12/1/2025 - High Schoolers Are Curing Lyme Disease. This Is the New Age of AI.

Transcript

1. 0:07Producer Brandon in the house. We're live. We're shaking. What's happening? Going to come up for a sec while Hello. No one's here. Hello.
2. 0:21What? It does it does tend to impede the audience when you tell everyone you're not going to go live and then
3. 0:28>> did I say I wouldn't go live? >> You said we were probably going to be off this week. I I did put in irregulars
4. 0:35that it was kind of up in the air and we weren't sure. But I think the running thought was that tonight was going to be
5. 0:42off and tomorrow we were going to do the uh salon presents. But >> Vicki Vicki contradicting you. Vicki says no you didn't. [laughter]
6. 0:51Maybe I said it to you. >> I don't know. Whatever. Um, you said you would try to go live. I I'm trying. I
7. 0:57have a headache and I'm on my um I'm on cell data. There's no Wi-Fi. So, I'm in a brand new a brand new New York City
8. 1:07studio apartment that a friend is loaning me. There's no furniture in it. I'm on a blowup bed. Um there's no
9. 1:13nothing on the walls. [laughter] So, oh man. All right, >> let me uh let me update uh everyone. The cat says hi.
10. 1:29>> All right, go. >> What's the update? Oh, you're gonna update everyone back there. >> Yes, on it. U on Okay, Gi, Gy, Gy, Gvy,
11. 1:44I'll see you out there. Okay, >> bye. Good day, good people. Good people. What's happening? Um, I guess I'll do this
12. 1:57since we're talking. How's the hair? God, always checking the hair. I do look like I'm in a bit of a good day. Some poor papaya guitar.
13. 2:09Some poor papaya guitar. Oh, looks great. All right, looking sharp. Cam Ken from Cleveland. I am in New York City. I'm in New York
14. 2:24City on 71st and Second Avenue. Um, all right. Hang on. I got to get myself situated here. I had I was sitting in a
15. 2:36different spot right before I went live and there was a giant light fixture above my head and it was weird. It was causing a weird glow. Um,
16. 2:47all right. So, so a couple of things. Um, see, he he just can't quit us. Um, I do have a headache. I am on shitty Wi-Fi or no
17. 3:01Wi-Fi. I'm on on cell bandwidth. So, if it goes crazy, that's what's going on. And I think I got a text from AT&T
18. 3:11two days ago saying, "Pay your damn bill." So, [laughter] which I haven't done yet. So, if uh if they make me go away, it's because I
19. 3:19didn't pay my damn bill. [laughter] Um okay. So, a couple of things. One is yesterday um was happy birthday. Happy third birthday.
20. 3:36It's late here. Hey, did you know this? Um, if you go live at 8:00 p.m. in Denver in New York City, that means your start
21. 3:50time is 1000 p.m. [laughter] That's Brandon knows this well because he's in even though he's in you're in my time zone now, baby. Even
22. 4:01though he's in uh Cleveland, which you think Cleveland full-on Midwest, right? You got Pennsylvania, you've got and you
23. 4:09got to drive across a big part of it to get there. It It's still East Coast time. It's still East Coast time. I don't know
24. 4:17how they pulled that one off. I think it was some sort of redistricting. There was gerrymandering involved.
25. 4:23Clearly a Midwest town, but no, East Coast time. So, producer Brandon knows this time very well. I don't. This is
26. 4:33exhausting. I don't know who came up with this idea. It's a shitty idea. Start Start a live at 1000 p.m. [sighs]
27. 4:43It's exhausting, people. Um, hang on. I got to put on shoes because the little the little stool I'm on has got foot
28. 4:51rests, but they are sharp. This is I'm in a very different I'm in a very different $20 million studio. [laughter] Um, [clears throat] but
29. 5:06okay. I know this very well. My doctor wants me to move west to learn AI. [laughter] Oh, now you know how we feel. I do. This
30. 5:19is horrible. It's just It's unacceptable on all levels. All right. Wait till we green screen you. Oh, yeah.
31. 5:26This is totally You can green screen me right up. Um, okay. So, a little bit of philosophical stuff here. Um, let me just get my screens
32. 5:40screens where they need to be. There we go. Um, [clears throat] I think I'm going to leave it like this for now until I show something.
33. 5:55I think. Yeah. Should I? I don't know. Um, you are a little little double chinny on
34. 6:10TikTok. Oh, I see what he's saying. I see what he's saying. He's [snorts] He's saying you're a fat [ __ ] [laughter]
35. 6:18I don't think Brandon would ever use those words, but that's what he meant. He's not wrong. Okay. And by the way, I can barely see the
36. 6:30post-it notes there, Brandon. So, and I'm going to turn off comments. [snorts] So, Brandon, if you want me to see a
37. 6:45comment, um, make it go live on screen. Leave it. This is fine. They see they don't mind my double chin. Why does chat
38. 6:57GPT voice only use 40 and not five or 5.1? Um, I'll tell you, Cam Katkin, even though we're we're uh celebrating Open AI's
39. 7:09third year anniversary this week. Um, and by the way, I don't know if you can hear that. I'm in I'm in New York City.
40. 7:16The sirens are going, the horns are blaring. Love this town. Um, before you continue, change it back because we don't have
41. 7:26your construction paper. They didn't it didn't make it through TSA. >> So, keyboard is uh finding everything.
