The Startup Ideas Podcast - I Ranked EVERY AI Tool Overrated or Underrated (Claude Code, Bolt, n8n, ManusAI, etc)

Episode Date: July 14, 2025

Join me as I chat with Ras Mic to discuss the current landscape of AI development tools, evaluating each as overrated or underrated based on their potential impact for both technical and non-technical users. We particularly highlight Claude Code as "extremely underrated" and discuss how tools like Devin and CodeRabbit are changing development workflows. The conversation also addresses my controversial viral tweet about building profitable SaaS businesses with AI tools. Timestamps: 00:00 - Intro 01:32 - n8n 06:35 - LindyAI 07:31 - Claude Code 12:09 - Devin/Code Rabbit 19:56 - Bolt/Lovable/Tempo 23:12 - Figma Make/Hostinger/Airtable 27:25 - ManusAI 30:10 - VAPI 31:08 - MCP 35:46 - Vibe Coding Profitable SaaS 51:11 - GitHub Key Points: • Ras Mic and I evaluate various AI tools as "overrated" or "underrated," including n8n, String, Claude Code, Devon, and CodeRabbit • They discuss the potential of MCP (Multi-Context Protocol) and its impact on non-technical users • They analyze Greg's viral tweet about building a $50K/month SaaS business using AI tools • Ras Mic provides a simple GitHub crash course explaining branching and workflow best practices Get Your Complete Financial OS at https://www.brex.com/sip LCA helps Fortune 500s and fast-growing startups build their future - from Warner Music to Fortnite to Dropbox. We turn 'what if' into reality with AI, apps, and next-gen products https://latecheckout.agency/ Boringmarketing - Vibe Marketing for Companies: http://boringmarketing.com/ The Vibe Marketer - Join the Community and Learn: http://thevibemarketer.com/ Startup Empire - a membership for builders who want to build cash-flowing businesses https://www.skool.com/startupempire/about FIND ME ON SOCIAL X/Twitter: https://twitter.com/gregisenberg Instagram: https://instagram.com/gregisenberg/ LinkedIn: https://www.linkedin.com/in/gisenberg/ FIND MIC ON SOCIAL X/Twitter: https://x.com/rasmickyy Youtube: https://www.youtube.com/@rasmic

So today we're going to go over the AI tools that people are talking about on X. And we're going to talk about if it's overrated or underrated. Okay. This is going to be spicy. Yeah, we'll be a little spicy. Like if any yet, somehow adds an affiliate. I don't know if they have an affiliate, but if they do, it's over. It's over, dude.

Please tell me they have what they make. Oh, yes. They're going to make so much money. for as of recording this it's underrated but we're gonna look back on this video in six months and be like why were we talking about menace Claude code underrated extremely underrated

Which is crazy because it's you know it's all over my feed right now Everyone's talking about it I think yeah cloud code if you're a non-technical person is more daunting But just just try You're not gonna single prompt a 50k a month SaaS but I do believe that we live in a world that it's completely possible to use things like cursor, to use some of these tools, to develop something that you can make $50,000 a month for minimum.

If there was any time to start something that has the slightest chance of being successful, there's no better time than today. Okay, so I share my screen and, okay, we'll do tools. And you tell me, let's start first tool. N8N. Overrated or underrated, Greg? I think, well, what do you think? I think underrated?

Only because if you know what you're doing, you can do well. But if you have like, you know, there's not an ounce of like, like, if you don't use modern tools, like you're not a heavy Google drive user, you're not on Slack always or you're not like setting up automatic. Email automation. You haven't used Zapier before, then yeah, this is basically useless. And I think we've all been seeing it on Twitter. It's the new drop shipping. Like I actually have, like if any yet, somehow adds an affiliate. I don't know if they have an affiliate, but if they do, it's over.

It's over, dude. Like it's, please tell me they have what they make. Oh, yes. They're going to make so much money. So, yeah, this is definitely the new Shopify because I see everyone talk about automation. So I think in that sense, it's overrated, but I have a feeling you think otherwise. Well, here's my take, because you come at it from like a developer perspective. So you look at it and you're like, yeah, I can go out and make these charts myself.

I'll just, you know, even if you don't want to do yourself, you go to chat GPT, you'll get the JSON, you'll copy and copy and paste. But for the vast majority of people, they copy these workflows and they don't work for them or something's broken or they don't know how to do it. So I think for the vast majority of people, it's overrated. Makes sense. But I do think for like somewhat technical or technical people, it's underrated. I get that. I call it non-technical and technical organizations. So you might not be physically writing code, but like you understand like technical concepts.

So I do agree. But yeah, if they somehow like start promoting this and do what Shopify did, this is a billion dollar company. we have the tab here we're actually playing with it you said so basically spring string or spring string yeah sorry i have string dot com so they're built on top of pipe drive which is basically an in any competitor yeah exactly so i feel like what string is doing i'm not involved in the company in any way but do you have to disclose that a lot nowadays i don't know the comment section is always like oh you're because i just did a tutorial on string and people were like this is a paper paid thing. And I was like, no, I literally message the founder like the night before. And I was like,

I saw your product. And I think it's really cool. And I actually had tweeted about that the future of automation is prompt to automation just because I think that people don't want to sit and make workflows all day, especially like non-technical people. No, that makes sense. So. And I can vouch that you probably message them because that's how I opted on the channel. You literally message me. And I'm like, and I thought it was a scammer first. It was someone else. So, I can vouch But I message him the night before He was like I'm leaving to Europe

For vacation But if you want to hop on tomorrow morning Let's do it So founder mode Yeah, founder mode. Yeah, founder mode. Yeah, I'm serious. Okay. Anyways, point is I think string.com Although parts of it are janky

Like they need it It is an alpha for a reason Yeah, they do mention alpha to their point Yeah, so yeah And by the way, I respect him in the tutorial he did with me He basically was like yeah, parts of it are janky I respect founders of AI companies that are saying stuff doesn't work all the time. Point is, I think that if you start using things like string.com and you start creating these

automations, like so much is going to unlock in your brain for how to like save time, make money that I think string.com is actually underrated. And I hope that, you know, the n8 ends of the world. I actually implement this. Yeah. I think it'd be very powerful if they did. And I don't see why they couldn't. They're so profitable.

