AI Announcer (00:10.072)
Welcome to Prompt This, the podcast for business leaders who've had it with AI hype and want the real deal. I'm this week's AI intro voice. Greg's your season go-to guy for scaling sales teams and cracking the Silicon Valley playbook. Clint is a successful startup veteran who turns big ideas into thriving ventures. They cut through the noise to bring you analysis and playbooks on using AI to launch new ideas, scale your business, and stop getting left behind. Now, here are your hosts.
Clint (00:40.194)
Welcome to the Prompt This Podcast. This is Clint.
Greg (00:43.106)
This is Greg.
Clint (00:44.942)
Today we're going somewhere most business leaders won't. Down to the command line. That's right, local AI. Real leverage on your machine.
Greg (00:55.542)
I'll admit when Clint said terminal window earlier, I was ready to tune out. That's where salespeople usually step back, let the engineer drive and we pick up our phones and start looking at LinkedIn, but not anymore.
Clint (01:10.072)
Well, that's where the shift is happening. In today's world of coding, you're not actually coding like you used to. You're managing coders. The AI does the execution and you direct the outcomes. A very different world.
Greg (01:23.566)
Yeah, and that's when it hit me. mean, one person can now operate just like an entire department. It's not theory. It's not things we're reading. Like maybe this will come. It's a reality now.
Clint (01:35.244)
Yeah, this isn't really about tools, to be honest. It's more about power. If you're still treating AI like a chatbot, you're playing small.
Greg (01:44.45)
This episode is going to change how you think about who does the work. So let's jump into it.
Greg (01:54.272)
All right. Well, we're back here again. We've got command line Clint for part two of his AI CMO tools that every marketer needs to know right now. got to say though, last week was really interesting. You finally answered a curiosity that I've been hearing Lovable about everything and you showed it to me. So yeah, you did mention a few more, you know, while we were talking and I kind of want to get into a few.
You have some more time to show me some of these
Clint (02:25.228)
Let's dig into Claude today. Let's go ahead and learn about Claude code and Claude co-work.
Greg (02:33.193)
Yeah. I've been, you know, copilot. You've been chat GPT now for almost his entire journey. suddenly out of nowhere, we were talking and I didn't hear you say ChatGPT once all you've been talking about is Claude this Claude that Claude for marketers. Claude's better. So I got to see this. I got to see what you're talking about.
Clint (02:52.782)
Yeah, yeah, it's been a fun journey these last couple of months as I've gone deep down this particular rabbit hole and I'm looking forward to sharing it with you. Let's go ahead and jump in. Let's take a look what's going on. And in fact, I think the best place to start would be just picking up on where we left off in our last discussion. Right? Okay. So you of course remember this website that we made together in Lovable.
This is your new personal website, Greg Rosenthal, the AI revenue leader. Behind this website, which is live today at this URL, grosenthal.lovable.app, any one of our audience members can go and check it out. This is the designer interface behind it where I built everything. You can see on the left-hand side, I had all of the commands that I was just typing in, just natural language.
coming out just saying, make it look like this, make it look like that. So that's how we built this particular website. Remember?
Greg (03:50.638)
It looks no different than, you know, the way I'm working with copilot or just asking prompting and driving together a website. That's what.
Clint (03:55.885)
That's exactly
I'm exactly right. So now we're going to, we're going to drop down a level, get a little subterranean get a little nerdy here and, and take it to a whole new place. So here's what we're going to do. So one of the things that Lovable has is integration with something called GitHub. So GitHub is, a source code repository. So it's kind of like Google drive source code. Here you bet.
Greg (04:20.277)
Good. Let me stop for one here. So what I'm starting to associate here is, right. So now that, so we're at the coat, we were kind of coding here. We're at the coding level, but what Claude code is allowing us to do is just like any other AI tool that we're working with, we're managing the coder Claude is the coder. We are kind of the manager, the project saying,
Go make these directories, go find information, stick them in folders, make folders, go code things. then cloud code is your coder. That's the, that's the machine is the coder and we're managing a coding process.
Clint (05:06.67)
is the exact right analogy.
Greg (05:09.422)
This maps to all the AI stuff we've been doing. This just opens up a new world to you. This reminds me a lot of when we set up our email accounts and we try
Clint (05:24.686)
Yeah, we had to set up our DNS configuration properly and get the error We didn't know what to do in there.
Greg (05:28.946)
We just kept putting error codes in and saying, what does this mean? What should I do? Now you're down to code level being able to do.
Clint (05:39.118)
That's exactly right. This is telling me that I needed to run this one command first before I could start the web server. And I'm going to run this command called npm install inside of the folder space. And now the web server is going and it's saying that it's available at this particular URL, which is a localhost colon 8080. So I'm going to put this into the browser that URL that it's kicking off.