42. 7:34>> Um, also um uh Sharon Crawford says you are uh looking like you're in the county jail. >> Yeah, I know. I can't I am I am like
43. 7:44literally so the the person that's lending me her apartment, she got it a week she got the keys a week ago and
44. 7:52it's so small that her sofa didn't fit into the door. Um so uh there's nothing there's literally nothing. So yes, I am
45. 8:01effectively in a county jail. >> Did you try the friend's pivot method? >> What's that? >> The for the couch to get it in.
46. 8:08>> Oh yeah, they tried everything. [laughter] We should be good. >> The Atlantic magazine had an article
47. 8:18about Microsoft being Sydney at the AI falling in love. Did they that was recently they had that? Because that's a
48. 8:24three-year-old story. Maybe that was a retrospective because of uh because of the the anniversary. Anyway, um it's three years.
49. 8:33So, I said this on Twitter. I think I feel like the last three years um Oh, I'll get to that question in a second,
50. 8:43Cam. I feel like the last three years have felt like three months and three decades simultaneously. We are
51. 8:50definitely in some sort of universe warp where I, you know, time is just weird and irrelevant at this point. um
52. 9:01like a lot has gone on and and in case you're not aware of it. Um there was a lot of AI in the world before chat
53. 9:11GPT came along, I started this channel and I started the AI salon uh and I started really paying attention to
54. 9:20this AI stuff. The week chat GPT came out. Right. To me, November 30th, 2022 marks a seminal moment in history, in
55. 9:29human history. Um, that's the moment the chat GPT launched. That's the moment that you didn't have to be an engineer
56. 9:38to be able to tap into the power of AI, right? So, we we had specialized AI for years and this general kind of, you
57. 9:48know, chatbot that we have now didn't exist. Time is weird. time is very weird right now. Um, so our ability to do all the stuff that
58. 10:00we've talked about on this channel has effectively um been in the world for three years. That's it. That's it. Um, if you think
59. 10:09that that's a long time, it's not. If you think that we are in a hype bubble, we're not. Like, I don't even think the
60. 10:16hype has begun yet. I think 2026 is probably going to be the year where we start seeing some madness hype kind of
61. 10:23stuff. Um, but it might be 2027, but I think 2026 is going to get weird. Um, so that's kind of fun. That's kind of
62. 10:35cool, right? Um, the why does chat GPT use um 4.0 40 instead of five or 5.1? Because from what I can gather
63. 10:51and this is pure chat TMZ. I don't have any I don't have any factual basis for this but from what I can gather
64. 11:00the culture at OpenAI is gather in small teams innovate [ __ ] and if what you innovate is cool we'll launch it.
65. 11:12So that's why you end up with things like we're going to launch custom GPTs and then we're going to launch the GPT
66. 11:18store and then Sam Alman says the GPT store is going to be everything unstable connection. If I'm going in and out it's
67. 11:25because I don't have Wi-Fi and then OpenAI just ignores the app store. For three years we haven't heard
68. 11:33or two years we haven't heard an update on it, right? Um, so things like advanced voice not using the latest
69. 11:41model probably because they didn't have a a team integrating it and they wanted to get five 50 5 and 51 out faster than competitors.
70. 11:54So I think it's just they're they're running very fast. I think that's what's going on there. Um, the other thing that I've been spending
71. 12:08a lot of time Tik Tok question, how many VL31 generations do you get in the VL? What's VL3.1? Oh, VO. VO probably.
72. 12:23Do you get on the yearly plan? They don't give you a number. Um, probably here's my prediction. The amount of
73. 12:34video generations you get for your annual plan is not enough. I don't know if you're going to actually
74. 12:42try to do anything with video. Um, video is really expensive still. Um, there's there's a new Chinese model that came
75. 12:50out, I don't know, last week or something like that that you would have to install yourself. um that I think is
76. 12:55much cheaper because you can run it on servers and it's just your compute time that you're paying for. You're not
77. 13:00paying a retail markup. Uh but all the video stuff's still really expensive. Um I use Chat GPT because of you and have
78. 13:07created prompts that save my company $130,000 in SA salary annually. That's great. That's awesome. Um thank you for the acknowledgement. Um,
79. 13:19and what's what's what I hope is is is the case is that that $130,000 a year salary can go toward, you know, keeping someone and
80. 13:31leveling them up to do something else, right? There's two ways you can when you have when you have a savings of 130
81. 13:39grand a year, there's two places there's two ways you can do that. One is you can let go of people. The other one is you
82. 13:44can do new things with the same amount of people. and you can do, you know, other kinds of uh activities for them.
83. 13:51So hopefully it's that. Is it raining in New York? It was raining yesterday. Yesterday was horrible. Today was fine.
84. 13:56Today was nice. Found out that AI Studio uses Gemini 3 to write code, but only ship ships apps that use 2.5 Flash. Fascinating. Oh,
85. 14:08that's why it always switches to 2.5 Flash when you're trying to write something. Absolutely. the sky is the
86. 14:14limit with more automations. Thank you. Well, thank you. Thank you for the acknowledgement. Um, you know, it's funny what what
87. 14:22you're what you're hitting on is a is a is is scratching the surface of something I've been thinking about a lot.