Real quick, I need to tell you about something that's helping me sleep better at night as an entrepreneur. I was running multiple companies with money scattered everywhere, zero spending visibility and making growth decisions pretty blind. I knew there had to be a better way and my friend told me about Brex. So I switched. And now I have mission control. for every dollar. Spend management and visibility, virtual cards for my team, and earning cash from the same day treasury product. I was leaving so much money on the table, but not anymore. I have a financial operating system for our businesses to make decisions so I can spend

smarter and move faster. And it's been a total game changer. I'm loving Brex so much. I reached out to them and ask them to sponsor the pod. If you're building something real, you need a complete financial operating system like Brex. It'll help you sleep better too. And you can't be mad at that. Lindy.a.I seems like to be something similar, but fair to say, is it more marketing-s? Yeah. It's, it's, I mean, N8N is marketing in a lot of ways. But Lindy, what I like about Lindy underrated is that it has the templates. So it's got like, I don't know, 100 plus templates that will be like, okay, you want to create like an automated email, you know, email outreach agent.

We're going to go and do you press a button. You can copy and paste that template and it works for you. And then, you know, for me, why I think it's underrated is I love seeing templates because it's like, it gives me ideas around like, okay, maybe I wouldn't use it for this use case, but I would use it for that use case. That makes sense. Yeah. That makes sense.

Okay. So I need to try out Lindy for some marketing stuff because I'm terrible at that. I just know how to make content. Claude code, underrated, extremely underrated. Which is crazy because it's, you know, it's all over my feed right now. Everyone's talking about it. But man, like I, like, if I had to pick one for the rest of my life, let's say like AI just stagnates here for the end of time. And you told me cursor, winters, all these tools.

I should probably add that to the list. Who would you stick with? It's cloud code. It is the first time. And I showed an example here earlier while we were chatting. You know, like, and there have been a lot of videos on, like, how to prompt, creating an MD file, like, having a PRD and giving that as context. And those skills are important. And I'm glad people are thinking like that methodically, like, like a product, like a product person.

I actually want to even talk about GitHub later just to give a simple crash course for most people understand how to use it. But Claudecote essentially showed us that they could build an agent that does all of that. So it'll have, it'll track its own task, it builds out a to-do list, it has its own plan, it executes on that, it writes tests. Most developers I would say, or most indie hackers, I would say, hate writing tests. I hate writing tests. But it's necessary. Sometimes. you should still do it. It's safe, right? You feel good about the product you're building. Cloud

code does it automatically. Right. So in my opinion, if you're a developer and you haven't at least even tried Claude code, what are you doing? Yeah. And if even if you're not, please, just like there's tons of tutorials. I'm sure Riley has a video. Like Riley always has a Claude or Curseer video. Like there's some video there. I would spin up cursor, Winstrap, whatever you like, and use Cloud Code because it's truly, in my opinion, the truest form of agentic coding we've ever experienced. So for a lot of people who are non-technical, even going into cursor is daunting. Your suggestion is non-technical people should spin up something like cursor, get clod code. If it was easy, everybody would have done it, right?

Honestly, I know it's difficult and I speak from a place where like these tools are easy to me, but I just, I think you would do yourself a disservice to not try. At least try. And if you get stuck, yes, what you can use to figure out what to do next, AI. Right. So yes, the tools are not built for non-technical people, but I would still use it. And one thing that's exciting too about Claude Code is they released, they also released an SDK. So a lot of like the AI platforms can now use Claude Cod's agent instead of building their own.

And I was at the Cloud Code conference and when I was talking with their team, they essentially were like, you know, we're going to advise like a lot of these AI agent, like especially coding agent builders to just use our agent because in a respectful sense, they're basically like, we'll smoke you. Like our agent will always be better. Right. And I think Claude has kind of proven in one and owns the sphere of like, programming. So I wouldn't be surprised if all the vibe coding tools and like any coding tool like Devon, Code Rabbit, the Loveables, the Tempals, the Bolts, Figma Make, hosting your air table, all of them, anything regarding code, I do see in your future where they probably just implement all this because why go against researchers who are now valued at like what $100 million a year?

So I think, yeah, cloud code, if you're a non-technical person, is more daunting. But just try. Okay, so under, yeah, I agree with you, by the way. Underrated for both technical people and non-technical people. Yeah. Somehow. Because once you get it, it's, and like there's this, I have a little graph here. I mean, it wouldn't be a Ross Mike, Greg Vitty if there's no graph. And I would put like clawed somewhere, cloud code somewhere here, where it is difficult, right?

It is difficult to use. It's not the cleanest U.S. They want you to use a terminal and a terminal can be terrifying. But there's nothing like it. So I would put cloud code right here. By the way, you know we should add to this list as the last one? I can't believe we didn't think about it. What did I miss?

MCP. Oh, no. Oh, no. You know you pissed my people off. I know. I pissed thousands of people off with my takes on MCP. But we'll get to that towards the end of the episode. And then, yeah.

And then there's these new class. of tools. I haven't used Devin. Devon and CodeRabbit. Okay. So I would basically, I would say these two are like the closest thing to an actual assistant. So Devin and CodeRabbit are meant for you to build projects from scratch. Show me Devin and Code Rabbit.

