
AI Learning Lab
05/11/2026 - OpenAI Launches a Four Billion Dollar Deployment Company to Help Enterprises Scale

Video2026-05-251:32:0417 views
Description
Kyle Shannon explores a pivotal week for OpenAI, highlighting the launch of their new deployment company and the Daybreak cybersecurity initiative. These developments suggest that powerful frontier models are imminent, shifting the focus toward large-scale enterprise integration and real-time, full-duplex interactions. Kyle also examines the "stealth" rollout of Google’s Gemini Nano and what it means for local AI processing on consumer devices.
The discussion moves beyond technology to address "The Great Repurpose" and how professionals can navigate the "Seven Economies" of AI adoption. Kyle explains that as traditional expertise becomes commoditized, a person’s unique journey and human identity become their most critical assets. He encourages viewers to prioritize community and personal storytelling to remain resilient as the professional landscape shifts under the weight of rapid automation.
#OpenAI,#ArtificialIntelligence,#FutureOfWork,#GenerativeAI,#DigitalTransformation,#Cybersecurity,#TechNews,#AIStrategy
Chapters:
00:00:00 Welcome Everyone
00:05:22 Tonight's Agenda
00:09:37 Codex Pets Bug
00:12:15 OpenAI Deployment Company
00:16:13 Large Scale Consulting
00:20:12 Service Demand Acceleration
00:23:13 Seven Economies Model
00:27:31 Economy Five Transition
00:30:54 Salon Community News
00:34:13 The Goop Phase
00:38:18 OpenAI Daybreak Security
00:43:22 Google Gemini Nano
00:48:07 Thinking Machines Demo
00:54:18 AI World Models
01:00:03 The Great Repurpose
01:03:32 Knowledge Work Future
01:07:59 Corporate Tech Lag
01:11:10 Generous Leadership Sharing
01:13:49 Crossing the Chasm
01:18:03 Episodic Short Form
01:22:32 Identity Over Expertise
01:27:49 Resilience and Value
01:31:30 Closing Thoughts
Chapters
0:00Welcome Everyone5:22Tonight's Agenda9:37Codex Pets Bug12:15OpenAI Deployment Company16:13Large Scale Consulting20:12Service Demand Acceleration23:13Seven Economies Model27:31Economy Five Transition30:54Salon Community News34:13The Goop Phase38:18OpenAI Daybreak Security43:22Google Gemini Nano48:07Thinking Machines Demo54:18AI World Models1:00:03The Great Repurpose1:03:32Knowledge Work Future1:07:59Corporate Tech Lag1:11:10Generous Leadership Sharing1:13:49Crossing the Chasm1:18:03Episodic Short Form1:22:32Identity Over Expertise1:27:49Resilience and Value1:31:30Closing Thoughts
Transcript
0:27 Heat. 0:46 Heat. 0:57 Welcome, welcome, welcome, welcome 1:00 everybody. How is everybody doing? 1:04 Welcome to the AI learning lab live. 1:07 What is shaking? What's going down good 1:09 people? 1:11 All right, we're in here. Kelly Camp is 1:13 in here. Sharon Crawford, just as I was 1:15 learning how to slow down the AI 1:16 business I've been waiting for for three 1:18 years just kicked in. I haven't seen the 1:21 video game in a while. That's awesome, 1:23 Kelly. I'm really thrilled for you, 1:24 Kelly. Scary, too, I imagine. Yeah. 1:28 Yeah, exactly. Just as you're getting 1:30 exhausted of not having enough business, 1:32 you're going to get exhausted of having 1:34 too much. Um, but by the way, Kelly, uh, 1:38 one of the things that we'll be talking 1:39 about tonight, 1:42 um, I think could be very interesting 1:45 for what you're up to. 1:47 Um, but we'll see. But I but I just I I 1:50 have a I have a sneaking suspicion. So, 1:53 it looks like Kelly has been paying 1:55 attention to what's going on with Open 1:57 AI. So, won't be a surprise to her. But, 2:01 um hang on a sec. Let me get all my 2:03 devices managed. And 2:07 >> you can make money with 2:11 >> that's that's about to happen. I think 2:16 we we are in we in just bonkers times. 2:20 Bonkers times. Bonkers. Bonkers. 2:22 Bonkers. Okay, welcome. So, it's Monday. 2:25 My name is Kyle Shannon. This is the AI 2:27 learning lab. Welcome. Hello, Ivan. How 2:29 are you? Um, welcome to the AI Salon. If 2:33 you're new here, if you're new to the 2:35 live, if you're new to the live 2:36 streaming in the AI salon, if you're new 2:38 to the AI salon, be sure to check out uh 2:40 I don't know which side of me it's on. 2:42 So, here's a question. For those of you 2:44 that have been here before, is all the 2:47 navigation for the AI salon, is it on 2:49 the side of where my name is, or is it 2:51 on the side where the AI learning lab 2:53 logo is? Where's the Where's the 2:56 left-hand navigation of the salon? 2:59 Which one? Right. Right doesn't help me 3:03 where my name is or where that thing is. 3:08 Problem is, I don't know right from 3:10 left. If my image is being flipped, it 3:12 might be backwards for me. 3:16 Anyone? Buler. Buler. Buler. 3:20 It's nowhere on an iPhone. 3:24 All right. I guess we got some people 3:26 from from iPhone sticking here. Chat on 3:29 your left. So over here on where my name 3:32 is. Your name is on the left. Icon is on 3:36 the right. Okay. Name's on the left. 3:38 Okay. So it's left. So, it's here. So, 3:42 if you're on if you're on desktop, go 3:44 check out all of the different spaces in 3:46 the AI salon at some point tonight. Um, 3:50 okay. We can read Monopoly. Oh, yeah. 3:52 That's Wait. Okay. If you can read 3:54 Monopoly, that's backwards for me. 4:01 Other left. So, the navigation is over 4:04 here. Is that correct? 4:07 Underneath the AI Learning Lab logo. 4:09 That's where the navigation is for 4:15 I I I 4:19 I'm exhausted. I'm exhausted. 4:23 Producer Brandon entered the studio. 4:25 Hey, producer Brandon. Okay. Todd Waller 4:27 is the first one. Right. Wait. Right. 4:30 Other left. Right. Correct. Okay. Over 4:33 here is the navigation for the AI salon. 4:35 Over here. 4:37 Who's on first? Yeah, exactly. This is 4:39 This is This is We need Producer Brandon 4:42 in here. Something terrible. 4:44 I was just trying to tell the good 4:46 people. Producer Brandon where where the 4:48 stuff is over here. Okay, so here's what 4:49 we're going to talk about. I have you on 4:51 my live. Very nice. It's like teaching 4:53 kindergarten. No, Kelly, it is much 4:56 worse than teaching kindergarten. That's 4:58 correct. 5:04 Oh, man. kindergarteners, you expect to 5:07 be playing with spitballs. 5:12 Upstanding professionals, you know, come 5:15 on, you are irregular. So, there's that. 5:18 Okay. So, so here's what we're going to 5:20 talk about tonight. 5:23 And we're going to dig into some some 5:24 detail. I want to go I want to go read 5:26 about this stuff and think about it. We 5:27 we'll read it together and we'll talk 5:29 about it. 5:32 But 5:33 so there there have been rumors that 5:35 there's going to be some new stuff this 5:37 week. It looks like there's going to be 5:38 some new stuff next week. So producer 5:40 Brandon, one thing that I didn't put in 5:42 our little X 5:44 channel, the direct channel, is I want 5:47 to go look at the video from the new 5:49 Gemini model, which if if the video is 5:52 in fact from the new Gemini video model, 5:55 it's flipping crazy 5:58 what it's doing. So it's so so we're 6:01 going to go we're going to go talk about 6:03 that. But 6:05 um OpenAI today made two major 6:08 announcements 6:10 that one is the um the announcement of a 6:18 a new company to deploy AI. 6:24 And so it's a combination of investment 6:26 banks and big consulting companies. Um 6:30 so we're going to go look at that in 6:32 detail and it's got some really 6:33 interesting components to it that I 6:36 think there could be a lot of 6:37 opportunity especially for people that 6:39 you know know how to talk about this 6:41 stuff, know how to deploy this stuff. I 6:43 think one of the things that's probably 6:45 going to be highly overlooked is the 6:46 kind of stuff we're talking about with 6:47 the great repurpose the the the me 6:50 meaning and identity stuff. 6:53 Um, 6:56 so I think there's some big opportunity 6:58 there. And then, yep, Gemini is getting 7:01 better. And then the other the other 7:03 thing that they announced today was a 7:06 cyber security 7:08 initiative. 7:09 Um, 7:12 that is I think the slicker version of 7:15 what Anthropic did with the Mythos 7:18 release, which I thought the the 7:20 Anthropic announcement of the Mythos 7:23 release was kind of 7:27 I don't know, cheesy. 7:30 It was cheesy. The Gemini Voice on iOS 7:33 app is great and free. Yeah, Gemini is 7:35 getting better. Um, 7:38 the new real time voice for chat GBT has 7:43 gotten better and it's really 7:44 interesting. Um, the new mirror moratti 7:49 company 7:50 has a real-time voice, real-time 7:52 translation. What what blows my mind 7:55 about all these frontier model companies 7:57 is they're all launching the exact same 7:59 features at exactly the same time. 8:03 So, and two months after that, China 8:05 will be there with the open- source free 8:07 version of it. It's been like that for 8:09 three and a half years. And I don't I 8:10 don't see it changing. I don't quite 8:12 know. 8:14 I assume this is because they just keep 8:16 hiring engineers and engineers keep 8:18 moving from from frontier company to 8:20 frontier company and they all just 8:22 >> their intellectual property is just all 8:24 over the place. So, 8:26 >> um Oh, yeah. Lovable has a new interface 8:29 builder, right? Like their their Yeah. 8:32 Can you make the Can you make the 8:33 purple? Oh, I can. Never mind. I got it. 8:36 Here. 8:40 Okay, good. Perfect. Perfect. 8:43 Wait, how do I click out of this? 8:47 Oh, you need to click off of Lovable 8:49 Aesthetics. Um, yeah, Lovable's got a 8:52 new uh a new thing out today. Um, that 8:55 apparently their interface is getting 8:57 better. I didn't think their interface 8:58 was all all that bad, but apparently 9:01 it's getting better. Um, I've heard that 9:04 there's going to be a new open AI design 9:06 tool. So, last week or the week before 9:09 we got the clawed design tool. So, 9:12 again, all these companies are releasing 9:15 all this stuff. But the thing that I 9:18 think is important for us to look at 9:19 tonight, 9:22 let me share my screen. 9:38 Oh, I did a cool little Look at my cute 9:40 little new avatar, dude. 9:42 So, so this is frustrating. I don't know 9:44 if there's anyone if if Gareth's in 9:46 here, if anyone knows what the hell 9:48 they're doing, but in OpenAI's codeex, 9:52 they added a new thing called pets. 9:55 And apparently you're supposed to just 9:58 be able to slash pets and your pet shows 10:00 up. Didn't work in mine. And I did all 10:04 this install stuff and I followed 10:05 tutorials and I even made myself a 10:07 custom pet that looks like me and and 10:10 Codex did this remarkable stuff. In 10:12 fact, let me show it to you. I have 10:14 Codex open here. 10:17 Um, 10:18 if I go to settings and then I go to 10:20 appearance, 10:22 then I scroll down here and I click on 10:24 pets, I scroll all the way to the 10:26 bottom, I've got this this pet, Kyle, a 10:29 fun, relaxed guy who celebrates when 10:31 good things happen. So, that's me. But 10:34 if I go over to 10:36 find a 10:39 go here and I go let's see 10:45 I go here and I go users and I go story 10:48 vine 10:51 and let's see 10:57 Here. 11:06 I lost something. 