88. 14:29And I saw two stories in the past day. Saw one last night and I saw one today. Um, and I've got them on on uh on X. I
89. 14:43can pull them up, but I don't think we need to. One is on 60 Minutes last night, there's a high school in Georgia
90. 14:51that has this program called GEM, GM, which stands for something. And so it's like an advanced science class. People are literally
91. 15:06moving from all over the country to be in that school district so that their kids can apply to be part of this
92. 15:15program that accepts 10 students. They're high school students. They're high school students. They're high school students
93. 15:26that used Crisper. you know, crisper, the gene editing technology, you know, where you can edit DNA. A group of high schoolers created a new
94. 15:43crisperbased technology that can detect Lyme disease in the blood two days after you get bit. The current technology is 2 weeks after
95. 15:54you get bit. And getting to Lyme disease quickly, trust me, I know, is really, really important. 2 days after someone gets bit, here's
96. 16:06the more remarkable thing. The same technology that they used to make the detection test is also going to be able to use to treat it
97. 16:20because they can identify the protein that is doing the damage. and therefore they can go in and do some editing and
98. 16:25[ __ ] up the bacteria that cause Lyme disease. They're in high school. They're in high school. Unstable connection.
99. 16:38Um then I wanted to show you this dude. This This one's worth showing. Yeah, this guy. You can see that, right? Here we go. All
100. 16:50right. Watch this. And so my novel mind controlled bionic arm was intended to address these shortcomings. So using
101. 16:58tiny non-invasive electrodes, I was able to pick up fuzzy residual electrical activity on the forehead. I theorized
102. 17:05that deciphering those obscure service level signals could potentially allow the interpretation of frontal lobe brain
103. 17:12wave activity. And so I con conducted this IRB approved study to collect thousands of brainwave data points from
104. 17:19a diverse group of human subjects. And by using this acquired data to train a complex artificial intelligence-based
105. 17:26mind readading algorithm, I was able to discern uh bonafide user thoughts from scalp electrical activity. And so my
106. 17:33final algorithm involves over 23,000 lines of handwritten code and over 900 pages of calculus intensive math. No
107. 17:40existing system has been able to achieve this. the the accuracy that my system has. And so in a series of rigorous
108. 17:47tests, my provisionally patented neuroprostesis has achieved a similar performance to that of the world's
109. 17:53industry-leading $450,000 plus invasive prosthetics. Uh the arm is extremely durable, lightweight, and has no
110. 18:00surgical requirement. And my entire system costs less than $300 to produce. He's 17. Now granted,
111. 18:14the dude's the [ __ ] genius, right? He's wicked level smart. 17. The current solution that does what his thing does is $450,000.
112. 18:28His solution, $300 without surgery. Right. [laughter] So, I assume that was a a spit in the eye of Elon Musk. >> [laughter] >> Um
113. 18:47the the acknowledgement from Dove Inc. The acknowledgement from Dove Inc. that, you know, they've figured out a way to prompt Chat GPT
114. 19:01to save 130 grand a year in their company is is all in a continuum. And three years into this generative AI
115. 19:14stuff, I'm more convinced than ever this ain't going away. And when I see stories like the Crisper story, the the the IGM
116. 19:24class out of Georgia, and I see this kid, what's his name? Doesn't say. Um creating a a brain interface that's
117. 19:35accurate enough to control a robotic hand for 300 bucks without surgery. You've got to understand that we're
118. 19:48entering, we're not in it. We're entering an era where you, you, you, me, you, all of us
119. 20:06can do anything we want. Can do anything we want. Now, I'm sure there's a troll or two in here going, "Wh the code's not really
120. 20:19good." You know, the the the physics in the video models are, you know, them. Pay attention to what's actually happening.
121. 20:35A high school class in Georgia may come up with a novel treatment for Lyme disease when no one else is paying the [ __ ] attention to
122. 20:44Lyme disease. This kid patented his thing there may solve mobility issues at a at an affordable price for anyone who's got them.
123. 21:00We are entering an era where anyone can do anything. You don't need to go to Stanford. You don't need to have
124. 21:1030 years of experience. You don't need to have a big lab. Right now, the high school lab in Georgia is very well funded. It's in
125. 21:21the most affluent school district in Georgia in in their part of Georgia. So there's resources there. But like I
126. 21:30said, we're at the we're at the beginning of this. And so so then the question becomes this. How do I get good at AI? What's the best
127. 21:38AI tool? How do I get good at AI? What's the best AI tool? What am I supposed to learn? If I go to school, what am I supposed to study?
128. 21:49Kyle, why isn't our lab as fancy? I know issue. You're absolutely right. Um. Ah, dove ink just hit it. Dove ink just hit it.
129. 22:02You just need ideas, practice, and to keep trying. It's easier than we thought. Our job moving forward, our job
130. 22:16is going to be come up with ideas. Our job is going to be to take an inventory of who you are, what your values are, what you want to
131. 22:29do in the world, who you want to impact, how you want to impact them, and then hold on for dear life as AI brings that concept
132. 22:44to life for you. That's going to be our job. I mean, I I I'm as convinced of this as anything that I've ever been.
133. 22:58I saw I saw a funny thing. Um, not a funny thing. I saw an interesting thing from Kiri who spoke at the AI salon last
134. 23:05week or last month or the month before, whatever. But she's created this no spoon automated video storytelling platform, right?
135. 23:16And someone on her Twitter asked her, "If everyone can just make movies, why would anyone watch anything? If every if if there's just movies
136. 23:27everywhere, why would anyone watch anything?" Well, because when everyone can make everything right now, the bottom the worst films in
137. 23:41the world live down here. And when AI gets really good, the bottom is going to elevate. So the the worst films in the
138. 23:50world are going to be better than the worst films in the world today. Right? So here we are today. Here we are
139. 23:56tomorrow, you know, two years from now. This is still the bottom even though it's deeper, right? It's closer to the
140. 24:05surface. Brilliant is still way up there. Brilliant is still off the screen, right? The shitty went from here
141. 24:13to here. Now, everyone can do at least this quality of storytelling. What rises above that? Human beings with ideas, actual storytellers,
142. 24:29people with something to say, people who know who their audience is, what will resonate with them or or what resonates
143. 24:35with them personally that they know their audience might like. and they will put some idea into these tools that
144. 24:42everyone has access to and they will do it in a way that's slightly different. Why do people pay attention to Kelly Bosch's work?
