1/14/2026 How AI Solved Unsolved Math Problems and Redefined Trust and Authority

Transcript

1. 0:02That's going to live. That's life. Okay. You ready, Champy? No.
2. 0:30It's not simple to say. Most days not recognize me. That place and its patience have taken more than I gave them.
3. 0:47It's not easy to know. I'm not anything like I used to be. Well, it's true. I was never attention sweet centered. I still remember that
4. 1:02girl. She's imperfect, but she tries. She is good, but she lies. She is hard on herself. She is broken and won't ask for help.
5. 1:19She is massive. She's kind. She's lonely most of the time. She's all of this mixed up and B beautiful. She's gone, but she used to be mine.
6. 2:02Good day, good people. I'm not perfect. Well, I I'm never perfect, but I'm a little under the weather, so this will either be genius or a
7. 2:16disaster, or I'll just be like, I'm tired. I'm going to bed. Could be that, too. Um, I can't tell at this point if
8. 2:23I'm feeling crappy because I'm still fighting off this sickness or because I've been essentially laying down for 48
9. 2:31hours straight. Probably both. Yeah, exactly. Thank you. Thank you. Welcome back. I appreciate that.
10. 3:05Oh yeah.
11. 3:20Oh yeah.
12. 3:54Bum bum bum. Um, Highway Man. Uh, I only know 17 songs or some amount. It's some amount that's small. So, requests and me don't go well
13. 4:10together. If you know depressing Americana, I I might have a chance of knowing it. Tik Tok Tik Tok came off.
14. 4:27Low. What? Oh, like this. That is that what you meant.
15. 4:48Oh, yeah. Danielle got a job. got a job in AI creating content with a creator who has three and a half million
16. 4:55followers. Danielle, congrats. That's awesome. First of all, you deserve it. Second of all, it sounds like it's in
17. 5:01line with who you are, what you're about, your values. Super cool. And like you're really good at this stuff, so
18. 5:09doesn't surprise me at all. I think that's awesome. Congrats, You know, one of the things we talk
19. 5:23about a lot in here is just keep putting yourself out there, play with this stuff, explore, learn, and let people know what you're up to.
20. 5:42Um, grandkitty's staring at the TV looking very puzzled. He had never heard champ apparently.
21. 6:02Yeah. Isn't that cool for Danielle? It's so good. It's so good. I'll tell you what. One of the things that makes that that makes me the
22. 6:11happiest in this channel is when people in this channel um you know are just end up doing things that are like like good and surprising
23. 6:23and a direct result of of the time and passion you put into just being a part of this community, but also whatever
24. 6:31you're doing for yourself and putting it in the world. You know, you just you become a model for other people in the community.
25. 7:32If I go into bouts of silent silence tonight, that's just because I just realized I was like just gone. I think that's how tonight might be.
26. 7:47What? What? Oh, Kake, I have a job, too. Awesome. Congrats. Personal int Google personal intelligence. What about it?
27. 8:01Could be something to look at. Oh, it launched. Okay. Is it just in the Gemini app or is it in Is it in regular Google stuff?
28. 8:20inside Gemini. Okay, cool. Um, yeah, there's there's there's lots of stuff launching. Um, you know, I I continue to be with all of you
29. 8:32overwhelmed with all the stuff that's launching. I can't tell what's a feature anymore, what's a major new model. Um,
30. 8:39one of the things I've been thinking about for the AI salon is um, you know, I'm I'm feeling more and more and more like um,
31. 8:53the thing that's most important for all of us is to really understand who we are, what we value, the the the things we want to do in the
32. 9:06world, you know the change we want to make if we if we want to make change and that that's the primary focus and
33. 9:14that the tools are secondary but I'm also realizing that there's a lot of people coming to the salon that are just
34. 9:19like how do I use chat GPT how do I get started so one of the things we're looking at now is how do we put together
35. 9:25content that's like basic intro content I did a I did a fiveday back to basics thing here I don't know four months or
36. 9:33so ago and I'm thinking of digging out that content and making that available in the salon. And then uh we're talking
37. 9:41about some other stuff that could be really exciting. So anyway, if you're if you if you have a if if you're an irregular
38. 9:49and you have a desire to do any kind of tutorial kind of content, um connect with producer Brandon or with
39. 9:57Vicki or with Andy, whoever you might have a relationship with and just let them know that might be a possibility cuz I think that could be
40. 10:06something that would be good for all of us that we have an area for uh true beginners. to to just how do I start? What's a prompt? What's
41. 10:18the best prompt? Those questions, you know, Silverf Fox, I've been narrowing down as well. Which ones you considering
42. 10:42Oh wow, KKE, awesome. That's amazing. I started cancing some sub subscriptions today. Narrowing it narrowing it down
43. 10:51for phase two. That's so good, Silver Fox. I think this idea I think this idea of of simplify it's it's a weird it's a weird thing.
44. 11:03It's it's about simplifying, but it's not just simplifying to simplify, is it? It's like start to think about
45. 11:12what it is you want. What do you want more of? And then what are the tools that give you more of that? And not just AI tools,
46. 11:24right? Tools could be things like doodling, crocheting, pottery, cooking. Tools could be like diving deeper on
47. 11:32midjourney. It could be like I'm going to learn really learn vibe coding and how to do a a truly scalable app with vibe coding
48. 11:42like the the implementation is way less significant. But that I think this idea of focus, this came up on the we had the the AI
49. 11:54readiness project podcast today, Ann Murphy and I and we were talking about this idea of she was just saying she feels like
50. 12:02people are showing up to she leads AI just like a level more serious. They're just like if they're coming in
51. 12:09they're like if I'm going to learn I'm going to [ __ ] go all in. I'm going to show up to all the meetings. I'm going
52. 12:13to watch all the videos. I'm going to do the [ __ ] So people are people are starting to realize they gotta they
53. 12:19gotta they got to step into it. VO3.1 ingredients to video is getting more expressive. Portrait mode has arrived.
