Changed my mind: AI is actually good for coding

Welcome to The Panel where bootstrap founders talk about building a better business and a better life. I'm Justin Jackson, the cofounder of Transistor.fm.

And I'm Brian Casel. I'm a product designer, developer, just launched Instrumental Components. And maybe today, we'll talk about what I'm doing next.

Yeah. Alright. So last episode was a lot about me and Transistor and the podcast industry. I had lots of great feedback from folks about that episode. We also talked about the format of the show, how difficult it was to get panelists every episode.

Yeah. I'm excited for this new format. I like, to be honest, like, I'm not expecting a whole lot of live interaction. There there might be a few people. I I see some familiar faces in the chat already.

That's right. Like, in the future, and maybe even this episode, if we get really crazy, we can add people as guests so I could DM somebody a link if if we're feeling like, hey, you know, you should be in this a part of this chat.

I also like the idea of the fact that we're recording it today, June 5, and we'll get this episode out, like, pretty quickly. Because, like, especially if we're talking about real time things that are happening in our world this week. So it's kind of annoying when that comes out, like, two weeks later. It's like, oh, I'm not even thinking about that anymore.

Yeah. I mean, we don't know what this is gonna become. But this could become a mashup of Bootstrapped Web, Build Your SaaS, and the panel all in one thing. We might even release more episodes. Maybe we can release weekly if we do it this way.

Yeah. I mean, it's always in like, for me, it it's always interesting how, like, there's what I'm actively working on. There's also what I'm actively publicly promoting. And for the past couple of weeks and months, that's been instrumental components. But the reality has been behind the scenes, maybe equally, if not more, I I've been plotting and and researching and exploring really what I plan to do next after instrumental components.

What is that like not you didn't have high expectations because are you trying to protect yourself a little bit? Or is that just like what you Probably

some of that. But but also just like being real about it. So people might be wondering, like, it's it's done just about 10,000 in revenue so far. Okay. In the first it's been so I started inviting customers, what, three three, four weeks ago. That's that's been okay, you know?

So $10,000 in revenue so far. And you you kind of rolled out the launch. And that's with a waiting list of how many people?

About 800.

Okay. That

So, like, that's that's where it feels like it underperformed. If if you just look at the numbers, like, I I don't know. Like, what's the number of customers on that? Like, 20? Something like that?

I mean, to me, that seems pretty reasonable. Did did you happen to catch Aaron and Ian on the Mostly Technical podcast talk about that much?

By the way, Aaron and Ian, I thought that was one of your best episodes yet.

Yeah.

Absolutely. As I mean, Aaron, as as believe me, I've I've been there, like, as as painful as, like, the, you know, the the flat launch thing goes. But I don't I don't see it as a failure anyway. That's a whole other story.

But I wonder if air if Ian's point because one of the things Ian was saying and Ian also just had this big session with Caleb who does Flux, which

is a similar product. That was almost like too long. I couldn't get into that.

But I'm wondering if there's anything there. Like, are you thinking about this as a info product, which typically you launch like a book or a course, and it's like a bunch of sales in the first three days, and then it goes down. And then you have to relaunch it every year. You have to do a

Yeah. Okay.

So that's

an interesting question. Because it doesn't Yeah.

And Ian was saying, maybe some of these products you need to consider, you know, you need to consider them to be actual products. So like a SaaS product, you launch it and, you know, out of the gate, you get a certain amount of sales, but then it builds gradually over time as you get all the channels kind of so which do you feel like instrumental components is you're gonna treat it more like a product, or what's your feel?

Great question. Because there's so much wrapped into that in terms of what were my expectations, what are my future plans for it, and also how I'm pricing it. Because I did change from the recurring model to lifetime. I just made that change a few days ago. Okay.

By that, you mean like AI or

anything AI and like how things are built. That that That's just like one thing that's in the landscape. But when I started building it, I I was like, I just gotta get back into the rhythm of building and launching new products. I'm coming off of four or five years of just being solely focused on Clarity Flow, my SaaS business. Right?

Okay.

It to me, the get go, it's always been like, let me try to just get back to building and shipping a small product. Let me go back to the first stair in the staircase. This is not a SaaS business where I'm like, let's launch a SaaS and focus on it for the next three years, whatever. So there's that. And then there's also just the expectation of how big is the Rails space and the market for this really, right?

Yeah. Yeah.

Yeah. So like there there might be people who like the look of instrumental components, but they're like, that's interesting. I just don't build in Rails. So it it wouldn't be even compatible with them.

Okay. Let let let's just pause here because I want to react to some of that stuff. I think some thoughts on the top of my head, and maybe on other people are thinking about this too, is one question would be, why not just go for projects that you think have the potential to be big? So why are you are you limiting yourself in saying, oh, I'm just going to do this. It's just small.

Good question. I think I think what I'm thinking about now in terms of like the the thing that I'm starting to really invest in going forward, that's more where my mindset is at. Like, whatever I'm starting now should be bigger, a bigger commitment, multi year investment. Right? But the last two years, I was in a reset mode.

