From Idea to App in Hours: How Non-Technical Leaders Can Build with AI Today
Episode 14 - From Idea to App in Hours: How Non-Technical Leaders Can Build with AI Today
===
[00:00:27] Tom Adams: Well, welcome to prompted Circumstance. No. Is this your call? Look, go for it. Mark. Sorry. This is yours. This is your
[00:00:36] Mark Redgrave: I, I'm just, it's not my show. Tom, you're the, you're the ringmaster of this thing. I just wanted say morning. I haven't seen your smiley happy faces for a little bit.
[00:00:44] Tom Adams: I know it's been weeks, it's been weeks since our last confession, so it's, uh, it's lovely to see you all back again. Uh, apparently Ryan, uh, Neiman our esteemed partner, uh, sometimes the maestro of innovation, we, we call him, uh, he's not with us today, so we're sorry that, uh, Ryan is not with us, but he's apparently out conquering the world in some strange fashion, so, uh, good
[00:01:08] Mark Redgrave: rumor, rumor had it that like he got like a proper job, like a, like a real job, you know? Like what, like, like what we used to have back in the day.
[00:01:19] Tom Adams: I don't think I've ever had a proper job.
[00:01:21] Mike Richardson: Tom hasn't actually.
[00:01:22] Tom Adams: I, I don't think I've ever had a proper job in my life, but
[00:01:25] Mark Redgrave: You know, you know, you know when you've got a proper job. It's like, I know when I've got a proper job 'cause there's somewhere for me to submit expenses. That's
[00:01:32] Mike Richardson: Oh, that's it.
[00:01:33] Mark Redgrave: whether it's a real job or not.
[00:01:34] Tom Adams: Right. Otherwise, you just spend and it just comes outta your pocket in, in these unpro jobs.
[00:01:41] Mark Redgrave: yeah. I hand my expenses to my, to my wife and she looks at me like, what do you want me to do with
[00:01:47] Tom Adams: what are you doing? What are you doing?
[00:01:52] Mark Redgrave: But like there used to be when you got a proper job, you like hand your expenses to someone and then, then like money appears in your account's real good, but I guess the money's gone outta your account. So it's kind of, but it feels good.
[00:02:03] Tom Adams: Right. It feels right. Feels right. Right. So, uh, Mike Richardson, the agility whisper to CEOs, mark Redgrave, the AI Advantage activator to, to C-suite leaders, uh, myself, Tom Adams. I don't know what I am, I'm definitely none of those. So, uh, we're, we're glad to be here and we're gonna just roll through a whole bunch of things that we haven't rolled through in a while.
Because we haven't, uh, actually been together for a while. So, uh, what's on your mind these days? Yeah, what's, what's catching your eye these days? What's, uh, what's, what do you find interesting, intriguing, frustrating? Uh, what's going on for you right now? What are you seeing?
[00:02:42] Mike Richardson: go for it, mark. Go for it.
[00:02:44] Mark Redgrave: You know what I am. So I, I had a, I've had a real, like, actually a really intense, like few weeks, but in a, in a good way. But something that has re like really blown my mind these last couple of weeks is I've been working on like a consumer app idea and I've got back into the whole Lovable kind of, uh, figma stack and something.
Guys that has just blown my mind is how much more capable these platforms are, like four months after I last touched it. And, and I know this is sort of a narrative that's everywhere. It's like, gosh, it's developing quickly, but it's like, it's really amazing how much like, and I'm working with a, um, a really good friend of mine actually in the uk who's, um, who's a developer and he's.
Looking at what I'm doing and he's going, mark, this is actually kind of amazing because in the background, right? It's like the database structures we need on superb base are already there, right? So don't worry about the terminology, ladies and gentlemen. If you're listening, it's like the point is like six months ago, Lovable was just a, a cosmetic aesthetic skin.
Now it's like. All the integrations are showing up. So like, uh, that's been really, really, really amazing actually. Um, as a, and, and very fun for me because I'm actually like, um, I'm actually doing, I'm bringing something to life, which has some technical, like backend Tom, which like, which like was way beyond me, right?
[00:04:19] Tom Adams: Yeah.
[00:04:19] Mark Redgrave: Like, like I, because I don't understand that stuff. Fortunately there's some really amazing people around that do, but like, but now I'm like, gosh, I'm, I do actually feel like I could almost build this thing. Is that me being still not true, but like I'm, that's how I'm feeling.
[00:04:36] Tom Adams: no, I, I, I a hundred percent believe that that's true now because the, the, you know, I, I think six months ago the tools could do interesting things and they would, you'd go, oh, this is cool. But now it's, it know, like all the major and Lovable just uses all the other major tools plus their own stuff. But at the back end, they know how to build software.
Now they absolutely know how to build software. So there's no issue with the software at this point. Um, you know, is it, is it a hundred percent there? I don't think so, but man, you can build really cool, solid stuff now.
[00:05:13] Mark Redgrave: And, and to give you an example, right, like I wanted a payment. Like, uh, a payment system in it, right? And like, uh, little, little check on ChatGPT, like probably Stripe's a good option. They're gonna, the fees are gonna be high, but honestly, like, just suck it up like, and you can make a change later. Um, Lovable built the Skype, the Stripe integration, like gave it to me, just said, put your API key in here and we're done.
