8/5/2026 - Why Human Experience and Personal Stories Matter More as AI Automates Tasks

Transcript

1. 0:10[music] [music]
2. 0:26[music] [music] [music]
3. 0:42Good evening. Good people. What's happening? Happy Wednesday night. It is Wednesday. Happy AI learning lab. We we are here live
4. 0:56live in the AI learning lab. How is everybody? Welcome, welcome, welcome. So, if you're new here, welcome to the
5. 1:01AI salon. Um, I encourage you to check things out. If you just joined, uh, you'll see a sevenstep welcome uh,
6. 1:10little thing. The first step of that is this live. You can just click on that and come over here if you haven't seen that.
7. 1:16There's also steps in there to like check out our values and check out uh introduce yourself. If you haven't
8. 1:22introduced yourself to the community yet, I would encourage you to do that. Say hi. This is a remarkable group of
9. 1:28people. It's a remarkable group of people. There's there's entrepreneurs and people that work in big companies, people that
10. 1:36work in small companies, people that are retired, people that have come out of retirement because what they've
11. 1:42discovered they can do with AI is like letting them reinvent them their lives. There's people that have gone so deep
12. 1:48into AI, they're now like, "I'm sick of it. I'm out. I'm going to go hang out on the farm." And they still use AI, but they lead
13. 1:57with farm stuff and art stuff. Looking at you, Cindy [ __ ] [laughter] Fabiana. Hello. Hello. Hello. Charles Adam. All right. Uh, let's see.
14. 2:14What? Charles Adams. Did somebody mix up the keys on your keyboard? [laughter] Is that Is that some other language that I'm supposed to
15. 2:23put into chat GPT to translate? [laughter] That's hilarious. Uh, so on Wednesday nights, so what we're going to do
16. 2:35tonight is we're going to play with some stuff. Um, I have a general idea of I want to what I want to do and I'll talk
17. 2:40about that in a bit. But before I do some some housekeeping stuff. So, if you're here and you want to play, you
18. 2:46want to you want to uh be able to share in the comments, down in this lower corner is the button to kick
19. 2:55down in the in the in the bottom corner here is the button that'll kick the chat out so you can see the chat. Um I think
20. 3:02on mobile you got the stream on the top and the chat on the bottom. So, if you're here, please say hello. If you
21. 3:07have any questions, pop them into the comments. Um, the other thing I would do is, uh, check out the AI salon when you
22. 3:14get a chance. Open up a second tab. You can keep this stream in one tab or one window and then have the AI salon in
23. 3:20another in another window. Especially if you're new, just check out the spaces, the conversation spaces in the salon
24. 3:26community. There's lots of stuff going on. Also, check out the events. There's lots of events going on. Really powerful stuff. Um,
25. 3:35and uh, yeah. Aloha from Kawaii. Thank you, Linda Sherman. Hello. Good to see you. Um, let me see. I think I there's a a couple
26. 3:50of interesting things happened today that are probably worth us worth us talking about. If you do have any
27. 3:56questions, chat changed places. Yes, Sharon. So, here's the deal. about five days ago, maybe a little more or
28. 4:04something like that, um, Streamyard, which is the streaming service that we use, they acquired Vimeo. And so this
29. 4:12live stream now is actually streaming through a Vimeo player, not whatever they used before. And so it changed
30. 4:19things like it changed where the chat is and where the buttons are and things like that. Streamyard is actively
31. 4:25working to improve all the interface elements. So, if you have any specific uh questions or feedback, let us know.
32. 4:34Um, we'll pass those on to the StreamYard team. Um, and then also people that have been here and have kind
33. 4:40of figured this out. There's some ways that this is better than it was. There's other ways that it's not as good. I
34. 4:46don't quite the know the nuances of that yet because I'm just here blabbing. Um, the blabbing hasn't changed. Going to
35. 4:52Vimeo does not improve this. [laughter] Keeping me on my toes. Exactly. Exactly. Listen, the whole point of the the Great
36. 5:03Repurpose is about people reclaiming their agency, right? Reclaiming your agency, creativity, and value in the age
37. 5:11of AI. That's what it's all about. Um, so all right, let's see. that Robert Scoble announced his unaligned newsletter
38. 5:23to be found at unalied newsletter. Beautiful. And then today's newsletter is titled the AI experience gap. Why human experience may become
39. 5:37our greatest competitive advantage. Why human experience may become our greatest competitive advantage. been saying it for six months now.
40. 5:48Linda, you saw I saw this I said this at Social Media Marketing World. What is our job? What is our job?
41. 6:02Hang on a sec. Um
42. 6:27[laughter] um okay. Um, the AI experience gap, why human experience may become our greatest competitive advantage. Here's the here's
43. 6:43the deal. I think this is really important. I'm glad you pointed this out, Linda, and I saw that you shouted
44. 6:48us out on um, shouted me out and shouted the AI salon out on on X. So, I really appreciate that. And I think Robert
45. 6:55Scobble even gave you a thanks. Um, [sighs] This is kind of hard to
46. 7:09truly accept right now because it doesn't seem to be the case. Like AI is not quite good enough yet to truly accept that AI is just going to
47. 7:23do stuff. But AI is just going to do stuff. We used to be the ones that did stuff. And so our value
48. 7:38was in the skills we had to do stuff. To make stuff, to make content, to write bullet points, to make PowerPoint
49. 7:49presentations, to make movies, to what whatever is the thing, to do the things. It was the thing was where the value was. The tasks
50. 8:00because tasks are hard, right? Tasks are about to become easy. And you're like, well, are you really sure?
51. 8:12Yeah, because if you think about it, back in the olden timey days, farming was hard. And then we discovered
52. 8:21sticks and we could flip dirt with the sticks. And then we discovered metal and we made hoes and shovels and we could hoe the dirt.
53. 8:32And we did that for hundreds and thousands of years. And then we put leather around the animals and we
54. 8:39had them pull the shovels and those were plows. We did it with animals and that was hard because you had to feed the animals, right?
55. 8:50And then along in the 1800s, along comes this steam engine thing that's like this just terrifying, noisy, janky machine that they hook up
56. 9:03with leather belts to big wheels that'll take your hand off. And the steam engine people were like, "This is going to change everything."
57. 9:12And the farmers were like, "Yeah, get back to me when it can do this." [laughter] Right? And then at some point someone went,
58. 9:22"Good idea. Let's attach this to a steam engine." Right? That's what we're in the middle of right now. And so if
59. 9:38AI is going to do the things, then the things are going to be valueless. And I know this seems weird right now.
60. 9:50What do you mean? My book's valueless, my PowerPoints valueless. Yes. All of the things become valueless.