145. 24:51She's using the same [ __ ] tools. In fact, she tells you what tools she's using in every video she does. She tells you
146. 25:00this was clinging. This was midjourney. The dancing was this. The song was every single video she tells you that.
147. 25:10Why is why does her work stand above? Because of her. So it's going to happen in the creative arts. It's going to happen in the
148. 25:18sciences. It's going to happen in business. Everyone can do everything. The only stuff that's going to stand out
149. 25:27is stuff where human beings say, "Oh, I've got an idea. I know how I could change the world. I know how how I could
150. 25:39you know what you know what really sucks right now because of the insurance companies doctors are legally prevented
151. 25:48from treating Lyme disease and therefore the scientific community has no interest in finding a cure. Wouldn't it be cool
152. 25:59if our little high school project was to cure Lyme disease? What? What? It doesn't make sense. And yet that's
153. 26:16the world we're entering. So I say this for a couple of couple of things. We we recently started in the AI
154. 26:23salon this thing called the AI um the AI salon mastermind practice and what it is it's a framework for each
155. 26:34of you to design a daily practice around how you use AI and the daily practice is is a practice it's a practice like you
156. 26:48practice you say you know who am I? How do I feel about this? That one makes me feel icky. I don't like that. What do
157. 26:59I want to do? What are my values? How how am I going to learn? How am I going to play with AI in such a way that I actually learn something
158. 27:08that's in line with my values? You can do that. And that's what it's about. And and here's what I don't know. And this
159. 27:18this is the part this is the part that that wigs me out a little bit. I know deep in my soul that the people
160. 27:27that are going to be rock stars, like [ __ ] rock stars using AI are going to be people who treat AI like a practice,
161. 27:37who really understand who they are, really understand what their values are, really understand what all these tools make possible,
162. 27:46and figure out a way to put stuff in the world. I know that's that's going to change the world. I know those are going to be the rock stars.
163. 27:52I don't know how long it's going to take for that broader realization to happen. But what I can tell you is if you start
164. 28:00right now doing a daily practice with AI and you know daily practices are rough. You know, you know why they call them
165. 28:12daily practices? Because you have to do [ __ ] daily. [snorts] And if you're not doing it daily, you're being left behind, right? And
166. 28:25doing it daily does not mean you've got to learn every AI out there. No, no, no. Quite the contrary. I think it's I think
167. 28:31it's actually quite the opposite. The daily thing is who am I? What do I want to do? Have I learned enough to
168. 28:37understand what's possible today? And and am I clear enough on who I am that I know what I want to do in the world?
169. 28:44That I know the change I want to make. And then, okay, so you decide you want to make the world safe for washing llamas commercially.
170. 28:57[laughter] It it literally doesn't matter, right? You can do whatever you want to do. And then it's going to be on you to
171. 29:06figure out, well, what do I need to learn? What AI tools do I need to be able to pull this thing off? What what's
172. 29:12my marketing? What's my business plan? What's my go to market strategy? How do I sell this thing? How do I do the ads?
173. 29:20I would think if I'm going to do a llama washing business, it probably make a good song. We can probably do some good songs
174. 29:31[laughter] and some good video ads of dirty llamas. Don't let your llama be dirty, right? So, you got to go learn all that
175. 29:40stuff. And and learning all that stuff is going to be icky, right? the ick. You're going to have to embrace the ick.
176. 29:48You're gonna have to embrace the fact that you may not you may not have the confidence to even think like who am I
177. 30:01to make the world safe for llama washing in public. The toughest part of the daily practice is not the technology. The toughest part
178. 30:11of the daily practice is you confronting who you are and what your insecurities are and the limitations that you know like know in your soul
179. 30:30you can't do right. We've all got these things. Every one of us has these things. We've got these things where we're like I know I'm
180. 30:39good at that. I [snorts] know I'm good at that. I'm okay at that. These things I can't do. Like it's fact, right?
181. 30:53We now live in a world or very soon [snorts] we'll live in a world where that's no longer the case. Which AI for medical advancements? It
182. 31:07It's too broad. Gravity jump. It's a great question. It's too broad a question. Um, the the Georgia high school team
183. 31:18that just came up with a novel uh detector for Lyme disease that's that's effective two days after you're bitten.
184. 31:26Um, that might also be the foundation for a cure. They're using crisper. So, crisper technology is DNA. It's a DNA
185. 31:36editing kit. It's a piece of software where you can just go in and I don't quite know how it works, but you can isolate parts of a gene.
186. 31:48And so what these kids did is they figured out that the the lime spyroet when it infects your blood it kicks off a certain protein.
187. 32:00And so they figured out what that DNA sequence was. And if they I don't know. I don't know if I don't know. They
188. 32:06edited some sequence in some DNA related to Lyme disease and then because of that they turned it into like it's literally
189. 32:15like a COVID test, you know, where it's like you put the the drop, I guess, saliva. I don't know. You do the drop,
190. 32:21you put the drop on the thing and then if you get a line it's positive. If you don't get a line, it's not positive. They made that. [snorts]
191. 32:28[clears throat] But for detecting Lyme disease with this DNA editor, um something that I would pay very close attention to
192. 32:38on the medical front is um Alphafold from DeepMind, Google DeepMind. So Alphafold um has now not only predicted
193. 32:49all of the proteins of life and how they fold, which is it's something like 400,000 proteins that they've predicted
194. 32:57how they fold. And to put that in context, in the olden timey days, um, three years ago, it would take one PhD researcher,
195. 33:13one year to predict one protein, how it folded. One PhD, one year, one protein. they just predicted 400,000 of them with
196. 33:26like 95% accuracy, some crazy thing like that. And then Alphafold two was even better. And then Alpha Fold 3 that just
197. 33:35came out is like not only is it that the proteins, it's like all of the molecules of life and how they work and how they interact.