Like seriously and guy, and.
[00:05:45] Tom Adams: Yep.
[00:05:48] Mark Redgrave: Like,
[00:05:49] Tom Adams: it's fascinating. Uh, I, I think I told you, um, since we last talked, I did a Lovable, uh, vibe coding lab with a hundred, with, with a hundred people who joined. And, um, nobody in the room was as far along technically as the three of us. So, complete nubs,
[00:06:08] Mike Richardson: Whoa, whoa me in that.
[00:06:09] Tom Adams: no, I even, I, I didn't even include you. Yeah, pretty low bar People could barely, in some cases, could barely get the, the Lovable code into the system, so that's how challenging it was.
Um, but in two hours. So I had a two hour lab and I taught them to how to think about, uh, how to think about planning an app. And my, my promise was, within two hours, you will leave with a functioning app for your business. And, um. Taught them how to think about it. And then from there I said, go. And the thing that was so mind blowing to me was I wasn't, after the first 20 minutes, there was no teaching.
I was just walking around a hundred people watching what they were doing. And what was so cool to watch was people getting to that point of, oh my goodness, this idea I had has actually come to life. And because, like you said, Lovable. Um, is got all the integration. So does Replit. I mean, uh, Ryan's a big Replit guy.
Um, Lovable is just the one who sponsored the event, but the, the reality is this stuff is so powerful right now to be able to produce stuff
[00:07:20] Mike Richardson: person? This was in person. 'cause
[00:07:22] Tom Adams: in person, a hundred people in a room.
[00:07:25] Mike Richardson: How did you, how did you come to get engaged to do this?
[00:07:30] Tom Adams: I proposed it, I proposed it to a conference, uh, that invited me to speak. And I said, well, what if we did it differently? They, they asked me to speak about AI and AI stuff and so I actually did a keynote on that, but I said, let, let's add to it. Let's do something really interesting. And um, and I proposed doing a vibe coding lab.
And at the end of it, um. And I, I, I was lucky I asked all the major platforms to support it and only Lovable responded. So that's why they got the, they got all the billing. But, um, it was, it was really cool to watch people who had never done anything like this before, never thought about coding, all of a sudden had these really cool apps sitting in front of them, which was
[00:08:15] Mike Richardson: So, so I have a question for you, Tom. Help me understand this, 'cause this is a question I have. Lovable Replit. Are these in the category of things you would call vibe coding platforms? And are there others?
[00:08:30] Tom Adams: Yeah, there's, there's really, I, I'm aware of three and then you can sort of see other ones. The three big ones right now are Lovable, Replit and Bolt. BOLT Bolt. Um, and those three tend to be the primary ones. Uh, Replit and Lovable, I think are significantly bigger than Bolt, but they, they're all trying to do the same thing, which is wrap up sort of vibe coding in a context.
But like Mark said, they've layered in so much backend capability now that all they say is, we'll do this and act this way and make this decision and get your stripe API key. The thing all works now. So what they've done a really nice job of, of, of is integrating where what I do is I tend to work in Claude Code and I build in Claude code and I go, I want this particular tool and I need this thing to work and, and I decide how they fit together.
When you're on Lovable and Replit it, they say, let's use this, or they've already got those integrations.
[00:09:26] Mike Richardson: so is Claude Code also in this same category? It's vibe Coding.
[00:09:32] Tom Adams: it, you can use it that way, but it's, it's, it's not designed like Lovable and Replit are Lovable and Replit are kind of designed to be vibe coders. Uh, whereas Claude Code generally is what, um, developers use.
[00:09:47] Mark Redgrave: Yeah, I think so. So let's talk about this for a second because, so with Vibe coding, Mike, right? Like the inspiration is your idea. That's it. Right? It starts with, have you got an idea? Just describe it. And I'm U by the way, I remember us talking a few months ago about like Tom, I think it was you who said, this is gonna be about voice.
I am using so much voice with this. So like my primary interface now is I'm just talking to these things, which is. So much easier because I can literally, like, I take my dog for a walk every morning, right? I'm literally, it sounds like it. I, I'm literally walking my dog talking into my phone, going, I'm thinking about this like, help me, how would I add a payment gateway?
Um, you know, and it's like, and it's just doing it. And I think so, so the start point, Mike is so, so, so let's talk about inclusivity for a moment. 'cause I think this is why it's so exciting. If you have an idea for something you want to do, like is it an app? Maybe, but it doesn't have to be an app, but like, if you, if you are a, a professional, a business person, a leader, and you say, it would be amazing if I could brackets describe the problem, put it into Lovable or Replit, and you are off to the races, right?
Like that. So that's the difference. Claude Code is for like, I've got a piece of code, help me improve it, debug it, speed it up. But, but Lovable and Replit are like imagineering tools
[00:11:23] Mike Richardson: Nice. Wow.
[00:11:26] Mark Redgrave: and, and I'll tell you like, like something that is amazing, right? And, and, and this is something that. Like real aha moment for me over the last couple of weeks is there are, there should be no restrictions on getting started for anyone who's listening to this because if you don't know how to get started, ask. Lovable Replit ChatGPT. Claude, just ask. Yeah, and this is why, where I think a lot of people are missing a trick. You do not need to sit at your desk and think, oh, I've got this really good idea. But like, I don't know where to start. There's no excuse for that because just go to Claude or any of the platforms and say, I've got a really good idea for an app.
