Learning Library

← Back to Library

AI Low-Code Market Evolution

Key Points

  • The low‑code “vibe coding” market is becoming crowded, with mainstream design tools like Canva entering and positioning themselves narrowly (e.g., as prototyping‑only) to differentiate from early innovators.
  • Early entrants such as Lovable and Replit are broadening their value propositions beyond simple prototypes by adding features like database integration, team collaboration, security scanning, and full‑stack web‑app capabilities.
  • Replit’s recent upgrade includes a planning‑aware AI builder (based on LangChain/LangGraph) that informs users of expected build times, improving user expectations and reducing frustration.
  • Despite progress, many users still encounter endless error‑loop problems when using these tools, highlighting a competency curve where the leading platforms are steadily improving but still require user guidance to resolve issues.

Full Transcript

# AI Low-Code Market Evolution **Source:** [https://www.youtube.com/watch?v=NGicTmhtygk](https://www.youtube.com/watch?v=NGicTmhtygk) **Duration:** 00:05:56 ## Summary - The low‑code “vibe coding” market is becoming crowded, with mainstream design tools like Canva entering and positioning themselves narrowly (e.g., as prototyping‑only) to differentiate from early innovators. - Early entrants such as Lovable and Replit are broadening their value propositions beyond simple prototypes by adding features like database integration, team collaboration, security scanning, and full‑stack web‑app capabilities. - Replit’s recent upgrade includes a planning‑aware AI builder (based on LangChain/LangGraph) that informs users of expected build times, improving user expectations and reducing frustration. - Despite progress, many users still encounter endless error‑loop problems when using these tools, highlighting a competency curve where the leading platforms are steadily improving but still require user guidance to resolve issues. ## Sections - [00:00:00](https://www.youtube.com/watch?v=NGicTmhtygk&t=0s) **Low‑Code AI Tool Market Trends** - The speaker outlines how newcomers like Canva are entering the AI‑driven low‑code space by positioning themselves narrowly (e.g., prototyping), while early players such as Replit and Lovable pursue broader functionality and enterprise features to maintain a distinct value wedge. - [00:03:45](https://www.youtube.com/watch?v=NGicTmhtygk&t=225s) **Affordable Kitchens Spark New Market** - The speaker likens the surge of developers embracing fully featured low‑code tools to suddenly giving everyone an affordable kitchen, creating a burgeoning market for complete apps while still coexisting with traditional professional development. ## Full Transcript
0:00I want to talk about where the vibe 0:01coding market is going. And yes, I think 0:03you can say vibe coding market. It's a 0:07weird world. So, what I mean by that is 0:10that there are a ton of tools that are 0:11jumping into the lovable bolt world. Uh 0:14the latest is Canva and they just jumped 0:17in with an also like, hey, you can 0:20actually type in and we'll create 0:21something for you. And then it was 0:23really funny because one of the leaders 0:25at Canva actually publicly said they 0:28think that there's only one good use 0:30case for it and that is prototyping and 0:32they think they're the best because they 0:33think they're a design tool and it's 0:36rumored that Figma is doing the same 0:38thing. I think the metos are going to be 0:41also rans at this point. You need to 0:45have a distinct value wedge to get 0:47people to use this stuff and people need 0:49to see the opportunity to deliver value. 0:52And what's interesting is the people who 0:54are late to the party are emphasizing 0:56less functionality as a value. So Canva 0:59emphasizing it's only prototyping. The 1:02people who are early to the party like 1:04Replet and like lovable are emphasizing 1:07more functionality. Lovable pulling in 1:10databases a while ago. today announcing 1:12lovable 2 where they're going to go and 1:15bring uh teams in. They're going to go 1:18and bring in uh security scanning. 