11:27 H. 11:32 Well, the thing I wanted to show you, I 11:34 don't know where it is now. It got It 11:36 got moved. It looks like Codex moved 11:37 something. 11:39 Anyway, it made all these animated 11:41 sprites for this little character. But 11:43 anyway, when I when I'm in when I'm in 11:45 Codeex, um, it I can't get that little 11:48 dude to show up. So, I don't know if 11:50 anyone knows how to do that. But if I 11:52 type in slash pets or slashpet and hit 11:55 return, nothing happens. I get no little 11:57 dude. I don't know how to get my little 11:59 dude. Anybody know how to get my little 12:01 dude? It's an acknowledged bug that it 12:03 doesn't show up, but I can't ever I 12:05 can't seem to figure out how to fix it. 12:07 Anyway, okay. Back here. So, let me go 12:12 to 12:15 I can see clearly now the rain is gone. 12:25 Okay, 12:27 so May 11th, 2026, OpenAI AI launches 12:31 the OpenAI deployment company to help 12:34 businesses build around intelligence. 12:37 OpenAI has agreed to acquire tomorrow. 12:40 Tomorrow, 12:42 giving the OpenAI deployment company 12:44 experienced forward deployed engineers 12:47 from day one. 12:50 OpenAI is launching the OpenAI 12:52 development or deployment company, a new 12:54 company here. Let me make these bigger 12:55 so you can read it. It's kind of crappy. 13:01 Let me try logging in. 13:07 That's fine. 13:08 Um, 13:12 the Open AI deployment company will 13:14 extend OpenAI's ability to embed 13:17 engineers um specialized in frontier AI 13:20 deployment known as forward deployed 13:23 engineers or FTEEs into organizations 13:25 working on complex problems in demanding 13:28 environments. These FDEEs will will work 13:30 closely with business leaders, operators 13:32 and frontline teams to identify where AI 13:35 can make the biggest impact. redesign 13:37 organiz organizational infrastructure 13:40 critical workflows around it and turn 13:42 those gains into durable systems. Okay, 13:45 so 13:48 we're now entering a new phase of chat 13:50 GPT where so here's here's 13:55 what this signals to me 13:57 is that 14:00 the models that are coming are about to 14:02 be really flipping good 14:05 like G GPT6 or you know whatever the you 14:09 know mythos from from anthropic right 14:12 it's it's so good it's so powerful we 14:14 can't give it to you. Um, so that's kind 14:17 of what this says to me and that and 14:20 that it's going to be worth companies 14:22 time 14:24 to really do this right and it's not 14:27 going to be obvious how to do it right. 14:29 And so they're going to put together 14:31 these these they've put together this 14:34 company to be the services component of 14:37 you know their token selling operation 14:40 in connection with the OpenAI deployment 14:42 company's launch. OpenAI has agreed to 14:44 acquire tomorrow, an applied AI 14:46 consulting and engineering firm that 14:48 helps enterprises turn AI into 14:50 operational advantage. The acquisition 14:53 will bring approximately 150 experienced 14:55 FTEEs and deployment specialists to the 14:58 OpenAI deployment company from day one. 15:00 So they're starting with 150 people. 15:03 That's going to grow very very quickly 15:06 because listen, the OpenAI deployment 15:08 company is a committed partnership 15:10 between OpenAI and 19 global investment 15:13 firms, consultancies, and systems 15:16 integrators. The partnership is led by 15:18 TPG with Advent, Bane Capital, 15:20 Brookfield as co-leading founding 15:22 partners and B Capital, BBVA, Emergent 15:26 Capital, Gana, Goldman Sachs, Soft Bank, 15:28 Warbert, Warberg, Pinkis, and WCAS as 15:32 founding partners. Investors also 15:34 include leading consulting and systems 15:37 integration firms including Bane, 15:41 Capgeemini, McKenzie. The deployment 15:43 company will work closely alongside 15:45 OpenAI's frontier alliance partners and 15:48 the broader industry to drive a AI 15:50 adoption. Um, 15:52 the OpenAI deployment company is a 15:55 majority owned and controlled by OpenAI. 15:58 Majority owned and controlled by OpenAI, 16:01 giving customers a unified experience 16:03 whether they work with OpenAI, the 16:05 deployment company, or both. 16:08 Now, here's the thing where I think 16:10 there's some interesting opportunity for 16:12 all of us. 16:14 It will launch with more than $4 billion 16:17 of initial investment, which it will use 16:19 to scale operations and acquire firms 16:22 that can accelerate our mission. So 16:24 they're starting with 150 people and 16:27 they've acquired this company and I 16:30 don't know if the acquisition of that 16:32 company comes out of the four billion if 16:34 the four billion is on top of that but 16:37 they're investing in acquiring firms 16:40 that can accelerate being able to deploy 16:43 um to different companies. Now couple of 16:46 interesting thoughts on that. 16:48 If you're dealing with companies like 16:50 Bane, Capgeemini, McKenzie, 16:53 these are big, big systems integration 16:55 firms, big consulting firms, they're not 16:58 going to wipe their butts for any 17:00 company that's less than, I don't know, 17:03 20,000 employees, 50,000 employees. 17:06 Like, these are going to be big 17:07 engagements, big money. 17:10 There are a lot of companies out there 17:12 between, you know, a,000 and 10,000 17:15 people, between 10,000 and 50,000 17:17 people. There's a tremendous amount of 17:19 opportunity below where these guys are 17:22 going to start. 17:25 My instinct 17:28 is that a lot of this $4 billion 17:32 will will go for filling in the 17:37 opportunity gap, right? So, they're 17:39 going to start with the high-end 17:40 customers, but everyone's going to need 17:42 AI, and everyone's going to need to 17:44 figure this out, and no one's got it 17:46 figured out, right? 17:50 Um, why deployment matters. Open AI was 17:53 founded as a research and deployment 17:55 company. From the beginning, we've 17:57 believed that building powerful AI 17:59 models is only part of the work. The 18:01 real impact comes from helping people 18:02 and organizi organizations use those 18:05 systems safely, effectively, and at 18:06 scale. Over the past several years, when 18:09 more than a million businesses have 18:10 adopted OpenAI's products and APIs, 18:13 across those deployments, one pattern 18:15 has become increasingly clear. The next 18:17 stage of enterprise AI will be defined 18:20 by how effectively businesses can deploy 18:22 this technology into real world use 18:25 cases and how well our alliance partner 18:27 ecosystem can help them. 18:30 Right? So, if a bunch of companies go, 18:34 "Hey, we bought you all chat GPT. Good 18:36 luck." which is essentially what's 18:38 happened up to this point. 18:41 And they're like, there's no ROI. Yeah. 18:43 No one knows what to do with it. 18:46 You dumb dumps. No one knows what to do 18:48 with it, right? Are you using it in the 18:51 seauite? Are you using it, you know, in 18:54 senior management? Do you even know what 18:56 this thing is that you've given to 18:57 people? 18:59 Um, building where the frontier is 19:01 headed. We launched OpenAI deployment 19:03 company as a standalone business unit. 19:05 So it can develop the operating model 19:07 pace and customer focus this work 19:10 requires the operating model pace and 19:14 customer focus this work requires. 19:16 Right? So if I'm reading that right, 19:18 it's like shit's going to be moving 19:20 faster than Fortune 500 companies can 19:23 deal with it. It's likely going to be 19:25 threatening business models, operating 19:27 models, 19:28 and it's probably going to change how 19:30 customers engage with companies, 19:34 right? So, this is this is a smart move, 19:38 wholly owned delivery, right? And this 19:40 is like we've seen this before, right? 19:42 IBM's done this for years. 19:45 At the same time, it will operate as an 19:47 extension of OpenAI, keeping customers 19:49 closely connected with research, 19:51 product, and in-house deployments. 19:52 Right. So, so their best customers will 19:54 get the best models. 19:57 Streamyard comment. Let's see. Let's 20:01 jump over here. 20:04 Um, Chris Bruno, I feel seen. I know, 20:07 right? But what does it all mean, Kyle? 20:12 This is exactly the focus of what I'm 20:15 doing now. Exactly. Exactly. Hi, Kelly. 20:18 This is Chris. So that's what I was sell 20:21 saying, Kelly, that 20:23 this is so just like you're seeing your 20:26 company accelerating 20:28 um or demand for your services 20:31 accelerating, if you do it right, you'll 20:33 likely bring on other consultants. If 20:35 you do that right, you'll likely, you 20:36 know, sort of ladder up to different 20:38 kinds of clients and things like that. 20:40 um 20:42 the demand for people who know how to 20:46 approach AI, 20:48 I think, is about to get insatiable, 20:50 right? So, as as the models get better 20:53 and better and better, people are going 20:56 to be like, "What do I do with this? 20:57 What do I do with this?" Anybody buler 20:59 little bit? Any any help? Any help? Um 21:02 so, I think there's really big 21:03 opportunity here. Um yeah, let me flip 21:08 back over there. Oh, it killed my killed 21:10 my screen share. 21:18 I need to change my career. Cam Katkin, 21:20 I'm late. What's the topic tonight? So, 21:22 the topic tonight 21:24 is 21:27 um this this Am I sharing this 21:31 correctly? Yes. 21:34 um this article from OpenAI or not this 21:37 article but they they basically are 21:39 launching a deployment company 21:44 so let's see let's go down here 21:46 accelerating deployment okay wait here 21:48 we go let's see all right we launched 21:51 the open AI deployment company as a 21:53 standalone business unit at the same 21:55 time open AI will operate as an 21:56 extension of open AI you know favoring 21:59 customers so basically 22:02 how open AI works right now is if you 22:05 don't spend whatever it is a million 22:07 dollars a month with them, they don't 22:09 care about you. It doesn't matter. 