Where do I start? And you'll start.
[00:12:17] Mike Richardson: Yes,
[00:12:18] Tom Adams: Well, and, and, and, and the cool thing is, Mike, you already say you are the, the sort of laggard in the room, but in since our last conversation, you actually created something. So what did you create?
[00:12:30] Mike Richardson: Yeah. So now we'll go into the slow lane of this conversation here. Everybody, we'll just pull over and travel at 60 miles an hour for a moment. Um, so yeah, inspired by one of my peer group members that I've, I've. Mentioned in the past that I'm getting them now to do, getting the members to do demos for each other.
And I had this member called Jim, um, do his second demo actually, and he is screen sharing. We are actually physically in a room together, so he's, you know, plugged into the projector and he's sharing a screen. He just takes us to Claude Cowork, and he shows us 16 projects that he's got in Claude Cowork.
He then switches over to PowerPoint and he says, one of the projects in Claude Cowork that I just showed you was I asked Claude Cowork to produce me a PowerPoint summarizing my 16 projects. Here it is. I haven't. Touched this PowerPoint at all. It's hot off the press from Claude Cowork. He steps us through half a dozen slides.
It's giving us an overview. He then goes back into Claude Cowork. He starts a new project and just says to everybody, you know what? Who's got an idea? What do you wanna work on? And this thing just starts chuntering away. So I took that inspiration. And I merged it with, you guys have been on my case for months and months and months.
Mike, get out of ChatGPT and get into Claude. And in particular Claude Cowork. So I pivoted to Claude and Claude Cowork. I still have ChatGPT and I. I'm, I'm in there and I'm thinking, okay, what am I gonna do as my first project? And I thought to myself, I know what I'll do. I said to Claude Cowork, I run peer groups of about 16 members.
I'm trying to inspire them all to go on this AI journey. Some of them are leaders, some of them are laggards. I want to be sure that none of them get left behind. Help me start thinking about that, you know, return. It asks me some questions. It, what a great idea. Ba baba, boom. Second prompt. How would I create that as an online platform?
Question mark. Hit return. It says, whoa, that's a great idea. Ba boom, boom. Three prompts later, it delivers me an H-T-M-L file. And it says, by the way, if you go over to, uh, Netlify, you can, you can load this up right now as, uh, an online URL and you can send it to your members, you know, in the most user friendly, um, familiar way possible.
So I did that, basically what I did is I asked Claude to create me an online, uh, Claude Cowork to create me an online platform. That would firstly explain to my members how to download Cloud cowork. Secondly, it would help them start to brainstorm projects that they could attack. Thirdly, it would encourage them to choose one of those projects and how would they get started?
With their first project, A to Z and I, I double down mark on what you've just said to all of my members. I said, look, don't overthink this. Start at a
[00:15:59] Tom Adams: Yeah.
[00:16:00] Mike Richardson: nothing. With nothing, which is just go to Claude Cowork and say, as you've been mimicking, Mark, I'm thinking about this. How on earth. Would I even get started?
Question mark. Hit return. And so I sent this, um, you know, we had our meeting Tuesday afternoon where this member was doing his demo by 7:00 AM Wednesday morning, I have an email teed up with that Netlify link in it. To this online platform that scrolls and scrolls and scrolls and takes the members A to Z into how they get into business with Claude Cowork.
I send it out to them all. They all loved it. That's project number one. Now I'm going back into Claude Cowork. I'm now going back into Claude Cowork and saying, okay, Jim had 16 projects. What are my 16 projects? You know where, where Claude can truly co-work with me and get work off my plate. So thanks for asking Tom.
[00:17:10] Tom Adams: That's beautiful. But that, that to me is the thing that is so evident now, which is, it, it, it's like Mark said, you do, you don't have to have this all figured out. What you have to do is start the process, because when you start the process, and that's. Kind of why I did the vibe code lab. 'cause I went, unless you see it and once you see it, you can't unsee it.
Right? Once you go in through the, once you go, you go through that doorway, you start going, oh my, I can do this and I can do that. Um, and I, I had emails afterwards going like, oh my gosh, it did this and now I'm doing that, and now I've added this thing. And now you fall in love with this crazy beast of AI, vibe coding, and really interesting stuff happens.
But. To me, the, the, the wonderful benefit is there's that weird moment. I don't know if either of you got it in the last little bit doing this thing, but all of a sudden you get this like, huh, this is unbelievable. Like, you get this little magic moment and, and.
[00:18:08] Mike Richardson: moly.
[00:18:09] Tom Adams: And in my vibe Code Lab, people were literally screaming out loud going, ah, like what is happening?
And they're seeing this idea. That's the thing we've never had before. And I mean, if you did it, it was, it takes so much time and it's generally with your hands, but now with with technology, you can actually see something emerge very quickly for you, which is where all the interesting magic is.
[00:18:34] Mike Richardson: But wait guys, but wait guys, I've got more.
[00:18:37] Tom Adams: Oh, oh,
[00:18:38] Mike Richardson: you, can I give you more? There's more guys. There is more. I wanted to come here and make you so proud of me today. You are gonna be proud. Trust me. Not only that,
[00:18:50] Mark Redgrave: You just turned up honestly.