Yeah, I can. So I'll start with Devin because I've been actively using it. So Devin looks like this. And first you think, oh, this is like, you know, any of the vibe coding tools. But it's actually meant to be integrated with like an existing codebase. And here's the powerful thing. There are three powerful things about Devin.

First, they have this thing called DeepWiki, which basically is like you can add like whatever code base is in your GitHub. Right. So like you can imagine for like someone building a serious project. You connect your repo and then I don't know how the AIA does it, but it contextualizes your entire project and breaks down exactly how it works. And I can't explain to you how valuable this is for anyone building something. You might not even be technical, but reading this document will actually help you understand how this system was architected, right? Like, for example,

by the way, for those looking at this, this is like an app I made for a few friends. And essentially what it does is it helps you create Apple passes for like your community. And, And it breaks down here how passes are created. And it will tell you here, for Apple Pasket Web Services implementation, device registration pass update, see the web. So it knows that I used an external service, Apple Pasket, and it tells me if I click here, I can read more about it. And then there's diagrams explaining how the APIs work, what the flow is, like what endpoint,

like, for example, device registration endpoint, it tells me it's this one right here, shows you even the back and forth. Like for anyone in an organization or even like maybe a small team, like three or four people building something serious, this alone is worth however much devonist charges. So this is number one. Number two, and this is also one of my favorite features is you can like start a session and you can basically tell it to either fix a bug, add a feature. And you can run multiple of these. here's the I think here's what's special about this is they have integrations if I go to settings and integrations you can integrate slack and linear and like in your linear ticket like let's

say you have a P and write out all the features and this is where I think yes it's for software engineer teams but even a non-technical person in a team can use this like imagine you have a detailed spec on linear on what this new feature needs to be you can literally add Devin, build this, and it will take all the linear information, and it will start to build a feature, and then it'll create a PR on GitHub. And I can show you one where I had an issue, and I asked it to fix it, and I can literally, can I click here, it literally wrote the PR. And mind you, files changed. It's one simple change. It was just removing this line. But I spent hours trying to figure out, why is this not working?

We've all been there. So like, and like the PR is detailed. Like there's reviews. I also use CodeRabbit as well. And then it shows my VERSEL deployment. And then I merged it in. So it's a different class, but might I say probably the most useful. Like the Devins and the CodeRabits and CodeRabbit is somewhat similar.

Like it does reviews and stuff like this. I don't know if they fully have this type of feature. But I would say they're sort of in the same realm. their actual assistance. And this for me, like if I was running, if I wanted to build something serious and I was working with two, three people, I think something like this

can actually be used by a product person who maybe knows how to write like a product document and they can ship features using Devon and I can review the code. And that's what a lot of my work has been at like now. I review more code than I write to be honest. Like I'll let Cloud Code do its thing I'll let Devin review, like, do something here.

And I'm just trying to understand what it's done. And if it makes sense, if it makes sense, I'll let it. I'll merge it. If not, I hold it back. So, and they have like a, I think, you pay $20 one time and then it's pay as you go. So, you know, just try it out. $20, a little fun. I think it's underrated.

It's not for everybody, but you won't know until you try. Like that's fair. And Code Rabbit? Yeah, so Code Rabbit is in the same realm. Like, these are like the assistant agents. I haven't given it a full spin, but where I've used CodeRabbit before is when you're doing these PR's pull requests, and I'll explain it when we get into like the mini GitHub crash course, it reviews your code. And there's actually, maybe I can find the tweet code with.

So there's this YouTuber called Code with Antonio, Super G guy. he is like the reason why most dev YouTubers no longer do like four or five hour tutorial videos. He's just the best at it. And we've all decided that he just runs this space. And what's interesting is he tweeted this. He said, in the last 90 days, I released three courses, which Code Rabbit reviewed 77 pull requests, left 42 comments, found 66 issues, and left 61 suggestions for me. Can I even say that I develop solo anymore? Just install it already.

Damn. So, and it's, and it's, and the thing is with Code Rabbit, what's cool is it'll identify an issue. And let's say, because someone I can already imagine someone's like, but I'm not technical. It'll give you a prompt to give whatever agent you're using to fix it. Context, code snippet and everything, right? So like, and when I tell you that Antonio is a G, like, if I, like, I just need to show you the stuff he builds, Greg, because. I feel like I'd appreciate it. He built a lovable clone the other day.

like and he does this himself like full on two billion dollar business he built it himself like and he goes line by line so like if you're someone who wants to learn how to code like he literally takes you like like this is what 10 hours and it's free and it's free yeah it's literally like it's free like he gives you everything and he and he and he talks pretty slow so you're not going to miss anything he's not a fast talker like me like so If someone like this, and all his videos are like 10 hours, what's this one, one hour 40s, this is actually his shortest video, 11 hours, 8 hours. This is crazy.

Yeah, and this is all free, bro. Like, this is all free. Like he, like, we've, I told you, as like, like, developer YouTubers, we have all decided that we will never do a video like this because you don't beat this, right? So if he thinks it's changed his workflow, like 66 issues, especially in a production code base, like this is serious, right? Like I know vibe coding is exciting and you could build your own project. But if you plan on having real users, which I hope you do, security is something serious. You have to talk about.

You know, your project actually working, users information being safe, payments being safe. And yeah, so I would say the code rabbits and the devons of the world. are super underrated. All right, let's move on. By the way, we should give a disclosure. You do work at tempo. I do work at tempo. So I'm biased.

You know, I get a paycheck every two weeks. Yeah. But with that being said, what do you think, actually? I'm curious. I put bolt lovable and tempo all in one line, even though I think tempos is slightly different, but maybe that's a biased talking. I haven't used tempo, to be honest.