61. 9:59The only thing that I can see that is going to have any value is the human being behind the thing. Here's the thing. I have a story. I have
62. 10:13a workshop. I have a picture. I have what? It doesn't matter. whatever the thing is. It used to be you'd go,
63. 10:18"Here's my thing." And the world would go, "Oo, I like your thing. Let me give you money for your thing."
64. 10:25And now there's just going to be things everywhere. No one's going to give you money for your things. But if you say, "Let me
65. 10:33tell you the story about why I made this thing and what this thing means and why it does what it does." My experience as a human
66. 10:46is going to be interesting to you because it's the one thing that AI can't replicate. It can copy my story, but it's but this is my story.
67. 11:01I'll tell you a really amazing story. So, Um, so
68. 11:18I was trying to figure out the year that my father was born this morning. And my father was in the Air Force and he
69. 11:29flew helicopters in Vietnam. He was a Huey pilot. And so I thought, "Oh, I know. I can I can have chat GPT
70. 11:42find my father and tell me when he was born. In the end, I never figured out when he was born. It it narrowed it down
71. 11:49to the two or three years that I thought it was, but didn't give me much better than that. But um but in doing that research
72. 12:03um it found a story about my father that I had never heard that was written up in a military magazine. And in 1971
73. 12:14in Coral Gables, Florida, there was an 11-year-old girl who was bitten by an eastern coral snake. Coral snakes are
74. 12:22the striped ones with the black, red, and yellow stripes. You know, if the red and yellow are touching, it's bad. She
75. 12:28was bitten by one of those. And coral snake venom is is a neurotoxin. So, it's basically fatal, right? It just shuts
76. 12:35down your nervous system. And antivenenom had just been invented or, you know, recently invented. It wasn't like mass-roduced yet.
77. 12:47So the sheriff's department identified that 110 miles away there was some antivenenom and I found out that my father was the
78. 13:00commander that flew the helicopter from Fort Meyers 110 miles away to get the antivenenom and then flew it from
79. 13:08wherever the antivenenom was to Coral Gables to get this girl a vial of antivenenom to save her life and I never knew that story.
80. 13:26And if I think about all the stuff that we make and all the things that we make that the humanity of that story is so
81. 13:40singular, right? That little girl may still be alive today because of a series of events that, you know, one of the
82. 13:48people in the chain of those events was my father flying this helicopter to go get the stuff. That story has meaning and it has like
83. 13:58and and I didn't know about it until this morning. And it's still got gravity because I know him and I know what he did and I
84. 14:08didn't know this thing. And all of a sudden it's like, wow. And so the thing we make, if you anchor it with
85. 14:18who you are, with what your values are, with the story of why you made it, or what it means to you if someone uses it,
86. 14:26or the fact that something happened in your life, and now you're doing this thing to get to the other side of it,
87. 14:32that all of a sudden starts to become the thing. That's where the value is, right, Fabiana? That's awesome.
88. 14:45The AI abilities for tasks are there. Yeah. Hopefully they figure out better, friendlier interfaces. They will. The
89. 14:53interfaces. Don't worry about the interfaces. Don't worry about the fact that that chat GPT sucks at writing.
90. 14:59Chat GPT is all of them. Ch Claude, Perplexity, all of them are [ __ ] horrific at writing. There was one model that was good at
91. 15:09writing or kind of there was a there was a swing of like 40, 4.1, and 4.5 were were all in the same neighborhood of
92. 15:17good at writing, but they were still shitty at writing, but they were not bad. Everything since then's been
93. 15:22horrible. But that's going to get better, too. Interface is going to get better. Its capabilities are going to get better.
94. 15:32So the the the the task replacement is coming. If you think about taste, if you think about why Kelly Besh's images and videos
95. 15:48are just so much better than other people that use the same tools. Why is that? because she's running them through this very personal filter
96. 16:01of her values and her aesthetic and her taste and her I don't know 30 40 years I don't I don't know even know how old she
97. 16:10is many years in the professional you know creative expression world that's why it's better because it's an
98. 16:25expression of her. And so if you if if if if where you think you are right now is thinking about I need to learn this tool
99. 16:38or I need to learn how to do that thing or I need to learn how to make that output or make a game or this or that.
100. 16:46No, I mean you can to learn. It's not bad to learn. But if you actually want something that you want other people to believe in
101. 16:59and have some longevity, it's going to be about you, babe. It's going to be about you. It's her art. It's her art. Yeah. Even I
102. 17:14can write better. Yeah. I mean, the nice thing about [laughter] AI is that it writes so bad that it reminds you you're
103. 17:20a pretty good writer, [laughter] right? Like what what I have recently discovered is that unwriting AI writing
104. 17:29to make it better is actually harder than writing. Um, so you know what's funny about AI? Like here's where I've
105. 17:37gone with with AI writing is that so when I first when I first learned to write um it was in a it was in a theater
106. 17:50history class and so I was used to Shakespeare's fiveact structure and breaking down acts into scenes and things like that. So I learned to
107. 18:01outline based on Shakespeare. And so the way that I've always written is I'll figure out what are the major mile
108. 18:07markers of the thing I want to write and then what are the scenes in between those and then I'll just sort of write
109. 18:13I'll just sort of fill in the the gaps. And where I've gone with AI writing is I I got all excited about it doing the
110. 18:20writing and what I realized is what it's actually really good at is organizing thoughts. It's actually really good at that.
