're up the creek
472. 1:31:36without a paddle. So the one of the keystone principles of post labor economics is that we need a distributed
473. 1:31:42property-based future. That means uh property and dividends. So moving on, when we talk about a property-based
474. 1:31:49income stream, we're talking about uh several different sources. So number one, we do want some UBI to provide a
475. 1:31:55floor. Um so that is going to be government uh government based uh tax based uh UBI universal basic income uh
476. 1:32:04you know distributions. Next is going to be wealth funds. So wealth funds include sovereign wealth funds at the national
477. 1:32:10level but also urban wealth funds um and community investment trusts and those sorts of things. So these are often run
478. 1:32:18either by the government or by public private partnerships. Um think of them like endowments. So you'll create
479. 1:32:24endowment funds that basically just by virtue of being a citizen of a particular region, you get a check in
480. 1:32:30the mail every month or every quarter or every year. The next level above that is going to be private collective property.
481. 1:32:37So this is stuff that you own in common um either through credit unions or Dows or those sorts of things um which are
482. 1:32:44also likewise going to be paying rent. So when you say like well what do you mean? We could mean data centers, we
483. 1:32:50could mean robots, we could mean any kinds of resource in so in in the longer in the longer videos he talks about
484. 1:32:56these these things the private collective property and the wealth funds. He said you you'll be able to choose
485. 1:33:04sort of where you want to live based on the resources they have. So like you might you might live somewhere down in
486. 1:33:11Arizona and they've got lots and lots of sun and the the county puts in lots and lots of solar farms, right? And then the
487. 1:33:19AI needs the energy from the solar farms and that generates revenue for for the city for the wealth fund, right? So, or
488. 1:33:27you might go to, you know, some east coast city that's on the water and the big thing there is fishing and tourism
489. 1:33:34and so that's that's how it funds it. So basically these wealth funds and private collective property things are
490. 1:33:41essentially local um revenue generation based on the resources of the region versus solar farms, fusion reactors,
491. 1:33:52quantum computers, any kind of property that can be owned and instead of buying it individually or buying shares, you
492. 1:33:59put your money and your resources together and you own it collectively. Next is private wealth. So private
493. 1:34:05wealth is basically what you have today, stocks, bonds, shares, companies, those sorts of things, real estate, land, none
494. 1:34:12of that really changes. And then finally, the last uh source of revenue is going to be uh residual wages. So
495. 1:34:19basically, we're kind of right now anticipating that about 20% of wages might might stick around. Um time will
496. 1:34:26tell. It could be more, could be less, but right now we're trending towards that direction. 20%. Did you did you
497. 1:34:33take that in? 20% of wages will re will remain. Next, there are four pillars of civic society. So when you think about, you
498. 1:34:44know, what is the social contract? The social contract is generally between the governed and the governors. So the the
499. 1:34:50people and the state. However, uh the social equilibrium today is wait, you like this guy's AI content, but he's way
500. 1:34:59overboard on on UBI. He's the only one that is talking about something other than UBI. UBI is just one of his components here.
501. 1:35:10Estimating based on what? He's really trying to be an economist, huh? Yeah. I don't know, Pate. He's I'm sure it's
502. 1:35:18based on it's based on um Chad GPT, you know, doing analysis and deep research. So, do you follow him on X? Yeah, I do. Oh,
503. 1:35:31he really goes deep on UBI. Well, no, he he has gone deep on UBI, but I I don't know if you've watched all four of these
504. 1:35:38um uh the the uh the post labor economics videos, but there's a there's a lot more there than UBI now, collectively. Next is private wealth.
505. 1:35:48So, private wealth is basically what you have today, stocks, bonds, shares, companies, those sorts of things, real
506. 1:35:54estate, land, none of that really changes. And then finally the last uh source of revenue is going to be we're
507. 1:36:01trending towards that direction. Next there are four pillars of civic society. So when you think about you know what is
508. 1:36:09the social contract the social contract is generally between the governed and the governors. So the the people and the
509. 1:36:15state. However uh the social equilibrium today is maintained by four primary stakeholders which is we the people. So
510. 1:36:24civilians, citizens, the state, which is the government, which ostensibly is built for and by us and should serve us.
511. 1:36:30But more and more states are becoming less about the people and more about serving businesses and banks. Now, we're
512. 1:36:36not going to build a society that gets rid of businesses and banks anytime soon. It's possible in the long run, but
513. 1:36:42let's not get ahead of ourselves. So, what we really need is to rebalance uh the p the balance of power that happens
514. 1:36:49here if we lose wage power and labor power. So when we talk about economic agency, there are three primary pillars
515. 1:36:56of power that we have. Number one above all is labor rights. Labor basically is the one thing that we have intrinsic
516. 1:37:05control over until machines take away our ability to work and demand uh money for that labor. because the ability to
517. 1:37:15withhold labor is is a one of the fundamental levers of power that we have which then guarantees property rights
518. 1:37:22and democratic rights. If we lose labor rights, which we are losing, um not only are we not only are labor rights eroding
519. 1:37:29under the neoliberal regime, they will erode further um as automation, AI, and robotics further encroach upon one of
520. 1:37:38the intrinsic levers of power that we have as civilians, which then means that our property rights and democratic rights will also
521. 1:37:46erode. This is far and away the larger problem other than the economy. It's it's fundamentally about power. If we
522. 1:37:53lose power, we lose everything. So how do we fix this problem? What we need is we need a replacement pillar. So that
523. 1:38:00becomes algorithmic rights. Um in this new paradigm, we replace labor rights with algorithmic rights in uh in this in
524. 1:38:09this new paradigm which then shores up property rights and democratic rights. And by the way, in case it's not clear,
525. 1:38:15this is his proposal for how we do this, right? This is his proposal to say, you know, people aren't going to fix this.