I still think folks might be interested in that because really, the tension I'm hearing here, and I think this is just what we all feel is I just was listening to the Steve Ballmer interview on the Acquired podcast. And Steve Ballmer's like, listen, every Microsoft success like Azure, everyone thinks that just popped off in the last few years. And now it's made Microsoft the most valuable company in the world. He said we were working on that for eight years before anybody even heard about It he says things just take time. And a lot of the stuff that we see that's a success.

What I would say to that is that that would that would be an excellent point to be made if I was in year one or year two, and maybe in maybe in year three of Clarity Flow. I did not come to the conclusion because because for the first three plus years, I did not do anything else. I I sold everything else.

Yeah.

Yeah. I I literally sold AudienceOps. I sold Productize.

Yeah.

Like like the course thing. I I sold that to someone else. I literally did nothing else for for over three years. Right? I did no consulting. I did I did no info products. I did no other software tools.

Mhmm.

It it it came to a point around between year three and year four. This is like now beginning of twenty twenty four. Where it's like, okay, like, what are my choices? My choices are raise more investment to to to, like, extend out the runway. Or maybe have some, like, magical thinking of, like, sometime in the next six months, we're going to rocket ship our growth where there's just no evidence that that's going to happen.

I think this doesn't get talked about enough, which is there is some amount of I guess luck would be the word. But, like, the the biggest threat when you're bootstrapping is that you're gonna burn out or you're gonna go broke. So you either burn out psychologically. You know, people get the shin people get shingles. People get like

Which I did.

Wish you did.

Around that time.

Around that time. Yep. Or you run out of money. And, you know, with it with Transistor, looking back on it now, especially if you listen to those episodes that my business partner, John, and I were recording at the time, we were close to breaking. Like, it was like, this thing has to work, you know?

So if we Yeah.

We we built it from together from like January 2018 to 08/02/2018 when we launched. And then it took us one year after that to both go full time on it. And you look at Adam Wavin and Steve and Steve Shoger. You know, they had this financial engine that was working with their course. And that was just, you know, pulling the whole business. And that gave them time to go and explore this other stuff.

Yeah.

And so this is it's hard. Like, you never know how long it's gonna take for that financial engine to get going. And and then you also just don't know, you know, like with Clarity Flow, for example, like it might be still growing slowly, but it's like, you don't know when it's gonna get to the point where it might be able to be Mhmm. The full time thing.

It's just hard. Exactly. It it absolutely is hard. So anyway, like, that's to me, like, that whole decision logic flow, like, that that's decided and I'm way past that. That was like two years ago. Right? And so I spent the last two years, the year of 2024 and now into 2025, sort of picking up the pieces and and figuring and in the meantime, doing some consulting projects like MVPs as a service.

Mhmm.

And that's been good for cash flow, like keeping keeping the lights on and and that's you know, that covers more than half of it. But Clarity Flow still is still just something in my portfolio. It's not but it's just not enough to be my only thing. Yeah. But so then so then it was like, okay.

Mhmm.

You know? SaaS businesses are great. And I'm not against And maybe maybe I will hack on something that that becomes sort of like a SaaS business again. But like And technically, I What I build in software with Rails and whatnot, they These are SaaS applications. But I'm sort of over the idea of like, I'm gonna launch a thing that's $19.49 dollars a month.

Mhmm.

But what that would have required is more investment. Yeah. There's just no way around that. Right? Because because I was spending a lot on marketing.

And

I I tried all the experiments. Like, tens of thousands on ads, on SEO, on all of it. Some things sort of worked, and some things still work for us today. We get customers mostly from SEO. Even our success, even our wins in marketing, they it still just gets us, like, x number of customers a month. That's it. Yeah. You know?

I I'm, like, of two minds about this because I think from the outside, there's part of me that's like, man, like, whenever you see anybody launch something and you're like, you think about the potential of that thing. And sometimes it feels like, oh, if they just doubled down on this and spent another, for example, instrumental components. You just doubled down on that, went to every Rails conference, got on every Rails mailing list, get you just immersed yourself just in that world. It feels like that's the that's the maybe the only or best way to truly find out what the appetite for something is in a market. Yep.

The

engine that drives my whole career, you know. So I think it's hard in those moments.

Yeah. And I don't wanna I really just don't wanna spend too much time talking about, like, the Clarity Flow business. But one thing that did go into that decision back in 2024 was, look, I am close with lots of other SaaS founders. Yeah. And I and I know their story and their MRR graphs from before they even started their businesses.

The the way to think about it too is I mean, you and I have used this metaphor of home runs versus line drives. And we've kind of joked that maybe we're both kind of like line drive people. But the difference between like going up to bat and like hitting something like maybe Clarity Flow is is a line drive. It's just like you launched it. It has customers.

Yeah. And that's that's what I came to. And then so with instrumental components, I the I was I was never expecting it to be, again, like an all in type of product or business. I Yeah. Really, my only intention in even starting it starting it is, like, I need to I need to just start and launch a new product again just to get back into that rhythm, get that muscle.

And that muscle is is super important. Yeah. This might be a good time to just bring in some of the chat here because I think it's I think it's interesting. We got we got Dave Junta in the comments here. One thing that Dave said was regardless of whether or not a business reaches escape velocity, you still need the drive and motivation to take it all the way.