22:10 Whatever you just, oh yeah, we promised 22:13 you a chat GP custom GPT store. Sorry 22:15 about that. Piss off. Right. So if 22:18 they're if you're spending 10 $10 22:20 million a month or more, they'll pay 22:22 attention to you. Now, what they're 22:23 laying layering on top of that is if you 22:26 spend 10 million or more a month and 22:29 you've you're using our deployment 22:30 company, then we'll really pay attention 22:33 to you. So, so we've seen this model 22:35 before. Um, the connection is a major 22:38 advantage. The OpenAI deployment 22:40 companies, FDEEs, forward deployed 22:42 engineers will be able to build for 22:46 where OpenAI's frontier capabilities are 22:49 headed, giving customers systems 22:52 designed to improve new models 22:55 improve as new models, tools, and 22:56 deployment patterns come online. 22:59 customers can move faster from day one, 23:01 spend c capital on durable systems, and 23:04 stay ahead of competitors by building 23:06 around the capabilities of what's coming 23:08 next. So, I want to I want to pop 23:10 something up here that I've talked about 23:14 a fair amount of times, but I still 23:16 think it is worth looking at again. 23:25 and and why why again I think this is 23:27 smart. 23:38 Okay, so the seven economies. 23:47 So if you don't know the seven economies 23:49 the this is this is basically a model 23:51 that got invented here on the AI 23:53 learning lab live 23:56 that was the answer to the question I I 23:58 said things are about to get weird and 24:00 Kelly Camp said but Kyle what does weird 24:02 look like and this the result was this 24:06 that as we've got AI accelerating away 24:09 from us and as most people don't even 24:11 know what AI is that that the world is 24:13 going to kind of shatter splinter into 24:15 these different economies 24:17 largely based on level of AI adoption. 24:19 So one of the things one of the places 24:22 my head has been is so so just quickly 24:27 the four economies to the left are 24:29 basically business as usual with a 24:31 little bit of AI adoption. So the analog 24:33 economy is screw it I'm going to go grow 24:36 blueberries. The legacy economy is I'm 24:38 not going to deal with AI AI at all. 24:40 We're just going to run it how we've 24:41 we've always run it. Economy three is 24:44 the efficiency economy. We're going to 24:45 use AI a little bit to make what we 24:47 currently do more efficient. And then 24:49 the transition economy is we're going to 24:51 do some innovation, but we're basically 24:53 going to keep doing what we were doing 24:55 before in the same model. 24:58 Economy five is a really interesting 25:00 economy because what economy five says 25:03 is I'm going to take a going concern 25:05 like I'm going to take an existing 25:07 company that's already got revenue. It's 25:10 already got customers. It might even 25:11 have profit and I'm going to burn down 25:13 the org chart. I'm going I'm going to 25:16 try to anticipate what AI makes 25:18 possible. I'm I'm going to try to 25:20 anticipate how I'm going to survive and 25:23 flourish with AI moving forward. Um, the 25:27 only company that I know that is 25:29 aggressively doing this right now is 25:31 Jack Dorsey's Block. So, the guy that 25:33 founded Twitter is he he laid off 4,000 25:37 of 10,000 people. He's replacing middle 25:42 management function with AI and then 25:44 basically putting all the people around 25:45 the edges of the organization to 25:48 interface with other human beings and 25:50 use the human beings for intuition and 25:52 and work and use the AI as a traffic cop 25:56 basically. Um, 26:00 so the thing that I've been trying to 26:02 get my head around is how are these econ 26:04 how are these companies in economies 2, 26:06 three, and four going to have the 26:09 foresight, the insight to really 26:11 understand what AI makes possible and 26:14 what's coming because economy six are 26:17 like small little nimble companies of 26:19 like 5 to 10 to 20 person um full AI AI 26:24 native companies that are going to 26:26 completely reinvent how economies 2, 26:28 three, and four work, right? They're 26:30 going to go compete with them. So, at 26:32 some point, if if the companies in 26:34 economies 2, three, and four really want 26:36 to survive, they're going to have to 26:38 flip into economy 5. So, does that take 26:42 them a decade to do or can they do that 26:44 in two years? And and what I would argue 26:46 is that what OpenAI 26:49 announced today with with the deployment 26:51 company is is that in theory, if they do 26:56 this right, and if this company's right, 26:58 in theory, if your company is working 27:01 with this deployment company and you 27:03 have access, you know what's coming 27:04 sooner than other people know what's 27:05 coming and these people have their [ __ ] 27:07 together, they're going to help you say, 27:09 "Here's what management looks like for 27:11 companies that are all AI. Here's what 27:13 business models look like, right? Here's 27:15 here's all of the things that are coming 27:17 and here's where you are. So, how do we 27:19 get you from economy 3 to economy 5? 27:24 All right, let me go look at a pin. 27:27 There's a pinned comment. Uh, let's see. 27:32 Hide. What's that say? Oh, I want to 27:36 consult for economy 5 companies and 27:39 build an economy 6 company. Come join 27:41 me. There you go. Mary Carol is is a big 27:45 systems thinker, understands this world, 27:48 understands all this stuff. Here's the 27:50 thing, Mary Carol, about economy. Five 27:53 companies are going to be really rare 27:54 right now. They're going to be very 27:56 rare, right? It's taking an existing 27:59 company that's got some amount of 28:01 resources 28:04 and and having leadership that's willing 28:06 to burn down the org chart. So my my 28:09 advice would be aspirationally shoot for 28:12 economy five, 28:14 but start with economy three and four. I 28:16 think most companies right now, if 28:18 they're using a AI at all, are probably 28:20 economy 3 companies. and how you get 28:23 them to transition from that to even 28:25 thinking about economy 5. That feels 28:28 like that could be a year just to figure 28:31 out where they are, where their pockets 28:33 of AI brilliance are, 28:36 what competition's going to look like in 28:38 the future, what an economy 5 company 28:41 would look like, how we would, you know, 28:42 what would need to transition. And then 28:45 to your other point, the economy six 28:47 companies, those are going to be a blast 28:49 because they can just invent, you know, 28:51 that sector from the ground up and just 28:54 say, "Here's a new way to do it." 28:57 All right. And I'll take care of all the 28:59 business that OpenAI doesn't care about. 29:00 That's that's my point is I think that 29:02 because they're starting with Bane and 29:04 McKenzie and these big ass consulting 29:06 firms, 29:08 they're going to start with the Fortune 29:09 500, Fortune 1000. 29:12 there's a lot more small, medium, and 29:14 big companies that that do not hit that 29:17 target market. 29:19 Um, I mean, this is this this the 29:22 initial customers of that deployment 29:24 company are going to be like, are you 29:26 willing to spend $500 million a year, a 29:29 billion dollars a year, then we'll then 29:31 we'll really come help you. If not, good 29:34 luck. Here's chatgbt.com. 29:38 Just ask it yourself. 29:42 Where can I learn more about the 29:43 economies? Um, I've got Champ. 29:50 What's going on? 29:54 Um, there's an article on LinkedIn, 29:57 Ivan, that I wrote about the seven 29:59 economies, and I've got a couple of 30:01 videos about it. I think there's a video 30:02 in the AI salon. I don't know if if 30:04 producer Brandon if if did I put a clip 30:07 of the seven economies in the salon? I 30:10 think I may have but I can't remember. 30:15 Oh man. 30:18 Okay, 30:19 so this looks different on mobile. 30:28 Why can I not see 30:33 I can't see pinned comments. 30:40 Strange. 30:42 Andy's in the house. What's happening? 30:45 There's so much opportunity right now. 30:48 Commercial break. Pin up on screen. 30:52 Oh, pin equals up on screen. Ah, got it. 30:54 Okay. Um, okay. So, it's 8 o'clock. So, 31:00 um, 31:02 one of the things that 31:05 I would strongly encourage you to do 31:08 is 31:11 now that you're in the AI salon, start 31:13 participating. If you haven't introduced 31:15 yourself, over here toward the top is a 31:19 space called introduce yourself. 31:22 Go do that. Introduce yourself to the 31:25 the to the community. What I would also 31:27 request if you've been in here a while 31:29 also go to introduce yourself and scroll 31:32 down through people who've introduced 31:34 themselves and welcome them. One of the 31:37 amazing things about the irregulars and 31:39 the people that show up to this channel 31:41 night after night after night is how 31:43 welcoming and accommodating and generous 31:45 you all are. And so what I'd like to do 31:47 is start to get that energy happening 31:48 inside the AI salon. So a fair amount of 31:51 people have have joined recently. Some 31:53 have introduced themselves. So go into 31:55 the introduce yourself section and and 31:56 join. Um the other thing that you should 32:00 think about doing is creating work and 32:02 sharing it. Sharing sharing it in look 32:04 what I made or in the community feed or 32:06 the community chat. Just get your work 32:08 out there. Start putting yourself out 32:10 there. 32:12 You'll also notice there's a space um 32:14 called the AI salon mastermind. And the 32:16 mastermind is a subscription area of the 32:19 salon. It's got a lot of stuff going on 32:21 it. We've got the practice the the AI 32:23 salon mastermind practice which is a 32:25 daily practice about um being more 32:30 intentional about how you use AI. Um one 32:34 of the things that was was mentioned 32:36 both in the pre-show and in the show 32:38 here is that things are moving so fast 32:42 right now. It is so overwhelming right 32:44 now. The people in the AI salon 32:47 mastermind that are doing this daily 32:48 practice. So we meet once a week um to 32:51 work with each other about how you doing 32:54 on your practice, what's your practice 32:55 looks like, what does it feel like. Um 32:59 all of those people are are slowing 33:02 down, right? They they are not 33:05 incapacitated 33:07 by the insane productivity gains of AI, 33:11 right? I think we all have been. I think 33:13 one of the things that's happening in 33:14 the AI salon pretty consistently is 33:17 people come in, they get excited, 33:19 they're like, "Holy [ __ ] I can do all 33:20 this stuff. Look at all this stuff I can 33:21 do. I'll do this stuff. I'll do that 33:23 stuff." I was one of those. I was 33:24 leading the charge. I was the little 33:26 pied piper. 