[00:18:52] Mike Richardson: not only that, not only that. Myself and a colleague in the uk, we are using an AI empowered.
Platform called Book Magic to write a book together. And we, we've, we've begun to build that out and, uh, we're gonna come out with a, a, a minimum viable product, you know, 1.0 by 1st of July. Not only that, there's more yet guys, I am in Lovable. You've just been talking about that. And we are creating, uh, that will go hand in hand with the book.
We're creating a, a sort of self-assessment, uh, success model. Uh, and actually there are two lanes to it, each of which have 12 components. So it's actually gonna be a fairly sophisticated self-assessment model. And, uh,
we are
[00:19:46] Tom Adams: we bow down. This is
where we bow
[00:19:47] Mike Richardson: with that and loving, Lovable, and, uh, again, all part of this sort of 1.0 minimum viable product that we'll be releasing on July 1st.
So, you know, guys, I just hope you're so
[00:19:59] Tom Adams: Not. Not even worthy. Not
[00:20:01] Mike Richardson: see what you've given birth to here. You've given birth to this monster now.
[00:20:05] Mark Redgrave: We've created a monster. A monster with purpose. Oh, crikey. Not sure we can handle it.
On, on, um, you, you know, like, like there's a step, there's a step above. Like, before, like, help me think about Right, which is, and I don't want to gloss over this because if, if people are listening and they literally don't know where to start.
You can start there, right? Like if I was to put into Claude or ChatGPT query that says I run a plumbing company with 30 people, right? And we, we specialize in this. I would love to start using AI in my business. Like, help me start return. You're off to the races guys. Like, like, and, and then, you know, like, and so, so there really is now.
For a lot of people, and I totally respect this, it's like it is not interesting for them, right? Like, and, and it's, it's called, I, I meet plenty of people who go, mark. The fact that you get all hot and heavy and sweaty over this stuff, I don't care for it. That's cool. But you know, like when I, I, I certainly, my own experiences is like, it is so damn helpful guys.
If you are, if you are a business leader and you have a challenge, a problem, a thing to think about, like this is a thought partner. That knows the collective knowledge of everything everywhere, all at once. So like, use it. It's like, it's like it's um, it's a resource that I really think you can't ignore.
We're not talking about building apps here. We're just talking about if you're running a business and you're trying to think about a problem, and we've all got plenty of those. Like this is, this is a $20 a month thought partner that knows everything.
[00:22:00] Mike Richardson: And by the way, there's always been two kinds of people in this world. The haves and the have nots, the leaders and the laggards, the victims and the victors. Right? So this is just gonna be the same again. Yeah. Okay. If you think this isn't for you. Okay. Good luck with that. Um, uh, but you know, this, this is just gonna sort out who's who.
So much bigger, faster, bolder than ever before.
[00:22:25] Mark Redgrave: Yeah. You know, uh, a final thing about this, Tom, you'll love this, right? So my consumer app, and by the way, part of the mission here is like I have been trying for the last 10 years to find something that will sell while I sleep.
For millions of people, right? Like I, my job, I sell high ticket things to a few people.
I want to sell low cost things to a lot of people. So that's the motivation for me. But I was working it through, and of course it's a little bit, but, um, I was doing this in ChatGPT and they're like, and I'm starting to get really detailed, right? And, and chapter two said, Mark, yeah, I appreciate you raising this, but like, this is not important. And it said to me, it said, what's important here is whether you get to market, we can talk about, uh, we can talk about intellectual property and protection, or we can get to market. And he said, you could be, are you gonna be the one in 10,000 people that actually does something or are you gonna be in the other group that has a good idea and does nothing?
It said that to me and I.
[00:23:38] Tom Adams: Beautiful.
[00:23:38] Mike Richardson: it say? Did it say, mark, listen a-hole, did it that to you?
[00:23:43] Mark Redgrave: It was. And I was like, that is a, but that is a really, that is a really like powerful piece of advice, right? Because that is right. Yeah. Like, uh, uh, like you, and, and it was like, it was basically saying, stop obsessing over details that don't matter. And I, I was like, wow.
[00:24:00] Tom Adams: yeah, that's beautiful. So I, I had the opposite, uh, kind of, uh, experience. So, um, I was, uh, I was at another event and did a event podcast. So basically during the course of this conference, I, I hosted a podcast and interviewed people during the whole week or the, the whole conference event, but in advance I set up and they had a sponsor.
And so I worked with a sponsor and we had we, and so behind the scenes I built this little app, um, in order to schedule people so I didn't have to do all this work. Right. Just they could schedule the, the conference group could say to all of the. All of the vendors, hey, schedule here, get in their slot, things like that.
Um. So I did all this work in the last, you know, in the last couple of months I built it out and it was tested. I tested it myself. And so, um, the event, or no, the podcast sponsor was the first interview. And, um, so, um, 20 minutes before he walks into this little podcast studio we had set up, and, and I said, oh, you know, really excited to.
To your, your, your interview in, you know, 45 minutes or 15, whatever it was. And he goes, well, I got the 20 notifications, thank you. And, and I went, oh dear. Oh dear. And he says, I didn't need 20 just one would've been enough. And I went, oh dear. What is going on? So I jump on, I immediately pull up my laptop and go into Claude code and I said, uh, what's going on here?