54. 12:29Yeah, I saw that. I I thought that VO3.1 had portrait mode, but I guess they didn't. Um, I'm still waiting for someone to do long
55. 12:41form uh character animations that's better than Hedra because Hedra is not Hedra's aging not great and I haven't
56. 12:49seen anything better than it. So, I don't know if anyone knows of anything where you can take like a long form
57. 12:54recording like a three minute recording of someone talking or a song and then have the the thing animate it. Who Who's
58. 13:01doing that well right now? Does anyone know? Mimi. Yeah. Yeah. When I join Sheile Leads AI, it's going to be for laser
59. 13:09focus on certification. There you go. Exactly.
60. 14:00I'm currently translating a caregiver cove app into French, Spanish, Filipino, and Russian. Oi, it's work.
61. 14:12Do you speak all those or are you using AI to do that?
62. 14:28AI shleing. It's still work. It is still work. Yeah, the translation stuff is really cool with AI, but but it's a lot of work to
63. 14:39like get it validated and certified.
64. 15:07I can speak English, American, and Australian. Steo, what's happening, Steo? What are you going to do tonight?
65. 15:22I just There's something about completely unplugging. Like I uh I just did a couple of meetings yesterday from bed, camera off, and then
66. 15:34I did little of that this morning with a big long nap and then did the uh podcast with Ann with a couple of bit story meetings.
67. 15:44Uh and then what did I do? I think I just got dinner and waited for this. Oh, I watched a shitty TV show.
68. 15:57I beat Steo. I speak English, American, Australian, and sarcasm. That's like a vacation day for you,
69. 16:09Kyle. Well, it is Silver Fox. And that's why I'm like I'm I'm learning something about myself right now. that
70. 16:21part of how I manage to do so much is continuity, right? If I don't take a break, I can maintain all these parallel
71. 16:32conversations in my head. You know, here's the AI salon conversation, here's my book, here's my play, here's my
72. 16:38company, here's my, you know, this like I've got all these different things. And so when I completely unplug, I kind of
73. 16:45come back and I'm like, what was I doing? God, that's a lot. Oh, man. If you stop spinning the
74. 17:02plates. Yeah, exactly. It's like right now the plates are all wobbling and I'm kind of looking at him like, "Huh?"
75. 17:11Oh, good lord. Good lord. Good lord. I'm the opposite if I don't speak it out loud for a week. Oh yeah, that's interesting. Yeah,
76. 17:30I've been going down a massive massive rabbit hole re researching the AI proved Liam lean math problems. Are those the
77. 17:40uh the propesed whatever those are called? Hang on. Let me see what they're called. It seems pretty fascinating. Math unsolved. Oops.
78. 18:08Erdos Erdos problem 728. For context, Paul Erdos left 1500 plus
79. 18:23unsolved math problem. Unsolved problems with cash prizes. Is that the stuff you were running down? Yes, the ones you mentioned the other
80. 18:33night. What did you learn in your rabbit hole? cuz I didn't I didn't run down the ma rabbit hole largely because I don't
81. 18:40understand math well enough to understand what they are. But actually, now that we come to think about it, okay, this is
82. 18:48is this going to be my cheat night? So, so this week I took a night off and I might have a cheat night tonight. So,
83. 18:54rather than doing my practice, let's just go [ __ ] around. Um, all right. Let me I got to share my screen because apparently y'all want to
84. 19:02see what I'm doing while I'm doing it. It's exhausting, but I'll do it for you. I'll do it. Okay.
85. 19:31Okay. I just I have to go on record and say when when your talent is sick and you have a producer who's not sick, that
86. 19:43producer can appear to the talent like a bully. He's he's right on it. The minute I don't have the black bar there, he's
87. 19:54writing pathy piffy little jokey comments. You gonna consider solving the black bar problem? I'm not quick enough to respond.
88. 20:05You don't get special treatment. That's good. Brandon knows. There's no crying in baseball. Okay, here we are. We're at GPT 5.2.
89. 20:16So, we're going to say, hey, three of the airdos math problems were solved this week.
90. 20:32Here's a tweet on one of them. Bang. So then what I'm going to say is what were the other two
91. 20:48and explain the I don't even know how to prompt this. Explain the the significance
92. 21:04of each problem in a way that even a guy with an acting degree
93. 21:19could understand. Listen, I'm not saying people with acting degrees could not possibly be high-end mathematicians. It's possible.
94. 21:33I'm just saying you could generalize to say that someone with an acting degree might not even have heard of the 1500 plus airdose problems.
95. 21:46If you're new here, welcome. I don't know what we're doing tonight. I don't think you'll learn anything, but please join us.
96. 21:58Ocean's here. Ocean will ask good questions. Ocean, come up with a good one. You always ask good questions. You
97. 22:03sent me off on a on a tangent the other night. It was good. Nor nor has an English major heard of these these things.
98. 22:15Liberal arts majors, unite. unite against the mathematicians and their fancy theorems and proofs. Oh man,
99. 22:33the other two between besides 728 are 729 and 397. Uh oh, my phone's dying. Hang on, I gotta charge. Man, I I can Have you ever been
100. 22:48Well, here's what's weird about my sickness. I don't have a fever, but I've got like crawly skin like you have when you have
101. 22:57a fever, but I don't have a fever. I'm not congested. I don't have a cough. I don't have any chest congestion. But my brain is it's like the
102. 23:10it's like the synapses have been filled with marshmallow marshmallow paste. So, it's like I have a thought and I can literally see it going
103. 23:21across the marshmallow paste across my brain. My brain is acting really slow tonight. So, so producer Brandon fully
104. 23:28exploiting that. You all can too. Um, my brain's felt like that for the past three years. It points to a reality
105. 23:38where AI assisted outputs that society can trust verify are auditable. Well, yeah. I listen I think that I think that
106. 23:49well I think I think how it probably goes Cam is that right now because AI is so janky like we don't trust it at all.