Yeah. So I can't rate it. But so I'm leaving that out of there. Okay. To me, both Bolt and Lovable are very similar. Yeah. So, I mean, if you haven't tried Bold and Lovable, it's underrated.

Yeah. If you have tried Bolt and Lovable, it's overrated if you don't know how to prompt properly. And it's underrated if you know how to prompt it and give it context. Yeah, I agree. I think like anyone who's used Bolt and Lovable seriously will, will start to appreciate like how difficult it is to build software. Because like there's a certain level you reach and it starts to get difficult, right?

So to me, they're underrated in a sense where that initial joy everyone experiences when they prompt something and like result shows up. Yeah. Is like I've seen it and like a lot of people like it's, it's amazing to see people experience that. unlock. But then it gets to a point where you want to take it really, really far. I actually think the next big wave of businesses are infrastructure companies. Like your super,

I mean, super bases are, you're killing it. But like your super bases, your convex is like these backend as a service companies that will basically handle all that nitty gritty stuff and like can easily connect to some of these vibe coding tools. I think those are the big, big business plays.

And that's where like, you know, anyone who says, oh, we need to replace, like AI replaces. engineers. Now you need engineers to build that. Yeah. Because that stuff is hard and there are even certain things that even cloud code as amazing as it is can't do. But I would say for what it it unlocks and what it's done. And I think how it's inspired folks, I think it's rated fair. I think the people will love it,

love it for a reason. Yeah. And if you haven't used it and you use it, it feels like a new world. Yeah. And then in terms of like, is bold better or is lovable better? I think It goes down to like, you know, is X-Calad draw better or TLG-L-L-D-Raw, right? That's how I feel about cursor, Winsterf. I know I didn't put it here too. Like, I get it all the time. Which one's better at cursor, windsurf? And it's like, they both release updates at such a pace.

Just pick your camp and stay with it. Right. I also think, like, there's nothing wrong with prompting both. Yeah. I mean, yeah, like you could literally have both, right? So I agree with you. I don't think there's like a notable, because it's always like, I feel like it swings in pendulums. Like one week, bolt is better. The next week, lovable is better. The next week,

this is better. That is better. And I think for the most part, unless there's like specific features that one has that the other doesn't, I think as a non-technical person, you're living. You want to move on to number seven? Yeah. So these are FigmaMake, Hostinger, Airtable. I haven't tried and I probably will never try. Because these are companies that, I think, Figma doing it makes sense. Okay, explain. People don't even know. Yeah, so Figma's like the designer toolkit, right?

Designers. They're going IPO. Yeah, they're IPO. And is it a smart thing to buy it IPO? Because Circle, you know, I read on Google, oh, you're not supposed to buy an IPO. That's when the investors dump it. Circle went like Forex. But this is financial advice, so maybe we'll talk.

I'm actually a Figma shareholder. Oh. Hold disclosure. This guy's cooking. I'll disclosure, you know. But I keep it real. I'll be honest with Figma Make. I think it's overrated.

Yeah, I think so too. I think, do you think it was more so for optics versus like they actually want to like maybe compete with the tempos, the bolts, the lovable? Because to me, it's like, I didn't think they had to do that. I mean, I remember when I sold my company to WeWork, it was a mobile app. and we work needed plays like that because they wanted to be valued as a tech company, right? So before you go IPO, it's really, really important to be like on trend. Makes sense. And that adds to like the, you know, the valuation and the excitement of whatever it is you're doing.

So my guess is that the, you know, I don't have. have any inside information about Figma, just to be clear. Like, I'm a small shareholder via, I invested in a company, that company got bought by Figma and I got some shares. So I know nothing like inside about it. But that being said, like I would imagine that they're like, oh, wow, the bolts and lovables are crushing it and tempos are crushing it. You know, how do we get a little of that pie so that the market gets excited about our IPO? Makes sense. Makes sense. Yeah. So that that seems like Hostinger has one too. Airtable has one too. And I mean, if you want to have one too, you can go on code on Tony's video and follow his thing.

So I definitely think like, I don't know, do you think anyone can come in the space right now and be as big as the bullets and the lovable? Like, because like it just seems like every week there's a new one. Well, yesterday, I think it was leaked that lovable is raising. at a $2 billion valuation. Yeah. Kudos to... Totally. And we talked about investing in lovable.

Do you remember? Yeah. And we didn't. You know? I told one of my bros about it. And I was like, yo, I think they just raised $2 billion. I was like, at the time, I think it was this much. And he was just like looking at me like, bro, do it.

It was just like, man, at the time, it just, you know? Yeah. But kudos to them. You know, stellar execution. 100%. Like just world class. Yeah. You know, irregardless, like one thing I will say, and I think we share the same testament,

competitor or not, no other person or not, I genuinely just love seeing people win. I love, like, excellence winning. That's what I like, I like that about you. Yeah. Like, there's no, like, I don't be for no word. We're just here to, you know, do great things. So it's like, I love, like, you know, as much of, like, you know, competitor or, you know, whatever they are, like excellence is excellence, man.

And they've executed and it's been inspiring. And the pie is huge, right? Everyone can win, man. Everyone can win, man. Everyone can win. I'm a big believer. We all can win. Nobody has to tear down nobody to win. That's right.

Yeah. That's right. So, yeah, I don't even, I'll be honest. I haven't used Figma Make. I haven't used the Hostinger one or the Airtable one. Although Hostinger is a great VPS provider, you want to, I think they even have a template to host your N8N. They do.

And so like, cool for them, but I haven't used any of these and probably won't. Cool. So how about Manus? I know you're the biggest video on the channel is a Manus video. Not the biggest, but it was one of though. Yeah, it's a top video, you know, because I think Manus was one of the first, like, agent. Yeah, like it felt like, oh, this is like a Jarvis. Like, it could do stuff for me.