This is alright. So I just wanna I don't wanna come off as, like, too defensive on this or anything like that. But there's a general not in not not necessarily just here, just in general. I've I've heard I hear you know, talking in public about about businesses on podcasts or on Twitter.

Mhmm.

It's it's just impossible for anyone on the outside looking in to know what goes into every individual Oh,

absolutely.

Day to day decision. Yeah. That's number one. You you will never get even whatever you hear me or anyone else talk about on a podcast, you are hearing like 5% of the story. Mhmm. You know, as as transparent as I wanna be, like, there's just so much more. The other thing is is for me and and just thinking about new businesses going forward, I really am how do I put this? Because I there's there's a there's a balance between, like, thinking big and trying to do the next big home run

Mhmm.

With taking too big of a risk. Mhmm. And Yeah. And also, like, I'm just not of the mindset anymore, especially I'm just not under the of the mindset that, like, if I'm doing one product, like let's say instrumental components, that does not bar me from doing another product. That's what is so challenging about marketing and being public about these products.

That's such a great point about, like, what people can see publicly and what, you know, what's actually going on. I mean, I'll just speak for myself. Like, people had a sense that I was stressed out while I was building transistor. They did not know how really stressed out I was. Like, I some of that came out publicly.

I think I think doubling down, like, it becomes obvious

Yeah.

When you should be doubling down. When when you are because I one one thing that I've noticed and observed with most of the successful SaaS companies, I think probably Transistor is included in this. At a certain point, it's they are successful despite their shortcomings. Mhmm. They are they are successful even even with gaps in their product.

Yeah.

I I do wanna talk about what what I'm starting up Yeah. Going forward. Just just to tie the bow on the instrumental components thing. Yeah. I I wanted to sort of report on the pricing because we talked about that a bit in in the last in the last episode.

Yeah.

You know, they paid $299 for the first year. But like the thing said like, first year, and then it'll renew at $499 And most people bought at that price. So meaning I had subscriptions in Stripe. I went in and I turned off those subscriptions. They are not gonna get renewed. And then there've been a bunch of sales since then. And so now it's promoted as like it's $2.90 it's $4.99 normal price. Right now, promoted as $2.99 Yeah. Lifetime.

And what and what was the thinking behind that? What what precipitated the change?

I think I think number one, just honestly, it just the recurring thing just did not really feel right, if I'm honest, about it. I felt a little bit like, what I subscribe? Would I actually pay $500 a year for a components library? Like, if I'm if I'm honest, like, probably not. Mhmm. You know? The the only case that I would make for paying annually would be if you're an agency and you're doing lots of apps for clients

Yeah.

On on a recurring basis, then I could see the case for a recurring thing. So there's that. And then and then the other thing is, like, the commitment factor. People are gonna obviously hear this the wrong way. But I'm I'm not ending my investment or commitment to instrumental components.

And again, like, that the launch you've done so far is nothing to sneeze at. That's that's a good that seems to me like a good line drive, especially with a waiting list of about 800 people. Yeah. That that seems about in line with what I would expect.

And also what I like about it is that, like, people find it through YouTube. Most most of the buyers for this don't know me, don't listen to this podcast. Yeah. They just found so it's so again, like, Rails is not a super popular framework as much as it used to be. Yep. But there's still a little bit of traffic that comes through YouTube, and they find my components. And and several of them have actually bought. Like

Yeah.

That's so that's a pretty cool thing that could continue to trickle in throughout the year. It's just a little thing in the portfolio.

Yeah. Yeah. Yeah. I love it. And I think for anyone listening or watching, let's if you've if there's ways we can help Brian out, like, if you have a Rails newsletter, feature it in the Rails newsletter. If you know some people in the Rails community, reach out to them. Let's get some kind of community action going here. Community action was very helpful for me at Transistor. Just people on their own promoting us and or referring us to people with newsletters and YouTube channels and podcasts.

Alright. Let me tell you about a really interesting thing that happened this week. Okay. This is this is about going forward.

This is the new thing.

I was in the shower thinking my thinking my my typical business strategy shower thoughts.

Mhmm. Mhmm.

And it's not like this just started two days ago. I've been thinking about this for for a year or more now. But like the direction that I'm taking is, like, helping builders build with AI. Yeah. And if you are an experienced full stack developer, I wanna help you transition to the new way that software is built and that is using tools like Cursor and specific new approaches to the whole software development process now collaborating with AI.

Yeah.

If you wanna deny that, you can. It's just that that's just the reality. And and so I I'm actually super optimistic about this future for how we design and build software. I think the tooling is incredible, and I and I don't think that we lose any sense of craft or taste or ambition. I think it actually maximizes that stuff.

Mhmm.

And and getting the speed benefits and getting the the product benefits. And and I was like, maybe there's a a some sort of, like maybe I can go into companies and and offer personalized workshops or or personalized training or coaching for for teams. Yeah. And and you can, like, outsource to me. Like, I'm I'm the one who's staying on top of the new approaches to building at a professional level, building with AI, and and everything that you can do with the cursors and and whatnot.

I love this for you, by the way. Just to interject, I think this is you've been talking about this forever. Of all the people in my mind in the community that have been a long time proponent of building with AI, it's you. So I think this matches up with your profile, your passion, your energy.