107. 23:58Like if if you trust anything AI says you should just not you should just stop that. And so when when these things say
108. 24:06they solved a math problem, there's a guy in this case who who validates these proofs and and says whether they actually solved them or
109. 24:15not. Um but there's going to be a point at which they're going to be good enough they're going to be like, "Yeah, that's probably
110. 24:24just right. We're not there yet." But okay, 729, why it matters, the actor friendly version. Erdos Erdos provided or proved
111. 24:37a kind of speed limit for when two factorials can fit inside another factorial. If a bang bang divided by nbang, then a plus b can't be much
112. 24:49bigger than n at at at most about n plus small cushion. Jesus Christ,
113. 25:04I'm so bad at math that I can't understand the actor friendly version of the explanation. That's pretty bad. Oh, good lord.
114. 25:20Why it's significant? Think of N prime. I It's not N prime. N exclamation point. What do they call that?
115. 25:28as a huge costume warehouse stocked with prime factors twos threes 57s in massive quantities divisibility questions are basically do
116. 25:40we have enough of every costume to build out oh piece costume piece to build out this outfit small primes are like black
117. 25:49t-shirts and jeans they're everywhere so the worry is are we being fooled by the abundance of common wardrobe items
118. 25:56rather other than the rare stuff. This is one This is one of the things where mathematics and me just aren't
119. 26:09friends. Like there's they're just these bizarre hypothetical worlds that mathematicians just swim in.
120. 26:21And then if you if you take the time to understand even the premise of the problem where I always go is like yeah.
121. 26:33So but it's that kind of thinking that led to the transformer which gave us all of generative AI. So I understand that
122. 26:42it's important but man it would sure be nice if I had some sort of bridge to this world. Tik Tok pin.
123. 26:50ask it what this means for the future of AI and human interactions. Okay, great. Um, rather than explain it,
124. 27:00tell me what you think this means for the future of AI and human interactions. That's good,
125. 27:16Cam. It means we're walking into a new social relationship where talking is no longer the the main event. Joint problem solving is. And
126. 27:31that changes how humans treat each other, not just how we use AI. Here's here's what I think it implies for AI and human interactions.
127. 27:40Trust becomes earned, not assumed. When you stop trusting AI because it sounds smart and start trusting it because it
128. 27:48can show work in ways that we can verify formal checks, re reproduct reproducible steps, audit trails, the vibe based era fades
129. 28:00and the receipts era begins. Authority gets weird. This is where this is the place I think it's going. When an
130. 28:09AI can beat experts at certain kinds of reasonings, humans won't just defer to credentials. We'll defer to systems.
131. 28:18That shifts status from who are you to can you validate that? Oh, that's fascinating. Expect more fights over who controls the
132. 28:28validation layer. I like that. H that's fascinating. collaboration becomes asymmetrical and we'll normalize it. The normalizing
133. 28:40thing we I think we've experienced that on this channel over three years that as remarkable as something is we we
134. 28:48normalize it very quickly, right? Um collaboration becomes asymmetric. Most people won't use AI like a tool.
135. 28:57They'll work with it like a partner who has infinite patience, no social fatigue, and frightening recall. Humans
136. 29:04will bring bring taste, intent, context, ethics, and narrative. Geez, kind of like the practice, right? Um, afterlive
137. 29:14party on Corey's live tonight. She's going to show us. Cool. This is cool. The Caregiver Cove new app. Oh, very
138. 29:21cool. So, so maybe what Hey Corey, if you want to pop up on screen here, um, feel free to just reach out to producer Brandon and
139. 29:34he can pop you up on on Streamyard and you can tell me about it and then we'll send people over there. If not, that's
140. 29:39cool, too. But if you want to pop up here, that'd be cool. Um, most people will not use AI like a tool. They'll work with it like a
141. 29:48partner. Yeah, humans bring taste. The new skill is directing, not doing. Totally agree with that. Human
142. 29:54communication becomes more intentional. This is the answer to this is sounding like the AI salon mastermind practice.
143. 30:06If you can offload drafting, summarizing, and planning, the scarce thing becomes meaning. what you actually want, what you value, what
144. 30:14you're willing to trade off. People who can articulate intent clearly will feel like wizards. This is important stuff, people. Like,
145. 30:25even though this is answering what's significant about these math problems, this list right here is actually really important.
146. 30:36People who can articulate intent clearly will feel like wizards. People who can't will feel increasingly powerless even
147. 30:43with superpowers in their pocket. Right? Because if you're someone who's really clear on who you are, what your values
148. 30:50are, and the change you want to make in the world, you're going to be able to look at AI capabilities and go, "Okay, I wanted to do this."
149. 31:00And you know what this is and you know what success looks like and AI will do its thing and you'll go like, "No, no,
150. 31:06no." and you'll hone it and hone it and hone it and you'll end up at the point you want to get to. Someone else is just
151. 31:12like, I want AI to do something cool and it'll do something cool, but there's no direction. There's no intentionality to
152. 31:19it. That that feels very right to me that that's where we're headed. We'll bond with AIS, but the real impact is how we bond with
153. 31:28humans. You'll see more people practice hard conversations with AI first, refine apologies, rehearse boundaries, workshop
154. 31:34empathy. This could mean human relationships, this could make human relationships better or make them feel more curated and less raw.
155. 31:45Some people will prefer the safe version. I think the counter that to that is some people are going to prefer
156. 31:53raw shitty interactions that aren't cleaned up. Important for whom? Kyle, let's see. Wait, wait, wait. What did Andy say?