And this is your website now running on my local machine. And like just a few steps, right? I took the code out of Lovable by pushing it to GitHub, download it from GitHub to my local machine and got it up and running in seconds where I tell you this whole exercise when I was doing web development 15 years ago was days of getting things set up.
Greg (06:34.886)
Like from a sales standpoint, if you're thinking of a sales team in a room and a customer scenario, this is where all of us business and salespeople tune out and let the, S E talk to the tech guys in there. We, we just, we, there's no interest for the first time. I'm actually following along and super interested because I know at any point there's, there's,
a window over here that can just tell me what's going on. That's very interesting.
Clint (07:06.99)
And that's, you know, I got to tell you, that's the most important pro tip about AI is if you don't know what to do, you just ask AI, what am I supposed to do here?
AI Announcer (07:09.358)
single
Greg (07:17.602)
We learned that in like the second week and we laughed it off and now it's second nature. mean, that's what we do. That's how you get through the day now, you know.
Clint (07:26.99)
That's exactly right. So let's go ahead and change up your website a little bit. Remember how when we built it in level, we added this section on fun and talked about your deep appreciation for old school rap and that sort of thing. Let's go further on the Greg Rosenthal fun path here. So we're going to go back over to the Claude interface and I'm going to type in expand on the fun section of the website.
to include Craig's love for boating. You have a boat, right? You do have a boat, that's right. And tie it back.
Greg (08:02.766)
Did you have a boat?
Clint (08:10.808)
to how this makes him.
a better revenue leader. I don't know. I don't know how it's going to make that.
Greg (08:18.926)
Yeah, let's give it a little twist.
Clint (08:24.27)
So it's doing its thing now and it's digging in and writing code and it kind of spits things out at the screen along the way. A Claude will prompt you saying, Hey, do you want me to modify this file? And of course your response is sure. I don't know what you're doing. Just so you say yes to everything. Right. So if it comes back in it, in it asks you, you want to do that?
Greg (08:49.806)
Why enter, why enter, why enter?
Clint (08:52.256)
That's exactly right. So here it is. It's asking me, do you want to make this edit to fun.ts.
Greg (08:59.308)
Absolutely. I've been coding for like four minutes. I know that that's what you choose.
Clint (09:00.716)
Absolute.
Clint (09:07.118)
That's exactly right. It's almost kind of like that's your one of your most important skills as a cloud coder is just to say yes.
Greg (09:15.437)
Right?
Clint (09:16.206)
All right. And so it's going through and modifying stuff and telling you how many tokens it's used. We're going to go ahead and let this play out here. And now it comes back and it's added a new section and it, uh, in the cloud code interface and the text interface, it's telling me what it changed, but I'm to go just right to the website, right? So here's the website and, um, I'm going to.
Greg (09:42.03)
Look at that, says boats and boom bap. Now yeah, boats.
Clint (09:46.498)
Beats, bars, boats, and boom bap. I like that.
Greg (09:50.704)
wow. So yeah, I did do some of the kind of Lovable style designing at the code level, but when you scroll through on the website, it looks just like we're designing a website.
Clint (10:04.814)
This is all new content that it just... Life on the open water. When Greg isn't building revenue engines or breaking down bars... Breaking down bars? What does that mean? You'll find him on the water.
Greg (10:07.128)
We have life on the open.
Greg (10:18.889)
is that the college gray? Yeah, I think that was.
Clint (10:21.688)
College. Boating isn't just a weekend escape, it's where some of his clearest thinking happens.
Greg (10:29.142)
It's kind of right though, actually, it's kind of right. Yeah.
Clint (10:33.134)
That is your meditation time?
Greg (10:35.51)
Yeah. Yeah. The farther I moved up the ladder, the more intense it got. And if I just got in a boat and started driving things just started to release for me. got, I down answers to big questions. Just popping my head.
Clint (10:51.256)
There's your quick introduction to Cloud Code. Is that what you were thinking it might be?
Greg (10:56.694)
No, actually that kind of blew me away. I was thinking something different, because I just didn't have a frame of reference. so I thought you were like actually building code and, and, know, building libraries of code. I don't know the right word for it, but that you would then go and use in some other applications, but you're, know, you're, you're, making code that you're like saving off, like a document or something.
something that you used to
Clint (11:28.306)
I you, was the world you lived in before this, just like creating documents with Microsoft Copilot or ChatGPT or Claude has just changed dramatically. It's the same thing with code. In fact, this is where it all kind of started. The people who built the LLMs were setting out to make their own lives easier first by building an LLM that allowed them to code better and faster.