Thank you. I appreciate that. So I'm thinking about that on Tuesday in the shower.

Yeah.

One hour later, get an email from a friend. I I I won't share his name. I don't know if he wants to be public about this. But he runs a a successful SaaS company. He's got a small team of, like, I don't know, eight to 10 people on the team.

Mhmm.

There have been a few step changes in the tooling around AI. Yep. And I think there there have been like three big ones so far, and we're we're entering the third one right now. The first one was like the just tab completion. Right? Like you're you're writing a line of code, and then it it's auto completed. You press tab. Cool. Like, that was kind of cool when it started two, three years ago.

Yep.

Then then we have, like, agent mode, which is like cursor for the past year, And that's kind of where we've been. That's where I've been now. But now there's this third mode, which sort of builds on agent mode, but it's it's a more task oriented. Like, you build out this PRD, you know, project requirements document to sort of map out, like, here's the whole product or here's the whole big feature. Let's break it out into tasks.

Yeah.

Literally a check a checklist. And then you have your AI build each feature and check off the thing in the in the checklist in an orderly professional like, this is how a professional project gets built in collaboration with the AI. Like, rather than, like, running in circles with these old random prompts all day and and, like Yeah. This is a more, you know like, so I I think and, like, so Ryan Carson just published a really great thing on this. There's a tool called Taskmaster, which which is built around the same concept.

I think and the headlines kind of write themselves for this stuff. The like, you know, one thing a lot of devs say is I feel like I just have constantly babysit the AI. And so I would, you know, if you were offering some sort of service that's stop babysitting your copilot or whatever, I could see that being attractive. Do you have a specific offer in mind? Or is it kind of coalescing around coaching or some sort of productized service? What what do you think?

I I have no idea. It's so up in the air. I that's been my problem is, like, overthinking this for the past six months.

Yeah.

Like like, this week is the first week that I'm really starting to, like now I'm working on this direction. Yep. But I've been thinking about it for six, eight months and thinking, like, alright. Well, what does this turn into? Is it a is it a course? Is it like one of those, like, cohort based courses? Is it coaching? Is it a book? Is it here's what I know for sure. The top of the funnel is is going to probably be mostly YouTube.

Mhmm.

Around today's version of how you build because today's version is gonna be different three, six months from now. So I think it's a more fluid, like, month to month. Like, here's here's where we are in the here's what's here's what's happening on on the cutting edge. But maybe one step back from the cutting edge, like, here's here's how to make this more mainstream.

Mhmm.

Like, there's always gonna be the cutting edge stuff and, like, there's no way anyone can keep up with that.

Yeah.

But but then there's the layer of, like, okay. AI building for the rest of us. Like, I don't wanna I wanna I don't wanna completely fall behind. But I wanna make sure that whatever we are investing in is tried and tested and and sort of, like, trusted at this point. Yeah.

You know? This would be a great place for us to bring in the panel of people that have been watching live, I think. We got Adam Wavin in the chat here. We got all sorts of great people in the chat. But I'm wondering for people out there right now who are on Teams or who are building things themselves, even folks like Adam Wavin, AI is affecting everything.

I I had a great, like, hour long conversation with with this guy today with with a friend of mine today about working with him and his team on on an engagement. And I was it was just a discovery call for me. Like, just to learn about, like, what what is the actual need? Because he approached me with this. What is the benefit or the pain point?

Mhmm.

And frankly, like, if you're a paid developer at at any company, like, you're making the correct choice to just work on the stuff that you're being paid to do. Like, build build stuff the way that you know how.

Yep.

So the value of bringing someone like me in, or eventually this turns into a bigger company, the value is like, we're doing that that, like, workflow improvement and staying up on the latest approaches and tooling. And and so, like, I can come in and work with your team and, like, sort of observe how things are going and and maybe find some low hanging fruit, make some recommendations, offer some ideas. But just know that, like, you know, it's it's like a like a gym trainer. Right? Like, someone somebody who's just coming in, like, there's no specific, like, deliverable.

Okay. So the the chat kind of erupted while you were talking. I I don't know how delayed they are, but we've got Andy Miller's, like, % this. I it sounds like he's struggling with which tools to use, what AI to use. Zach Gilbert's saying as a solo dev, he feels like it's hard to keep up with everything.

That's the thing.

Yeah. Yeah. See, maybe I can just throw out one vision for what this could look like. I'm sure you've thought about this. But to me, there does need to be something like an AI for the rest of us newsletter that Yep.

Yeah. I I really like that. So first thing was the newsletter and just like yeah. Like condensing what actually matters. What what should I spend time thinking and focusing on versus there's just so many new tools, new startups coming up every single day. I personally, like, try to block out and ignore most of like the newest cutting edge. Like this is a startup that just came up this week. Probably flame out by next week. Yeah. Right?

I invited Adam to I invited Adam to talk about what he's he's doing on the tail end side. But finish your thought, Brian, and then we'll get to Adam.

Yeah. So like like, cursor has has become sort of, like, pretty solid in the in the workflow. But, like, one thing I'm playing around with today is this, like, v zero from from Vercel, which is like I think they're probably trying to build their own version of cursor. But but what it's really good at is building like UI front end stuff. So so the workflow of like building the the front end design there first and then taking it into cursor to build out like the rest of the app is is sort of a that again, like, just looking for these, like, directional patterns that that people are doing that maybe teams can start to adopt.