157. 32:04important. How is this revoly in any way? Um, okay, hang on. Well, it might not be re revvelatory, but this feels in
158. 32:13line with where I think things are headed. Trust becomes earned. We stop trusting AI because it sounds smart and
159. 32:21we trust because it can show work in way so it can be it can be validated. That that's fine. Whatever. Authority gets
160. 32:28weird. When AI can beat experts with certain kinds of reasonings, human won't just refer to credentials will defer to systems. The the shift
161. 32:39that shifts status from who are you to can you validate that? I think that's important. I think right in right now's
162. 32:45in today's world, who are you is a way that we we sort of imply trust, right? If you went to the right school,
163. 32:54we're going to trust you. That that will start to fade. I think that's right. Collaboration becomes asymmetrical. Most
164. 33:01people won't use AI like a tool. They'll work with it like a partner. No social humans will bring so I don't know if
165. 33:08this is revoly but this is this is what I believe deeply. Humans will bring taste, intent, context, ethics and
166. 33:15narrative to AI and that our value will be bringing those things. Um human communications will become more intentional. If you can offload
167. 33:28drafting, summarizing, and planning, the scarce thing becomes meaning. What you actually want, what you value,
168. 33:35what you're willing to trade off. People who can articulate intent clearly will feel like wizards. People who can't will feel increasingly
169. 33:43powerless. That feels that feels I don't know if it's revvely to me. People who can articulate intent will
170. 33:50feel like wizards. That feels right to me. Corey's here. Oh, cool. Is Corey backstage? Yes. >> Hey. >> Hey everybody. How you doing?
171. 34:01>> What's happening? >> Well, I've been going talking about rabbit holes. Holy hump and holy holy. So, >> I've I've done this this app for
172. 34:12caregivers who are taking care of people with Alzheimer's and dementia. >> And it's called Caregiver Cove. And I've done it in Lovable.
173. 34:20And I have sort of stretched the lovable bounds kind of like a very big balloon. You know how we like circus balloons.
174. 34:30>> I'm kind of asking it to do a whole lot. >> And there there's that there's that aching four minutes where you keep
175. 34:36blowing them up knowing it's about to blow, but it hasn't quite blown yet. >> But it's doing it which is incredible.
176. 34:42Like it's absolutely insane how much I'm packing into this thing. So, for example, yesterday I decided, so I have
177. 34:50a section on here and I'll I won't take off all of this time, but I'll just tell you a couple of the components and then
178. 34:54how I got into this mess that I'm in now. So, I have this one component that allows for a user to take their phone,
179. 35:03put it under their chin, and go into a breathing sanctuary >> where there'll be a brain stimulation of
180. 35:11sound going from side to side. And it it gives the user which is the caregiver all these lovely thoughts and lovely
181. 35:19sounds so they can just have a minute of respit almost. >> Uhhuh. >> And then I thought wait a second I also
182. 35:25should give them a full proper guided meditation. So then I went into garage I went into chat GPT and I made a custom
183. 35:34GPT back when we were making all those GPTs >> and it was called Mindmate and another one called Zen Friend. And what they did
184. 35:42was you could put in anything that you needed to work on, anything you were upset about, anything that was sore, if
185. 35:47you had a locked jaw, if you had a busted shoulder. And I like fairies and I like deer and I like babbling brooks. And it would
186. 35:55create a meditation for you and work on those parts of you that really needed work. So I went, I'm going to go into
187. 36:02that and I'm going to say, I want to create a guided meditation for caregivers of Alzheimer's and dementia people. Mhm.
188. 36:10>> I want it to be generic enough that a man, a woman, any nationality can understand it and enjoy it, but it was
189. 36:17in English. So, I got that written down. Then I took it back into chat GPT, my project, and I said, "Clean this up. I
190. 36:25want it to be even better. You're the, you know, the marketing manager here. You're the blah blah blah blah blah." It
191. 36:31did that. I then spoke that into GarageBand. I went into Sunno. I made my own music to lie underneath it as a lovely meditative music.
192. 36:42>> I I compressed all of that and brought it back in. So now you click on a little area and it's a guided meditation and
193. 36:50you then hear Corey talking through your guided meditation. Then I realized I was talking to a cousin of
194. 36:58>> Oh, you recorded it. You recorded you saying it in Garage Band. So it's you being your soulful self. Oh, that's
195. 37:04beautiful. Yeah, with my choice of music because I went looking for music and there's lots of it out there, right?
196. 37:10>> Yeah. Yeah. Lot lots of wave files I could have downloaded. I hope you got some cello in the meditation music.
197. 37:17>> Well, actually there wasn't cello in this either, but there was very gentle, calm, lovely Coreyish music, right? So
198. 37:25brought that in, put that in there. And then I realized that at least in Canada and in my own family, there are a lot of
199. 37:34caregivers that are supportive of family members who come from other places. We have a very large Filipino community,
200. 37:44a large Russian community um that end up doing that sort of caregiving work in our hospitals and whatnot as well.
201. 37:52>> Mhm. So, I thought I should be translating this. So, then I went down, how do I translate this? >> Wow.
202. 37:59>> So, I put it in the chat that there's a a framework and and I found this out. Thank you, Chat GPT. And I'm using I go
203. 38:08back and forth between chat GPT and Gemini and Claude >> because I'm in the middle of a project and there's like there's so many
204. 38:18sections in the project that I don't want to miss where I am right now because I still have to go back to this
205. 38:23one and I don't so then I go over to this other >> Yeah, that makes sense. >> LLM, right? Yeah. >> So, what I've just done is I it said get
206. 38:31the framework in first. And the framework is called I whatever I just I I typed it in there. I forget what it's called now, but it's
207. 38:39some framework. Um, so get that started. And what it did is it and and I gave it those four Spanish, English, French,
208. 38:47because French is our national language as well. >> Although many of us don't speak it. Hands up. However, I want to make this
209. 38:54accessible. I want to make this a really important app that'll be used in maybe in schools like u maybe in nursing,
210. 39:02maybe in facilities that they can then give out to their primary care teams for families, for a daughter, for a spouse, for a anybody,
211. 39:13for anybody who's caregiving or who's around somebody with Alzheimer's in our hour. I say our like I'm talking to
212. 39:20everybody else there. There are many of us who have this color hair who we are a growing group of people who are losing our [ __ ] minds.