Greg (11:55.374)
This is really cool. I'll tell you one observation. Um, you know, I've sat at many, um, meetings. I've sat in, you know, many,
Urgency problems that we're having with the cloud that we were running our production solutions in, over thousands of customers. And I've sat next to the guys that actually have to check in and check out code and fix things. And, and they never look as calm as you just did while executing through code. And they usually type 10,000 times faster than I could ever type.
I'm just watching them. look, look worried here. You're like, Hey, why don't you just, make me an entire website and, be sure that you put it the right directory on my machine. Okay. It was sad.
Clint (12:49.262)
a mental shift because that's the way it used to be, right? You had to be aware of every single detail of what you're doing when you're coding. And it was a very focused kind of game on type mindset when you were doing it. And now you can spin off three or four different terminal windows and have three or four different instances of Clod all modifying the code base at the same time. instead of managing one coder,
I could easily be managing three, four, five coders who are doing stuff at the same time in parallel. It just completely changes your entire approach to development. Again, it's much more of a managing exercise than a dive into the deep end of the pool and try and figure it all yourself.
Greg (13:37.038)
Well, I'll tell you Clint from, you know, before we started all of this, was a very good sales person and in any of the seats in the sales org, was getting sales management, was a business level executive. And now that we've started this, I'm now a marketer. I can do food, I can do websites.
Clint (13:59.672)
You are about to be a-
Greg (14:03.764)
I can, I'm PR, I can do PR now. I can do all of these things and we're only like three months in, four months in, something like that. And, you know, we're scratching the surface with some of this, but this is incredible. This is the side I never thought I'd see deep coding.
Clint (14:23.284)
This point you just made, think is probably the single most important aha moment coming out of all this AI stuff, which is one person can suddenly be like a whole small business operator themselves. They can be the marketer. They can be the website developer. They can be all these different roles. It really just kind of empowers you to do things that you never thought you could do.
And that to me is the most exciting part is I'm in there just doing stuff that it's not just faster on the things that I already know how to do. It's doing new things that I didn't know how to do. That's the crazy part.
Greg (15:01.782)
Yeah. I mean, it's like for me and me for, you know, mirror, I don't know how much this costs per month. Probably not that much different than some of the other solutions. just got a computer science degree and I didn't have to go through what I watched you go through years ago to get that degree.
Clint (15:22.552)
Just think about what business is going to be like when anybody can be a decent coder, anybody can be a decent marketer, anybody can be a decent salesperson with the help of AI. Now, I tell you, there's still a difference between being an expert and being decent at it. Absolutely. Right? yes. Even in the case of what I'm doing here in terms of the code that's happening, I got to have an understanding of the background principles in order to put it to work. But
That's just a fraction of the knowledge compared to what I needed to know in the past to make all this happen.
Greg (15:54.264)
That's something I like, you know, I could pay for a training class or online class or do some research, some YouTube subscriptions and could get the basic knowledge down. And let's say, you know, if I'm going to tackle something that's very important and building something for my business, I could learn that in hours to days, then be able to jump in and be effective with cloud code.
This is interesting.
Clint (16:24.566)
Let's shift gears out of the code and go to Claude Cowork. You ready for that?
Greg (16:31.416)
So maybe from the top level real quick before we dive in, what is cod coward? Like what's it's.
Clint (16:37.458)
Absolutely. So first off, to access Claude Cowork, you need to download the client-side application that runs on your local computer. You can't just go to Claude.ai and interact with Claude Cowork. You actually have to download the application. And more than anything else out of the gates, it started off as just another interface to Claude. Like ChatGPT has and Copilot just have a local application. And then they started expanding on it.
and the lead developer at the company Anthropic, which is the company behind Claude, a guy named Boris. What he discovered is that when he created Claude code, people started using Claude code for doing lots of things other than coding, right? All the kind of AI stuff that you would do. And he said, well, you know, that's a great idea. Let me go ahead and expand the user interface of the Claude application.
so that you don't have to be, you can be doing the code stuff, but you don't have to be in a command line interface. And so you can see here how they added in Claude code into the application so that it's a standard web-based interface style as opposed to a command line style. And then they added in cowork like a month ago. This is new stuff, right? This is only like six weeks old, right? So in there, they're still kind of figuring out everything that Claude cowork can be.
And I'm gonna walk you through some steps here, but I'll tell you, Greg, there's a lot of hype around Claude Co-Work right now. And I think it's frankly a bit more hype than reality at this particular moment in time. There's a lot of great ideas in there, and I think it's gonna turn into this much more powerful tool over time. Again, it's only six weeks old at this point, but it's a bit hyped at the moment.