198. 33:44It's it it is mind-bogglingly complex what it what what they have done. Well, all the scientists that would have spent
199. 33:55a year figuring out how a protein folds can now just go into this this database or this this model and say, "Oh, I want
200. 34:05to deal with this particular protein. Show me how it folds. Now, let's design some molecules that can dock with that
201. 34:11thing." So, I would pay attention to those those big science projects um because the research labs that are
202. 34:19adopting those are going to really start accelerating. [snorts] Um and and there are probably there are probably specific AI tools
203. 34:27that I don't know enough about to to even mention. So, like I don't I don't know I don't know that world at all. Um
204. 34:34let me go back here and see if I'm missing anything. Um, I'm just going to put on SC screen really quickly here
205. 34:49the mastermind. [clears throat] There you go. Um, bit.ly-salon-mastermind.
206. 35:03So, so we were talking about daily practice before. Um, go there. If you've not joined the AI Salon mastermind, you should do it. Um,
207. 35:13the other thing that you should do, I don't know if you have have Festivus up and running, if if you've got a banner there. Uh,
208. 35:21there you go. If you go to community.thesalon.ai. Um, or if you go to aifestivist.com. Um, oh, my little vertical thing is I'm
209. 35:33now a little horizontal dude over there. All right. Um, AI festivist. And if if I sound like I'm talking like I'm trying
210. 35:44to be quiet, I am because I'm in a New York City apartment and I I [snorts] keep hearing people nearby
211. 35:50and uh so I'm trying not to be too loud uh because I am a I don't know if you knew this about me. I'm a loud talker.
212. 35:58Um but anyway, AI Festivus December 26th and December 27th. If you don't know about AIFest, go to aifestivist.com.
213. 36:08Register right now. It is a free event. Free f free. Not three. Free. It's a free event. Um it's 12 hours on Friday and 12 hours on Saturday
214. 36:24of free. Did I mention it was free? Um AI education. So, we're basically asking, you know, the smartest people we
215. 36:34know who are doing really cool things in AI, would they be willing to come and talk for an hour to our to our
216. 36:40community? So, we're putting this on with um Ann Murphy. Um and we're putting this on with uh the AI salon. So, Ann
217. 36:49Murphy and she leads AI in the AI salon community. And yes, Mr. IT, it is free. So, so go sign up right now and do me a
218. 36:59favor. Share that with friends of yours. Share that with friends of yours who are like, "I don't like AI." I'll tell you
219. 37:06right now, I don't like it. It uses too much water. It uses too much energy. It's going to the robots are going to kill us.
220. 37:16I don't like it one bit. It's the world's greatest plagiarism machine. Um, send it to them. Because what they will discover when
221. 37:30they come to AI Fest, what we discovered last year is that the people that are doing really interesting stuff with AI
222. 37:36are not talking about AI. They're not they're talking about what they care about in the world that they happen to be using AI to help
223. 37:50them pull off in really remarkable ways. You know, that was the comment that started the the the live tonight. You
224. 37:59know, hey, thank you for teaching me to use chat GPT with, you know, a handful of prompts. We're saving 130 grand a year in our business.
225. 38:09That that is tangentially related to AI. What's more important there is someone took the time to understand what it was,
226. 38:16understood what their business was, and made a difference. Right? I can't wait. last year was amazing. So yeah, please go register for that.
227. 38:27Festivus is guaranteed to be a hundred times more informative than the Tony Robbins seminar. There was a part of me
228. 38:33that was like, I wish I would have presented at that. There's a part of me that's like I'm glad I didn't. Um the
229. 38:38other thing tomorrow we've also got AI Salon presents. Um I'm going to be in a different space than this one. Um we've
230. 38:45got Danny Newman from Denver. He's an entrepreneur in Denver. He's a tech startup founder. He's had multiple
231. 38:52exits. He's saved businesses in Denver that were basically going out of business. He rescued them. Um, and he's
232. 39:00got a new AI space in Denver called ID345. And we're talking to him about uh maybe doing something with AI salon. So,
233. 39:08that's really exciting. Um, so he's going to be speaking tomorrow night. So, tomorrow night at 700 p.m. East Coast time, 5:00 p.m. Mountain
234. 39:18um is AI Salon Presents. Okay. So, if you go to the salon.ai or go to community.salon.ai, um go to our events calendar and you can
235. 39:29uh RSVP for that as well. All right. Beautiful. It's also really hot in here. One of the things about New York City apartments is
236. 39:41all of these buildings, all the the they're called pre war buildings, so preWorld War II, I suppose, they all tap into a massive
237. 39:54underground steam network, you know, like from the 1800s. [laughter] [clears throat] And every apartment in New York City
238. 40:06that's got that taps into that that steam infrastructure has has radiators that are full of steam, which is hotter than boiling
239. 40:16water cuz it's steam. And so and there's no way to regulate them. You can't really turn it off. It's
240. 40:23like either cold or like ripping hot. Um so you have to open your windows. The only way you can regulate temperature is
241. 40:30open your windows. So, so I thought I had to balance about right. No, it's getting really hot in here. So, that's
242. 40:38pretty funny. Um, AIDNA editing ask about AlphaFold. There's our AI bot. Who wrote the AI bot that comes here every