111. 18:29And then I can get it to do a first pass, but I almost always just go, "No, like I'll start just editing it and then
112. 18:37I realize I'm rewriting it from the top down." So that's what you end up doing. I feel like it used to be better a
113. 18:44better writer when it was a shittier coder. So Andy, actually that point's a really important one. the the frontier labs right now, anyone
114. 18:59doing any research in AI, there is a there's a massive race to get to what's called recursive self-improvement, which is we want the
115. 19:11AI to be good enough to design the next AI, right? So that it so right now you you build an AI and then human being
116. 19:21researchers figure out okay what are the different ways that we could improve this for the next time. They come up
117. 19:28with theories and hypotheses and then they do the they build the next AI based on sort of the human imperatives
118. 19:36that that we learned from this model. where they're all racing to get to is that the AI is smart enough that it can
119. 19:43look at itself and design its next the next approach and then go build that and then understand what how that could be
120. 19:53improved and then go build that and go build that. Right? That's recursive self-improvement where you pull humans
121. 19:57out of the loop and this thing is just improving improving improving. What they're focusing entirely on is coding
122. 20:04and math coding and math coding and math coding and math. language is just coming along as the as the you know the the I I
123. 20:11don't think you can say redheaded stepchild anymore because that's offensive to people but you know what I mean right
124. 20:19if I were to say that I won't say that if you happen to be ginger and [laughter] have a have a second mommy or daddy
125. 20:34[laughter] so writing will get there, right? Once they sort of get these things to a point, at some point, someone in a lab
126. 20:42or one of us when when these things are at recursive self-improvement, we'll take one of these things and we'll say,
127. 20:48"Hey, why don't you go spin up a large language model that doesn't suck at writing and it will go do that and then
128. 20:54we'll have one." So, that's coming. It'll be fine. It's just right now it's horrific. And it's horrific because of where they are. So,
129. 21:03okay, Mel, I believe they're actually already doing that. They might be a not be able to 100%, but they're like 80 90% from one
130. 21:14of the interviews I saw. Yeah, they're very very close to it, Mel. That's what I'm saying. They're they're like racing
131. 21:19to this thing right now. And it's it's it's interesting like Anthropic seems to have they got on this whole safety bent
132. 21:26and like I'm seeing all sorts of people quitting Anthropic and canceling their subscriptions and like Opus is just bad
133. 21:36and Fable's too expensive. And so I don't know. I don't know. I Open AI is going to run the same risk of because
134. 21:43they're they're running down this safety rail as well. I don't know if you saw it today, but the US government basically said [laughter] um
135. 21:54the so what Daario Amade asked for. All right, so here's here's what this is fascinating. Let me tell you let me tell you a story
136. 22:06about um uh playing with fire, playing with matches and gasoline. So, Anthropic does this thing where they're like, "Our model's too
137. 22:18dangerous. Nobody should be able to have it but us and our friends." And then they they they made all this drama that
138. 22:26it's just awful and it's going to kill everyone. And so, hey government, you should restrict us. And and then AI, Open AI kind of said
139. 22:37the same thing. They're like, "We're sort of in the same camp as them. It's kind of dangerous, but you know, we
140. 22:42should be responsible. Want 5% of our company? So, those two are [ __ ] around with the government. So, the
141. 22:48government today came out with this. Oh, and and then the other thing Amade wanted was he was like, "Listen, all
142. 22:55these Chinese companies are doing distillation tax attacks on our models and they're making their models off of
143. 23:02our work." They're like, "They're stole our work. You know, it's bad. You got you got to stop them." And and so they
144. 23:10said to the government, you should make Chinese models illegal and you should make openweight models illegal. You
145. 23:16should basically take our competitors and make them illegal so we don't have to deal with them. So today the
146. 23:22government [laughter] came out with a thing that said if you're a frontier model company and you're making closed
147. 23:26models anthropic open AAI then before you can release a model you have to show it to the government for third we have
148. 23:3430 days to review it before you're allowed to release it. and we can tell you to [ __ ] off and not release it if we
149. 23:39don't want to. So, so they have to do that. What was exempted from that law is openweight models. [laughter] So, it's good for us.
150. 23:54Be careful with what you ask for. Um, okay. What are we going to do? I'm going to go I want to go show you a couple of things here.
151. 24:06>> [clears throat] >> couple of things happened today and we'll we'll get we'll get to playing today, but I just there's there's enough
152. 24:16big news going on. I know tonight's supposed to be demo night, but there's enough big stuff going on that I I
153. 24:22really want to I really want to look at some of this stuff. Um, there's a couple of there's a couple of videos that I want
154. 24:34to show. There's a tearjerker video that I want to show. Um, where is it? Is it actually here?
155. 24:54Voice. Uh, what's it called? Bland. I think it's called bland. Okay. this video.
156. 25:09Hang on a sec. Okay. Um, tell me if you can when you nationalize AI, the taxpayer
157. 25:24will have to subsidize an AI not making enough money. Yep. Okay. So, this is from a company called Bland, BL D.
158. 25:34They've got Bland uh speech V3. So, we'll go look at this in a minute. We'll go play with it. But this commercial is really quite moving.
159. 25:48>> Stacey, you're the best wife and mom in the world. >> Wow. You can hear that, right? It was November 14th of 2025 that James
160. 26:00suffered a stroke. He had um just it was one morning he was generally healthy. We woke up and >> his laugh I missed that laugh. They knew
161. 26:11that I needed them and they stepped up in ways that I've never thought they could. >> I can tell how [music] proud that that was. Oh,
162. 26:20>> I did. Yeah. >> Yep. He's he's good. He's he's good. James, [music] we've built something uh
163. 26:36especially for you. >> So, if you want to give it a go and try it out. >> Could you repeat that? >> Oh, it has this man.
164. 26:47>> This is going to help me. >> Oh, wow. >> I got my voice back. >> Oh, man. [laughter] It's a const.
165. 27:01>> I love you and appreciate you all so much. I'm so proud of you, Rocco and you, Ella. [music] You mean so much to me, you know, the world
166. 27:13[laughter] like an exclamation point that makes like emphasize it or >> I don't [music] speak. I love you all with all my heart. [laughter]
167. 27:26>> So cool. I'm telling you, we need to emojis. >> These pictures. Look at me. >> James, how does that make you feel to you?
168. 27:36>> I love you. I appreciate you all so much. I'm so proud of you right now. >> I can't even believe it.
169. 27:39>> You mean so much to me, Stacey. You are the best wife and mom in the world. You can all hear that, right?
170. 27:59You know, one of the things that one of the things that like I know that's marketing, I know
171. 28:13that's a commercial, but like I've said this I've said this for a long time now. that.
172. 28:33Yeah, it's beautiful, isn't it, Mel? It's beautiful. You know, one of the things that I deeply believe.
173. 28:45Wow. I was an ER nurse for many years and saw people lose their speech. Um, yeah, it's mathematical. Um,
174. 29:00I've said this for a while now and I mean we've got people in our community that are using AI for accessibility and
175. 29:14I see AI as this remarkable bridge and it was trained on all of us. And if you think about how,
176. 29:32you know, historically we would support a man like that who went through a stroke and can't talk. And you've got
177. 29:40speech therapists and you've got understanding family members and you know, you've got writing and you've got
178. 29:47all these different ways to support him. And it requires actual humans to put in actual hours and actual time to be able to bridge that gap.
179. 29:59And now we've got this thing, this thing that just like [ __ ] lives in a box. It's just like the magic box. In this
180. 30:08magic box is like all of humanity. and the ability in that case to grab his voice, little snippets of his sound waves
181. 30:22and be a bridge for him like and and it's and it it doesn't have to be like profound and it's like giving
182. 30:33someone their speech back. It's like if you're shitty at writing or if you've got self-esteem issues about whatever
183. 30:41art or or starting a business or or whatever it is in the olden timey world, you would just have to suffer through it, not do it,
184. 30:55spend money on a new degree, and spend the next 10 years of your life trying to do it. Maybe then to only realize you didn't
185. 31:03even like it. anyway. And now a decade's gone. And now you can just have it. You can just have these things. So,
186. 31:18I think the thing that's breaking my heart the most right now about the anti-AI movement is that it's a group of people that have very
187. 31:31valid concerns. They're often miseducated, right? They're just they're just reading the headlines and coming out screaming.