213. 39:29>> Yeah. >> And there's going to be a lot of us. I my own family my my father passed away from it. >> My stepfather has it like
214. 39:38>> Yeah. My father did too. So yeah. >> Okay. Yeah. >> If I would have had what I'm making I think it would have been well not me
215. 39:47personally. I shouldn't say that. If I was indirect like every day with the person, >> it has a scenario section, it has a help
216. 39:55section, it has a resource section that takes you based on where you live to the Alzheimer's society of your country
217. 40:02city, wherever that is. Um, it has a whole bunch of like you're doing fine or here, take a breath here or it's it's really quite robust.
218. 40:11>> That's cool. And so what you're going to do later is you're gonna you're going to show people you're going to walk people
219. 40:15through it. Well, Mimi wanted to know about the how I'm doing this translation now because I have to
220. 40:22>> I could only get the the first section in which is the the um the menu. >> Mhm. >> See if anything broke. Nothing broke. So
221. 40:30now what I'm doing is I'm taking the TS file of each of these huge databases of like 300 stories that are dripping every week on this thing.
222. 40:39>> It starts with about 120, but then there's going to be more. So, I'm going to take the TS files of all of the
223. 40:46different database parts that I have. I'm bringing them into Gemini. I'm saying, "Break these down into batches so you don't screw up."
224. 40:56>> Mhm. >> Then I want you to translate this batch. Uh let's say there's five batches in a category. >> Yep.
225. 41:05>> Translate each of those so that I can then have JSON files like I'm a potter. I don't know this stuff. bring all of these things.
226. 41:14>> This is like I this we're hearing this again and again. I mean this Liz Miller Gersfeld was like this. She's like I'm
227. 41:18not technical and then now she's you know she uses Claude code and GitHub and you know like we're all figuring this stuff out. It's so cool
228. 41:27>> and it just goes on just rabbit holes. Yeah, >> rabbit holes and rabbit holes. You can have you played with translating the
229. 41:32guided meditation. >> I have not yet but I pro because I am not using I've never used 11 labs and those kinds of things.
230. 41:41>> Okay. Yeah, I was just going to say 11 Labs and Hey Jen can probably do it wholesale where you just upload your
231. 41:47thing and it'll just translate it in your voice in those languages. >> I probably should do that with 11 Labs. >> Yeah,
232. 41:53>> I don't need to do I'm not doing any visuals during that because I'm hoping people will close their eyes.
233. 41:58>> Yeah. Yeah. Yeah, that makes sense to me. >> I thought about it. >> Yeah, but I 11 Labs should be able to do that just wholesale.
234. 42:05>> Yeah, that's a great idea. I I hadn't done that yet. I hadn't gone to that place yet. >> Yeah. No, leave it to me to add more
235. 42:12stuff to your plate. >> Oh my god. Wait. What I'd love though, Kyle, if if you're up for it and given that you've had a family member with
236. 42:19Alzheimer's. >> I'd be happy if you want to be one of the testers of this and try and break it. Say this doesn't work.
237. 42:25>> I I am famous for finding bugs quickly. >> There will be a lot of them, I'm sure. >> Yeah. You know, >> and what's your what's what's the
238. 42:32ultimate goal? When when do you want to launch it? Do you know like and and is there like a business model? Do you want
239. 42:37to give it away? What's your what's your goal with it? My goal is for a monthly and a yearly subscription and then a
240. 42:44premium tier for agency, schools, >> um, long-term care facilities that they can then give it >> to people who are there
241. 42:54>> as well as suggest it to families. >> So, I'm I'm thinking way bigger than I have dare the right to think about. But I think
242. 43:02>> No, no, you so you where where your mind is going. I mean, I'm telling you, man, this is the future we're coming into. If your
243. 43:12instinct is I want to think bigger, think bigger. >> Yeah. >> Why not? >> Why not? >> It's a It's an odd place for me because
244. 43:19I'm very used to keeping things very, you know, just >> make stuff for me because I like it. >> Keep it safe. >> Keep it safe. Yeah.
245. 43:27>> But it's your fault. I just want you to know that this is entirely and utterly your fault. >> Good. Well, I'm glad I could inspire you
246. 43:37to think bigger. >> Wear that well >> the crap out of me on a regular basis. Probably 10 times a day, probably more. So,
247. 43:45>> I talk a good game. >> You talk a great game. >> Thankfully, >> if anybody wants to because Mimi asked,
248. 43:53I'm just going to show how I'm copying and moving files so that they're prepared to then move back into Lovable.
249. 44:00So, Lovable has a database in each of those languages to use to make that framework work. >> That's that's super valuable. Yeah. If
250. 44:11anyone's playing and lovable and doesn't even quite understand what what Corey's talking about, that's Go hang out with
251. 44:17her after this. I'll try to I'll try to end this like in 15 minutes or so, just so you >> Oh, don't do it on my account.
252. 44:24>> Well, I just, you know, I don't I don't want everyone to be up super late and I'm also half under the weather. My
253. 44:29brain is totally unfunctional at this point. So, >> all right. Awesome. Congrats, man. That's that's amazing.
254. 44:35>> Thank you. Really, it is your fault. >> Yeah. You're welcome. >> And I'll see you tomorrow. >> Okay.
255. 44:42That's great. All right. Let's go back to Tik Tok pin. Corey UGC. And just like that, Corey is a founder. Exactly.