Yeah. Yeah. And you could see how you could coach somebody on all that stuff, especially like, again, transistor, like, we're not doing very much in that right now. What Adam, what are what are you doing? You I I I couldn't quite get from chat.

Yeah. I think we're just noticing a lot of places where there's, like, things, projects where we've historically said, oh, yeah, that would be good to do. But like Jesus Christ, that would be like so much work. Someone would have to spend three months doing the same thing over and over and over and over and over again. And people are gonna quit if I ask them to do that.

Oh, wow.

Right out of there. You know what I mean? And that's a fun project, you know?

Another example is, like, this week on on instrumental components, one of those, like, post launch things that I that it was on my to do list was, like, I gotta make these components inter internationalized

Mhmm.

For for, you know, multi language translation. Right? I had a couple European customers buy, and they emailed in. They're like, I can't translate these. You you don't have the localization tag.

Mhmm.

Right? Which, you know, in Rails, if if if you don't if you don't build that from the get go, it's like it's sort of a pain in the ass to to go through and the tedious work of doing it. So I wrote a prompt, and and I and I built a sort of a flow that is reusable. So now it literally took me one day to internationalize 30 plus components. And and it's reusable so that every time I I create a new component, I can like, one of the last steps in the in the launch process for a new component is, like, internationalize it.

Yeah.

So, like, that's that's the type of thing. Like, these tedious projects.

Yeah. And, Adam, is there other ways you're using it? This is actually the first time I've talked I've heard you talk about using AI at all. I I couldn't I didn't know if you were, like, for it or against it.

Or I used to be more skeptical than I am now. And I think the thing for me is I was seeing everyone talk about, like, one shotting things and, like, vibe coding things. And to me, that feels like bullshit and is just not not interesting. To me, it felt like, okay. I like programming.

Like, just the fact that it can it can build multiple files.

Yeah. And you can. What it feels like is it feels like having your feet up on the desk pair programming with someone next to you that you can just, like, talk to like a human being that understands what you want them to do, and they do. You know?

And also, like, there's a lot of the skeptics will say, like, oh, but it's just gonna spit back, like, like code slop that I can't use.

Yeah. Then you're not being specific enough with what you're asking it to do.

You're you're still approving every page and every change. So you're so experienced developers can read the code that it that it spits back and you can improve it. But the other thing is that, like, this stuff is improving so fast. I mean, I'm using today, I'm using mostly Claude Sonnet four. Even 3.7 was so much better than the models from like six months ago.

Yeah. Yeah. I I started doing that on the very first day. Basically, anytime I found myself correcting something that it did, I added it to a rules file to make sure I never had to ask it to do that again. So just like little opinions in, like, our tailwind projects, if it used, like, a bottom margin, I always use top margins instead.

of stuff that I was

doing before. It's just like so, like, to me, the the real unlock has been realizing that this actually lets me, like, program more, not less.

We should be spending our mental cycles on design. Designing the designing the structure of the architecture, the the user experience. And our skills as experienced full stack designers, developers is now to direct AI to carry out our design. You know? Mhmm.

I've tried to use it with like things I don't know how to do and it's a fucking disaster. Like I was I was trying to like prototype like a swift UI app the other day, which I've never done in my life. And you instantly fall into that trap of, like, AI being, like, confidently wrong, but you can't tell because you don't know how to do it yourself. But when you're so I I've until, like, I really this really clicked for me, I always kinda thought the idea with AI was like, okay. This is gonna, like, let me do stuff I didn't know how to do before.

For for people building new stuff, though, I I I do think that there's an element of, like, now that I build with AI, I can be more ambitious about the types of things that I build. I actually think that instrumental components in the way that I built it and and and the extensive functionality that's built into it would not definitely would not have happened without this style of building. Or I or I would have built a much simpler version of it.

Yeah. I found it useful to separate those two things, though. Like, the using AI to learn and using AI to code as, like, discrete tasks. Like, I've never had success trying to learn get Cursor to do something don't know how to do. Like, I just I get bad results with bugs that I don't understand how to fix.

I've been doing that pattern where I I I go to ChatGPT, which the nice thing about ChatGPT is that it has the memory built in. Like, has all my past chats, so it knows a lot about me and my preferences. So I use that as, like, my strategist that where where I bounce ideas and, like, how would I go how would I architect this? Like, that like, so I'll, like, get the approach there. And then I might I might also double check-in Claude.

Is it similar for you, Adam? You've got Cursor for coding stuff and then ChatGPT for the learning stuff?

Yeah. Yeah. Yeah. Right now, although I do wanna try some other stuff, a lot of people on our team are using Claude code and liking that a lot. And I've heard a lot of people have been having better results with that in Cursor and candidly kind of like appeals to my sensibilities as, you know, it's like a terminal based thing and I kind of like that it's separate from the editor.

Don't use the, like, the inline, like, pop up things that you see, like, oh, that overlay the the current file. I'm I'm all about the chat sidebar, you know. Interesting. Also think that that is a pattern that is going to become super common across all apps. Right? So, like

Every every app needs to have a chat now.