243. 40:46night? Who wrote it? Please tell me. I'd like to thank you. It's pretty cool. The furnaces. Yeah, exactly. Cam Captain,
244. 40:55where are those furnaces? Aren't they down? They're downtown, right? Or I guess maybe they're all over. I don't
245. 41:00know. They're probably part of Khed. Anyway, it was cool just walking walking the streets in New York tonight. Pretty cool. Um,
246. 41:10I was in the Bronx. I was in the Bronx for three days. The Bronx. I was in the Bronx. Went to Patricia's. Patricia's in
247. 41:19the Bronx. Had uh chicken parm. Uh, rigetony bologine. Bolognesei. Um, it was good. It was good. And this
248. 41:33[laughter] guy, this guy, Patricia's, this guy came in. What did he say? He goes, "Kyle, I haven't heard of a
249. 41:41Kyle in the Bronx for [laughter] 30 years." And he said there was some other neighborhood I should have been in where
250. 41:49all the all the Irish were. [laughter] I got called out for my for my uh h for my name. Yeah, he was
251. 42:06um go to Sylvia's in in Manhattan tripast. Did you go to the New York Botanical Gardens? It's in the Bronx. So
252. 42:16beautiful. No, I didn't. I just I just hung out. I hadn't seen it was it was my wife's family and I hadn't seen her
253. 42:21family in like six years. Like four human beings were created in the time since I last saw them. [laughter]
254. 42:35So anyway, um who's got any questions? Anybody have any questions? My husband went to for him there. Oh, cool. Nice.
255. 42:44Beautiful. Beautiful. Beautiful. Beautiful. Beautiful. Beautiful. >> [singing] >> Peter Bernson. The Tony Robbins AI
256. 42:58Summit was so bad. It was so bad. Oh. [laughter] Oh man. I credit you often in our AI council.
257. 43:13Oh, that's very sweet. I appreciate that. Irish, you can try Smith and Luisi. Um, AlphaFold has open access for
258. 43:25non-commercial use. There you go. Yeah. Okay, there you go. So, do you know how to do protein stuff?
259. 43:37Neither do I. But there's this thing now called Alphafold where we have chat GPT. We now have Gem uh Claude Opus 4.5,
260. 43:47which of the three big big new ones, you've got um Chat GPT 5.1, you've got Gemini 3, and then you have Claude Opus
261. 43:574.5. It appears that Opus 4.5 is the winner of the of the big models that were just released. Um, producer Brandon wanted to say to
262. 44:10everyone who bought his book and gave reviews, thank you. Um, I got to tell you, I'm very impressed with producer
263. 44:16Brandon's. So, producer Brandon, like I haven't done with the last two books I've written, what producer
264. 44:24Brandon did, this this is the equivalent of reading the manual for me. So, I understand why I didn't do it, but I
265. 44:30should probably do this for the next one. Um [clears throat] he went and he figured out using chat jeep how to market a book and he's like
266. 44:39following the advice and he's doing great. So he's getting his book out there. It's called the 10 minute 10,000 minute
267. 44:48practice. No 10,000 minute mindset. 10,000 minute mindset. So go go to go to Amazon. They can go to Amazon right now
268. 44:58and pre-order, right? Or can they buy it? buy it. It's live. Yeah, he he sold on the on the 30th anniversary.
269. 45:09Um he kicked it off. So So yesterday and it's live for 99 cents on Kindle and his request is get it for 99 cents for
270. 45:20Kindle. Read it and give it a review for him being swell. Right. Ain't that cool? I think that's really cool. really cool. And again,
271. 45:32this lives in that same in that same neighborhood of we're going to be living in a world we if you're resourceful right now, we live
272. 45:45in the world right now. The AI tools are going to get so good over the next three years that three years from now,
273. 45:53we will literally live in a world where anyone can do anything. Might be some accessibility questions, right?
274. 46:07But for the most part, anyone can do anything. And then my question to you and your question to you should be what do I want
275. 46:18to do? That's the harder question. That's the hardest question cuz we've always had excuses, right? Well, you know, I would sure like to
276. 46:29find a cure for Lyme disease, but you know, I'm not a scientist. Neither were those high school kids that
277. 46:36just found a cure for Lyme disease. That's chapter three of this book. There you go. There you go. The stuff I'm talking about here is not
278. 46:51rocket science. This is the [ __ ] we talked about in the early days of the web. The technology is ultimately not all
279. 46:57that interesting. What's more interesting is what do you do with it? And with AI, it is such a revolutionarily powerful technology.
280. 47:06That question of what do you do with it is can be really profound. >> [snorts] >> And so that's what I want you to start
281. 47:11preparing for, thinking about. Go to AI festivist, join the AI salon, join the mastermind. If you join the mastermind,
282. 47:19come to our Thursday mastermind practice lab meetings where you get to hang out with other people who are designing their daily practice.
283. 47:32Because what you're going to find is what I'm finding with daily practice is daily practice is hard. And so, how do you make it not hard?
284. 47:43Well, maybe you don't. Maybe it's always going to be hard for you. But, but you can get into a habit and you can
285. 47:49understand that other people are dealing with this, too. And here's how they've figured out how to make it work for
286. 47:55them. And one of those conversations is going to lead you to go, "Oh, that's really cool." Um, yeah,
287. 48:11my [snorts] AI song on Apple Music. Welcome, welcome home. Please sample it. Gravity jump. Do me a favor. Drop the link to
288. 48:20that in the AI salon in look what I made or the community feed. That's great. I have a theme day playbook for days. Nice. Love it.