256. 44:58Listen, man. This is this is the world we're entering. I mean, quite honestly, it's the world we're in, but right now, you have to be um
257. 45:11you have to be really entrepreneurial right now because you have to glue a bunch of [ __ ] together. like all the
258. 45:16stuff Corey's talking about about having to manually move stuff, you know, out of lovable back into Lovable, all that
259. 45:24stuff. Like in the future, all that's going to be transparent and instant. Um, but then there won't be as much
260. 45:29opportunity. There's so much opportunity right now to really just invent yourself. So anyway, all right. So, back
261. 45:36to this the ma math problem solving thing. So, Kyle, can you post some of this to to the AI salon? Yeah, I think
262. 45:42what I'll do is I might have chat EPT tighten this up a bit. Um, and then Andy, sorry I'm spicy. You can be spicy.
263. 45:50I frankly I love it when people challenge me. However, tonight my brain is not all that functional, so I can't
264. 45:56hold more than half an idea in my head at the time. Um, what does do it in 11 Labs wholes sale mean? Oh, just Todd
265. 46:05that um I'm pretty sure within 11 Labs um she can just take her audio file in English, pick a language and say translate this into
266. 46:16Spanish or French or Portuguese. I forget what the languages were that she was using. Um she can just choose the
267. 46:24language and say just recreate the the the audio file in her voice in a new language. It's I think it's when I say
268. 46:33wholesale I mean it's just like a single function. You don't have to jump through a lot of hoops. Um but I haven't done it
269. 46:39in a while so I could be wrong on that. And again you know brain brain is a little mushy right now. All right. We'll
270. 46:46bond with AIs. Um the real impact is how we'll bond with AIS but the real impact is how we bond with humans. This is a
271. 46:54thing that I also think is going to happen that the that the robots in a weird way force us to be better humans.
272. 47:02Um, you'll see more people practice hard conversations with AI first. Oh, this is the thing where it was like refine apologies
273. 47:10and so so conversations become curated. That that's again that's one where I think that I think some people will go
274. 47:17to this more curated kind of human interaction. I think that the resist the the push back against AI will be that
275. 47:26people will just be raw with one another. A new etiquette will emerge. Um, drop that into Notebook LM and map
276. 47:34it out. Oh, that's a good idea. Okay. A new etiquette will emerge. AI disclosure and credit hygiene. In the same way we learned norms around
277. 47:44Photoshop and filters will develop norms around AI assistance, when to say AI helped, what counts as authorship, what
278. 47:53counts as cheating, what counts as collaboration. Social trust will hinge on these norms. I actually don't agree with that. New
279. 48:03Tik Tok pin. I got Perplexity Pro for one year. It's amazing. Agent mode. Can browse the website and even write posts. Yeah, Perplexity
280. 48:14is quite good. And I haven't used Perplexity in ages. So, I'd be curious to know how it is now. Mentorship
281. 48:21becomes scalable. That changes identity. If everyone can have access to a tireless coach, critic editor, then
282. 48:28talent looks different. More people will ship more work. The identity crisis will be if I can do it in a day, if if I can
283. 48:37do in a day what used to take me a month, who am I now? I think that's a puritanical industrial revolution hangover. I don't
284. 48:47think that's going to last that long. That guilt. The conflict shifts from AI versus humans to AI, humans with AI versus humans with AI.
285. 49:00Interesting. The most important interactions won't be you versus the machine. That I agree with. It'll be
286. 49:06which people, teams, orgs have better AI partners. That I disagree with. I think it's it's not who have better AI
287. 49:14partners. I think it's which people, teams, or orgs have the most clear point of view. and better ideas that leverage
288. 49:23the AI to do better work, better workflows, better evaluation, better values. And that competition will pressure society to define what good
289. 49:33means. New Tik Tok pin. I don't see it. Let's see.
290. 49:47Why is my phone all crooked? Can someone explain how Apple's creative apps are going to compete with how Apple's creative apps are going to
291. 50:02compete with when they don't do their own AI? No, I can't I can't explain to you why Apple has their head so far up their ass
292. 50:14with AI. I I don't get it. I I mean, their creative apps I don't whatever this thing they just launched was. I think it's just a
293. 50:23bundling of some of their most popular apps. Maybe in a way to keep them semi-relevant. I I don't know. I
294. 50:36I mean they announced a deal with Google yesterday that Gemini is going to be the official force behind,
295. 50:45you know, Apple. it's going to be their their AI of choice because because they didn't figure out how to do it on their
296. 50:50own. Um, no, I don't know. I I mean, frankly, I I think that tools like the Adobe suite and a lot of the Apple
297. 51:01professional creative tools, I think a lot of these tools, if they don't deeply integrate AI into them, I think they're going to be in
298. 51:11trouble. I I I don't know. I don't know. I'm flumxed. Apple Apple has me flumxed and apppopleleic about about how how bad
299. 51:22they've screwed the pooch on this thing. All right. If I had to put it in one sentence, AI stops being a thing you
300. 51:28query and becomes a thing you relate to. And that forces humans to renegotiate trust, authorship, authority, and intimacy.
301. 51:45H. Um, write this up as a an article that
302. 52:02talks about what math problems were solved,
303. 52:16how they were validated. But more importantly, how this
304. 52:32more importantly, what the future, what this means
305. 52:47for the future of AI, of humans. and relationships to each other and AI or IIA.
306. 53:02All right. I'm sure Apple will figure out the AI eventually, likely just using Gemini temporarily. I agree with that. I
307. 53:09mean, they'll eventually get it. I mean, here's the thing about Apple. They've got buckets of money and they're not a
308. 53:15company full of stupid people, right? They've got a bunch of smart people there. Um, it's just like I just I the thing that
309. 53:25I'm flung by is like how did the corporate machine like so miss this? Like how did someone in there not go uh you know,
310. 53:38red alert, red alert, we've got to get our [ __ ] together. It just hasn't happened. It's amazing. Three artist's
311. 53:43problem fill in a week. The math is cool. The human part is the real plot. In the first two weeks of January 2026,
312. 53:50there's the airdose problem. Okay. How are they validated?