33:28 Let's go make musicals. Let's go do 33:29 this. Let's go do that. 33:34 It gets to the point that you literally 33:36 can't keep up with it. Even your own 33:38 ideas because it is a friction-free 33:41 environment. AI is frictionf free, 33:44 right? You have an idea, boom, there it 33:46 is. That's one of the brilliant things 33:48 about it. It's also one of the curses of 33:50 it. So, if you haven't joined the 33:52 mastermind, please do. Um Andy 33:54 Scarantino, who's here tonight. Yeah. 33:56 Productive to the point of 33:57 incapacitation. Um she's midway through 34:01 a a workshop on um decoupling your 34:05 identity from your job. And tomorrow is 34:09 week three and we're talking about the 34:11 goop phase. 34:14 So, so if you haven't joined that 34:16 workshop, if you're in the mastermind, 34:19 um, tomorrow's going to be a fun one. 34:22 And the goop phase, if you don't know, 34:24 you got you got yourself a caterpillar 34:26 and then they put themselves in the 34:28 cocoon and then they emerge as a 34:30 butterfly. 34:32 Well, I didn't know this. I should I 34:36 don't know why I didn't know this, but I 34:37 didn't know this. But in between those 34:38 two forms of life, they basically 34:42 disintegrate into goop. They are 34:44 nothing. They're no longer caterpillar. 34:47 They're no longer They're not yet 34:49 butterfly. They are goop. So, as you 34:52 decouple your identity from your work, 34:55 it's going to feel like goop. It's not 34:57 going to feel great. It's going to feel 34:58 weird and unfamiliar and things like 34:59 that. So, that's what we're talking 35:00 about tomorrow. That's what Andy's going 35:02 to lead us through tomorrow is the goop 35:04 phase, which I'm really excited about. 35:06 So, if you haven't joined the 35:07 mastermind, good time to do it. The 35:09 great repurpose, there's a lot of 35:10 content in there. There's a lot of stuff 35:12 coming. Um, and anyway, and just just 35:15 whether you join the mastermind or not, 35:17 that's irrelevant to this conversation. 35:19 My request, my my deep request, start 35:23 participating, start connecting with one 35:25 another, start building trust, start 35:27 building relationships. 35:30 As AI accelerates, 35:33 people are going to get increasingly 35:34 panicked. What do I do? What do I do? 35:36 What do I do? What do I do? What do I 35:37 do? There's going to be all this noise. 35:38 Just like there's all this noise of 35:40 like, I've made too much crap. There's 35:42 going to be all this noise of people 35:43 going, what do I do? What do I do? Get 35:46 in community. build some trusted 35:48 relationshipships, form small little 35:50 cohorts within the AI salon. All right, 35:53 that's a lot of what the purpose of of 35:56 the AI learning lab has served as is a 35:58 little sort of cohort incubator. So 36:01 there's a lot of really powerful 36:02 relationships in the AI salon that were 36:04 largely formed in the comment section of 36:07 of these lives. So okay with that 36:11 the other thing that we should talk 36:13 about is we have a monthly theme for the 36:15 AI salon and the theme for May is 36:18 momentum. So, if you go into um the 36:22 community feed and then I think also 36:24 where else is it is it's it's in 36:26 community feed and it's in 36:27 announcements, 36:29 Brandon. 36:30 But the the theme of the month for for 36:33 May is momentum. And Brandon puts out a 36:35 weekly prompt about go prompt 36:40 your things you're doing. Start thinking 36:42 about momentum and what that looks like. 36:46 All right? We're building [ __ ] We are. 36:49 Oh, that's right. And then Brandon 36:52 there's there's so much stuff. The other 36:55 thing in the mastermind is Brandon is 36:58 now he did his kickoff week. He's got a 37:00 four-week sprint um teaching lovable. So 37:04 if you want to learn about vibe coding, 37:06 if you haven't ever vibe coded your own 37:08 application, it is insanely cool. It is 37:12 insanely cool. You're like, I I have 37:16 some idea of some application I could 37:18 make. And then you go to lovable and you 37:19 put it in there, it's like out it comes. 37:22 It's mindblowing. What's what's powerful 37:26 about doing it with someone like Brandon 37:28 who's made probably, I don't know, 30 or 37:30 40 or 50 apps, 37:32 is that he can tell you all of the 37:34 things that are going to be frustrating 37:36 for you and help preempt that 37:38 frustration. In fact, session one of the 37:41 lovable workshops, they didn't make 37:43 anything. They didn't make any apps. He 37:46 spent the entire hour getting people to 37:48 think about what do you want to do? Be 37:51 intentional about what you're trying to 37:53 create. Um, AI life hacks is on 37:56 Wednesday. There's just a bunch of 37:57 stuff. So, go in the upper right hand 38:00 corner over here is the events tab. So, 38:02 go look at all the events coming up. And 38:05 then if you're not in the mastermind, 38:06 I'd consider joining. There's a bunch of 38:07 stuff there. Okay. 38:10 Woo. Beautiful. Um, okay. Let's let's 38:15 flip over to another 38:19 another moment, 38:21 another thing that was announced today 38:23 that I think is worth us looking at. 38:28 Okay. 38:30 So, not only did open We all good. Does 38:34 anyone have any questions? Any questions 38:36 about the salon? Any thoughts about the 38:38 salon? Anything anyone wants to talk 38:39 about before I move on? 38:45 You're all good. Cool. Beautiful. Okay, 38:48 back to this. 38:50 So, not only did OpenAI today announce 38:54 the deployment company, the OpenAI 38:57 deployment company, we're going to help 38:59 you deploy AI. They also announced 39:04 Daybreak Frontier AI for cyber 39:07 defenders. So 39:12 introducing Daybreak, Frontier AI for 39:15 cyber defenders. Dayreak brings the most 39:18 capable AI models, codecs, and our 39:21 security partners to accelerate cyber 39:24 defense and continuously secure 39:27 software. 39:28 So this, as far as I can tell, 39:32 Let's see. Find and fix any 39:34 vulnerabilities earlier with Daybreak. 39:36 Right? So, if you remember, 39:41 two weeks ago, three weeks ago, 39:44 Anthropic launched Mythos. They didn't 39:48 launch it. They said, "We've got this 39:50 new model, Mythos, 39:52 that is so good, you can't have it." 39:56 Right? It's so good. what it can do is 39:58 it can find all the vulnerabilities and 40:00 software and so we're going to give it 40:02 to these 52 companies to go patch all 40:04 their bugs. 40:07 This is kind of a less cheesy version of 40:09 the same announcement from OpenAI, 40:12 right? And so in a week where we're 40:15 expecting some announcements from OpenAI 40:17 and next week we're expecting some 40:19 announcements from 40:21 uh from from Google. We just had some 40:24 from Anthropic. So we're we're in an 40:26 announcement. We're in an announcement 40:28 little parade right here. 40:32 This is kind of a preempting preemptive 40:34 thing that says to me something is 40:36 coming from OpenAI that's an incredibly 40:38 powerful model. What Daybreak is is it's 40:40 going to give early access to the new 40:43 models to cyber defenders and people 40:46 that work on cyber security. Um so if 40:49 you work in that field uh if you want to 40:51 do that I think this is probably 40:53 something to look into. Um, 40:57 but again, both of these announcements 41:00 from OpenAI today 41:03 say to me that something is about to 41:07 come in the next I'm guessing it could 41:10 be this week, but probably in the next 41:12 three months, but probably within the 41:14 next three months. So, where are we 41:16 today? We're we're kind of miday, right? 41:19 So, mid June, mid July. 41:23 Yeah, probably by mid July, maybe mid 41:26 August, 41:28 um, 41:32 something's going to happen. Some model 41:34 is going to drop that's going to be so 41:36 good that all of a sudden it kind of 41:39 opens up these floodgates of like, holy 41:41 crap, we got to deal with this. We got 41:42 to and and by deal with this, what I 41:45 mean is big companies that are still 41:47 kind of sitting on the sideline are all 41:49 of a sudden going to jump in. Um, and 41:52 companies like Kelly's are going to get 41:55 busier fast. And I think it sounds like 41:57 from what Kelly's talking about, at 41:59 least locally in Dallas, that those 42:03 floodgates are starting to open. If 42:06 Anthropic drops Mythos 42:09 and Open AAI drops GPT6 and um, Google 42:14 drops whatever their new Gemini, I think 42:18 it was called, Omnis or Omni or 42:20 something like that. 42:21 some big ass new model from them. If 42:23 you've got three major companies 42:24 dropping three major models, um 42:30 we're going to need to deal with that. 42:32 And so I think these moves are to do 42:33 that. Google dropped a local small model 42:35 inside of Chrome and people are worried 42:37 about it. Yeah, I saw that too. So So 42:40 let me let me jump back over here for a 42:42 second. 42:43 Um 42:45 there's a there's an interesting post on 42:47 LinkedIn from from Brent Peterson. And 42:49 if you don't know Brent Peterson, you 42:51 should follow him. Uh he's got a company 42:54 called uh signal and cipher. Um he's 42:58 really smart on this stuff. He's been 42:59 building um AI agent memory models for 43:03 about two and a half years. So he's been 43:05 anticipating 43:07 what's just showed up, all these agents 43:09 that everyone's talking about. Brent has 43:11 been anticipating that for two and a 43:13 half years and building a memory system 43:14 for how to manage those things. 43:17 Um he put a post out today about this 43:20 new thing that Google did which is 43:23 sleazy. I know it it's just surprising, 43:27 isn't it? Cuz Google would never do 43:28 anything sleazy. 43:31 Um 43:33 they um 43:36 they're dropping in the installs of 43:38 Chrome now a four is it a four billion 43:42 parameter model? a really small large 43:45 language model 43:47 as a downloadable file local to your 43:50 computer 43:52 and 43:55 yeah 4 billion. Yeah, it's it's it's 43:57 Gemini Nano I think is what it's called. 44:00 So they're dropping Gemini Nano into 44:03 your computer and and the significance 44:05 of it is this. 44:08 If you have that local model as part of 44:10 Chrome, when you do things like, hey, 44:12 summarize this page for me. Hey, you 44:14 know, tell me about this, that, or the 44:16 other, rather than having to go out to 44:18 Gemini and use their big expensive GPUs, 44:21 they can just run this local model 44:23 that's right on your computer and do all 44:26 of your AI stuff right there. But what 44:28 they but what can also happen is other 44:31 people can now use your model and 44:34 basically steal the cycles on your 44:37 computer to run inference run AI cycles 44:41 on your little model on your computer. 44:43 So Brent Peterson has a pretty good uh 44:46 in fact let me go see if I can find it 44:49 now. I'll let you find it or maybe maybe 44:51 maybe uh producer Brandon can find it. 44:53 If you can find Brent Peterson's article 44:56 on that little model, I thought it was 44:58 quite uh educational today about about 45:02 the implications of of what Google's 45:05 doing in this non-transparent way. 45:07 They're just sort of sneaking it into 45:09 the into the download, which is kind of 45:10 crappy. 