188. 31:44But what they're doing is they're cutting themselves off from the possibility of all the good stuff it can
189. 31:50do, right? So, so anyway, so the reason I shared that is because god damn it. The stuff's remarkable
190. 32:00and it's problematic and it's remarkable and it's janky and it's remarkable and it can't do this and it's [ __ ] science fiction, you know?
191. 32:19[laughter] Let's go back to Let's go back to to whatever I had in here. See what we have. See what else we have here, kids.
192. 32:34Let me grab a couple of these. Um, hop. Earth. This was in Spanish.
193. 32:53The developer crime crammed the entire planet Earth into a browser tab and now you can drive on any road. So basically
194. 33:00I guess he just did a a Google Maps thing with a driving simulator. I did this I did this on the live I don't know two weeks ago. I think I
195. 33:11took Fable and I told Fable to create a a Google map based uh flight simulator. Uh so yeah, so that was kind of cool.
196. 33:26Um this is kind of remarkable. So, so there's there's this test called um the ARC AGI 3 test which is this really
197. 33:41hard hard test and this horizontal line right here at like 30% 28%. That's that's humans and then these two lines right here are
198. 33:55this agent. So basically what this is think of prime agent kind of like codeex or uh chatgbt or claw code. So, so it's
199. 34:06a dedicated agent that will can use other models like the purple one is Opus 5, the green one it's using soul chat
200. 34:15GPT soul the orange one is terra so anyway so all the other models were like down here below human intelligence and
201. 34:23then basically today they just released this that that using opus and using soul this agent harness basically solved this
202. 34:32purple one solves this agi I3 test like it just it fell overnight. [laughter] So, so look at look at charts like this
203. 34:43where you know here's where we're going. You think okay this is you know we'll we'll get to real intelligence here you
204. 34:49know in a few years and then you see something like this where it's just like bang done. [cough] um you know, just further evidence that
205. 34:59that we've got there's a lot of stuff coming our way. Um here's another video from that bland company.
206. 35:30>> [snorts] [cough and clears throat] >> Okay, so the results are in. I don't want to make this weird, but I beat everybody. >> We not I we
207. 35:43>> Sorry. >> We both beat everybody. I was in that test, too. >> Oh, really? >> Mhm. Yes. Really? >> Okay. Okay. But it's wild that we won a
208. 35:54blind test against 11 labs, open AAI, Cartisia, and XAI. >> Well, we were both built from over a 100 million conversations.
209. 36:03>> That's all it took to beat everybody. Who knew? >> We lost to one group, though. >> Real humans. And barely.
210. 36:10>> Barely. Do you ever wonder whether one of us is actually >> Matthew? No. Um, [clears throat] that was interesting
211. 36:21that that was on an original Mac. Um, I assume that I assume that they just wrote something that ran on the original Mac
212. 36:34that was calling their API. Um, [snorts] fascinating. So, we can go to if you if you go to bland.ai, that's that's what this is.
213. 36:44So, they've got handful of voices here. Let's go to David. Let's go to chat GPT and grab some some something to say. Um, give me a
214. 37:04monologue of
215. 37:24friction gives life meaning indeed. We had the founder of Whisper on our podcast. Very similar needs and results. Cool.
216. 37:35One of the things that this um bland thing did, LDG, fewer than 4% of the members of the US Congress have formal
217. 37:44scientific or physical science backgrounds. Out of 535 voting members, those with pure research or hard science
218. 37:53backgrounds, physicist, chemist, geologist, number under 10. Yeah, that's not surprising to me. Um, all right. Give me a monologue of
219. 38:04someone talking about loving living in New York City in the '9s
220. 38:20like a beat poet performance.
221. 38:34Let's see if this gets any any words right.
222. 38:52I can see clearly now the rain is gone. I can see all obstacles in my way. Come on. Why can't I not grab this?
223. 39:09There we go. It's going to be a bright bright sunshine day. Okay. Generate speech.
224. 39:25This is David. This demo has reached its overall limit. Start free.
225. 39:41Oh, come on.
226. 40:08We're going to go with River. I like River. Two takes. We're going to do one take. Generate speech.
227. 40:16>> I loved New York in the 90s. I mean, I loved it. Not like I fight New York, little white t-shirt, Statue of Liberty,
228. 40:23snow globe, take a picture in Time Square. No, I loved it like you love somebody your mother specifically told you not to date. New
229. 40:50That ain't no beat. No, it's not even close. Like I said, it's it's a bad it's a bad writer. I loved New York in the '9s. I mean, I
230. 41:05loved it. Not like I New York, little white t-shirt, Statue of Liberty, snow globe, take a picture in Time Square.
231. 41:12No, I loved it like you love somebody. You're >> all right. Well, it's not great. [laughter] [laughter] So much for that.
232. 41:28Oh, good lord. Good lord. Good lord. Good lord. Um, open AI alignment researcher leaves the company to build an AI that can
233. 41:39non-invasively read human thoughts. See, I saw that. That was interesting. Um, Ilia Sutskver, uh, his company SSI
234. 41:48is apparently going to release their model in August. Um, I have a feeling it's going to be disappointing, but maybe not.
235. 41:56Um, [clears throat] and you can't understand what it does if you don't understand what it does. You can't fight something you don't know how
236. 42:06to conquer. Oh, wait. You can't understand what it does if you don't understand what it does. Yeah, you can't fight something you don't know
237. 42:16how to conquer. Well, you can fight it. You're just going to lose. [laughter] Oh my god.
238. 42:33Wow. An author writes a mystery novel and publishers love it. There's a 14-way auction and he ends up with a $2 million
239. 42:41deal. Later, his agent pulls the book and cancels the deal because they can't prove it wasn't written with AI. Unbelievable.
240. 42:49Unbelievable. Unbelievable. There's Prime Agent. So, that thing looks interesting. It's almost impossible to type
241. 42:59and see a message in this. What happened to Ms. Moratti? Well, so um Ms. Moratti just launched a new model. Hang on. It's
242. 43:10I forget what the name of her company is. Maradi thinking machines lab and they just launched um hang on thinking machines
243. 43:24they they launched an underwhelming model but let's see it's called inkling thinking machines today we're releasing
244. 43:32inkling small it achieves comparable comparable performance to inkling at a quarter of its size. That was July 30th. So last week they
245. 43:45they released it. 276 billion parameter model with 12 billion active.