313. 54:19This is cool. An AI generated the mathematical argument, not just found a paper, produced a workable line of reasoning. A proof
314. 54:30assistant pipeline forced it to be exact. Lean, which is a methodology, doesn't kind of accept a proof. It
315. 54:37either complies or it doesn't. Both 728 and 729 are marked as leanverified and 397 is leanverified as well. Human
316. 54:47mathematicians still played gatekeeper reviewing whether the result answers the intended question whether it's
317. 54:54non-trivial and whether it's genuinely new or a restatement. Ta's comments in the 397 thread show a kind of that kind of scrutiny in
318. 55:05action. At least in one case, someone wrote a human readable translation of the lean proof and explained the method
319. 55:12in conventional math language for the rest of the world to evaluate and build on. The combo creative generation and formal verification
320. 55:22plus expert social acceptance is the new tripod. And it's the tripod that changes everything. All right. And then it gets into
321. 55:32we're moving from trust the speaker to trust the process. authority gets renegotiated. Actually, this moving from trust the
322. 55:41speaker to trust the process. This gets into blockchain like you know um blockchain verified like you know you
323. 55:51can you can trust the trust the blockchain. It's all there all that transparency stuff. That's fascinating.
324. 55:58Authority gets renegotiated. Some people will not enjoy the negotiation. I agree with that. centers of power are gonna
325. 56:04[ __ ] hate this this next 10 years because the power is gonna shift. It's gonna it's gonna suck. For the people in power that are like,
326. 56:16"Well, this is the way it's always been, son." Uh, Tik Tok, I think we need to consider the AI access inequality issue so we don't
327. 56:30get locked into just one AI. I agree with that. I absolutely agree with that. Well, what's fascinating, Cam, one of
328. 56:37the things that I've I've heard some of the people I respect talk about with these AI models is as they're getting
329. 56:43stronger, they're getting more convergent. They're they're converging into kind of a single point of view. Not that
330. 56:51not that they're all the same, but that that basically that they're coming to the same universal truths. Um, so it's
331. 56:58possible that that all AIs converge into a single model. Um, which is the truth, but I don't [ __ ] know. That's very
332. 57:07science science fiction for me still. Um, collaboration stops being a metaphor. Right now, most people still
333. 57:13treat AI like a fancy search tool. All right, so I'm just gonna copy this and I'm gonna go post it over. The
334. 57:21headline I can't stop thinking about there. The trailer, the real movie is this. We are entering an era where AI
335. 57:27can generate novel work and machines can verify it. Which means human relationships will reorganize around trust, proof, credit, and intent.
336. 57:40Oh. Um, you can remove any references to acting. Um, just give it to me straight.
337. 58:06Why is my phone crooked? Hang on. Is that better? I think that's better.
338. 58:21Yeah, that's better. All right. No matter how much they train against it, you need to rotate two degrees. Yeah, I did.
339. 58:54Bottom line, we now have a credible pattern for AI assisted discovery that can be mechanically verified. Most
340. 59:02people experience a collaborator
341. 59:33I don't like that headline. I'm gonna I'm gonna not include that. Okay. Okay. Okay. So, where I'm going, I'm heading over to the AI salon.
342. 59:49If you don't know the AI salon, you should. So, you come to the AI salon, you can learn all about us and the cycle of AI
343. 59:58readiness. Play first, create excellence, generously lead. And then I'm going to put this in community feed. Yeah. I'm gonna say um
344. 1:00:16this was an interesting enough take from chat GPT on the three
345. 1:00:36air dodos um math problems that were solved last week with chat.
346. 1:00:53Okay. With chat GPT 5.2 to Pro and what the implications are for the future
347. 1:01:11of human AI interaction. Um, I did not edit this.
348. 1:01:27It is just this side of AI slop, but interesting. There you go. Boom. Okay.
349. 1:01:44Oh, wait. It [ __ ] up the formatting really bad. Hang on a sec. Oh, no. Let's see.
350. 1:02:17Still [ __ ] it up, didn't it? H Oh, let me try something else. No. Oh, maybe.
351. 1:02:33Yeah, that's better. Okay.
352. 1:03:15There was [ __ ] up formatting. Let me let me try to find it.
353. 1:03:39Maybe paste it as an article. Yeah, I might need to paste it as an article. Andy, are you still watching? Any any
354. 1:03:53advice? I don't know what I'm doing. Um, how do I paste it as an article? Share post. How do I make this an article? Oh, post.
355. 1:04:08Is it here? I'm not in water cooler. I'm in community feed. Hit create in the upper right. Ah, cool. Beautiful.
356. 1:04:24Okay. Did this do better? No, it still [ __ ] all this stuff up. The statement roughly here. Hang on. Let me try something
357. 1:04:35here. 728. The statement roughly. Let me copy this here. Go back here.
358. 1:04:52So, that worked. Okay. Yeah, the Apple note [ __ ] up that paragraph, too.
359. 1:05:09But hang on. The archive write up. Let's see. The archive right up. Okay. Proved lean. Okay. So now, let me get rid of all this crap.
360. 1:05:30Okay, that's good now. Beautiful. Beautiful. Beautiful. Beautiful. Okay. So now going to pull that out. We're going to
361. 1:05:44drop a title in here. We're going say um this was an inter I'm going to put in square brackets
362. 1:05:58interesting um interaction with chat GPT about the three eros math
363. 1:06:15problems. s that were solved last week. Uh I did not edit this so it is
364. 1:06:36just beyond AI slop but is interesting. What are
365. 1:06:50your thoughts? Question mark. Okay. All right. Post
366. 1:07:06done. Thank you. We don't need to notify people of that. Okay. Um I told Corey I would get out of here sooner than later. I'm starting to
367. 1:07:19get like flop sweaty because look what I made before you bounce. Okay. So in AI salon also if you haven't played with it
368. 1:07:28if you go into look what I made people make cool things. Brandon made something. Let's see what he made. It's
369. 1:07:35officially Girl Scout time. Why just order cookies when you can take a personality quiz to help you decide?