45:13 All right. Um 45:18 got two other things that I want to show 45:19 you and talk about. 45:33 All right, 45:38 we have ringside seats. 45:42 I saw that today in GitHub. Cool. 45:46 Marlene Paul, I made an app this weekend 45:49 on lovable 45:51 for the build for mom. Oh, cool. Nice. 45:55 Beautiful. Todd Waller, they want us to 45:58 run mini data centers in our houses, 46:00 too. Yeah, shit's gonna get weird, man. 46:04 Oh. Oh. Oh. Oh, I've got I read this. I 46:10 read this thing. I can go find the I can 46:12 go find the tweet. Hang on. This is 46:15 crazy. Um, let's see. Uh, what's it 46:18 called? 46:20 Um, 5.2. to 46:23 um 46:25 data center. 46:29 Okay. 46:30 Can am I sharing? Let's see. No. 46:41 Okay. Um this is this is brilliant. I 46:45 don't know if it's true, but let's let's 46:46 assume it's true because it's it if he's 46:49 even done this once, it's [ __ ] 46:51 brilliant. I met a guy making $5.2 46:53 million a year. Super simple business. 46:56 He finds a small town. He buys a plot of 47:00 land for cheap. And then he announces 47:02 that he's building a data center there. 47:05 And then he lets the residents whip up a 47:07 frenzy and start pushing back on the 47:09 data center. He then negotiates with the 47:12 city to buy back the land for double 47:14 what he paid for it. He does he does one 47:17 of these a week. 47:22 So I don't know if that's true, but if 47:24 it is like I would love to have a brain 47:26 that just thinks like this diabolically. 47:30 It's just crazy. 47:32 Um, you know, hey Podunk, hey Podunk, 47:36 Iowa, I'm going to put a data center in 47:38 your cute little town. The hell you are. 47:42 And then, you know, city council's like, 47:44 just just buy him out. Buy him out. Get 47:46 him out of here. 47:48 So So anyway, if you want a quick 47:49 money-making scheme, there you go, Ivan. 47:51 You're trying to figure out what to do. 47:53 Go buy some land in a small town. 47:58 All right, let's see. Let's go back to 48:05 So, 48:07 thinking machines, this is Mera Marott's 48:10 company. 48:12 Um, 48:14 and and uh Mera Moratti was the CTO of 48:19 Open AI and uh didn't didn't get on so 48:24 well with with Mr. Alman. She was there 48:25 for a while though and uh she went off 48:28 to do thinking machines and so they put 48:30 out a video today. So, let me make sure 48:32 that I'm sharing my audio. 48:52 Swing. 48:55 All right. 48:58 Hey, I need your help with something 48:59 today. You ready? 49:02 Absolutely. I'm ready. What's up? 49:04 >> Yeah. So, we're giving an announcement 49:05 today and I've got two of my friends 49:07 coming to help. Every time one of them 49:09 enters the frame, I need you to say 49:11 friend. 49:14 >> Got it. I'll say friend whenever one of 49:16 them walks in. 49:17 >> Cool. So, we've got a new system for 49:19 full duplex audio and video, which means 49:22 that uh you can stream input into it in 49:24 real time and it can respond to you even 49:27 while you're speaking to it 49:28 simultaneously. How does that sound? 49:31 Sounds like a solid setup. Full duplex 49:33 with real-time interaction is super 49:35 useful. 49:40 >> Friend, 49:40 >> hey Roan, I heard you're talking about 49:42 our amazing interaction model. I have a 49:44 few things to add, but to make it 49:46 interesting, I'll do it in Hindi. Can 49:48 you translate in to in English in real 49:51 time for my friend and for audience? 49:54 >> Absolutely. I'll translate as you go. 49:58 model. 50:00 Today we're taking a look at our preview 50:02 model 50:05 to release it which makes conversation 50:07 between humans and AI easier. It has 50:10 many features like web search and 50:11 artifacts 50:16 >> friend. 50:18 >> Hey guys. Uh I just heard something 50:20 really cool about search and artifact 50:22 generation. Uh, perfect timing for me 50:24 because just now I was wondering what 50:26 are a typical humans simple reaction 50:29 time for auditory, visual and tactile 50:32 communication cues. Could you search for 50:34 me? 50:35 >> Let me find those typical reaction times 50:37 for you. Got it. So tactile is fastest 50:40 at around 150 milliseconds, then 50:42 auditory at 140 to 170, and visual the 50:45 slowest at 180 to 250. 50:47 >> Oh, interesting. Can you visual 50:48 visualize them in a bar chart for me? 50:51 >> Absolutely. Let me generate a quick bar 50:52 chart for those reaction times. 50:54 >> In the meantime, could you quickly 50:56 explain to me why is auditory even 50:57 faster than visual? That's a little bit 50:59 unexpected. 51:02 >> Sure thing. Auditory is faster because 51:04 sound signals travel a shorter, more 51:06 direct neural path to the brain than 51:07 visual info does. 51:09 >> Makes sense. Thank you so much. Great 51:11 chart, by the way. 51:12 >> You're welcome. 51:14 So, and there's there's some other 51:16 videos there you can go watch. But, um, 51:23 hang on a sec. Um, 51:26 okay. So, so what's important about 51:28 this? 51:31 We now have 51:34 full duplex 51:36 where you can talk and it can hear you. 51:39 While it's talking, you can talk to it 51:41 and it can hear what you're saying. 51:43 uh it can translate in real time. We've 51:46 now got some version of that from 51:47 Google, 51:49 from Open AI. Um they just they just 51:52 came out with theirs a week or two ago. 51:54 Now from thinking machines, we're going 51:56 to start to see this everywhere. This so 51:57 this idea of full duplex 51:59 um real time interaction that are tied 52:02 to increasingly smart models, right? So 52:07 I would think that the quality of 52:10 interactions are going up. It's funny. 52:12 There's a Tik Tok video. There's a Tik 52:13 Tok channel right now of this dude who 52:17 just talks to Open AI's voice and he and 52:20 it's it's just he he basically talks to 52:23 it and it's just doing idiotic things 52:25 like he has it count to a thousand and 52:27 it can't or he has it tell tell him how 52:31 many A's are in the word 14, whatever, 52:34 and it just fails miserably. 52:36 um we're probably going to start seeing 52:40 those embarrassing fails start to fade 52:42 away. So, just like we saw pretzel 52:44 fingers start to fade away from the 52:46 image gen models and now they've gotten 52:48 to the point that we're doing full-on 52:51 um reasoning with visual stuff. Um that 52:56 demo from Thinking Machines was again 52:58 this this full duplex while I'm talking 53:01 it's recognizing what's happening in the 53:03 video. Another friend just showed up. 53:05 another friend just showed up. It's 53:07 rendering, you know, from whatever it 53:11 whatever the system is that's running. 53:12 It's spawning new windows and generating 53:15 charts and graphs while you're talking 53:16 to it. Um, 53:19 so that's coming. 53:22 I've talked for a while now about the 53:25 idea that world models are coming. And 53:28 if you think about world models, where 53:30 my mind goes 53:32 is things like Unreal Engine 5, where 53:36 you've got these powerful 53:40 um 3D world builders that have physics 53:43 and game engines and fluid dynamics and 53:46 part particle systems and all that sort 53:48 of stuff built in. 53:50 my mind kind of goes to a gaming 53:53 environment, which is really just, you 53:55 know, what what Meta spent eight billion 53:58 dollars on, you know, trying to create 54:00 the metaverse and they just created like 54:02 shitty version of Second Life with for8 54:05 billion. 54:07 Um, 54:10 I think that the world models are going 54:13 to get way more sophisticated than maybe 54:15 we're that maybe than I'm anticipating. 54:18 And so a video came out today from from 54:22 uh rumored to be from Google from the 54:25 new Gemini video model. And so we'll 54:28 watch the video and then on the other 54:30 side we'll talk about it. Now the 54:32 disclaimer here is 54:34 this may not be the real thing from 54:38 Google. This might be a fake thing, but 54:41 we'll see. But if this is real, um, holy 54:46 [ __ ] what's coming? Um, and it starts 54:49 to shift my perception of of what world 54:52 models make possible. So, so let me let 54:55 me go find it and then we will. 54:59 Okay. 55:02 Gemini app cooked. Gemini Omni new video 55:06 model. Here's the first output. 55:09 And see the text coherence. If this is 55:12 not the nano banana moment for video, 55:14 what is? Okay, so here we go. 55:16 >> We start with the fundamental identity. 55:18 We start with the fundamental identity 55:20 sin^ 2 + cosine^ 2 = 1. Now if we divide 55:24 every term by cosine^ squar, we arrive 55:26 at the identity for tangent. 55:30 So, so if this is indeed what's coming, 55:33 wait, did I share my screen? Yeah. um if 55:36 this is indeed what's coming 55:40 where you know not only is the guy 55:42 talking but he's he's he's 55:47 talking about a math formula that he 55:49 understands the math formula and is 55:51 writing it on the board right now is 55:55 this real time is this going to take 30 55:57 seconds is it going to take two seconds 56:00 none of that I know 56:02 but imagine a world model 56:07 where 56:09 you can interact with data in in in in 56:14 just imagine interacting with data or 56:17 knowledge in any way that works for you. 56:21 So like 56:23 you know if you're an academic you love 56:26 professors in front of green green 56:29 boards writing in chalk and you could 56:31 just say make me a professor that's 56:33 that's explaining this concept to me. If 56:35 you like Bollywood, you could say, "I 56:37 want to learn this same concept but as a 56:39 Bollywood dance." Or, "I want to learn 56:42 the same concept sitting swinging in a 56:45 hammock by a pond and and you know, 56:47 someone is sort of meditating uh these 56:50 concepts to me and the formulas are 56:52 being written in the clouds in the sky, 56:57 right? And then within that, you're 56:59 going to be able to interact with other 57:01 people. And within that, you're going to 57:03 be able to 57:06 create things for yourself and share 57:08 those with other people. Um, 57:16 I think that 57:19 our 57:25 our interface 57:28 for for how we do AI 57:32 is about to change pretty radically. I 57:34 think by the end of the year, it will be 57:37 it will be apparent 57:41 the ways it will be different. I don't 57:43 think it'll fully change for three 57:44 years, but I think by the end of 2026, 57:48 we're going to be able to like this 57:50 group in particular because we're in 57:51 this conversation. Play. Exactly, Lori. 57:54 Play. 57:56 The reason in the AI salon, we we lead 57:59 with play. The reason we say play first 58:05 is that it is hard for me to imagine 58:10 the implications 58:12 of if if I can just spin up a physics 58:16 professor teaching me things about 58:18 physics like I'm looking at a a lecture 58:22 and at any point I can just shift that 58:25 reality or shift the lecture or say hey 58:28 I don't like looking at the formulas I 58:30 want to see shapes so draw concepts for 58:33 me instead of writing formulas and the 58:36 professor will be like sure I got you. 58:38 And I'll start drawing cool shapes and 58:41 drawing arrows and things like that. 58:44 Like what are the implications of that 58:46 for everything? For business, for 58:48 consulting, for right 58:51 those people that can understand what's 58:53 coming 58:56 and can be the translators of okay, 58:58 here's how the world works today. 