You can leave it. Yeah. Yeah. Yeah. And you were able to watch it browse in real time. Yeah, yeah, yeah, yeah. Which was also really cool. I still use Manus a lot. I think it's underrated. I agree. I think I've used it a few times, like for research stuff. And it was pretty good. Now, my only concern, like if I was part of the Manus team is, you know, with Open AIs deep research and all that stuff, you think they'll go into that direction or you think Manus is a specialized tool enough where it's like, there's still a place for them.

I think for as of recording this, it's underrated. But we're going to look back on this video in six months and be like, why were we? we talking about madness. I feel you. Yeah, because like it's like, I mean, Sam said it in like the most, you know, gangster way I can never think he's like, just make sure you align your growth with us. Otherwise, we'll steamroll you, right? So it seems like with operator and with like deep, deep research, even perplexities pivot to like a browser.

Like, you know, I wouldn't want to, I wouldn't want to battle those guys. No, I will also say with madness is I was using it the other day and it was like, connect your Instagram. No, they need to. I was like, I don't think so, you know, the Chinese government, you know. Let me just get your information. No, okay. Yeah. Yeah.

So be careful on madness with, you know, giving your information. Yeah. I genuinely, like, I actually have like, like, not a theory, but like a hypothesis that every site should just be sign up with Google. Yeah. Like done with the email password stuff. Because if you sign in with Google, there's no way you can really do anything

heinous with my information because Google will take care of that. Yeah. They give you just enough. So yeah, having to give IG access is pretty wild. Also, like we added, you know, on ideabrowser.com, the new thing I'm working on, we added sign up with Google.

It was just email before. Then we added sign up with Google. I was surprised most people sign up with Google. Yeah. Like it's just because like even for myself, Like maybe this is me being a developer. Like the fact, like I have to give you my password and remember it. And, you know, like password could potentially leak.

I think a couple billion passwords leaked not too long ago. Versus with Google, click, click, boom, done. Right? So I'm very like most apps I build will probably only be sign in with Google. And I like that. Yeah. Okay. This is more you.

I've heard like it's a voice agent or something like that, right? Yeah. Yeah. So it's voice agents. That's what it is. So it allows you to create voice agents. Oh. Yeah.

I think it's most people have never used it. So it's underrated. Like the fact that you're able to like. Oh, dang. Yeah. 400,000 calls. Yeah. The fact that you can create workflows that allow you to like create a Twilio number and then

start calls. calling, you know, add an Excel list of like, or Google Drive list of all these phone numbers. And then reach out to them and have, you know, have basically an intelligent, you know, voice agent speak to these people, I think is. Not a underrated. I just pin this. I'm going to play with their API later. Cool. Last one, not least.

Come on, man. Come on, man. Okay. You trip the whole internet. Show them the Theo. Let's hold up the Theo tweet. Theo is not happy at all. So Theo is like, you look up to Theam, right? Theo's my guy. Like I, like, I will say he was one of the inspoles, like, one of the biggest

inspoles for like tech YouTube, like for dev YouTube. Like his, like, and like I can, I can honestly like give credit where credit is do. Like, I grew a lot as an inch because of his videos. Where did he? He tweets a lot. Dang, Theo, he had me. Okay. So, Let's see. So while you're pulling that app, I tweeted that I think that. Here, once you understand MCP, you never see the internet the same way. By the way, the context there is I was talking about non-developers. Yeah.

When non-developers understand what MCP is, it's like a light bulb moment around, wow, I can use LLMs in this way. That's all I was trying to say. Yeah. Yeah. And, yeah, I think, like, again, Matt, Thiel's a very smart. developer, but this is the classic case. And Theo, if you're watching this, I'm wrong, let me know. But this is a classic case of a developer thinking from a developer's mind.

Like, I read this and I understood because I have a lot of, I would say 90% of my friend group are non-developers. Yeah. I spend most of my day with non-developers versus developers aside from my workplace, right? So I get what you mean when this blows your mind. Like when a friend of mine on, I think it was Claude Desktop, set. up like I think it was a Gmail connector or something and he was able to ask something about his emails blew his mind away it felt like Jarvis him so yeah but it's your next tweet that made me laugh

wait wait go back go back because uh yeah scroll down this guy relax dude it's just a JSON scheme yeah yeah like yeah currently building an AI agent like brad you see GitHub yeah like it's devs thinking about devs and then I responded it's not what it is but what it unlocks Like, that's how I think about it. Yeah. Given I understand this protocol better than 90% of the LARPers on Twitter. I could tell you for most of it. Like, it's just, uh.

And then look how long of it's free to rights. Yeah, I'm not reading all that. God bless you and I hope you go. It's like, yeah, what, you see, like, even like the fact that there's still people who don't know what MCP means. So unlocking MCP unlocks freedom. Okay, so why do you think MCP is, is a huge un-mock.

Yeah, because the thing is, it's like you're for like a non-technical person, the fact that my Claude desktop app can now have access to external information with almost minimal configuration on my end is powerful. It's like, to me, like the way I look at N8N, like I can build all those automations. And to be honest, for the most part, I will because I enjoy writing the code. I think I can maintain it better. and like I as a developer would feel more at ease having control over the code versus trusting an aden.

But for someone who's not technical, who might have some technical knowledge, what it unlocks is huge. And that's the same way I think about it too. So, and if you remember the video we did, I told you like from a debt perspective, it's kind of just an API at the end of the day. But for a non-technical person, being able to integrate these third party services into like an MCP client of your choice. it's money as game changer so yeah the devs the devs taken out on this one the the mcc

stands for money the money context protocol yeah i'm with that bro so i would i would definitely say this is underrated underrated yeah because like most people most people using chat gbt still don't have does chat gbt have mcp i'm more of a cloud guy yeah so like once that happens it's like imagine people i don't think they do i've never i've i've i've i think i've i think think they mentioned it that they were going to start building chat gbt i think they mention but i don't know if it's in there yet yeah yeah yeah but has mcp support oh they have oh but like specific okay yeah so and like outlook dude yeah like no thank you teams teams right yeah so yeah i think mcp's under rate i think devs being devs um

And that was interesting. And I think you definitely followed it up with some other banger tweets that picked up in Steve. To say the least. Like the 50K, I need to show people that the 50K, what was it? Is it not technical? There it is. Right here. Man.