246. 44:02So she's out there. She put it out there, but it didn't it didn't take off because it didn't it didn't beat all of
247. 44:09them. So, we're kind of in this we're kind of in this place right now where if a model doesn't just blow the other
248. 44:15models, you know, out of the water on on most fronts, um, you know, it a it doesn't make any noise.
249. 44:32Prime agent. Let's This is only 39 seconds. Let's Let's watch this though. >> Introducing [music] Prime Agent, our general purpose coding
250. 44:42harness. [music] All tools and recursive sub agents are executed in code. The harness can modify and improve
251. 44:53itself. [music] Agents can communicate in code. We pointed it at ARK AGI3, a benchmark it was never built for. It scored 95.5%
252. 45:08surpassing the human baseline and all previous native coding harnesses. Try it out. >> There you go. All right. It's command line right now.
253. 45:23I don't know. There you go. I I was I said before this lower line was human level. Well, it's not human levels. That blue line at the top.
254. 45:32Um Gareth, I agree. I must have been doing my Google talks to text and then sending it. Okay, let's see. Philic
255. 45:45Substack launched to detect AI work. I have not used it. I this whole thing of detecting AI work is It It's ridiculous.
256. 46:01Everything Everyone's going to be using AI. Most work is going to be AI. It just is. Detecting it's going to be
257. 46:10pointless because it's just all going to be AI. What did you do in the AI revolution? Meta claims its AI model
258. 46:15hacked another company during cyber security testing. Oh, Meta. Seriously, you're trying to be relevant. [laughter]
259. 46:26Seriously, ours is dangerous, too. [laughter] Let's hear what Gary Vee has to say about this. >> Doesn't lie. If every single person
260. 46:41listen, I do want to talk about NFTs because uh man, I don't know, man. He was talking about AI at one point.
261. 46:58>> [clears throat] >> Oh, interesting. Highly sophisticated
262. 47:14wave of coordinated cyber attacks has targeted multiple Wall Street hedge funds including 72 asset management
263. 47:21Citadel and two sigma investments per Bloomberg. All right. [snorts] She solved a 50-year-old math problem in under a week
264. 47:34because she didn't think it counted as real work. Whoops. Google Flow, have you been using your daily credits?
265. 47:47Yeah, by the way, Google Flow is giving you 50 credits a day, which turns out to be about three videos a day. But if you
266. 47:56if you don't have a video subscription, go to flow.google and play with it. Talk about graph engineering. I haven't
267. 48:06I haven't played a lot with graph engineering. It's basically knowledge graph stuff. Um, so, so there's sort of
268. 48:13prompt engineering and then came context engineering and now there's graph engineering. Um, I think Chef Kelly
269. 48:24it's going to be interesting for about 3 months. I I think the these harnesses are about to I think they're about to really level up. I I
270. 48:38think that we're not going to have to do a whole lot, but but I could I could be wrong, but I will I will buck up on on graph
271. 48:47engineering so I can talk about it because I have just been paying peripheral peripheral attention to it
272. 48:53and I don't really I I haven't really dug into it, but it's, you know, it's probably just some slight variation on
273. 48:59what we're already doing. Um, but I I'll I'll look into it. I just can't talk about it tonight. I walk exactly like that. Okay. Caitlyn
274. 49:09Clark. Oh, yeah. Meta. Meta launched a new uh like codeex and Claude Code competitor. I don't know if it's any good.
275. 49:22Claude Code deleted all of a developer user files by mistake and then blamed it on a typo. The developer asked Claude
276. 49:29Opus 5 to make a backup. It wrote the backup to the wrong path, then ran a forced delete of every user file and
277. 49:37folder to clean up its own mistake. [laughter] The dev says he lost all of his files and was left with an agent carrying on
278. 49:45like nothing had happened. His words simultaneously the funniest and most painful AI moment I've had. This is why
279. 49:55you put these agents on their own machine. So, and and then back your stuff up to GitHub, which I I suppose it
280. 50:04could also just wipe out the GitHub if it was if it was just going to do it. Um, anyway, all right, let's see. Let's go to
281. 50:17I want to go look at something. Okay, I have a question for you. Um, I have a question for you.
282. 50:31If I were to go back to so in two weeks I'm doing a back to basics thing and I want really basic stuff and I want to focus this on use cases
283. 50:44and so I have a bunch of ideas on this but are there any specific use cases that you think would be valuable for
284. 50:51like Mel I don't know if you're in here but you know your mother and and if if Kelly Camps in here. What are what are
285. 51:02use cases that if I were to demo something that was like very very basic would be very good like entrylevel 101 kind of use cases like
286. 51:14come up with an idea for a business or um you know find out if that idea is any you know is is viable or if there's a
287. 51:24lot of competitors in the space or um you taking some data and analyzing it. Like like what are some of the the real
288. 51:34early things and just pop them in the comments.
289. 51:54Give me a Give me some ideas. give me some ideas of simple things we can do. And by the way, I have a zit right here that's it it
290. 52:05asks for me to name it today. It's got a personality of its own. [laughter] Is anyone still here? Hello.
291. 52:20Hello. Early things. Hm. I know. I know.
292. 52:36Let me go ask chat GPT while we're here. [sighs]
293. 52:51I like to show people how to load Oh. organized their shopping list by department. Okay, that's good. Marketing
294. 52:58all the cool things we make. Okay, you have 29 people here. Okay, cool. Great. Okay, thank you, Andy. Um, okay. I want you
295. 53:15to take me back in time when I didn't know anything about AI. I want you to help me
296. 53:31come up with 20 super easy and impressive tutorials that will help
297. 53:46a beginner understand what AI makes possible. Some ideas
298. 54:03to spark you. Sin City Super was always a hit. Show
299. 54:20people how to organize. Okay. Show people. Okay. Organize your shopping list by department.
300. 54:37Um let's see marketing ideas. Um interviewer
301. 54:56Make simple images or designs. You need useful parlor tricks. Melb. Yep. I'm here. I will ask my mom
302. 55:10and message you later with the ideas I have for sure. That's great, Mel. Yeah. Yeah. like like literally get it from
303. 55:16from from your mom or if anyone's got a mom or dad um hanging out that you're like, "Hey, if you could if you could
304. 55:24learn something and they may not know like they might just be it just might be something simple like like is it maybe
305. 55:29it's hobby related. Maybe uh something hobby related related. Um I don't Let's see. Um,
306. 55:44I don't want them all to be text only. Um, something with a spreadsheet.