Uh, event sponsor, first interview, uh, is gotten 20 of these kind of things. And it went, uh, let me think, let me think. And it said, shut it down. That's what it said. It said, shut it down. So literally. It said, I'm not, I'm not prepared to fix this right now. Shut the whole thing down.
[00:25:59] Mark Redgrave: I love that.
[00:26:00] Tom Adams: the very funny part of it all was the whole thing.
Like instead of going, Hey, make cool stuff happen, it's going, no, I don't, uh, we can't fix this. Shut it down. So literally had to turn the whole thing off. So nobody got their reminders. Nobody got an advance notification. All the things I had built in it basically said, yep, sorry, I can't help you.
[00:26:24] Mark Redgrave: Yeah, Tom, I've reviewed, I've reviewed the process flow here and the triggers and our recommendation is Kill it.
[00:26:30] Tom Adams: Kill it. And I, and I literally went, are you serious? You helped me build this whole thing from scratch. I don't know how it works. I literally don't know how it works. 'cause I don't know how to
[00:26:42] Mark Redgrave: This is how the end, this is the end, guys. This is how it ends.
[00:26:45] Tom Adams: No. So I went, eh, shut it down. Okay. I shut it down. Which is just hilarious. But it's so funny now 'cause it's like, like it's.
Inanimate, it's not human. And yet weirdly, it starts communicating like I do to Mike. Mike, you know, my coaching clients. Sometimes I have to say, Hey, think bigger, or other times I go shut it down, and now it's telling us the same thing. So it's quite
[00:27:11] Mark Redgrave: But if we're, if, if we're nice to it, does that matter? Like, because I still say please and thank you. I still say, I, I still say, oh, that, yeah, that would be great. Would you like me to do a one page summary? Yeah, that would be great. Thanks. I'm like, I dunno. Does it like, does that help?
[00:27:31] Tom Adams: Well, I, I have, uh, the, a layer for me, and I was gonna say, this is one of the things, um. You know, it Lovable and Replit and those tools do it in the background. But one of the things that is emerging right now is layers of capability that are layered on top now of like Claude Code or Codex or some of those, um, those more programming stuff.
And the guy who runs, um, the guy who's the CEO, Gary Tan of. One of the big, um, one of the big, uh, angel investment funds. I can't remember which one, but he's developed this whole thing where you download a, a layer of open source code into your Claude Code environment. And now it does things like, it gives you all of this brain extra.
So first there's Anthropic or Claude or OpenAI that gives you the, the context of the world. Um, but these layers now, so there's, there's this, what, what's called like CEO review and you do slash CEO review in your. Terminal. And what it does is it goes through a series of 20 questions and says, are you thinking about this?
Have you thought about this? Are you aware of this? Before you move forward, it, it like forces, uh, whatever your, whatever your LLM is, to actually think through the process. And then it says, okay, if you've got to there, then it could say you're thinking too small, you're thinking too big. Um. Narrow in on this and it gives you this layer of stuff, and I'm, what I'm starting to find now is there are all these layers of capability that are coming with people who've played with the tools that are being factored in that, to me now, I, I'm using them all the time, like I'm constantly using these skills that come along for the journey and I'm asking.
My environment to add those skills from these third parties. So I, I feel like it goes back to Mike, your concept of. Um, of artificial intelligence plus human intelligence. Where I'm seeing such cool magic start to happen is when you layer in what Anthropic or, or OpenAI brings, or Gemini, but you layer in these layers that humans are developing to make the the tool better.
And every time we do it, I go, this is Mike's model coming to play
[00:29:53] Mike Richardson: collective intelligence.
[00:29:54] Tom Adams: intelligence. It's beautiful to watch it unfold in front of you.
[00:29:58] Mike Richardson: So, Tom, I have another question for you, and then by the way, there's, there's still more. I've got one more piece of news that I do wanna share with you.
[00:30:05] Mark Redgrave: It's like show off territory now, isn't it?
[00:30:07] Mike Richardson: but I have a question for you, Tom and I, and I think this is very common in the early phases of blazing a new technology trail like this, right?
For your average person, myself included, it gets very confusing when you are trying to sort of understand. The categorizations of what's emerging. Right. That's why I asked you, know, are Replit and, and Lovable and Bolt are in one kind of bucket. They're one at one kind of category of something. Cowork, Co-Pilot is another kind of category of something.
ChatGPT is another kind of category of something. What, what is the best. Uh, what are the best attempts, Tom, that you've seen to sort of start revealing and educating people about how that sort of categorization map is emerging and unfolding, and no doubt morphing? It's all morphing, right? Because things mash into each other and they, they conflate and they get disassembled and all of that.
How, how are you understanding that landscape, Tom?
[00:31:19] Tom Adams: Yeah, and I, I don't know if there's a great tool because, and so the first question was, is there, is there a thing out there that sort of categorize this? And I'm not sure there's a thing that categorizes it, but what I see is there's sort of the chat tool, which is I interact with you. Um, and there's that interactive thing, and that's not that Lovable and Bolt don't have that.
But the, the first level is you just chat generally to get a, a thinking partner, a uh, search partner, a research partner, a document support, an editor, that kind of thing. Like you're working, it's almost like in, in software. What, what are that called? Pair programming. Um, there's this concept of pair programming.