1:20There's a couple of other features. If 1:23you look at the direction of the build 1:24of that product, it is not just for 1:27prototyping. They are envisioning web 1:30apps being buildable and lovable. If you 1:33look at Replet, I wanted to love them so 1:35much six months ago. I am actually 1:38loving them now. And the reason why is 1:41they have finally figured out the 1:43planning function with their building 1:44agent. They which I think is built on 1:47Langchain and Langraph at least it was a 1:49couple of months ago. Uh but they've 1:51figured out the planning function. This 1:52agent now takes time to sit down, think, 1:55plan the whole workflow. And then they 1:57give you a notification that says, "Hey, 1:59this is going to take 10 minutes. Go 2:00make some coffee. We're not in a rush 2:02here." And you set the expectation with 2:04the user so that the user knows that 2:07they need to be prepared to let the 2:11builder build. I saw this firsthand 2:14because my kid was coding an app that 2:16she wanted to build in Lovable today and 2:18she saw, hey, it's going to take 10 2:21minutes and she could figure it out as a 2:22nine-year-old. Oh, all right. I got to I 2:25got to wait. And what was interesting is 2:28six months ago, Lovable and Bolt both 2:31and and Replet would have sort of died 2:36at that point where they like build 2:38something and it's functionally not 2:40workable and the error fixing goes into 2:42a loop. And I'm not here to say that 2:44doesn't still happen. Every single day 2:46people slide into my DMs and they're 2:47like, I tried lovable and I built this 2:50thing and I got into an endless error 2:52loop. I believe it. I've had it happen 2:54to me, too. There are ways to fix it. 2:55That's not what this video is for. But 2:58the point is from a competency curve 3:01point of view, the products in the head 3:04of the pack like Lovable and Replet are 3:08pushing up the sigmoid curve really, 3:10really fast on capability because they 3:13believe that AI intelligence is going to 3:15be good enough to allow you to build at 3:16least some light web 3:18apps for like almost no money completely 3:23inside the product. And so fundamentally 3:26like the value proposition of Lovable is 3:28it's the app to end apps, right? Like 3:30it's the app where you build all the 3:31other apps you need. That is a big 3:33vision. I'm not saying they're there 3:34yet. I'm not saying they're going to get 3:36there. I'm not sure I buy that vision, 3:38but that's the 3:39idea. 3:41And what I am impressed by and noticing 3:45is that vibe coders are flocking to this 3:48and they are leaning into fully featured 3:51products and they are leaning into the 3:52promise of finishing their products and 3:54they are demanding that and expressing 3:56frustration when it doesn't work. I 3:58would not be getting this many DMs about 4:01using a product and it not working and 4:04people being frustrated if people didn't 4:05have a real desire to build a real app. 4:08The analogy that I gave to a friend 4:10recently is that what we are seeing in 4:12vibe coding right now is like what would 4:15happen if you had a a city and all that 4:19they had up until last year was 4:23restaurants. Restaurant restaurant. 4:25Nobody had a 4:26kitchen. Imagine in that world suddenly 4:30you figured out a way for everybody to 4:32have kitchens affordably. Boom. Suddenly 4:36you can have kitchens. Is everybody 4:38going to cook at home? 4:40No, definitely not. There's absolutely 4:43still going to be a market for 4:44restaurants and we all know that 4:46intuitively because we have kitchens in 4:47our home and we still go out to eat. But 4:51there's going to be a ton of interest in 4:53cooking at home and that's going to be 4:54an entire industry. And it's incorrect 4:57both to compare that to the restaurants, 5:00to the professional engineering jobs, 5:02and also incorrect to say it doesn't 5:04exist. It's a real market. It really 5:06exists. 5:07It's going to lean into fully featured 5:10apps. No one wants to cook their food 5:11and have it half 5:13done, but it's different than 5:16restaurants and full-scale engineering. 5:17And I think that we're missing that 5:19nuance. And I think that the companies 5:21that come in and sort of publicly 5:23um say that the products they're 5:25bringing to market are fine if they're 5:27half featured like Canva did, they're 5:30not going to succeed long term. I don't 5:32think that that's actually going to take 5:33market share from the companies that are 5:35committed to delivering the home coding 5:38home kitchen experience. I had to use 5:40the scare quotes. Let me know what you 5:42think. I think this is a topic that I'm 5:45really curious about because I see a ton 5:46of interest in my in my social feeds 5:49and I I think there's appetite here that 5:53is not fully satisfied by the Good.