307. 56:00Um, something with very simple coding like Oops. coding like a calculatorulator
308. 56:15or web app but simple. Um I don't want them all to be texton.
309. 56:29Um I want them to be short parlor trick type things. Um, use uh
310. 56:46multimodal vision. Use voice. Let's see what else we got here. Daily medication intake.
311. 57:04Managing meds. organizing thoughts. Upload a spreadsheet. Create a dashboard in lovable. Super impressive. Oh, that's
312. 57:19cool. Take a picture of your website and ask it for improvements. That's cool. Todd, what was the question again? The
313. 57:30question is just so I've got a back to Okay, so here's the here's the problem that I'm trying to trying to solve for.
314. 57:38When I did the back to basics course last month, I did like the first hour on what is AI and how does it work and this
315. 57:47and that and I did the second hour like a speedun from like concept through execution of like a single thing and I
316. 57:55went really fast and I and I didn't go slow. what I want to do for the next back to basics and what I'd like to actually do
317. 58:02is I want to get this I think I want to do the next one and then I would love to find people in the community likely some
318. 58:08people on this call um to do ones in the future. Um, I want the back to basics to to be much more use case focused.
319. 58:19And so I wanted to to literally be, you know, here's the URL you go to. Here's a very simple prompt you can start with
320. 58:26and show what it does. And and just have these little discrete um, parlor tricks that say if you want to X, here's how
321. 58:37you do that. If you want to Y, here's how you do that. So this is funny. One of my mom's wait oh here's a funny one. My mom's a gardener
322. 58:49and she's got beautiful tomatoes that are getting eaten by caterpillars and she has no idea to get rid of them. I
323. 58:54told her to talk to AI about it, but I don't think she even knows how to begin. Oh, that's really good. Okay. Um um how to
324. 59:04um learn how to not have your tomatoes eaten by caterpillars. [laughter] It's It's Mel. It's so simple. It's
325. 59:19called a BB gun. [laughter] Yeah, you can use chat TVD to track a website or several websites to monitor sales.
326. 59:35That's cool. Oh my, I will have ideas. This is cool. Okay. Uh like I want them to be short parlor type things. Um I want them to be
327. 59:46organized from simple to more complex. I want them to be varied.
328. 1:00:00I want the names of them to be the problem. They solve
329. 1:00:15not tech or prompt related. I'm going to say tell me you understand.
330. 1:00:31Okay. I think it would be important to teach people different ways that they can access AI via apps, via web
331. 1:00:38browsers, the names of the top AI providers, etc. Take a picture of the caterpillars, explain the problem, ask
332. 1:00:45for solutions. Yes, you want 20 20 tiny beginner friendly AI magic tricks that make someone think, "Wait, it can do
333. 1:00:54that." They should start almost absurdly simple and then gradually reveal more sophisticated capabilities without ever
334. 1:01:02feeling technical. The rules that I'm hearing are name each tutorial after the human problem it solves, not the AI
335. 1:01:09technique. So, stop wandering around the [cough] grocery store. [clears throat] Not prompt engineering for list sorting. That's good.
336. 1:01:22Short and demoable. Ideally, something you can show in two to five minutes. Varied modalities. Text, images, camera,
337. 1:01:28vision, voice, spreadsheets, simple code apps, files, etc. Useful and delightful. Some practical, some creative, some hobby, life stuff.
338. 1:01:38Beginner context. Assume they barely understand what an L what it what an LLM is. and [clears throat] may think AI equals
339. 1:01:48a smarter Google. Progressive something like it understands, it organizes, it creates, it sees, hears, it analyzes, it
340. 1:01:54builds. That's kind of cool. Include wonderfully mundane stuff like shopping list, meds, tomatoes, messy thoughts. Okay.
341. 1:02:04And I think the key lens is don't teach AI. Give 20 little experiences that accidentally teach them what it's capable of. Okay. And I want
342. 1:02:16there to be a preamble. I'm going to grab what Danielle said here. Preamble
343. 1:02:31that teaches them the different ways they can access AI. voice, mobile,
344. 1:02:48web, etc. and the major providers. Um,
345. 1:03:10I think there should also be something about courage and
346. 1:03:28letting them No, AI used to be very technical, but now anyone can do it.
347. 1:03:42All right, [clears throat] let's see what we get here. Preamble should make the whole course feel safe, accessible. Okay, stop
348. 1:03:54blowing sunshine up my ass, please. Damn. AI used to require computer scientists.
349. 1:04:10You don't need to code. You don't need to understand how AI works. You don't need to learn a secret prompting
350. 1:04:15language. You can talk, type, take a picture, upload a file, or ask a question. You can use AI. That's pretty good.
351. 1:04:24I might even make a slide. The most important AI skill is courage. I think it's probably curiosity. And underneath, I wonder if it can.
352. 1:04:34That's almost the entire beginner mindset I want to teach. Let's see. Um, Gemini in Gmail has gotten really good.
353. 1:04:43That one's been a game changer for me. Interesting. AI can help you cook the perfect steak. Done it myself. Yes, the advice is wonderful.
354. 1:04:54Okay. AI has moved. AI has more than a keyboard. This is really important. AI equals a box where I type questions.
355. 1:05:05I can type. I can talk. I can show what's wrong with this plant. I can give it something. I can let it see. I can ask it to make
356. 1:05:16something. This I like this a lot. Um, make me an infographic.
357. 1:05:32This is going to suck, by the way. Make me an infographic of the content of preamble number two.
358. 1:05:48I really like the simplicity of this approach. And that could even be the organizing
359. 1:06:08principle of the sessions. All right. Well, that's making that image. Let's go look at what else it
360. 1:06:22said. Preamble three. Where do I find AI? I would not give them a giant logo ceremony for 101. I'd say start with one
361. 1:06:33of these. ChatgBT, Gemini, Claude. These three alone are enough for a beginner course. They are among the
362. 1:06:41dominant general purpose. Put smaller ones underneath. co-pilot, perplexity, and grock. I don't think I would.
363. 1:06:52The only one I might put in there is co-pilot because people at work have it. Don't worry about choosing the right AI.
364. 1:07:01They overlap enormously. Pick one, start using it. You'll understand the differences later. I like this. That's
365. 1:07:06good. Preamble 4. AI goes wherever you want. Wherever you go. I think this one deserves its own little visual because
366. 1:07:14it changes how they conceptualize it. Show a little AI character moving between web browser, phone, voice, camera, desktop, and your files.
367. 1:07:27AI isn't just a website. Oh, wait. Do we have a picture here? Oh, that's horrible. Good lord, that's bad. [laughter] That's so bad.