Clint (18:34.862)
Let's take a deeper look here at Claude Cowork. And what I've done in the background is I've downloaded the MP3 file and the transcript of the podcast that we did last week. let's do a little work on the podcast. some things I do this, have to do this every week when I create a podcast. I create the show notes for the podcast. So let's, let's see how I would do that in Claude Cowork here.
First step, I'm going to go to prompt this AI and I'm going to open up one of our previous episodes, which has show notes in it. Right. It's got the episode summary, guest introduction, this week's AI challenge, all that kind of stuff. And I'm going to tell Claude Cowork, first I'm going to ask it, do you have access to the?
MP3 transcript in this folder and it should say yes.
Greg (19:31.096)
That's an important step. You know, I've, I've seen that even in co-pilot. If it doesn't, it'll in co-pilot, it'll pretend it does and keep answering you. And then finally there's a context clue. I'm like, you can't even see this. Can you really see it? And it goes, no, I can't see that file. So I usually start with that. Can you see this file? Can you see this screen? Is this something you can interact with once it says yes, it is. Then we go deep with it and totally works.
Clint (19:55.694)
That's a good pro tip.
Clint (20:00.184)
Cause you can spin your wheels trying to make something work and not realize it. Let's can't see the file. So here we are. It gave me a quick summary of the podcast transcript. This is the one that we did last week about Lovable. And I'm to tell it to create show notes for this episode following the format of a previous episode at this link. And I'm giving it the URL of
the previous episode. So it's saying it's working on it. It's updating its to-do list, whatever.
Greg (20:35.806)
That's interesting. It keeps its own to-do list. So this is kind of like it works differently.
Clint (20:40.77)
Works differently than Microsoft Co-Pilot? All right. So I'm ready to show you Claude Cowork. you know, in the background, I had some, I had a few issues getting it all set up properly. Claude Cowork requires a Chrome extension plugin that has to be configured. And, and things didn't work right for a few minutes. And so I had to go to debug mode and figure it all out. I think that just kind of represents the fact that Claude Cowork is a new product and, and it's still
Greg (20:42.594)
Yeah, all right.
Clint (21:11.054)
going through the kind of the burn-in phase, if you will. So I would not necessarily recommend Claude Cowork right now as a production-ready tool for all our audience members to jump into. This is more of just kind of a basically a little teaser of what Claude Cowork will be able to do in the future. Just not quite there. Yeah, not quite there yet.
Greg (21:33.496)
Devling.
Greg (21:38.126)
Okay, it's time for this week's AI Challenge. Now the AI Challenge is a takeaway assignment for our listeners to get their hands on some AI tools and do some exercises.
Clint (21:48.398)
Today's AI challenge is about pushing yourself. I pushed myself recently. I went beyond ChatGPT and into cloud code inside the terminal, inside a real source code repo. And I'll tell you, it stretched me. I some help getting started and you absolutely need some technical fluency to unlock the full value of cloud code. But that's the point. There's a ceiling to AI literacy if all you do is prompt in a browser. So this episode is a challenge.
Pick one tool beyond the obvious, Gamma, Clay NotebookLM Cursor, Lovable, something that makes you slightly uncomfortable. Two hours, real output, ship something. That stretch is where your next level of AI capability actually gets built.
Greg (22:38.178)
Yeah, look down into the show notes and you'll find the link that goes right to the blog. It's got all the instructions that you're going to need. Now, if you've just finished an AI deployment within your business, we'd love to hear about your experience, the good, the bad, all of it. So go to www.promptthis.ai, go to the contact us page, fill out the form and we'll be in touch to talk about the story.
Greg (23:06.318)
All right. Well, that was pretty interesting. And now I see where you're, where you've been spending all of your time over the last couple of weeks. And, again, those were AI wow moments for me. We've been talking about it. you know, we've been learning a lot, kind of talking about the same things and different layered over different people's crafts out in the tech world and, business world. Well, this just took the tools right. Right. Yeah. Right home. And.
That was a wow moment. Anyone can be a coder
Clint (23:38.397)
Anybody can be a coder
Greg (23:40.632)
can execute like, you know, multi tasks and, and, this was just a sample. Anyone can, you know, can have Claude cowork basically execute your machine and you just have to manage it. That was a wow moment for me, Clint. So yeah, thanks for bringing that demo.
Clint (23:58.37)
is so changing the way we work. It's exciting. It's amazing. A little bit terrifying. But the way we work is going through a dramatic change right now. And I'm personally excited.
And that's another episode of Prompt This.
AI Announcer (24:21.486)
Thank you for participating in today's audio-based human interaction featuring Clint and Greg. To access all previously recorded Prompt This data units as well as extended knowledge modules, please navigate to www.promptthis.ai. Directive, engage the follow mechanism located below this transmission to ensure future episode delivery. System, message, we have logged your presence. We anticipate your return.