And they are. Right? Like, so Descript Descript now has a cursor like I think they call it underlord.

Wait. Are they good?

It's like it's Are these

are these sidebar chat AI chat things good?

Because I've I played with the Descript one yesterday, and it it's not great yet. But it but, like, that's and and Figma is now launching their chat thing.

I haven't tried the one in Figma, but me me and Steve always talk about this with Figma. Do you remember when Figma announced their first, like, AI stuff? It was, like, at their Figma config event last year. It was all, like, press a button and get, like, a mock up of, like, your UI, and then all the designers, like, revolted against them, which to me totally makes sense. Because when I'm working with Steve in Figma, the opportunities to use AI, which I think maybe it does a better job at now.

Figma is the one where, like, I've been wait I haven't I haven't played with their new AI stuff yet, but I've been waiting for it for so long because, like, physically, like, my body and my back hurt after I do a ton of design work in Figma. Because it like, unlike code where even, like, the tedious grunt work in code, at least I can use keyboard shortcuts to do a lot of it. In Figma, it's all mouse. So it's like, that's where I lit I'm physically in pain from, like, doing all these, like, oh, I wanna round the corners just a little bit more and all these shapes. Like, just do it for me.

So here's an unlock for me is there's part of me that has been fighting this whole AI transition because naturally, I learn and I develop and I experience and explore through relationship. So having guys like you that I could talk to and chat and bug and get screen share with has always been the way I learn. It feels like the way to learn AI is similar to that. Like if the unlock for me was if I had an Adam Wathen or Brian Cassel on a screen share with me, and I'm like, okay, going in there. And Adam goes, no, no, no.

Well, that's that's why I'm I'm really into this new shift that's I think kind of happening right now, which is like we're going like, there's there's been this agent mode for the past year.

Yeah.

But even that is like like, my fingers are tired from typing prompts. Like, you know, like like and and especially because, like, I do wanna give it super detailed instructions because I, like, I have a very specific way I want it to to carry out my instructions. So sometimes my prompts are like three paragraphs long.

Yeah. Yeah.

So there's like, I I'll even like draft prompts and other files before I bring them into Cursor. Right? And that's a pain. And I'm doing it repeatedly throughout the day, you know. And so that's where this more cohesive, strict strategic look, develop like, put your energy and your creativity into this PRD process upfront.

Yeah.

We're we're front loading the strategic plan. Like, either it's a product or like a big feature. Yep. Let's I I could type in a three paragraph prompt that says like, here's what I wanna build today. Now now turn that into a long detailed PRD, project requirement document.

through it.

It does it. I review it. I I approve it. It and like and also, like, taking a test driven development process with this really helps. Like, I I haven't been a super TDD person in the past.

Well, you can imagine to get started, like, if I had someone to coach like, here's the best PRD for your guys' Rails project. Like, let me help you write that. And because someone who's done it and done a bunch of work in these like like Adam was saying, he he every time he ran into something were you adding a rule in the PRD, Adam? Was that what you were is that what you're doing?

With the cursor rule stuff, cursor has this concept of, like, a rules file that lives in, like, a dot cursor for dot cursor rules folder or something in your project. And you can just drop, like, markdown files in there to kind of give it to kinda tune it and point it in, like, the right direction for, like, every instruction.

So there's, like, overall rules for the the platform, and then there's PRD is, like, project specific.

Yes.

Got

it. Yeah. Yeah. But then you can even have like a cursor rule for like, every time I want to build a PRD, follow this process.

Mhmm.

You know? And then and then there's a cursor rule for how to turn a PRD into a task list. And then how to turn your and and then how to how to create a checklist for that. And, you know

May so, Adam, for you, I'm just trying to think, how do you think this is going to affect your business?

Like Yeah. I I think it already is. You know, there's a lot of, like, vZero, you know, as a competitor with our business at the end of the day. Yeah. I'm I'm not sure.

I do I on on that point about about traffic to your website, I I just I heard this interview with Rand Fishkin on on on like the Wistia podcast like two weeks ago with Chris Savage. Really, really great interview. Just talking all about and and this is a thing that he's been saying for years. This, like, zero traffic or zero click Internet is happening. Mhmm.

I'm not sure what my takeaways are from that really or my conclusion from it. I think one thing is that I think it's gonna be more important than ever to have like an audience.

I yeah. That that is like the conclusion that I keep coming back to. It's like if if it's zero, if if traffic if organic traffic from search is on the decline, the answer to that is personal influence.

Yeah. Yeah. And then yeah. And then the other thing for for us is just to I don't know. Like, I think if you're building a tool that costs money and it solves a problem, then or you still can find customers.

But directed by you and Steve.

You know what I mean? We we do need to

have another Adam Wathen episode to, like, to unpack your because, like, the the thing that comes to mind now is, like, yeah, you're you're saying things that, like, like, v v zero or our competitors, but, like, it's also, like, everything is using Tailwind now.

Yeah. And it's frustrating because everything on Earth is using Tailwind, and there's, like, nothing we can do to benefit from it. It's purely a burden in a lot of ways. You know? It's just, like, I have like over a million dollars in salary obligations that pay people just to fix bugs in a free tool that people don't pay me back

for.