59:01 Here's how AI works today, but here's 59:04 what's coming for AI. and I'm going to 59:07 start thinking over here and bring that 59:09 back to the to this reality. Right? So, 59:12 some people are going to be bringing 59:13 people who know nothing about AI into 59:15 the AI world. Then there's going to be 59:17 other people out in the front edge of 59:18 this thing going, "Holy [ __ ] you should 59:20 see what's coming." And they're going to 59:21 bring that back here and sort of pull 59:23 everyone forward. 59:28 I can't be in a regular if I don't show 59:29 up. I'm tardy. H It's exhausting. 59:33 Can't be late in AI. There's no sleeping 59:35 in AI. 59:37 Um, welcome. 59:40 we're talking about. So, if you're new 59:43 here, if you're just catching up or you 59:44 came in late, unacceptably late 59:50 um 59:54 the next two weeks look like like 59:55 there's going to be some fairly major 59:57 announcements um 1:00:00 and some some new capabilities and I and 1:00:03 I just Okay, so so here's 1:00:07 the great repurpose that was born here. 1:00:09 The seven economies was born here. Those 1:00:12 largely came out of 1:00:15 what happened to programmers in December 1:00:17 of 2025. So in December of 2025, 1:00:22 Anthropic launches Claude Opus 4.6 1:00:27 and it's just their next model. It's 1:00:30 their they went from point 4.5 to 4.6. 1:00:34 There was something about 4.6 that just 1:00:36 did coding better than anything else. 1:00:41 and like so good that developers were 1:00:43 like, "Oh [ __ ] 1:00:45 it can just do the work now. I still 1:00:48 have to be there this and that, but like 1:00:50 I don't need to code anymore." And you 1:00:52 know, we've we've heard we've heard 1:00:54 early adopters throughout 2025 like 1:00:57 Andre Carpath early in 2025 said, "I 1:01:00 don't really code anymore. I let it do 1:01:02 the code. I just I vibe code." He's the 1:01:04 one that coined the term vibe coding. 1:01:06 And in 2025, he said, "Ah, I don't 1:01:08 really do that anymore. I just I just 1:01:10 talked to it. 1:01:12 Opus 4.6 kind of delivered that 1:01:15 capability to coders in December of 1:01:18 2025. 1:01:21 It seems like in the next two weeks, 1:01:23 probably between now and June 15th, 1:01:27 the next month, 1:01:31 we're going to have glimpses into models 1:01:33 that do that for every other profession 1:01:35 or most other professions. 1:01:40 and and having just come back from 1:01:42 social media marketing world where most 1:01:44 of the social media marketing 1:01:46 professionals 1:01:49 were just kind of like, "Yeah, I know 1:01:50 AI's coming. 1:01:53 I'll deal with it." 1:01:56 Like 1:01:57 we're likely going to see something in 1:01:59 the next, you know, month that makes it 1:02:02 very very apparent that kind of any kind 1:02:05 of job that you can do is going to be 1:02:09 doable in in this really 1:02:13 mystical 1:02:16 way that we we don't know how to value 1:02:18 it. We don't know how to how it devalues 1:02:21 current work. We don't like it's not all 1:02:24 going to change in the next month, but I 1:02:26 think we're going to start to see models 1:02:28 where it becomes crystal clear that you 1:02:30 can't hide from it anymore. 1:02:33 And so then the question becomes, what 1:02:35 do we do? What do we do? 1:02:39 Let's see. Todd Waller, just implement 1:02:42 AI for me equals needs AI implementation 1:02:46 specialists. I will just implement AI 1:02:48 myself. Needs AI education. Oh, that's 1:02:50 that's good. Yeah. And and you know 1:02:53 what's interesting, Todd? There's 1:02:55 probably another category here, which is 1:02:57 I don't want to deal with AI at all. 1:03:00 Right. And then that's that's going to 1:03:03 be like, well, let's sell your company 1:03:06 then. Let's let's get new management in 1:03:09 here that that does have an appetite for 1:03:12 that. or 1:03:15 let's find a way for you to sort of 1:03:16 gently glide this company down to a 1:03:20 landing, right? Because the customer 1:03:22 base is going to start disintegrating. 1:03:25 It's crazy. 1:03:27 I'm very excited to live off the 1:03:29 government. I 1:03:33 this is 1:03:35 like 1:03:39 it is hard for me to conceive. 1:03:43 I've I've been like this for three 1:03:44 years, but I feel like this year we're 1:03:46 going to get to experience the beginning 1:03:48 of it. It's hard for me to conceive 1:03:52 how almost all existing businesses keep 1:03:55 doing what they're doing. 1:03:59 You know, if you're manufacturing stuff, 1:04:02 you've got a few years, but then the 1:04:05 robots are going to come and do that and 1:04:06 completely transform that anyway. But 1:04:09 anything that's in the neighborhood of 1:04:11 knowledge work just feels like it's 1:04:12 about to go. 1:04:15 And it, you know, what's weird is I 1:04:17 don't think about it like going like 1:04:19 it's going to go 1:04:22 and we're not going to need those 1:04:24 people. It's just that what those people 1:04:25 do is going to be a very different 1:04:27 thing. And none of us can anticipate 1:04:30 that right now. And that's freaky, 1:04:32 freaky, freaky, freaky. The age of 1:04:34 abundance. But this community will 1:04:37 figure it out. Listen, 1:04:42 it's why this channel exists. It's why 1:04:44 the AI learning lab exists. The one 1:04:47 thing I knew there there's there's one 1:04:50 thing I know. If I'm an expert at one 1:04:52 thing, 1:04:53 it was that if AI is not going to go 1:04:56 away, 1:04:58 then there's going to be a lot of people 1:05:00 trying to figure it out and it's going 1:05:01 to move really fast and no one's going 1:05:03 to know anything. And the only way 1:05:06 you can be successful in that kind of 1:05:08 environment is to be with other people 1:05:10 trying to figure it out at the same 1:05:11 time. It's like the only thing I know. 1:05:14 and and it's like I feel like I feel 1:05:17 like we went through a bit of a lull 1:05:20 with the with the salon and with me 1:05:22 doing the lives and like I couldn't I 1:05:23 kind of lost my footing. So So like two 1:05:26 and a half years we were generally 1:05:28 talking about tools and I I would say 1:05:30 from 1:05:31 I don't know end of 2024 through 2025 1:05:35 it's kind of like what what is this? 1:05:36 What are we doing? What's what's the 1:05:38 point? What's the point? What's the 1:05:40 point? 1:05:41 I think we're coming the the the 1:05:45 first of all, thank you to Andy and for 1:05:47 Brandon for pushing me to reconentrate 1:05:51 the energy of this channel back into the 1:05:54 salon. 1:05:56 Um, 1:05:59 as these tools get better, I think it's 1:06:01 going to get more depressing on the 1:06:04 surface 1:06:06 because it's like, oh yeah, shit's about 1:06:07 to really change and I think people are 1:06:09 going to 1:06:11 you know, 1:06:12 pull back and isolate. 1:06:15 Feels scary, feels existential. [ __ ] it. 1:06:18 I'm going to go sit in my cave. 1:06:21 Um, 1:06:23 I think it's more important than ever 1:06:25 that we don't do that, that we just keep 1:06:28 being in the conversation. Yeah, it's 1:06:29 going to get weird. Yeah, 1:06:32 a video model is going to be able to 1:06:34 have a professor writing actual formulas 1:06:36 on an actual whiteboard that are 1:06:38 correct. 1:06:40 And what are the implications of that? I 1:06:43 don't know. 1:06:45 I don't know. I know that Jim Ross, as 1:06:48 soon as that comes out, we're going to 1:06:50 have the the self storage version of a 1:06:53 video that that talks about, you know, 1:06:56 you can take a snapshot of the floor 1:06:58 plan of your uh of your storage facility 1:07:01 and it'll create a five minute movie of 1:07:04 the security protocol for your for your 1:07:07 units. I know Jim Ross is going to do 1:07:09 that. That I know. But beyond that, what 1:07:12 do we do? I don't know. But when we 1:07:15 watch what Jim Ross does with these 1:07:17 tools, 1:07:19 we're going to go, "Oh, oh, I could go 1:07:22 do that thing over there." 1:07:26 Right. Valerie Cox, I work for a public 1:07:28 utility where we typically run slow and 1:07:31 we are even in a rush and we are even in 1:07:33 a rush to figure this out. 1:07:38 And you know, Vicky Baptiste in office 1:07:41 hours last Friday 1:07:45 said her IT organization has spent the 1:07:49 last year building an AI system that 1:07:53 they're ready to roll out. 1:07:56 It's irrelevant. 1:08:00 The tech has moved on. It it it took 1:08:02 them, you know, so long to 1:08:06 so long relatively like a year in a big 1:08:09 corporation is probably 3 to 5x normal 1:08:12 speed, but a year in AI is just 1:08:16 ridiculously slow. 1:08:19 So, what do you do with that? I don't 1:08:21 know. Well, what's cool about the salon? 1:08:24 We have people like Vicki and Chris Vuno 1:08:27 and a and a bunch of folks that work in 1:08:29 big companies 1:08:32 that understand how they work and can 1:08:35 experience what they're doing and then 1:08:36 they can look at small little nimble 1:08:38 companies nipping at the heels of those 1:08:40 big companies and then as these new 1:08:42 techs come technologies come along 1:08:45 you know we'll be in the conversation. 1:08:47 If you're worried about layoffs or 1:08:49 losing your career I'm here to support 1:08:51 you through it. It's beautiful. What did 1:08:54 you What did you say? Something 1:08:56 something Mary Mary Carol about um 1:09:01 that that you you couldn't have 1:09:03 anticipated the what did you ask for? 1:09:05 You wanted some sort of change in your 1:09:07 career or something like that and you 1:09:08 couldn't have imagined how much you've 1:09:10 been blessed with the answer to that to 1:09:12 that challenge to the universe. 1:09:18 Oh man. I had an interview with a 1:09:20 company today, a big national company, 1:09:21 and the hiring manager described their 1:09:23 challenges and said, "But you know, it 1:09:26 will take us over a year to get a system 1:09:28 in place." I laughed. Yeah. 1:09:32 And this, listen, 1:09:34 this goes back to 1:09:40 this, 1:09:42 right? 