You sit beside each other and you look over each other's shoulder and you're back and forth. That, that to me is the entry. That's the doorway drug. Um, and then to me, the next, the next evolution of that has been these vibe coding tools now, which kind of go back and forth, but you're doing a thing. You're not just in interactive zone, you're actually building something.
So the interaction goes into build, um. And that
[00:32:27] Mark Redgrave: Think thinking, thinking, building. It's almost like you've, there's two, two categories. Yep.
[00:32:33] Tom Adams: Yeah. And then I think the next layer from that, that I'm, I'm seeing is what are, and what's emerging now is this next layer, which is more agentic, and we've talked about that before, which is. You, you do some of that work in advance. Um, and a lot of times that's not done in Lovable, but it can be in Replit.
There's this next level where you're doing something in order to get this activity or action happening that continues to work on your behalf. Uh, and so that's where you're kind of building ag agentic, and that's generally done in more coding based platforms. Or you're hiring or you're like sa, it's a SaaS product you're buying now.
Like you can go get things and you talked about it the last time we talked, you B, you got a tool to check your email, right? That's running your email pro. Like I kind of call those AI assistance.
[00:33:25] Mike Richardson: yeah. And therefore, where does Cowork fit in that schema?
[00:33:31] Tom Adams: Yeah. Co Cowork to me is getting into this schema that I'm talking about right now. It's kind of like an assistant layer, which is you, what you did is you go into it and say, chat, chat, chat, chat, chat. But what I want is that cowork to co-work with me now. So it's kind of a gen in its approach. Right. But there's, there's an, there's an activity that includes Chat, but it's.
It's building something that's recurring, like generally in cowork. You're saying, I want you to do this on a weekly basis or a daily basis or something. Um, and then you can kind of categorize it into, I've got these documents, or I have this stuff, but then that the,
[00:34:11] Mark Redgrave: ag will come to life for the 98% through UI and like Wizzywig, like, uh. Platforms, right? There's just gonna be natively built into browsers and stuff like that. Because honestly, like I think agentic is way too much for most people to handle intellectually, me included. Like, it won't be long though. It won't be long.
It'll be, it'll turn up as a button on your browser, right?
[00:34:38] Tom Adams: But I think that's what you're getting with Cowork. And I don't know if you got the latest update yesterday, but the latest update to Cowork yesterday has new layers. 'cause both Opus, um, O Opus 4.7 and Sonnet 4.7 came out yesterday. And so this new layer has some interesting new capabilities that have been added to it.
Um, and then I think you get into the final layer, which is pure agentic work, which is, it's, so it's basically software development and then it just runs on its own independent of you. Then that's a, to me, a step past, um, Claude Cowork, but Cowork has been really cool. Um, Codex this morning, or OpenAI dropped this morning.
Its version of Cowork, so it's now live. If you have, if you're using the Codex tool, uh, which is basically the desktop version of coding or Claude Code, uh, from, uh, OpenAI, it now does that same kind of stuff. It does co-work. It basically showed up this morning.
[00:35:42] Mark Redgrave: I thought that was a pill I took for toothache. But that's codeine, isn't it? That's different. That's different.
[00:35:48] Mike Richardson: Actually I realize I've got two extra things to say,
actually, not just one, two. So, uh, in addition, at one of my other peer group meetings, I had a couple of members, um, do AI demos and, and one had just come back from a weekend long AI training class. He put himself on. And, uh, he had said about using Claude Cowork to build himself what he calls a command center. So basically, uh, a control center, almost like a cockpit for his, for his business. And, uh, he's, you know, plumbing in, you know, connectivity to this system and QuickBooks and his ERP and all that kind of stuff. And he was really excited about the progress he'd made. And then the other member that actually shared his screen and, and, and, and took us on a real time tour, uh, he's in a specialist business and he could never find a CRM that was close enough to his sort of business model and business process. So he set about building one himself and he gave us the speed tour of this thing. And it was remarkable.
[00:36:59] Tom Adams: Yeah.
[00:36:59] Mike Richardson: And, and everybody was like, you got to be kidding me. I mean, you built this from nothing. Absolutely. Using Claude Cowork.
And then he said, yeah, and then I went out. And did a bit of work in Replit and I did a bit of work here and I wanted to double check this and I wanna double check that. So it was a little bit more, but essentially the 80% of it he did, he did include cowork, and then he sort of went out and sort of bullet proofed it, you know, in some of these other places.
So I, it was just like so eye-opening. It was just unbelievable.
[00:37:31] Mark Redgrave: these SaaS companies are gonna get murdered, right? They're gonna get murdered.
[00:37:37] Mike Richardson: And then the other bit of news is, do you remember we had a guest, Riley Strickland? Um, from a company called Cadre AI in a previous episode. I forget which episode it was, but it's back there in, in the long history of this podcast.
And, uh, well, one of my peer group members who's in my sales leadership forum, um, he decided to quit his previous job and he's now become the VP of growth for Cadre AI.
[00:38:09] Tom Adams: Way. That's so cool.
[00:38:11] Mike Richardson: is gonna be on the leading edge of this thing. And, uh, Cadre AI has been growing, you know, very rapidly and just doing phenomenal work and really sort of taking its game to the next level.