You know what I mean? Yeah. Yeah. So it's kind of brutal. Like I'm I'm very ready for, like, my next act in a lot of ways, honestly.

I mean, I I think that sounds direct directionally correct. But, you know, like, the thing that that I I think I'm thinking about, and maybe if if you're thinking about, like, what what is the next act or the next phase of of Tailwind Labs, what you were talking about last time with, like, building up the the YouTube studio.

Mhmm. Yeah.

Like, I'm I'm sort of I'm trying to restrain myself from jumping too far ahead to, like, what is the next product? Like, is it a course?

Is it

a is it a this? Or is it a that? Like, I'm just, like, let me just get into motion. And if all that means is I'm releasing new YouTube content every week and listening to the feedback and trying to spot the trends of of where, you know, what what are the solid patterns that I can build off of, then then whatever products or business direction, like, that's going to make itself obvious at some point. But but for now, if all I really should be concerned with, maybe maybe you, Adam, too, is like, just be in motion on the on on what would need to be true for the next phase to happen, and that probably means, like, weekly YouTube videos.

Yeah. For sure. Yeah. We're still bullish on doing that stuff. Like, neck I'm right now, I'm working on, like, a Tailwind course, like, official paid Tailwind course that, you know, where the the, you know, unique value proposition is, like, it's taught by me.

I love it. I think that makes total sense for for your business.

And then, yeah. And then on the side, me and Steve are designing video editing software right now because that's like a pain point that we have, like, when we make this content, you know? And I'm and that I'm getting pretty excited about that.

So Talk about something that's outside of your normal like, what you know. You must be having to learn a lot there.

Technology wise, yes. But it's kinda funny because, like, it's a very full circle moment for me because when I was in, like, college and, you know, committing to becoming, like, a professional programmer, I went in with like, the thing I wanna be able to do is build my own like audio editing tool to use instead of Pro Tools or Logic or whatever because I just have so many workflow opinions about all this stuff. Because that's what I was doing before is recording bands for a living and I there's just so many things that could have been like faster, you know? And Steve has the same background, like he went to school for audio engineering.

I I I got I got a degree in that too and and and we all ended up being like web developers. So there's that.

Yeah. So to get into like, yeah, designing, like, software for compositing video tracks and doing edits and

I do wanna hear about that. It sounds a little bit out of left field for you, but also, like, I I don't know what your road map is or plan for that is, but I feel like it has to be dude, like, it has to be AI, dude. Like, the the idea of, like, magical

I think there has to be some AI stuff in there. There's so many things that AI can do there now.

Yeah. So, like, I'm using Descript now. I I go back and forth between ScreenFlow, Descript, and a little bit of Screen Studio whenever I'm doing videos. And I'm just always on the lookout for, like, how can I remove the grunt work? And

Yeah. Like, how can I just get it to delete everything except the good takes, which is always the last take? And you can tell because it transcribed everything. Yeah. I've seen

Descript has that feature, but it, like, only works, like, 80% of of the time.

And Yeah. And even, like, captions. Like, do you remember, like, Justin? I mean, both of you guys will remember this, but, like, was a time back in, like, 2015 where, like, we all felt guilty about our podcasts not having transcriptions. And we were all debating what service We're asking people at MicroConf, like, do you pay to transcribe your podcast and what's their turnaround time and whatever. That is such ancient history. That whole idea, it's just an instant free thing that's perfect.

I just

did a

voice note an hour ago and my iPhone transcribed it. Done. Like Uh-huh.

Yeah. It's insane.

Okay. So here's to wrap to wrap this up, I I'm hearing some things. This is

AI is cool. Like AI. Threat threatened by AI, empowered by AI. We're really just, like, trying

to lay lay, like, the groundwork so that, like, when the LLMs read the transcript of this episode, like, we're we're on their

good side. We're spared. We know.

To me, it sounds like the first theme I'm taking away here. And again, the chat's blown up here. I think a lot of people participated.

So much for trying to keep up with Really appreciate it.

I think it's I think it's awesome. But I I think expertise is one thread we have here. Like, there's still nobody that can build UI components like Adam and Steve. Like, there's something about that that is magical. But also expertise in terms of using AI, knowing when it's wrong, knowing how to direct it.

That that was that's the same exact sentiment that I heard from my friend today, the guy who who reached out asking for that. And that's what I'm learning. And I I I think that there's just aside from, like, a direct, like, work with companies on that problem, I think that there's a larger opportunity that I'm kind of focused on.

I think there's a massive appetite here. Even look at the chat. Listen to Adam. Listen to every everybody's just saying like, everybody's grappling this with this in real time. It's clearly a paradigm shifting. It's clearly like the tectonic plates of the industry are moving. Everyone's trying to figure it out.

And it's like one of those things that like with anything with web development, software development, I've always had this massive level of impostor syndrome of, like, I who am I to to to teach anything related to programming or coding? Because there are so many people who have who are so so much more experienced than me and and are smarter than me when it comes to how to write programming logic. Right? But I think that when it comes to development workflow and building with AI and using tools like Cursor, I think right now we're still in this, like, phase where it's like, we're all just figuring this out. Yeah.

Right? Yeah.

You know?

