The Best AI Coding Tools to Use in 2025 | Colin Matthews
Summary
The podcast episode, featuring Colin Matthews, delves into the landscape of AI coding tools anticipated for 2025, focusing on their capabilities across prototyping and full-stack app development. It highlights a significant convergence among AI tools like Bolt, V0, Lovable, and Replicit, where previously unique features such as Figma imports and backend integration are now broadly supported, marking a matured market. However, despite Google's leading AI coding model, tool quality and user experience vary widely, underscoring that effective implementation matters as much as underlying AI power. The discussion features a detailed comparison of text-to-prototype tools using a unified 'LinkedIn clone' prompt, revealing that Replicit stands out with its support for both client and server-side code including databases, whereas others mainly focus on client UI with differing fidelity and responsiveness. V0's prototyping approach, while consistent and fast, results in visually uniform applications which is a double-edged sword for rapid prototyping versus unique branding. Lovable's distinctive reliance on GitHub synchronization for code editing highlights trade-offs between collaboration workflows and direct prototyping agility. Cursor is spotlighted for pioneering a chat-based IDE interface that previews real-time diff changes, improving transparency and user control compared to earlier tab-completion tools like GitHub Copilot. The episode further addresses the divided user base: engineers demand granular control and safety in changes, often viewing large AI-generated diffs as risky, while non-engineers prefer conversational chat interfaces despite potential inefficiencies. Skepticism remains around AI’s current ability to enable non-technical users to build production-grade applications independently, emphasizing that human engineering expertise remains vital for robustness and maintainability. Prompt engineering emerges as a critical skill to maximize AI tools’ utility, demonstrating the ongoing need for clarity and precision in user inputs. The episode concludes with reflections on market dynamics, including enterprise preferences for established vendors, the challenges AI startups face in differentiation amidst feature parity, and the outlook that AI will augment but not replace professional software engineers in the near term.
Key Takeaways
- 1AI coding tools have matured into a convergent market where feature sets such as Figma imports and backend integrations like Supabase are now standard across platforms, shifting differentiation away from isolated capabilities toward overall user experience and customizability.
- 2Cursor's innovation of integrating a chat interface within IDEs that shows real-time code diffs is a game-changer for AI-assisted coding, offering developers transparent insight into live code modifications and replacing less informative tab-completion models like GitHub Copilot.
- 3While AI coding tools facilitate rapid prototyping for non-technical users, these users currently cannot reliably build fully production-ready and maintainable applications without substantial engineering expertise.
- 4There is a distinct divide in user preferences between chat-based AI coding tools favored by non-engineers for ease of use and traditional tab-based code completions preferred by engineers prioritizing small, verifiable changes and workflow safety.
- 5Text-to-prototype AI tools exhibit divergent strengths and trade-offs in first-shot outputs, with Replicit leading in full-stack code generation including backend and database support, while others like Bolt and V0 focus on rapid, visually consistent client-side UI prototyping.
- 6V0’s reliance on default UI styling and frameworks like ShadStandUI and Next.js results in visually uniform applications, providing rapid, consistent prototyping at the cost of creative uniqueness and brand differentiation.
- 7Effective prompt engineering remains a decisive factor in the success of AI-assisted coding, requiring users to convey technical requirements precisely to generate desired behaviors such as routing or data management.
- 8Lovable’s approach of providing ‘read only’ code previews and requiring GitHub synchronization for editing reflects a trade-off between maintaining structured version control workflows and limiting fluid direct prototyping edits.
- 9Enterprise adoption of AI prototype tools is influenced heavily by vendor maturity, integration depth, and security guarantees, providing established players an advantage over newer entrants despite comparable feature sets.
Notable Quotes
"It's kind of funny, like as time has gone on, the tools have seemingly converged in their functionality. So, you know, at one point in time, like, you know, only one of them had a Figma import or only one of them had a Superbase integration. But now they all can really support all the same features. And so this is funny because Google has the best coding model right now, but I guess their tool is underwhelming."
"So like the view in, in cursor where you can actually see the, the files that's changing. Yeah. In real time, like the minus and plus, because at the time GitHub copilot was just doing like more tab based changes. So like you'd be writing and it would recommend something. Oh, okay. Like in the, in the code itself. Yeah. In the code itself. But in cursor kind of, I guess, popularized the idea of having a chat in an IDE and then you type in something and it shows you what's going to change."
"And like, it's kind of like, uh, it's kind of crazy how, how baddie, uh, GitHub copilot dropped the ball here. Cause they used to be the most popular. Right. Yeah. Yeah. It honestly, it all started with the diffs."
"Like it's like, here's 2,500 lines of new code, figure out what happened and do you want it or not? And it can take you just as long to go through that as it is to just do it yourself. So, you know, yeah."
"Like can regular people build production apps with these tools? And like, if you go on, on X or Twitter, like the answer is like, well, obviously yes. And you'd be dumb to not think so. Um, but in my experience, like I, you know, I've literally had hundreds of people go through my course. So like start from zero and try to learn these tools as well as like myself as someone who's more technical. And I think the answer is like pretty definitively no right now."
"So TRDR, I don't think a proper engineer is going to get this disrupted anytime soon. Yeah. Agreed. Yeah. Cool."
"So for the text to prototype one, you can see that we have a couple different examples. It's all going to be the same prompt. So you can see at the top here, I have clone LinkedIn. We're going to use the same prompt for all of them. And this is all like the first shot, meaning I didn't change anything. I didn't iterate over time. I just kind of let the tool do its thing."
"Replicit actually has a database. Yeah, yeah. If we take a look at the code here, you can see that we have the client side codes. That's what we're seeing in the UI. And then we also have some server side codes. We have API routes. We could theoretically make updates or make changes."
"One thing you'll probably notice is that all V0 applications kind of look similar or the same. And it's kind of not that hard to tell when an application was built with V0. This is Firebase Studio. So there's nothing. I can't move. There's a red box over here. This is it. This is the whole thing."
"Lovable has this kind of code preview here. And you can see at the top, it says, read only. If you want to update the code, set up sync with GitHub. I believe they recently shipped a feature for, maybe it's in beta or something like that, where they are trying to make it so you can edit it. Yeah."
""Personally, I still think it's Bolt. I think Bolt gives you the most flexibility. I think Bolt kind of has some good defaults. And then when you kind of want to go off in your own direction, Bolt doesn't really stop you from doing that. Yeah, I think so.""
""Most tools at this point support Figma imports. At a certain point in time, it was only some of them, but now they're pretty consistent that they all support that. Yeah, cool.""
""It's obviously not responsive, right? Because you can see it going off to the far right-hand side. So it's not responsive.""
""So v0 actually works a little bit differently. V0 does it themselves. But as far as I can tell, they do it based on a screenshot. And so they'll work differently.""
""Magic Patterns has some unique features, right? Like probably one of the more interesting ones is the ability to put this on a canvas and then put different chats side by side. So this from like an organizational perspective is actually pretty useful because you can see like I've imported this design here. Then if I want to test a different version of it, I can just fork it and get the copy and then hop into this one and have like, you know, the kind of the side by side view or like, this is my new one here.""