Two and a half million views. Okay. You want to read it? I don't think you understand. You vibe code of sass. You can't read it. Oh. Yeah, you just, you just jump to 50K.

Your net margins are at least 50% because you don't need a big team. You're making 25KM1, 300K a year. You could sell business for 10.50 next day to do all day long. Your business network, your business worth is 3 to 5. You own 100%. You keep growing or selling. You double down the business. All of a sudden, six to 10.

Your customers love the product. Why are you raising venture capital again? I don't think you understand. I think people, you know, this was just such a. great jump that people for like you know this was your point why are you raising venture this is obviously best case scenario you're just on a hot streak right but i think people just saw this i mean where am i wrong in this besides besides okay i go down if you scroll down they said you correct yourself here i correct i corrected myself five to ten x uh i bidah which by the way you're

making millions of dollars building a SaaS. But tell me where am I wrong in this, in this. You know what it is? Because they, because they, because folks don't have the context of like the content you create. So this, I'll say this. If this wasn't a Greg Eisenberg post, I would think this is someone selling me like some course. So I get it. But the context of your content and the ideas you give and the point you're trying to make here, you've always been like, I remember, you're always the anti-VC guy. And people don't know this. You invest in companies all the time, actually.

So it's like, but you're telling them you don't really need to raise money. And I believe this too, unless you're building something where there's no way you can be profitable because like you're, you know, you're using these new AI tools and the credits are like anthropic is expensive. Like I've seen the tempo anthropic bill. It's high, right? So company like that, like the lovables, the bolts, they have to raise money because. If they're somewhat profitable right now, that's a miracle. I don't think they are. So for that type of business, it makes sense.

But if you're building like, man, your content generator, you don't need like Sequoia Capital to invest money in you, right? So I get the point you're making here. But people just read this. And can I tell you, you were on every engineer. Like you were co-tweeted by all my favorite developers. Yeah. Like PrimeGen, like even the small accounts. I was like, my Gregs, that tool.

Twitter check is going to be an amazing steak dinner. Totally. But I get your point. I get you're talking about raising capital here. And now, question, have you heard, like, of a vibe code? Like, I've heard some people, like, for example, this guy, I think his name is Alex. He used my template, like my SaaS template. And he's not really a developer.

He's not. I've had him on the pod. Yeah. So 300 bandos. Dude. And he was, I think he responded. He was like, yeah, absolutely. Yeah, we, no, we got a burger together in SF.

And I'm like, I'm like, brother, you use my template and you make more than me for my job. Like, you're cooking. Yeah. Right. So it's like, technically, I mean, he got some ways to go. But Alex is, he's on the way there. Now, again, this doesn't mean everyone's going to do it. Just like every, like I know so many talented engineers, so many talented engineers.

They probably won't reach that number. Right? So that part is on you. So I get what you're saying. But I also understand that when you wrote this, you kind of knew this was going to. Yeah, I know. I know. I know. I know. I thought it might ruffle some feathers. I know. Yeah. Because a lot of people are like, oh, how can you write this, this, this, this, that? I'm like, I know he smirked when he hit post. Yeah. Yeah. I mean, I truly believe in this post, honestly. Like, I truly believe that, okay, it's going to be hard. You know, you're not just going to like... And don't quit your job. Don't quit your job. You know? You're not going to single prompt a 50K a month SaaS.

But I do believe that we live in a world, that it's completely possible to use things like cursor, to use some of these tools to develop something that you can make $50,000 a month for minimum. Like, or $10,000, okay? Or $20,000. But like, $50,000 is... even that I think 50 best people

The 50 piss people off. I think if you did 10, it would have been calm. I don't think the tweet would have gone this viral. But to be honest, like if I'm being honest with you, I think that, you know, especially in a world where now you can like partner with creators and do TikTok ads, you know? Look at like what the Cal AI guys are doing.

And they're printing. Printing. Like $50,000 a month is less money than a convenience store. I will say this. Even I, you saying, that there's a part of me that's a bit uncomfortable even imagining that but that just shows like man like you just got to you got to switch your thinking well it's a limiting belief yeah yeah yeah it's possible it's possible i'm not saying it's probable but it's possible yeah and it's never

been more easier than compared to like five 10 years ago yeah agreed so that's my point of this post is that and that it could be life changing and that was my point and i and i and i and i leave this. You know what? Let me give it a retweet too. I'm going to get a retweet too. Because I was crying, bro. I was laughing at all the responses. But no, I think if there was any time to start something that has the slightest chance of being successful, there's no better time than today. And that's a fact. There's no better time. Like you can't say it was better five years ago, 10 years ago, especially in our like, I would say world, like tech, SaaS, whatever. Like,

In fact, I'll be honest, as a developer, like, you know, back in the day was better. The jobs, like, you would be hired for less skill. You can overcharge people to build the simplest apps, right? Like, I've heard stories of people pay like 50, 60K for, like, a simple crud app that you can like vibe code now, right? So there isn't a better time than now, and I fully agree. And stop hating on my boy, Greg. I mean, you can keep it real with me, you know? Like people.