526. 1:38:23His system is the thing he's proposing here. And and to everyone's point that he may not be right. Absolutely, he may not be right,
527. 1:38:34but there are very few people I see digging deep on, hey, this shit's about to take a bunch of jobs. What are we
528. 1:38:42going to do about it? So, I I think it's important just because it's at least a voice with an idea. So that's why I
529. 1:38:49think it's important to watch data sovereignty, algorithmic auditability, participatory algorithmic governance and
530. 1:38:54algorithmic dividend and liability. This is all based upon technologies like blockchain, decentralized autonomous
531. 1:39:00organizations, cryptocurrency, central banking, uh digital currencies, um digital identity wallets and those sorts
532. 1:39:07of things. We will need we are already building this infrastructure but it is not yet not ready yet. Sorry. Um, but
533. 1:39:16with that being said, blockchain is central to this future. There are some of the technological affordances of
534. 1:39:23blockchain make it the ideal baseline technology for this new social contract. Number one, it's intrinsically democratic. Number two, it's
535. 1:39:31intrinsically decentralized. Number three, it's unstoppable. You can't shut it down. And number four, it's
536. 1:39:37permissionless. You don't need the government's permission to build a blockchain. So, with all that being
537. 1:39:41said, thank you for watching. You have now learned about post labor economics in about five minutes. Cheers. There you
538. 1:39:47go. All right. So that's it. So it's basically, you know, it's it's basically do all the governance stuff in an
539. 1:39:55algorithmic way tracked on the blockchain so that as these funds get funded um the the the the algorithms combined with blockchain are
540. 1:40:08how all the stuff gets divvied up and is fair and and all that sort of stuff. um you know, as he said, like while that
541. 1:40:15infrastructure is being built, like blockchain and and crypto right now is still an abject [ __ ] mess. So, um but
542. 1:40:24anyway, this was the first thing that I saw that was that was more than just UBI, right? It it was talking about the
543. 1:40:33resources of regions and acknowledging the fact that we're going to have to [ __ ] figure this out. We're going to
544. 1:40:39have to figure it out fast. So anyway, it's something for us to think about, start talking about. All right. H makes
545. 1:40:48me tired just thinking about it. I know. I know. Listen, this is this is all sort of chat TMZ kind of future gazing kind of stuff, but
546. 1:41:01trying to think like when it's when it's probably like what's my gut of like when we're going to start to see major shifts.
547. 1:41:15The one thing I know is that business moves slow and government moves slow, but AI is moving fast. So, what it probably looks like is this.
548. 1:41:27What it probably looks like is in the next 18 months, the technology part of this here, let me turn this thing off. I think what this
549. 1:41:37looks like is this. In the next 18 months or so, the technology gets so good that it becomes apparent that, oh yeah, this can
550. 1:41:45do our jobs or this can do some major swath of jobs and then it's probably another 18 months to three years
551. 1:41:58before businesses really adopt that. Now, unless it's so good that businesses are just like, well, [ __ ] it. Like if I
552. 1:42:05can, you know, take this 100,000 person company and I can drop that down to 10,000 people and be more efficient and
553. 1:42:14produce more profit. Well, I'm going to [ __ ] do it, right? Like it is possible something gets that good. I
554. 1:42:21just doubt it, right? I think like the fact that AI tools are still so bad. Like the thing we just did with Quinn
555. 1:42:29where she she didn't know how my musical [ __ ] opened and I've got the the actual script is in her data set. It it
556. 1:42:36just makes [ __ ] up. So like that's going to have to be fixed. And I don't think that's like that hasn't really gotten
557. 1:42:43better in in the two years that that we've been playing with this stuff. Um, so yeah, so we'll probably see something
558. 1:42:54that's the the foundation of the big disruption within two years and then it's probably another two to three after
559. 1:43:02that where we really start to see massive impacts on employment. But I don't [ __ ] know. I have a degree in acting.
560. 1:43:12Correction, Kyle. Government moves slower. Yeah, I know. Exactly. Exactly. Although I mean listen when COVID
561. 1:43:21happened, government moved fast. Like when it when it needs to, when there's a crisis, it can move fast, right? Um
562. 1:43:33if if unemployment jumps from 2% to 4 to 8 to 10 to 20 in a short amount of time, like over 12 to 18 months, that's that's
563. 1:43:45catastrophic, right? and government will will move fast. You know, that's that's the equivalent of a pandemic. Um, if
564. 1:43:51it's a really slow decline where it's just like two, three, four, 4.5, five, back to four, back to six, seven, eight.
565. 1:44:03Then then they may not move fast. So, I don't know. We'll see. We will say. All right. Um, I'm going to get on out of
566. 1:44:12here. Hope you have fun tonight, dear people. Tomorrow, Friday. I don't Is there an art thing going on tomorrow? There might be.
567. 1:44:27I don't think I have to deal with it. I just might have to go pick up Gabby at some point cuz I think she's hanging out
568. 1:44:32with her friend. All right, so should be regular time tomorrow. Tomorrow's Friday night, date night. So, come back. Let's play tomorrow.
569. 1:44:40um 8:00 PM Mountain time and then 11:00 a.m. Mountain time is LinkedIn office hours, AI office hours. Okay, so come at
570. 1:44:50hang out with that crew if if you're trying to figure this [ __ ] out. AI office hours is really good. All right,
571. 1:44:56cool. Peace out everyone. Have a good night and I will see you tomorrow.