289. 48:36[clears throat] [singing] Um,
290. 48:50I'm trying to think if I have anything else I want to talk about. There were a bunch of new video models
291. 48:57today. I think two or three video models today came out. Um, Google's VO 3.1
292. 49:15and Sora, is it Sora 4? Like, I don't even remember the numbers anymore. It doesn't matter. But the Sora model and the VO model are
293. 49:28are really they were pretty revolutionary and the and Nano Banana Pro the new image gen models pretty pretty remarkable. So the sense that I
294. 49:38get is the new models that came out today. One came out from Runway Gen 4.5. One came out from another company, Clling
295. 49:50came out with another one and that's now on a bunch of different platforms. Uh that one was called XO or 101 01. It
296. 50:00doesn't [ __ ] matter. Doesn't it literally doesn't matter. Um, I think they're all just incremental. Um, so I'm I'm not really paying too
297. 50:12much attention to them. It is possible I'm wrong. Deepseek just dropped a new model, a video model, Danielle,
298. 50:21if DeepSync dropped a video model that I'd be interested in. What did they drop? Danielle, everybody wants to know.
299. 50:35No, not a video model. I bought a $4 $49 smart lamp that was designed by the people who made
300. 50:50Pixar. Oh, I saw that thing. That was pretty cool. That was only 49 bucks. That was pretty cool looking.
301. 50:58Yeah, the people, one of the designers of the Pixar, you know, the lamp that's in the title of the Pixar thing, they
302. 51:06made a uh an actual swing arm lamp that like moves its head and looks at [ __ ] for you. Sora and Vio are still the best
303. 51:14video models. Yeah, I agree with that. I agree with that. Um but what's going to happen is the video models right now
304. 51:25um Midjourney Midjourney is a sneaky company. Um MidJourney is also I feel like MidJourney is slipping a bit. I feel like Nano Banana
305. 51:35um Nano Banana's ability to do style transfer is pretty close to mid Journeys um without all the complexity. Um, so I think MidJourney
306. 51:48runs the very real risk of being like Napster where they came in, they did all the [ __ ] wrong, but they they did all
307. 51:57the [ __ ] right, but they did all the [ __ ] kind of unethically and illegally. And they're going to take the heat for
308. 52:02that. They're going to take the hit for that. And they're probably not going to be resourced enough to keep up with all
309. 52:08the other models like that are that are coming at them from all sides. Um, but that said, Midjourney's video model is not a video
310. 52:20model. It's a component of a 3D world building model. So, where all of these video models are headed is understanding and replicating
311. 52:33simulating the world. And we'll be able to do that in real time in 3D with dynamic code generation. So you can
312. 52:45enter worlds, interact with worlds, generate new ones, interact with those. That's what's coming. So until I start to see signs of that,
313. 52:55like like at this point, the the video models are great. All right. So the wheels now look actually like they're
314. 53:02turning like a normal car, right? We no longer have the weird hands. We can do dancers now. We couldn't do
315. 53:10dancers for a while. Um, but I think where this gets really interesting, and I'm I'm pretty confident we see this in
316. 53:162026, probably mid 2026, is text to environment. And it ain't just about movies. It ain't just about movies. You're going to be
317. 53:31able to create a world and within that you'll be able to create a movie or you'll be able to create a game or
318. 53:38you'll be able to create some sort of data visualization that brings a concept to life that you've never been able to
319. 53:44do before. So, there's all sorts of implications for it. Google anti-gravity deleted someone's entire drive by accident. Wow.
320. 53:56Wow. Yeah. Be careful of these agentic tools. They just might delete all of your data. I would think that wouldn't
321. 54:05Google have backup of that somewhere. It skipped. Oh, it's skip the trash can and just hard deleted.
322. 54:21Nice. Oh, their local drive. That's bad. They had it backed up. I go, "What could possibly go wrong? What's the worst one of these agents
323. 54:36could do? Erase your whole hard drive?" Oh [ __ ] it just erased my whole hard drive. If that person had their hard drive
324. 54:44backed up, I have a feeling they were doing sneaky nasty stuff that they were trying to get it to break like that.
325. 54:50[snorts] Um, I grabbed a screenshot off my phone onto my computer and one image generator was able to read it. Oh, yeah. The
326. 55:00it it's amazing what these models can now see and understand. No, it was a novice user vibe coding. They admitted
327. 55:10they were clueless. They went Oh. Oh, Google admitted they went went wrong. Oh, they had no idea what they were doing.
328. 55:23Well, yeah, that's that's on Google, quite frankly. You know, if you make vibe coding tools that have a lot of power in them and you
329. 55:35make them accessible to people that don't know what the [ __ ] they're doing and you don't have safeguards in there
330. 55:41for clueless people, then that's kind of on Google. And I know the engineers are going to be like, "Well, you should read the
331. 55:47documentation. You You should understand how coding works. Yeah, you should. Who's got 10 years? Who's got 10 years to get good at coding
332. 55:58and understand what coding means? I don't. I just want [ __ ] I just want [ __ ] to work. Yeah, that's
333. 56:13we're probably the next three to five years is going to be a lot of bad growing pains with AI where AI uh is
334. 56:25promising on the surface but kind of shitty underneath. I saw a really good who is this from? It was from Ali K.