No, I'll tell you this. Like I said, it's that 50K number. it made me feel uncomfortable. But because I know you and I know your content and I know the mindset you have, this wasn't you like gloating or whatever. But for an engineer, like I said, if I, if I didn't see your name, I would think this is someone trying to promo a course or something. If I didn't see your name and I didn't read the last two lines, right? So if it was, I don't think you understand your customers love your product. If it ended there, I would have been like, yeah, one of those probably has a course link in bio. But why are you raising?

venture capital again, meaning like there's no point in you doing this. And it being you and your content is when I saw this and I was like, I get what he's trying to say. But like you also have to learn in the in the deaf space, we're very much tired of like the fake guru. Well, it wasn't everyone. Like no, but like I think we have like an exceptional like like we despise it, especially because we see people fall for it. And that fake guru make like millions and millions of dollars. So there's like this deep. Like I know we all are, but with devs, it's deeper. It's too bad because I, the point

of the post was really to inspire people. And man, like, I told you, it made me feel like, I texted you this. It made me feel uncomfortable. I'm like, 50K month? Like. Jack Fritz. Did you see Jack Fritz? Yeah, he's like, easy money, right? What did he say? Go to his his ex? He's actually, uh, I've talked with him before, Jack. He's like making, what, 20? Something like that. I think it's in replies.

right no no he quote retweeted it he has a lot a lot of tweets so you gotta really go down read a little bit of that it's actually not out of the question for someone vibe coders sask to get 50k there are plenty of people who know 50k month is not a ridiculously high number it's entirely economically

feasible that 100 more vibe coders could be making 50k a month and wouldn't impact your chance of success either and he's young too right yeah mostly the limits you set on yourself are the limits you tend to get stuck on. So remember the next time you think 50K a month is not doable for you or some other Vibe Code or for it sure for for sure it is it's not easy but definitely doable.

Keep going. I'm 19k a month. I told my mom my goal was 30K a month and I knew she thought that was insane a bit. It was beyond her reality and just in reach of mine at the time making one K a month from my mobile app. But now I'm where I never thought I'd be and I could tell I held it back in these false limits just because I thought it was so impossible mounting. No plenty. No plenty. And I mean, a lot of people make 50K a month. Why not? Maybe it won't be. Why not you? A lot of people are making

a lot of people make 50K a month. Why not you? That's true. You vibe code of sars. Okay, Craig. Yeah, Sean is one of those devs. He's funny. But like, what did he say? He's like, you vibe code of sass. You get 50K a month. Okay, Craig. 519,000 views. I told you. I told You were on my feed. And it wasn't even your tweet that I saw. It was everyone's quote tweet of you. I'm like, what's going on? But I think everyone just skipped this part.

This was your conclusion. Your conclusion is you don't need venture no more. My conclusion, honestly, is that. And also, you can, you can do this. At least try. Oh, so there's an agreement here. He says, can you build a profitable SaaS? Absolutely.

Is it as simple as Greg makes it sound? Come on, that's LinkedIn successful. The I heard sass takes to the TV. agree with that. The average shast takes, okay, you know what? Maybe I agree with the average SaaS. Okay, great. Are we trying to create average products here? Yeah, this is an oversimplication of the process and misleads. I think, again, I think your post was taken as this is now the new reality. Yeah. Versus you're saying this is difficult. It's not easy, but this can be attainable. At least people didn't get that from the post, but I understand what you're

trying to say, right? Like, if 50K a month was easy, we all be making 50K a month. But it's not. But it doesn't mean it's not possible. That's right. Also, the paradigm has shifted. Like, yes, the average SaaS took, you know, what did he say? Two to three years to make... No, things have definitely got faster. Things have gone faster.

And I think a lot of people, like, even the YC companies pivot so much nowadays. Like, you're not sticking to something two to three years. Like, you ideally want some sort of product market fit much quicker. But, again, I've never grown a successful SaaS. He says, pretending the hard parts don't exist isn't abundance of mindset. It's lying to yourself. Real abundance is knowing the game is hard and knowing there are still tons of waste to win. I disagree.

I disagree. So I'll talk about that. Yeah. I fully disagree. It's lying to yourself. Real abundance is knowing the game is hard and knowing there are still tons of ways to win. So like what would you, okay, so where do you disagree here? Because I'm not sure.

Pretend you the hard person. I feel like that statement is leading from a place of fear and not leading from a place of courage and abundance. So if you want something, you have to picture like yourself at the finish line, right? Like if you're running a race, you picture yourself number one. Like think of a horse track, like the horses, right? The horse that has the blinders on and is just like picturing itself getting to, you know, finishing that line is more likely to be number one, even if it isn't the biggest horse and have the strongest legs and stuff like that.

we underestimate, going back to overestimate under or overrated, underrated. I think what's underrated is how powerful our minds are when we could actually say, when we actually could like think about the mat, you know, getting to the finish line and winning. Like when I start a project, I'm not like I'm trying to get to 5K a month MRR or I'm trying to get five or 10 customers. Like I'm picturing like, you know, the CNBC video and I'm picturing this or I'm picturing whatever is success for that project. And if you do that, I just find that you're more likely to win. You know what it is? I have this theory that in order for a company to succeed, there needs to be two types of people.

One with the rhythm and one with the tism, right? So CTOs are often like, you know, the more, I would say tism short for autism. more like technical, more predictable more. We have to do this. This is achievable. We can build this. And the CEO has to sort of be a dreamer that pushes everyone to feel out comfortable. And to me, this looks like a classic CTO versus CEO where like I think a lot of engineers and even myself, like, I feel like I'm like in two worlds where I get him and I get you.