370. 1:07:42Play along with our family's custom GPT here. That's hilarious. You You did this last year, too, didn't you, Brandon?
371. 1:07:53This is updated for this year's cookies. All right, good. Hey, Brandon, you want to you want to hop up on stage and talk
372. 1:07:59about your LOL since since I'm brain dead and you can talk about this, too? This is pretty cool. Yeah, happy to happy to hop up. Uh, so
373. 1:08:12first of all, the cookie thing. Um, so my daughter Naomi selling Girl Scout cookies and uh last year when we were in
374. 1:08:19the throws of making GPT, custom GPTs, I thought, well, I wonder if we could have a GPT figure out which Girl Scout cookie
375. 1:08:28aligns with your personality. I asked Chachi to do it and it came up with a five question quiz that scans
376. 1:08:39your personality for like what type of adventure do you like? Describe your perfect weekend. And then at the end it
377. 1:08:45tells you what Girl Scout cookie most resembles you. It also has advice for like if you like chocolate, which Girl
378. 1:08:52Scout cookie is the recommendation. And of course, uh, because it is girls cookie season, the app ends with a link
379. 1:08:59out to my daughter's, uh, page if you want. >> That's genius and evil. It's It's evil genius. That's good. Awesome. And then
380. 1:09:10what are you what are you gonna teach us on Saturday? Um, so, so while you're eat while you're uh when you're done
381. 1:09:17ordering your Girl Scout cookies on Saturday, I am also doing uh a Vicki Larn out loud to go over my feat of strength, which was
382. 1:09:28going from thought to or from prompt to publication in six hours using Amazon Kindle Direct Publishing. So, we're
383. 1:09:36going to go through that entire process in in delve deep, if you will. I feel like we're reclaiming the word delve. I
384. 1:09:42feel like Delve is back on the list of words. >> We can use Delve again, I think. >> Yeah. Yeah, that's the book.
385. 1:09:48>> That's the book. That's the book. >> And so we we are going to go over kind of stem to Stern if you've never
386. 1:09:58published anything before. No prior publishing experience needed. And um you know, hopefully we'll come out of that
387. 1:10:04session with uh an action plan that'll allow you to self-publish your own book. That's cool. That's very, very cool. Beautiful.
388. 1:10:15>> And while I'm here, I am uh pressed to tell you about the uh mastermind uh program tomorrow. >> Yes. Yes. What you said.
389. 1:10:28>> Yeah. So, if you're not in the mastermind, you should join the mastermind because it's awesome. And if
390. 1:10:31you are in the mastermind, we'll be having our continuing uh session of the practice lab tomorrow with um Kyle and uh uh
391. 1:10:44>> I think you're Liz Miller Roshchel uh at noon. Uh and that if you are jumping in late because we've got a bunch of new
392. 1:10:54people after festivist, you can go back and look at all of the p past sessions on demand. >> Yep. Yeah. And so the the practice lab
393. 1:11:02session is at noon Eastern and it's the group of people that are starting their own daily practices centered around AI,
394. 1:11:10how they use AI. Uh and and it's it's it's a pretty remarkable group of people that are willing to put
395. 1:11:19themselves out there and and just, you know, be uh be vulnerable and and figure out how to how to take their AI game to
396. 1:11:28the next level. It's really really powerful. So if you haven't joined the mastermind yet, do that. We have lots of
397. 1:11:32cool stuff coming this year and uh and the this group's a really powerful group. So beautiful. Thank you, sir.
398. 1:11:39>> Yeah, of course. And while I'm doing shameless self-promotion tomorrow at 8 Eastern, Rick McCaulay and I are going
399. 1:11:45to be talking about ShiaPT health and doing a deep dive on using AI in your own personal health journey over on
400. 1:11:52Rick's YouTube channel and will be live right before I go live. Right. Beautiful. So, we're we're just starting
401. 1:12:00to bookend you. We've got a a content calendar. We've got me me and Rick before Corey's doing the after party.
402. 1:12:06>> Exactly. Exactly. I like it. >> We're doing 24 hours of AI all the time. >> We could do it. We can have Steo do one,
403. 1:12:14you know, out of Australia. Beautiful. All right. Beautiful. Everybody, um, thin mints. I'm a thin mint. The only
404. 1:12:25way I'm a thin mint here, here's my only thing about thin mints. Thin mints, I consider a serving size of thin mints
405. 1:12:33one sleeve, and I consider myself well constrained if I don't eat both sleeves. So, there you have it.
406. 1:12:46How they try to make a practice. Yeah, exactly. Well, listen, man. A daily practice is no it's it's it's no easy
407. 1:12:54feat. I'll tell you that. Um, it requires some it requires some moxy. So, it's a cool group. All right. Beautiful.
408. 1:13:05Um, Brother 52, I pitched a three-week course on a me memoir writing with an AI course to 200 of the 50 plus crowd.
409. 1:13:13Super cool. I hope you get it. That's beautiful. Love it. All right, KakeF's out of here. All right, I'm out of here. You guys go hang
410. 1:13:23out with Corey, learn how she's doing what she's doing. Then Saturday, go hang out with Brandon. If you're in the
411. 1:13:28mastermind, come hang out tomorrow. If you're not in the mastermind, go join it. Just go to the AI salon in the first
412. 1:13:35four little sections there. The last one you'll see is learn about the mastermind. It'll take you right there.
413. 1:13:40Okay, beautiful. Peace out, everybody. and I will see you tomorrow night regular uh regular