Working with some of the big, you know, companies out there and doing great stuff. So, yeah, it's gonna be interesting to have a sort of behind the scenes set of eyes and ears inside of Cadre AI.
[00:38:34] Tom Adams: Wow, that's very cool.
[00:38:36] Mark Redgrave: just, just on the success. Do you remember like in previous episodes, do you remember we talked about the single person, billion dollar company back in episode two? Okay. Its exists. Like there's a guy, there's a company in, I believe in the Bay Area called MEDVi and run by a guy called Matthew Gallagher.
Um, you know, they've basically generated a year after he does GLP one drug logistics. Basically he set up his, his platform to do that. They did, um, they did $400 million in like their first year right, and valuation. Now he hasn't raised any money, but the valuation's well over a billion dollars.
One
[00:39:18] Mike Richardson: name of the company again? Med V.
[00:39:20] Mark Redgrave: Med, M-E-D-V-I-M-E-D-V-I. So like it exists. We, we said a year ago, like this is gonna happen, it's to happen
[00:39:29] Tom Adams: well Mark what I didn't. What I didn't tell you was I, I've already made a billion dollars in, you know, I, I mean, I, I know you, you find it hard to believe that I still come and hang out with you guys now that I've made a billion and, uh, I didn't share this with you at the last time, but we've had a literally a billion downloads on this podcast.
'cause I think
[00:39:52] Mike Richardson: I tell
[00:39:52] Tom Adams: the number one podcast about AI in the world right now. Like I, I, I just wanted to let you guys know, and, but I've
[00:40:00] Mike Richardson: what we call in the AI world, we call that hallucination? Is that what that is?
[00:40:04] Tom Adams: But I I've been kind of taking the money. Sorry.
[00:40:06] Mark Redgrave: it's billion dollars. Not billion Thai bart, or, yeah, like dollars.
[00:40:14] Tom Adams: Yes. So I have, based on what we've learned, based on what you are learning individually. 'cause the three of us on this call now are sort of the unemployable and we tend to help other people a lot. Um, what are you advising today? You weren't advising three months ago? What's different in what you're saying?
What based on, yeah, based on what you're actually experiencing now.
[00:40:38] Mike Richardson: Yeah, number one. Number one. Pivot to Claude Cowork, at least for the foreseeable. Period of time. I mean, who knows what will come along and leapfrog Claude Cowork, who knows? But for the foreseeable period of time, I'm so glad I did that. I'm hearing from so many other people, not least of all, both of you, months and months and months ago, and, and Ryan as well.
So I'm so glad I did that. And then really for me, it's what we've been saying in different ways. Yeah. You want to go on an A to Z journey? Just start at a. With I'm, I have this idea, or I run this kind of business. I have no, I want to start using Ai. I have no idea whatsoever on earth to get started, hit return and you'll be on your way.
[00:41:26] Tom Adams: Love it. Love it.
[00:41:27] Mark Redgrave: I had hope. It doesn't say shut it down.
[00:41:31] Tom Adams: Shut her down.
[00:41:31] Mike Richardson: Can't help you.
[00:41:33] Tom Adams: Shut her down. I can't
[00:41:34] Mark Redgrave: lemme, lemme, I'm, I'm gonna upload my, uh, you know, my, uh, my, um, key KPI dashboards for the business. And you get back, shut it down.
[00:41:43] Tom Adams: it down. Yeah. Yeah.
[00:41:48] Mark Redgrave: Not the answer we wanted, but anyway.
[00:41:49] Tom Adams: How about you, mark? What are you saying now that you haven't been saying in the LA like literally three months ago? Are you saying anything
[00:41:55] Mark Redgrave: yeah,
you know what? So. There's a, there's a difference, like most of my conversations with clients and as an, as an adv, like in my role as an advisor, okay. The thing that has really, um, changed in the last three months is this concept of, um, like. Having a man, A CEO mandate. Okay. As the answer to what? Pilot Purgatory.
So lemme just explain that. It's like the lots of people, there's a huge amount of energy, effort, noise, investment. Very little of it is still turning up in real business results. And one of the reasons for that is these projects are buried in companies at. A layer below where the CEO cares, that's a problem.
So a lot of the work I'm doing at the moment is helping people close that gap because it will own, it will only ever be a pilot unless you can connect the strategy of the business to the CEO's, like to get a mandate from the CEO. So, so honestly, like the last three months, that's really galvanized for me 'cause a lot of people are stuck there.
[00:43:06] Tom Adams: That's really good. Uh, the, the thing that I'm, I'm starting to say now because we're all experiencing the cool capabilities of Lovable, Replit, Claude OpenAI to develop conceptual pieces and toolkit and stuff like that. But the thing that I'm starting to notice, uh, a because I ran into it, but it's. The concept in theory, and, and you can build this really cool app that does all this cool things.
Uh, a you have to get it out in the marketplace if you're gonna sell it. But the second thing that I'm seeing is a lot of this work is. One step removed from full integration, right? So Lovable can create a super base database, but how's that connecting to your data? And the thing that I'm starting to see is a lot of people have vibe coded this thing, so it's really cool.