1:09:44 if the transition economy. So if a 1:09:47 Fortune 500 company 1:09:50 right now today, they're like, "Okay, AI 1:09:52 is a big thing. We're going to commit to 1:09:54 this. We're going to we're going to 1:09:55 figure it out. We're going to build a 1:09:57 system that we can roll out to all of 1:09:59 our all of our constituents. 1:10:03 And if they do it really fast, it'll 1:10:05 take them a year. And so by the time 1:10:08 they roll it out, companies in economy 6 1:10:12 will have iterated 1:10:15 a dozen different systems that are that 1:10:17 are like 12 generations ahead of what 1:10:20 what the big company just did. 1:10:25 And like on the surface you're like, 1:10:26 well that yeah, but that 10erson company 1:10:29 could never compete with this 50,000 1:10:32 person company. But what's going to 1:10:33 happen is this 10-person company is 1:10:35 going to figure something out where they 1:10:37 get a customer or two where all of a 1:10:39 sudden they're now not a 10erson 1:10:40 company, but they're a 200 person 1:10:42 company. All amplified by AI, 1:10:46 iterating at the speed of AI 1:10:50 and and in fairly short order, an 1:10:52 economy 6 company's going to be over 1:10:54 here attacking an economy 4 company 1:10:56 because the economy 4 company thinks 1:10:59 they're moving fast. 1:11:06 It's gonna get weird. It's going to get 1:11:08 weird. Gonna get weird. Weird. Weird. 1:11:10 Weird. Weird. Weird. Todd Waller, both 1:11:12 Rachel Woods 1:11:14 and and Beanie Hat Nate uh Nate Nate B. 1:11:17 Jones uh recently said that we should 1:11:19 try to auto automate ourselves out of 1:11:21 our jobs because you either be promoted 1:11:23 or another company will want you for 1:11:25 those skills. Yes, 1:11:28 someone in the AI salon, I got an email 1:11:33 3 months ago, their entire division was 1:11:36 laid off except for them 1:11:40 because they knew AI. So yeah, like 1:11:44 don't hide from this [ __ ] This I mean 1:11:47 this is the thing. Generously lead is 1:11:49 about 1:11:51 telling people here's what I'm doing. 1:11:53 Here's what I'm doing. 1:11:55 If you're in here, there's 33 people in 1:11:57 here right now. If you have not shared 1:11:59 something that you're working on, 1:12:02 whether it's a creative project or an 1:12:04 app or or some new strategy or some new 1:12:08 framework, whatever it is, if you have 1:12:11 not shared that in the AI salon, like go 1:12:13 over here, go to look what I made or go 1:12:16 to the community feed or go to the AI 1:12:19 learning lab live the irregulars 1:12:20 channel. 1:12:22 If you haven't shared in a while, just 1:12:24 go share. 1:12:26 And then tomorrow when you get up, do 1:12:28 the same thing on LinkedIn and do the 1:12:30 same thing on X and do the same thing on 1:12:32 Facebook. Whatever is your poison, 1:12:35 social media poison of choice. 1:12:38 Just go put yourself out in the world. 1:12:40 I'm doing this. I'm doing this. I'm 1:12:41 overwhelmed completely, but I did this. 1:12:46 MySpace. Go post to MySpace. 1:12:53 Kelly Camp, my ideal customer profile is 1:12:55 curious, forward thinking, and an early 1:12:58 adopter. The size of the company, 1:13:00 vertical is irrelevant. 1:13:03 So, Kelly Camp, I would love to know how 1:13:06 are you finding those people because 1:13:09 that like I go to the same place right 1:13:12 now that that I think talking about AI 1:13:15 on the level that we talk about it in 1:13:16 here which which is funny the level that 1:13:19 we talk about AI in here is a lot like 1:13:22 ah I don't know could be this could be 1:13:25 that 1:13:26 because we're not dealing with absolutes 1:13:28 we're dealing with shifting sands Right? 1:13:32 So being able to have that conversation 1:13:34 with someone in business means that you 1:13:37 have to find someone in business 1:13:40 who is an early early adopter. If you 1:13:42 have not read if if you all want to um 1:13:47 understand the dynamics of the world 1:13:49 we're in and will likely be in for the 1:13:51 next decade, go get the book Crossing 1:13:54 the Chasm. 1:13:56 Um, it's it's it's a little dated, but 1:14:00 it's still very very true. And it just 1:14:02 talks about early adopters and fast 1:14:04 followers. And the chasm is the chasm 1:14:06 between early adopters 1:14:09 and fast followers. The people that, 1:14:11 okay, I'm ready to do AI now. They're 1:14:15 going to come in the next two, three, 1:14:17 five years. 1:14:19 So if you're trying to sell into these 1:14:21 people having these early adopter 1:14:24 conversations, they literally can't hear 1:14:27 what you're talking about. So you have 1:14:30 to find those people. So I don't know 1:14:31 what I don't I would love to know your 1:14:33 trick for how you're finding those 1:14:34 people. Kelly, I embraced the ick today 1:14:37 and watched hours of tutorials on how to 1:14:39 use Git and GitHub. That's okay. Part of 1:14:44 play 1:14:46 is playing across domains. And part of 1:14:50 playing across domains 1:14:52 is embracing the ick, right? I will 1:14:56 admit that I know what GitHub is and I 1:14:59 kind of know it conceptually, 1:15:01 but for example, I had to ask my 1:15:03 engineer for for my company, Storyvine, 1:15:06 the other day. They were talking about a 1:15:08 pull pull request, and I was like, 1:15:10 "What's a pull request? Isn't it a 1:15:12 push?" And it's he's like you're pulling 1:15:14 things into the into this branch and it 1:15:17 sort of made sense to me. But if you 1:15:21 want to learn new things and learn the 1:15:22 new lingo, even with AI teaching you or 1:15:25 YouTube teaching you, you're going to be 1:15:27 in the ick for a while. Like, oh, I 1:15:29 don't want to learn a new thing. Go 1:15:30 learn a new thing because you now have 1:15:32 the capabilities to do it. Okay, Kelly 1:15:34 Camp. Actually, clients are finding me. 1:15:37 Ah, they've been Okay, wait. Keep that 1:15:41 up on screen. Keep that up on screen. 1:15:45 Everybody take a screenshot of this. 1:15:48 When we talk about 1:15:52 generously leading, when we talk about 1:15:53 play first, go learn, create excellence, 1:15:56 go build [ __ ] generously lead, go talk 1:16:00 about what you're building and what 1:16:02 you're learning along the way. Kelly's 1:16:04 been doing this for three and a half 1:16:05 years. My clients are finding me. 1:16:08 They've been secretly watching my AI 1:16:10 journey on LinkedIn and are now reaching 1:16:12 out. 1:16:16 There's your testimonial. That's the 1:16:17 cycle of AI readiness. 1:16:20 Kelly is a three and a half year 1:16:22 overnight success. 1:16:28 But that's it. 1:16:31 Why are they going to Kelly 1:16:33 AI Journey on LinkedIn? 1:16:36 Not AI Journey. And maybe I'll tell 1:16:38 someone if they're not a jerk. 1:16:42 AI journey in public out loud. 1:16:48 Kelly Anderson, 1:16:50 I have a question. Wait, hang on. I have 1:16:52 10 Oh, wait. Go back to that one for a 1:16:54 second. 1:16:56 Mary Carol, I have 10 extra copies of 1:16:58 Crossing the Chasm if anyone wants one. 1:17:00 I invited a speaker to Xerox Park when I 1:17:02 was leading a women's organization eight 1:17:03 years ago. connect with MC if you want 1:17:06 to read Crossing the Chasm if you want 1:17:08 to understand 1:17:10 why conversations are weird right now 1:17:14 where sometimes when you're talking to 1:17:15 people about what you're learning with 1:17:17 AI 1:17:18 it like you just see 1:17:23 they're dead they're dead in the in the 1:17:26 eyes 1:17:29 they're in a completely different 1:17:31 universe there's this giant gap between 1:17:33 where we are and where they are. They 1:17:35 literally can't hear what you're talking 1:17:37 about. So, if if you're trying to figure 1:17:39 out how to do business, how to be in 1:17:43 this current moment in history, go read 1:17:45 the book Crossing the Chasm. It will 1:17:47 help. Okay. Kelly Anderson, I have a 1:17:49 question. A social media a social media 1:17:53 marketing Oh, at social media marketing. 1:17:55 Did you see any irrelevant or irreverent 1:17:58 innovative promotional gimmicks? 1:18:01 Yes. 1:18:03 the very last 1:18:07 talk that we saw and I assume that these 1:18:10 talks are going to be available at some 1:18:12 point on YouTube. I know as speaker as a 1:18:15 speaker I've got access to them right 1:18:16 now but the very last talk that we saw 1:18:21 um I forget the man's name. If Andy's 1:18:24 still on here she'll know it. She'll she 1:18:25 can pop it in the comments. 1:18:29 He gave this really remarkable talk on 1:18:32 his his journey to create 1:18:37 an incredible amount of income using 1:18:40 YouTube shorts 1:18:42 opening Pokemon packs. 1:18:46 Um, and what he was talking about the 1:18:49 the core concept that he was talking 1:18:51 about was short form episodic shows. 1:18:57 And so one of the examples that he 1:19:00 showed was a Tik Tok channel where you 1:19:01 never see the person's face. They start 1:19:04 with a rock about this big and they kick 1:19:08 it for a minute. And the point of the 1:19:11 channel is the guy is going to kick that 1:19:14 rock until it's a perfect sphere. And so 1:19:17 every day he drops the rock on the 1:19:19 sidewalk and kicks it for a minute and 1:19:22 then that's the video. 1:19:25 And then the other video that he showed 1:19:27 was the dude out of out of Portland with 1:19:30 his cat 1:19:32 that that quit his job at whatever some 1:19:38 tech thing 1:19:40 um or like a retail store. Um and he 1:19:43 bought a and he drained his 401k, bought 1:19:46 a sailboat and decided to teach himself 1:19:48 to sail and sail from Portland to 1:19:50 Hawaii. and he's going to sail around 1:19:52 the world. But the first leg was from 1:19:53 Portland to Hawaii. And so basically it 1:19:57 was the journey of him buying the boat, 1:19:59 learning how to sail with his cat, 1:20:02 and then going on this like 26 or 30 or 1:20:05 35day journey. 1:20:08 And by the time he got to Hawaii, he was 1:20:11 welcomed by the governor and he had all 1:20:12 these sponsorship deals and [ __ ] like 1:20:14 that. So, the the idea of doing these 1:20:19 daily short form videos where there's 1:20:21 something at stake. 