And I heard this saying one time. It was like, pessimists are often right, but optimists make money. And it's like, I've decided that I want to be an optimist and I want to make money. So this to me looks like this is a CTO talking and what you're saying sounds like a CEO talking. I feel like if the inverse happened where he was CEO and you were CTO, I don't think you'd work out. But there's a reason why I think this dynamic between two people is the best in terms of building like a really big company. Like someone needs to dream and someone needs to remind us that the reality is. crap and we need to fix things. But I think both without a good balance might not be the best. But again,

this is all thought. This is all theory. I haven't built a successful business yet. So I don't know. I think what's interesting is the non-technical people like myself are starting to build products now. So now we're going to see who's really about it. I agree with you. I agree with you. And as the models are getting better, We're like, you know, I agree with you. But yeah, I definitely do think it's like a lot of work starts in the mind. And, you know, I definitely prescribe to the fact that, you know, if you want to build excellent things, it starts with some sort of belief and confidence that like, okay, I might not have it all figured out. Yes, it's hard. But at some point, if I go hard enough, long enough, strategic enough, you know, right place, right time that I can make you happen.

So, you know, I accept the inspiration. 50K month, here we go. Here we go. Can we end with a three to four minute GitHub? The simplest, this is going to be the simplest tutorial of life. Like, this is going to be so simple. Think of GitHub as this one giant line. This is your project.

And think of, I'll maybe, like, draw, like, little lines here. Think of these as, like, major mild. milestone, product updates, things you've shipped, right? Let's say for the first one that this was an admin dashboard, right? We shipped an admin dashboard, Greg, and our customers are extremely happy. But then we got some new feedback that they want like a live tracker in the admin dashboard. We currently have a working project. We currently have working code. What most people have been doing, with their vibe projects. You build and you kick publish, and then it publishes right away.

What most people have been doing is they're adding a new feature on top of working code. Now, here's the problem. Most people set up is as follows. They have their vibe coding tool, vibe coding or IDE for the devs, right? So this is where you write or you prompt your code. And then you push this code to GitHub, right? And then think of GitHub again as this giant line of like features you're continuing to stack. And then the way most people have it set up is you are,

this code is probably getting pushed to Versel, right? Versel is, Vesel or Netlify. These are the big boys. Now here's the problem. The problem here and the reason why GitHub is so important is let's say people want the live tracker, but our app is working fine how it is right now.

If I prompt my IDEE or my Vibico tool, even myself as a developer, I build this live tracker feature. And let's say it didn't test well, or even let's say we tested, but there was some bug that we missed. And I pushed this to production. This pushes to Versel,

updates the site, updates the app. Now I've pushed a bug into my live code. Site can go down. User can mess. Like, there's a lot of things that can go wrong. GitHub works as such. So let's say I've built, I'll remove this right here.

So let's say we've shipped the admin dashboard. Perfect. Now we want to add the live tracker. GitHub has this terminology or this thing called branching for we can copy the code from its last working point and create a new branch. And I can give a name to this branch. And usually the naming convention for like a production branch is main. It used to be master, but apparently master is offensive now. it's main.

Did I just delete main? It's main. And for like the new branch, I can name it whatever I want. So in this case, let's say we're a large team. I'll name it Mickey slash live tracker. And the reason why I'll name it this way is you know that it's me who's working on this. And by the name, you can kind of tell what feature I'm working on. So what I'm going to do is I'm going to be building out the live tracker feature in this branch.

This code is now separate from this code. This branch is connected to Versel. So as long as I don't push, as long as I don't touch this branch, the live deployment is fine. And the reason why GitHub is so sick is I can continue to work on this branch, thoroughly tested. And then when it's all said and done and it's good, I'll merge this. Let me get my pencil. I'll merge this back into the branch. So now I've had a clean run where this app is working fine.

I spent enough time testing to make sure all is well. And once I'm confident that all is well, I merged it. And I showed an example earlier when we were talking about Devin. This is a classic example where the change that was made was this right here. This line was removed. You can see the branch name here. The branch name is Devin slash some number, fixed, pass, template expiration. right so kind of the same naming uh naming convention you can name it however you want but i find that

helpful and then the way in technical organizations it works is a manager or someone else reviews it in this case it could be you or code rabbit and then you see this option right here that's uh that's uh that's hidden right now where you click merge and then once you've merged it it's essentially into your main code base so although most people will not be into github head heavily, I think understanding that this is how it works. And I think as the tools continue to get better, we're going to stop, like vibe coders will have access to these type of tools, like being able to branch. And this is essentially how it works. I have a main branch. This is connected to my Vicellar Netlify. If I push code to here, it redeploys my site. And in order to

build a very sustainable piece of product that works, that's functional, every single, every single feature I want to build or change I want to make. I have to turn into a new branch, right? I vibe coded. I tested it. It works. I'm happy. I'm happy. I'm happy. Then I'm merging. And there are times where you start and it never gets merged. But guess what? You don't care because it does not affect the main branch, which is connected to your Versailla Netlify. And that's a GitHub crash course. I appreciate that. Yeah, I appreciate that. I think, well, one thing I, you know, we often forget is GitHub is owned by Microsoft. Yeah.

So you got to think that OpenAI with Codex and the integrations that they're going to be doing with GitHub. And I think like all the vibe coding tools. I mean like this is even stuff we're working on on tempo. Yeah. It's like if we really want people to build serious applications, these workflows are coming. Right. For sure. And I can imagine it almost being like every time you work on a new feature, like the tool automatically.

creates a new branch for you, right? Makes sense. Because directly pushing live into production is... You're crazy. Yeah, especially if you're making money, you can... Especially if you're making 50KM-on. 50XM.

Yeah, yeah, yeah, yeah. Dude, thank you so much for the chat. Greg, this is fun, man. People need a... I'll link it in the description, but follow you, subscribe, and you'll be a smarter person for it. I appreciate that.

Thank you so much. It's always a pleasure. And yeah, hope to do these more. Absolutely. Until next time. My man. Take care.