But it doesn't integrate anywhere. Like it's, it's, it, it's so, it it's kind of an island. It's kinda like Mark said, it's, it's a skunkworks project or a pilot, but the pilot doesn't actually connect. And so where I see all of the challenge now is. Integration with existing systems, and, uh, Riley said this on our call, is, is your database ri?
Like, do you have the right data to connect dots with? Those are the pieces now that I feel like I'm now starting to say, 'cause even three months ago it was really hard to make. These apps really connect together, but now they connect together and then you go, uh, but what do I do know? How do I, how do I connect the dot?
And that's where people are struggling right now, and that's what I'm focusing my attention on, is make sure your data. If, and, and I tend to work with the lower middle markets, um, on the lower end, but that's where they're all struggling is the ideas are a dime a dozen now. It's like, can you connect, can you make it connect to your stuff?
So, um, that's the focus that I'm taking right now.
[00:44:59] Mark Redgrave: Yeah, I concur with that. I think that's, um, but. The rate of development with these tools. Tom, as we spoke about at the start of this is like three months ago, it was like, like Lovable was a aesthetic skin to nothing. Okay. Now it's like, so, so who knows where we're gonna be in six
[00:45:17] Tom Adams: Oh, I know. That's, that's the
[00:45:19] Mark Redgrave: agreed. This is, this is the challenge. Like, and, and of course that just. From a risk point of view, now everyone's like, all the alarm bells ringing, right? It's like, hold on. Like we want everyone hitting new apps to our core database systems and to our ERPs, and everyone's gonna, everyone's gonna shit the bed, right?
So it's like, this is what's about to happen.
[00:45:38] Tom Adams: Okay, so, uh,
[00:45:40] Mark Redgrave: Can I say that? I can, I, I probably can't say
[00:45:42] Tom Adams: No, you can, I, I check, I check a button that says, uh, we are only for adults. Um, so one, one final question and I have to give it to Mark 'cause um, this is just Mark's thing. And then we'll, 'cause we've, we've hit our 45 minutes. Um, but Mark, uh, one of your favorite brands, Allbirds had a big, a big week.
And, um, I know that you pretty much dress in Allbirds and you, you know, you live in Allbirds, uh, their gear. But, um, something happened this week that I noticed on LinkedIn you kind of like scratched your head a little bit about. So, uh, tell me about your perspective of Allbirds.
[00:46:19] Mark Redgrave: I mean you like if you've been on LinkedIn in the last few days, you can't have missed this, but basically Allbirds, this is a shoe manufacturer. I lived in New Zealand for eight years, so they're a Kiwi shoe manufacturer. They have shops in very nice places all over the world, which is why they hit the skids, right?
$4 billion valuation. The retail market collapsed. Basically, they sold the assets three weeks ago for 30 something million bucks. These numbers are not right, close enough. And, um, and the, and it was actually like a real shame because it's a great brand. The product's good, not durable enough, but people loved the product,
[00:46:56] Tom Adams: I have many of them. I love their
product. I
[00:46:59] Mark Redgrave: footprint. Too expensive. A classic challenge for like modern retail, right? Like how on earth can you afford this stuff? And basically then they announced a couple of days ago that they were pivoting and a new company had been formed from the ashes of the one they just sold called New Bird Inc.
And is going to do AI data centers. And here's the thing, the stock that was worth nothing, shot up 600% in like four hours. So, so I'm like,
[00:47:30] Tom Adams: But did you
[00:47:31] Mark Redgrave: the whole internet
[00:47:32] Tom Adams: to that is yesterday it dropped the 398%. Yeah. So
[00:47:37] Mark Redgrave: I know, but it's like, but it's just, and, and every, the commentary's the same. And I'm not, and I'm not alone in this. The commentary is like, okay, that shows you where we're at with this, right? That from nowhere and like somebody with $36 million of cash. Can say, I'm gonna reinvent myself as an AI data center business.
Which by the way, is the business to be in. Like if, but, but you, I think you need 2 billion, not 30 million.
[00:48:04] Tom Adams: right.
[00:48:05] Mike Richardson: A little bit of experience
[00:48:06] Mark Redgrave: But, but it's like, but yeah, so, so I scratched my head Tom, and I just went, I just, this is so ridiculous. I think the internet responded, right? It's like, what is this? But did, I don't know, like.
[00:48:19] Tom Adams: I, what I didn't tell you was that the billions that I've sort of sucked outta the advertising pool from the Prompt and Circumstance podcast is now going into, we're calling it the Prompt and Circumstance da a Data Center Inc.
[00:48:33] Mark Redgrave: Very good.
[00:48:34] Tom Adams: And, and I just wanna let you know, you guys are, are minority partners in that business.
[00:48:39] Mike Richardson: Oh, okay.
[00:48:39] Tom Adams: you're gonna make billions. You're gonna make billions.
[00:48:42] Mark Redgrave: So, so who owns the liability to Oh, you do. Okay. Okay,
[00:48:46] Tom Adams: Full. Fully taking it. Fully taking it. Gentlemen, it's been delightful to catch up. Uh, love you deeply and uh, thanks for, uh, thanks for catching us up, and we'll talk in a few weeks.
[00:48:56] Mark Redgrave: Yeah.
[00:48:56] Mike Richardson: try not to be late next time. Mark,
[00:48:59] Mark Redgrave: Sorry. Yeah, I'll try. No promises.