Yeah. And the the other thread I'm getting and this has come up before, but, you know, Adam's working on this new video editor. It feels like looking at what AI could enable for your team for, you know, as a as a another bet during this time is another thing that's happening. Like people are saying, okay, well, what can we chew? What can we take our unique expertise in into that we wouldn't normally be able to do and start building some shit that might be the next act of our of our stuff.

Yeah. I think the important thing is just do stuff and make sure you're not being

a A Luddite?

Not swimming, like, yeah, up against the current. You know what I mean? Like, make sure you're informed before you take that stance or you're gonna find yourself a little behind the curve, I think. And I think the best way to learn the AI stuff, like, I I really learned a lot very quickly by just, like, doing that one project where I just didn't let myself write any code. And I don't think that should be, like, the goal.

The creativity and the craft, that happens upfront in at the prompting stage. But even before that, it's like the PRD, like the project plan stage. Like, that's where we as the experience, like, look, this is how this is what we need to build because this is what we wanna deliver for customers. Like, that's where all the creativity is happening.

Yeah.

You know? Because, like, even even, like, these, like, non technical, you know, idea guys and gals who, like, they might have, like, yeah, but like like, I mean, people can have great ideas about what a customer wants and that's extremely valuable. But then there's also incredible value in like, okay. Like, actually delivering a a solid product Yeah. You know, from from front to back.

Did you have one more thing you were gonna say, Adam, before we close here?

Yeah. There's two things I'll say quickly. On the vibe coding thing, I'm not interested in that for, like, professional work, but it is extremely cool that you can build custom tools for yourself to solve little problems extremely quickly. So, like, the I built a workout tracking thing for myself just so that it would tell me what plates to put on the bar and what my next set would be and calculate a couple things and stuff like that. And that, I entirely I've never read the code.

I'd love an app that would keep my camera.

That'll be fun. And then the other thing I'm excited about that I'm just, like, starting to get into now is just building custom AI, like automation flows. There's this really great article by Torsten Ball who works at Sourcegraph. It's called like how to build an agent or something. And as you're reading it, whole time you're just expecting to get to the hard part where it's like, oh, yeah.

work. Interesting.

And Yeah. Then I didn't have to

do But that's that is where we're going, and it's not even it it like, that is possible today.

Right? That's a % possible today. And the best part about it is if it sucks, I can just, like, say no and close it. Whereas right now, if a human being does that, I should be like, dude, I'm so sorry. I know you spent, like, like, fifteen hours on this. I don't have time to really review this right now. You know? It's just like it's a cold machine.

More more likely, it's gonna be like 80 or 90% there, and you're gonna go in and tweak a few things and then ship it. You know?

And and if I let it sit there for nine months, the AI isn't gonna get mad at me

or think I'm

ignoring it or, you know, it's there's so many elements to having, like, a soulless machine do work for you instead of human beings.

It is It's it's also this is such a, like, a bigger conversation, but, like, what what does it mean for our psyche to to have some like, an AI that we can just talk to or strategize with? Like, I like, I've I've been doing, like, like, business strategy and planning for this direction. Right? And how ChatGPT has my whole history and everything, and it's like, damn it. Like, knows what Mhmm.

You know? Like, it's insane, dude. Yeah. It's true. Alright. I gotta run. I'm gonna be late, drop my kids off at dance, but, thanks for having me on. It was fun.

Yeah. Thanks, man.

Alright. See you guys.

I mean, I think we should end it. But this to me, I think this worked out really well. We got we had this active chat. We had some old we got Junta in the chat. We had all sorts of people interacting.

The is a lot more active than I expected it would be.

We had Adam join us at the end. It was kind of like the perfect like, I could send a link to Adam because he was interacting in the chat, and I think we could do that more often. And it just this whole thing this is like one of those episodes where I am just chewing on all this AI stuff. Now I'm thinking in my head, like, what are the next things I need to go do? I think for sure, Brian, it feels like there's an opportunity for you here and some sort of like AI for the rest of us newsletter needs to happen.

I I like that. The the thing yeah. The the thought that I had there was like a monthly because it's also like, how do I I mean, maybe some of it I could use AI to help. But like like a monthly builder letters. Mhmm.

right now. I think it should be weekly, personally, because I feel like people need that drip. They might only catch, you know, one out of three emails. But

But I almost think I also think that, like, it's that's what makes it different. Because the the point is, like, you don't need to stay on top of every single latest greatest thing because there are 50 new things that drop every single week.

Yeah. Yeah.

Yeah. The the point is, like, you should ignore most of that.

Okay.

I know. That's the thought. And also, I just don't want to have the work of having to deliberate.

Alright. Well, so next Thursday, we'll go live again. I think you should have a this will be easy for you. A sign up page for that. Just even if it's just a waiting list, like I'm planning on writing something in this zone, you know.

I have a potential first client for this. I don't want this to turn into just a client service. I I do wanna start to think about how is this a product and an actual business that grows. But for right now, I'm I'm very much in the learning phase and the audience, like, launching phase.

Yeah. Definitely, folks. Reach out to Brian if you've got thoughts as you've been listening or watching. Again to everyone who showed up live. We'll see you next Thursday, 12:15PM Pacific on June 12. Yes, sir. See you.

Later, folks.