1:20:24 Um, and then he he talked about he was 1:20:27 trying to figure this out and he he did 1:20:29 a failed attempt and then he did another 1:20:31 failed attempt 1:20:33 and then he was doing something with his 1:20:34 kid and he and his kid were bonding over 1:20:38 Pokemon cards and they went into some 1:20:40 Pokemon card shops and and he said, "Oh, 1:20:43 maybe I could do videos of me opening 1:20:46 Pokemon decks and, you know, ripping 1:20:48 Pokemon decks and and seeing if we got 1:20:51 any good cards or not." And so they 1:20:53 started doing that and there there he 1:20:54 was getting no views and he for like two 1:20:56 months there was just nothing. And then 1:20:59 he started experimenting with like he he 1:21:01 wrote a jingle at some point like you 1:21:04 know should I open it or should I keep 1:21:06 it sealed? It was it was this really 1:21:09 cheesy sort of um thing. And basically 1:21:12 if he like opened a pack like maybe he 1:21:14 paid $50 for this rare pack and he 1:21:17 opened it up and then there was a there 1:21:18 was a no good cards in it. He was like, 1:21:21 "Oh, I should have kept it sealed." And 1:21:23 see, he had these little things. And the 1:21:25 minute he had a couple of these little 1:21:27 gimmicks that sort of built up the 1:21:29 anticipation of each episode was like, 1:21:32 "Is it going to be good or is it going 1:21:33 to be bad?" And people latched on to 1:21:36 that. Like his numbers just went through 1:21:38 the roof and his revenue went through 1:21:39 the roof. And he's making like I don't 1:21:41 know. I've seen that guy. Yeah, exactly. 1:21:44 Kelly Anderson. He's making like $2.5 1:21:46 million now. And like just ridiculous 1:21:48 stuff. And at the end of his 1:21:49 presentation, he actually had a $1,500 1:21:52 pack of Pokemon. 1:21:55 Um, and and he and he filmed an episode 1:21:58 live. It like it was his 400th episode 1:22:01 or his 4,000th I think 400, I don't 1:22:03 know, whatever it was. 4,000. 1:22:06 Um, and he did it live in the room and 1:22:08 and we all sang the song and and then 1:22:12 you know he he opened it and there was a 1:22:13 good card in there and then he ended up 1:22:15 giving the cards to someone a kid that 1:22:17 was at the event. So it was pretty cool. 1:22:19 So that idea of short form episodic um 1:22:23 episodic things 1:22:26 is really important and it ties in Kelly 1:22:29 with um 1:22:33 it ties in with this this thing that's 1:22:35 that's hit me 1:22:39 recently is that 1:22:42 any marketing that we do about the thing 1:22:47 so I've made a workshop or I've got this 1:22:50 expertise or I've got this training or 1:22:53 I've got this 1:22:56 I don't know this comic book. 1:23:00 If you lead with talking about the 1:23:02 thing, 1:23:04 no one's going to care, 1:23:08 right? 1:23:12 But like when Brandon in Festivus was 1:23:16 talking about making the books for his 1:23:18 kids, 1:23:22 the books are going to be easily copied. 1:23:25 Like he did a book that was about the 1:23:26 alphabets of, you know, AI terms, 1:23:30 but it was it was how Brandon thought 1:23:33 about making the book for his kid 1:23:36 that is the hook. And then you can talk 1:23:39 about the book after after you after 1:23:41 you're in it. But the one thing that 1:23:44 can't be commoditized if everything's 1:23:46 about to be commoditized, the one thing 1:23:48 that can't be commoditized is your 1:23:50 story. And so I think this idea of these 1:23:53 episodic things like I'm on a journey 1:23:56 like Mary Carol like your your journey 1:23:59 about hey I worked at all these cool 1:24:01 places. I've got this cool background. I 1:24:02 got all this expertise and I'm trying to 1:24:05 figure this [ __ ] out. like that journey 1:24:10 is likely going to be way more valuable 1:24:13 to you than any of the expertise that's 1:24:16 been valuable historically 1:24:20 because expertise is now a commodity 1:24:22 which is really [ __ ] depressing, 1:24:24 right? It's the whole point of the great 1:24:25 repurpose is all of this stuff we've 1:24:28 done all of our lives to make ourselves 1:24:31 valuable, that thing 1:24:36 that used to be our value is now not, 1:24:41 but we are. And to the extent that we're 1:24:45 comfortable or not comfortable 1:24:48 putting ourselves out there, I think 1:24:50 that starts to become I I think it 1:24:52 starts to become the only kind of 1:24:53 differentiator is is that kind of brand. 1:24:56 Like hook people into the story of your 1:25:00 expertise and then they will buy your 1:25:03 expertise, but they'll buy it because 1:25:04 it's your expertise. 1:25:07 Right? 1:25:11 We've all had AI conversations with 1:25:13 people and they go to sleep. Maybe I 1:25:16 should paint some AI robots and submit 1:25:19 to shows. 1:25:21 Well, I you know, listen, I think that I 1:25:24 think that Kelly Camp 1:25:28 just her her consistent 1:25:31 Kelly Camp does it in this very quiet 1:25:36 kind of humble way, right? She's just 1:25:38 like, "Oh, yeah. I was at a coffee shop 1:25:39 and I told people about this. Oh, I 1:25:42 talked in a room full of CEOs and only 1:25:44 two of them had used AI. Oh, I talked in 1:25:46 a room full of dentists. Oh, I I was 1:25:48 with a bar owner and I I showed him how 1:25:50 he could take a picture of his 1:25:51 inventory. It's it's been this very 1:25:53 quiet like literal drip campaign like 1:25:56 doop doop doop doop, which any one of 1:26:01 those individual contributions that she 1:26:03 did, you'd look at it and go, "Well, 1:26:04 that's insignificant." But when you 1:26:06 string those all together, 1:26:09 it starts to tell this story of how 1:26:11 thoughtfully and how patiently 1:26:14 she's been interacting with this stuff. 1:26:17 And so at the point at which you're 1:26:18 ready to talk, you're like, you know, I 1:26:21 I kind of like that energy like 1:26:23 everyone's like screaming, 1:26:25 >> you could make money with 1:26:28 >> and here's this Kelly Campwoman who's 1:26:30 just she's she's just very consistent 1:26:32 and she's just there. Yeah, I want to 1:26:34 work with that. Right. It's not her 1:26:36 expertise. 1:26:40 Everyone will assume you've got the 1:26:42 expertise because everyone has access to 1:26:44 chat GPT. So, we can all be experts. But 1:26:47 why would I work with Kelly? Well, 1:26:51 because of 1:26:52 the herness. 1:26:55 Lori Blair. Maybe it's valuable though. 1:26:57 Um because who we are has led us to be 1:27:00 experts. 1:27:01 Oh, Lori, don't get me wrong. I What I'm 1:27:04 talking about, the value is us. The 1:27:07 value is us. 1:27:11 How we gathered that expertise. 1:27:14 The work ethic that we had to have. Mary 1:27:16 Carol, you worked for these big 1:27:17 organizations. You know how to navigate 1:27:19 big organizations. 1:27:21 You have enough of the skills to know 1:27:24 what it takes to do that. 1:27:28 the expertise of what you do when you're 1:27:30 there, that may not be as valuable, 1:27:34 but how you do that is valuable. And 1:27:38 then all that expertise, the value of it 1:27:40 doesn't go away for you. It just goes 1:27:42 away for you marketing, right? 1:27:47 Let's see. 1:27:50 We're now in a world where my journey of 1:27:51 five layoffs and the resilience I've 1:27:53 cultivated is is more valuable than my 1:27:55 PhD. Yes. Yeah. Yeah. Yeah. My PhD from 1:28:00 Harvard Business School, my postoc from 1:28:01 Stanford Medical. Sigh. But it might be 1:28:04 the story of who I am is more valuable 1:28:06 than what I accomplished. And that might 1:28:07 be refreshing. I think that's what it 1:28:09 is. Mary Carolyn. Like we'll be talking 1:28:12 about this tomorrow in in Andy's 1:28:14 workshop about the goop and identity. 1:28:18 The goop is 1:28:20 we're exiting a world. We're exiting a 1:28:23 world 1:28:25 where you have a PhD from Harvard. 1:28:30 In in the world we're in today, you're 1:28:32 like, "I have a PhD from Harvard." 1:28:34 People are like, "Oh, damn. 1:28:38 We're entering a world where 1:28:41 it doesn't matter. 1:28:44 Everyone's going to have a PhD level 1:28:48 genius 1:28:50 in every subject. 1:28:53 And on this 1:28:58 wait on on this 1:29:02 your little rabbit R1. This is going to 1:29:05 be the equivalent of a thousand PhDs 1:29:08 right here in my breast pocket. I'll 1:29:10 just pin it on. I'll just pin it on like 1:29:12 this. 1:29:15 So why would I work with Mary Carol? 1:29:17 Well, who did she have to be to get 1:29:20 that? 1:29:21 Who did she have to be to get those 1:29:23 jobs? Who did she have to be to string 1:29:25 together all that expertise and figure 1:29:27 out how to talk and think and 1:29:30 oh, I want to work with that. Kyle, I 1:29:33 believe you were correct when you said 1:29:34 one of the biggest takeaways from the 1:29:35 last three years may be our ability to 1:29:38 emotionally bounce back from the AI 1:29:40 whiplash. Yeah. And I think it I think 1:29:42 it's from the AI whiplash and it's also 1:29:44 from this identity whiplash. 1:29:47 Like 1:29:49 you just said it, Mary Carol, like it's 1:29:52 really flipping depressing 1:29:55 that that those things that were so 1:29:58 valuable are are now just sort of 1:30:00 sliding in value and you're like 1:30:01 watching them slip away going, "No, it's 1:30:04 like a bad dream." Right? But I think 1:30:07 you're right. I think your experience 1:30:10 going through that 1:30:12 starts to become more valuable. how you 1:30:15 got there, what you did, what you've 1:30:17 learned as you've gone through those 1:30:19 layoffs. I don't know. I don't know what 1:30:21 any of this means. Anyway, we are at 1:30:23 time. This is a really good 1:30:24 conversation. 1:30:27 Bless the broken road. There you go. 1:30:30 Let's see. 1:30:32 All right. Fantastic. Let's see. Time to 1:30:35 wrap. So, Wednesday night uh is going to 1:30:39 be um here in the AI learning lab is 1:30:43 demo night. Um, I haven't decided what 1:30:46 I'm going to demo. I have a sneaking 1:30:48 suspicion that the OpenAI announcement 1:30:50 is probably going to be Thursday, so 1:30:52 we'll probably miss it Wednesday. But if 1:30:55 there's anything that launches between 1:30:56 now and Wednesday, it might be a new 1:30:58 thing. Um, if anyone has any requests 1:31:02 for for what you want me to play with on 1:31:05 Wednesday, um, feel free to DM producer 1:31:08 Brandon or um or just DM me or just post 1:31:12 it in the regulars channel. We'll take a 1:31:14 look. We'll we'll look at that. Um, 1:31:18 it might be a night. It could be a sunso 1:31:20 night, but we'll see. Um, thank you all. 1:31:23 It is time. I'm gonna I'm I'm trying to 1:31:26 get better at at at making these a bit 1:31:28 more predictable. We'll start at 7:30. 1:31:30 We'll end at 9. You can if you're on the 1:31:33 East Coast, it's 11. You can watch your 1:31:35 local news. 1:31:37 Have some have some normaly in your 1:31:39 life. Um and still do do some good 1:31:42 stuff. I hope this was valuable to you 1:31:44 and interesting tonight. Um and with 1:31:46 that, peace out. I really appreciate 1:31:49 y'all being here. Um I'm liking this new 1:31:52 format. I'm loving that we're we're kind 1:31:54 of in here in the salon together. So, 1:31:56 please go out and participate in the 1:31:57 salon and uh I will see you Wednesday. 1:32:00 Bye.