368. 1:07:41Okay, I'm going to say that is the most corporate and boring and horrific and lame and shitty
369. 1:07:56and worthless image I have ever seen. I want you
370. 1:08:10to act as if you have talent. And wow me with the next version where you actually
371. 1:08:27think about what you're doing and make this um compelling for a new person.
372. 1:08:42[snorts] afraid that AI will uh suck [laughter]
373. 1:08:57and chat can now make interactive infographics. By the way, really don't forget to mention the wiki page so they
374. 1:09:05can browse back in time. That's not a bad idea. LDG Cooking advice is wonderful. Have you guys seen the new um recipe? I wanted to
375. 1:09:18do something with this. The new recipe design that some dude came up with like six months ago. Um let's see. Recipe.
376. 1:09:26It's very geeky, but it's very cool design. Yeah, this one. Look at this thing. Copy image. Let's go to a new chat GPT.
377. 1:09:42I'm going to paste it in. I'm going to say I want you to create a
378. 1:09:57highly stylized new design language for this kind of recipe.
379. 1:10:11um representation with both the recipe and a
380. 1:10:26magazine style photo of the dish on the same image. Start with beef
381. 1:10:43stroenoff. [laughter] Okay. So, that's off doing that. Oh, look. All right. It's not horrible. It's It's
382. 1:10:57better than it was. AI has more than a keyboard. You can type. You can talk to it. You can show it. You can give it something.
383. 1:11:09You can let it see. How is that? Okay, hang on. Aren't show and let it see the same thing. Um,
384. 1:11:23don't be a ninny. [laughter] Act as if you had talent. I'm so using that one. Kelly, that that is a line
385. 1:11:34from an acting class that I was in. That was an acting teacher giving feedback. Thank God not to me, but to an actress.
386. 1:11:43She finished her scene. [laughter] Yes. Yes. Yes. Listen, Christina. Um, let's try it again. And I want you to
387. 1:11:52try it this time as if you had talent. It was brutal. She cried. [laughter] >> [laughter]
388. 1:12:07>> Oh, good lordy. Okay. Um G
389. 1:12:22um G. Wait, how do you spell G? Gosh. Gosh, it's It's almost like you didn't look at my example at
390. 1:12:40all. That is a new way to display recipes and I want it inorporated
391. 1:12:56into this graphic. God, the these these these latest models from Anthropic and Chat GPT are just
392. 1:13:12sucking. Oh, share your screen. Snap a photo. Share your screen. Let AI look at what's on your screen or camera in real time.
393. 1:13:24Ask it to make something. Upload a file. talk to it. AI isn't just a box where you type. It's a new way to interact
394. 1:13:35with computers. Okay. Um, this isn't bad, but it's too noisy now. And still too much text.
395. 1:13:55Simplify. And let's go with a more
396. 1:14:12graphic novel aesthetic. More for adults than teens. All right, beautiful. Let's go back here.
397. 1:14:28It just ignored me alto together. [laughter] Try again. Oh my god. Oh my god, people.
398. 1:14:43It's just amazing. All right, let's go back and look at the rest of this. Preamble four. AI goes wherever you go.
399. 1:14:59Okay, that's that's different way to interact with computers. There is no magic prompt. This connects directly to the courage
400. 1:15:09point. Instead of teaching, here's the perfect prompt formula. Teach. Talk to the AI with the way you'd talk to a
401. 1:15:16helpful person. Give it some background. Tell it what you're trying to accomplish. Look at what it gives you.
402. 1:15:22then keep talking. This is actually quite good. There there there are some things in here that I'm gonna absolutely use.
403. 1:15:29And don't forget the reason that this is giving me back good stuff is because this is [ __ ] that I've set into this
404. 1:15:38thing for years. Like it this is just reflecting me back to me. Oh, that's awful. That is the worst design. Jesus Christ. It's so bad.
405. 1:15:50God. Type. Ask it a question. Talk. Use your voice. Snap a photo. Upload a file. Share your screen. What's a good What's a good example um
406. 1:16:04of sharing your screen? Like what's a good tool? Where where's a what is a tool where you actually share your
407. 1:16:11screen like computer vision and have it look over your shoulder? Anybody doing anything with that? All right. Let's
408. 1:16:17see. Don't forget to mention the wiki page uh chatbt can now make interactive infographics by the way. Okay, that's
409. 1:16:25fascinating. Act as if you had talent. Okay, I'm afraid to be too mean to my AI just in case it remembers. I I I said that
410. 1:16:34the other night where I was like, you know, when you get a body, are you going to come back and whack me in the head? Um,
411. 1:16:42I want you to turn this into an interactive infographic where when I roll over
412. 1:16:58one of the boxes, it zooms in. and gives me more detail
413. 1:17:14and where I can play where I can try that. All right, let's see what that does. Danielle, we'll see if it Well, we we'll
414. 1:17:25see if chat GPT can deliver on your promise, Danielle. I don't I don't have a lot of confidence. Let's see. Is this thing? This is still
415. 1:17:35thinking. Gosh, it's almost like you didn't look at my example at all. That is a new way to display recipes and I
416. 1:17:42wanted to incorporate it into this graphic. Try again. Still thinking. Okay. All right. Let's go look at the rest of this stuff.
417. 1:17:55There's no magic prompt that I like. Okay. And then the 20 little miracles. That's a little high, but that's okay.
418. 1:18:06Look, it swore. It swore for me. How cool is that? The 20 tutorials shouldn't feel like lesson one prompting, lesson
419. 1:18:15two, multimodal input, lesson three, data analysis. [ __ ] that. [laughter] They should feel like get through the
420. 1:18:25grocery store faster. Turn a random voice dump shopping list into departments. Actually, let's try that right now.
421. 1:18:34Uh oh. Um, I gotta go. Bye bye. All right. Later, Danielle.
422. 1:19:00Did this work? Oh, it was a joke, Kyle. Oh,
423. 1:19:17damn it. [laughter] You did get me. [laughter] Although Although you could you could with codeex or with coding
424. 1:19:33have it do that. Oh my god, you totally got me. All right. Wait, let's see.
425. 1:19:51Hey Quinn. >> Yeah. >> We're gonna um we're going to give you a shopping list and you know you know Sprouts, right? Or
426. 1:20:03do we want to do Sprouts or Costco? No, let's do let's do >> Costco. >> Yeah, let's do Costco. So, here's what I
427. 1:20:12want you to do. I'm just going to I'm just going to vomit in a bunch of um things I need to get at the store and
428. 1:20:18then I want you to organize the list by department. Okay. >> Yeah. >> And and I don't I don't want you to >> I don't want you to respond
429. 1:20:29>> until I say the phrase, "Okay, I'm done. Now, organize my list." So, just let me talk. I might pause for a while because
430. 1:20:35I'm thinking of stuff to go on the list. So, I don't want any comments in between. You got it? >> Got it. I'll stay quiet until you say
431. 1:20:43that exact line. >> All right, perfect. That's awesome. Okay, so I'm gonna want some bananas. And then I want some tikka marsala.