335. 56:31Miller I think. Yeah, I think it was Ali K. Miller. She put out a tweet today that said um Claude Opus 4.5
336. 56:44point for her made a significant jump. She said all of the models to this point if she would be doing any sort of
337. 56:52creative writing they would get her to like 70%. And that's been about my experience. my experience, I I often use the the 80%,
338. 57:00but if you really want to like write something, it's it's always so broken it repeats stuff. It's like 70% is pretty good. If
339. 57:09you're using a good model, you can do creative writing at at about 70% there, right? And then it's up to you to
340. 57:18reprompt and edit and Frankenstein [ __ ] together to get the other 30%. She said claudopus 4.5 went in in the 85 to like 93% range
341. 57:33um for her. So if it's a if it if you do a jump from 70% good to 90% good that's really significant because once you're at 90%
342. 57:45like this is my opinion. My opinion is if you've got a writing tool that can get you to 70%. It's a coin toss whether it's faster to
343. 57:58just write it on your own or write it with AI and then fix it, right? It's a coin toss like like writing is much slower than chat
344. 58:10gptting. But chat gpting is so rife with problems that in order to fix it, you're probably looking at somewhere in the
345. 58:19neighborhood, unless it's short stuff. If you go from 70% to 90%, now you're just an editor. Now you're going in and
346. 58:28cleaning and polishing rather than, you know, structurally fixing [ __ ] [snorts] Um, that's a that's a big big deal. So,
347. 58:38it looks like it looks like Claude Opus 4.5 is pretty good. Japan unveils a new self-sizing sneaker pod. They claim will
348. 58:46change buying shoes forever. You know, I saw the video of that, Brandon, and it it's this weird thing where it looks
349. 58:53like it's automatically like weaving threads around someone's foot, but then it like seems like it fills with liquid
350. 59:00and then they pull their foot out and there's no liquid on it. So, I assume that that was an AI animation. So maybe it's just a prototype or
351. 59:09something like that. If that was the actual machine, it was really [ __ ] cool. But I don't think that was a real
352. 59:14machine. Claude helped with film script analysis regarding my plants. I I don't know what Joy Joy Party is now
353. 59:22making films about plants and she's got to do script analysis about the plants in her films. So Claude's helping her with that.
354. 59:31That's pretty cool. Joy Perty, she's such a trip. >> [clears throat] >> used to drive me crazy when it went off
355. 59:39the rails at 70%. Seems so much better now. Which Claude was it that he mentioned? It was Ali K. Miller. She um
356. 59:46Claude Opus 4.5 is the one she mentioned as being really good. Oh, not green plants. What kind of plants?
357. 1:00:01Like manufacturing plants. Oh, plants in the script for continuity. Oh, yeah. Yeah. Oh, good. Yeah, those
358. 1:00:13are those are hard to track. That's good. I like it. All right. Um, I'm going to get out of here because I
359. 1:00:20probably have blasted out of my data cap. Tomorrow, we've got the AI salon and I'm gonna It looks like uh Oh, what's going on?
360. 1:00:33Your live will end. uh physical dexterity test. Um I'm going to do the live I'm going to do the AI salon from Union Square. So 14th
361. 1:00:46Street. I'm on 71st Street. So after the AI salon, so I've got an all day AI salon strategy meeting tomorrow
362. 1:00:57and then in the evening I've got salon presents and then I'm going to try to get back to this lovely place. Um,
363. 1:01:06so I'm either going to be so wiped out I can't go live or I'll go live for a little bit, but it'll probably be later.
364. 1:01:13Probably be 10:30 or 11 cuz I got to get up here. And it's, you know, it's the big city, you know, it's it's the city so nice
365. 1:01:22they named it twice. It's like it's the city that never sleeps. I mean, if you can make it here, you can make it anywhere.
366. 1:01:34I love Union Square. When will you be there? In Union Square. I'm I'm I'm working down there tomorrow. So, Andy
367. 1:01:45Andy Scarantino and I are doing an all day strategy session somewhere near Union Square. Um, nice to see you again. Thank you,
368. 1:01:53Silver Fox. Thank you, Source Camp. Um, plants and payoffs. Sorry, I wasn't clear. That's okay. That's okay. I
369. 1:02:07I like obscure references to creative projects. You know what you meant? I [clears throat] like trying to figure
370. 1:02:14that [ __ ] out. I thought it was like, you know, your characters all had different plants in their apartments and
371. 1:02:21some were green thumbs and some were killing their plants and some were succulents and others were perennials
372. 1:02:27and some were annuals. Like there's a lot to manage when it comes to plants in movies. So, it seemed plausible to me. [laughter]
373. 1:02:42All right, we have a new AI salon partner leadpages.com. Cool. 30% off. Awesome. Beautiful. All right. Details in AI
374. 1:02:57Salon. Which which which uh space? Partnerships. That's right. There's a partnership page in the AI salon. So, if
375. 1:03:06you go in there, you can find out who we've got partnerships with and we've got deals. Okay. Robert Plant. I like
376. 1:03:12it. All right, I'm out of here. Peace out. There. Community.thesalon.ai. Here, do this. Go here. All right. Good easy, everybody. Peace
377. 1:03:29out. Um, what I've been watching I've been watching Gary Vee do do lives every day. That man's a [ __ ] machine. like he's
378. 1:03:40got because he's got a whole staff like he knows he knows all the Tik Tok tricks. He's like go down to the righthand
379. 1:03:46corner, do this, do that, join the fan club, go over here, do this, do that. And I'm like, hey everyone, this is cool. [laughter]
380. 1:03:59Oh man. All right. Someday. Someday. All right, everybody. Peace out. [clears throat] Beony, beauty. All right, I'll see y'all
381. 1:04:13later.