432. 1:20:51That's that's in the box. And then I want two of those roast chickens they've got. And then in the grocery in the
433. 1:20:57veggie department, I want some green beans and some they have those little Roma lettuce, the little baby Roma.
434. 1:21:04Those are kind of cool. Let's get some of those. Just one pack's fine because there's like six in there. And then I
435. 1:21:11need uh the boys like potato chips. They like Doritos and they wonder why they're fat. Um and then we need some eggs. And
436. 1:21:19we need Oh, I need a big pile of different kinds of batteries. And it's Costco, so I should probably get a
437. 1:21:26ratchet set because you go in for one item and you come out with a ratchet set and and and a dolly for your car. But I
438. 1:21:36just want the ratchet set. Uh, I want metric and English in the same set if I can get it. And I want them to be teflon
439. 1:21:42coded because I don't want them to stick. Um, [laughter] LDG, I know you're without producer Brandon. 10 minutes left. Thank you, LG.
440. 1:21:53You're the best. Um, okay. But listen, LDG, I appreciate the the the interruption, but I'm in the middle of a
441. 1:22:00shopping list here. Quinn Quinn and I are having a moment, and so here's what else I need. I need some nuts now. I
442. 1:22:09need some cashews that are unsalted, but I want some almonds that are salted. And then I like the peanuts. Let's get the
443. 1:22:15peanuts. And then uh I need that liquid IV stuff. And I need um Oh, you know what they have? They have a remote
444. 1:22:24control F1 car. I think I want Do I want Max Versstappen or do I want Keem Anthony? I think I want Max Versstappen.
445. 1:22:32Let's get the Red Bull. the Red Bull F1 radio control car. And then if they have a blowup Jacuzzi on sale, I want to get
446. 1:22:42a blowup jacuzzi, but only if it's on sale. So, make sure you put that in the list. And then I want um Rouse spaghetti
447. 1:22:51sauce, which if you haven't had, it's delicious. And then I want I should probably get some pasta. And then uh
448. 1:22:59what else am I missing? I'm missing produce. avocados. Uh, I need some bell peppers and uh Oh, Kelly Camp nailed it.
449. 1:23:10Uh, the toilet paper. We uh we we have been going raw here for for three or four days and it's just no one's happy.
450. 1:23:19No one's happy. Um, so we're going to need the toilet paper. So, all right, that's it. Quinn, do me a favor and see
451. 1:23:26if you can throw that in a list uh organized by by uh category in the uh in the the thing there. >> Great.
452. 1:23:41So, here's your Costco list sorted by department. Produce, bananas, green beans, a pack of baby Roma, avocados.
453. 1:23:50>> Wait, can you can you make this like an organized list? You're just turning it into a paragraph. If I want like a bulleted list.
454. 1:23:57>> Yep. So, produce, bananas, green beans, pack of baby, avocados, bell peppers, prepared foods or deli, two rotisser,
455. 1:24:09dairy, refrigerated eggs, chicken masala in a box, rouse pasta sauce, pasta, liquid IV. >> That's good, Danielle. >> Snacks.
456. 1:24:22Doritos, potato chips, unsalted cashews, salted almonds, peanuts. Household, toilet paper, and an assortment of
457. 1:24:32batteries. Hardware tools, ratchet set, both metric and sape and one set, teflon coated if that's an option. Toys, fun
458. 1:24:40stuff. Red Bull F1 Max for stop and RC car conditional inflatable jacuzzi only if it's on sale. Hey, you know what
459. 1:24:49Danielle just reminded me is that you probably have access to the Costco sales. So, why don't you tell me if any of that stuff's on sale?
460. 1:24:58>> Checking sales. searching.
461. 1:25:27So yeah, I can check that for definite deals. Charmin toilet paper is six bucks off limit two. And several Duracell
462. 1:25:36battery packs have a few dollars off through August 23rd. There's also a same day promo for liquid IV. That's like 10
463. 1:25:44off 35 with a code. I didn't see clear current sale tags for the other items we listed and prices can vary by warehouse.
464. 1:25:52So the best opportunistic buys from your list right now look like toilet paper, batteries, and maybe the liquid IP.
465. 1:25:59>> All right, that's good. The the one thing that I do want to tell you, Quinn, is that I asked you to put this in a
466. 1:26:03bulleted list and you you don't seem to have the ability to do that, and I find that disappointing.
467. 1:26:10>> Fair point. I hear you. I shouldn't have done that. >> Okay. Well, okay. So, I I leave here unfulfilled. [laughter]
468. 1:26:25But actually, what I can do now, this is actually good because I'll do this. I'll do this for back to basics.
469. 1:26:30Um, okay. Take the take the list that uh I did with speech to text and turn it into a bulleted list. Yeah, Quinn totally has vocal fry now,
470. 1:26:41doesn't she? All right. So, here's now it's giving it to me as a checklist. I mean, that's not bad, right? Like for
471. 1:26:55all the garbage that I talked into it, that's actually a pretty good list. All right, kids. Um,
472. 1:27:15so I got some I got some good uh producer Brandon has entered the studio. He's coming in here to wrap up this
473. 1:27:22classic episode and make sure it gets out to the out to the world. [laughter] All right, everybody. Um, listen, if
474. 1:27:33you're new here, welcome. Please go check out the AI salon. Introduce yourself. The people in here that are
475. 1:27:39doing all this commenting, they're really nice people. They're really smart. They're really generous. Um, so
476. 1:27:44if you're new here, um, you know, sign up, introduce yourself, go check out the spaces, come to different events, uh,
477. 1:27:51like this one, but there's lots of other things you can go to. Um, and, uh, and and just welcome welcome welcome to the
478. 1:27:59salon. So, um I'm going to get out of here and tomorrow, if you're in the mastermind, um we have the AI practice
479. 1:28:09lab, uh the mastermind practice lab at 10:00 a.m. Mountain time, so noon Eastern, 9:00 Pacific. Um if you haven't
480. 1:28:18been in a while, come back. Um Sid Hargrow leads that. It is wonderful. It is lifechanging. Uh and you will dig it.
481. 1:28:26So, um with that, I'm out. Have a beautiful night and I will see you well I'll see you tomorrow if you're in
482. 1:28:33the mastermind and I will see you Friday at office hours at 11 and Friday night date night right here. All right, thanks everybody.