Critical Thinking about IA app Builder

[haughty gulch]

I’m generally in favor of helping people build and create new projects, but I don’t think an AI app builder is a good direction for Wix.

Building a website with AI is one thing; building an app is something entirely different. App development is inherently more complex than simply writing code or prompting an AI to generate features.

I may be biased, but I’m concerned that allowing anyone without a minimum level of development experience to build apps will flood the marketplace with half-baked, buggy AI-generated low value apps. Over time, this could erode trust in the app market itself, leading users to become wary of installing apps altogether.

I know my opinion is unlikely to influence Wix’s broader strategy, but I don’t believe spreading resources even thinner—by adding yet another AI tool and editor—will benefit the community.

Personally, I would much rather see better integration of AI into existing tools. Or even no AI at all, but stronger MCP support and improved documentation, which would ultimately benefit the entire developer ecosystem.

When you think about it, today you can already build apps using:

• The Block Editor
• The (old) Wix CLI
• The Astro CLI
• Self-hosted solutions

Each of these approaches has its own limitations and feature gaps, and they don’t always align with one another. Adding yet another way to build apps only increases fragmentation.

And to be completely honest, I’ve never heard developers complain that apps are “too hard” to build and that they desperately need an easier, AI-driven way to do it. I genuinely struggle to understand the investment decision behind this.

Rather than lowering the entry bar with AI-generated apps, Wix should focus on better integrating AI into its existing products. For example, an AI site-building agent that deeply integrates with the App Market would be far more valuable....

...Such an agent could recommend relevant, high-quality apps based on real business needs—for instance, suggesting appropriate apps for a bakery based on its actual use cases.

We could even go further by enabling micro-MCPs for apps, allowing the Wix AI agent to install, configure, and interact with custom apps in a structured and reliable way.

Ultimately, using AI as a simple code or feature generator isn’t real innovation—it’s the new “old thing.” It’s been done, seen, and repeated many times already and does not bring a lot of value

[timber gazelle]

Hi Kentin,
Thank you for your thoughtful and detailed feedback, I really appreciate you taking the time to share it.
First of all you’re absolutely right: app development is inherently more complex than prompting an AI to generate code or features. That distinction is very clear to us as well. The main purpose of the AI here is not to replace real app development or lower the bar to no-code app creation. Rather, it’s meant to assist developers in getting started - an area where we consistently see friction. Many developers struggle with the initial setup: understanding which extensions to add, which webhooks and permissions are required, and how to structure a new app for a Wix site. The goal is to help with that starting point and reduce boilerplate and guesswork, not to bypass developer expertise.
It’s also important to clarify that this environment is intended for developers, not no-code users. We don’t believe an end to end, production-ready app that solves real business needs can be built “just by AI,” and the tooling is designed with that assumption in mind.
Regarding your point about using AI in more impactful ways such as recommending relevant & high quality apps for specific types of sites or business needs - there is ongoing work in that domain as well.
Thank you again for your feedback, it is greatly appreciated

[haughty gulch]

Helloc @timber gazelle Thank you for taking the time to reply 🙂

I’m glad to hear that we’re aligned on the end goal of this feature.

That said, while I understand your point, I don’t reach the same conclusion. If developers need assistance bootstrapping a project and understanding its structure, generating an entire project with AI will not teach them how to maintain it, nor will it help them build real expertise, in my opinion.

If the intent is to help developers bootstrap their projects and learn “the Wix way,” I believe tutorials and videos would be a more cost-efficient and sustainable approach than adding and maintaining yet another editor.

I also think a major source of confusion for developers lies in the poor quality of error messages returned by the APIs. Very often, the response is a simple 400 or 403 with no context. Most of the questions I see on Discord come from developers who can code, but don’t understand why their code isn’t working because they lack actionable feedback.

If Wix APIs returned more descriptive error messages, it would significantly reduce confusion. As a bonus, these errors would also be far easier for AI agents to debug and reason about.

For example, instead of:
403: permission denied

something like:
403: permission denied — this request requires the "READ_MEMBERS" permission

would help everyone and reduce the need to rely on support.

Finally, AI tools evolve extremely quickly. In six months, today’s AI agent will likely be outdated by newer models, which means it will require constant maintenance from Wix just to stay relevant.

So to summarize my position: I believe Wix should focus its AI efforts on delivering high-quality foundational tools that benefit all developers, along with AI-ready interfaces (such as MCP), rather than building yet another AI editor that risks being obsolete by the time it’s released.

[zealous crane]

True. And with harmony, they are sunsetting blocks it looks like. Before product is evolved and polished, they are moving their attention to new things

[haughty gulch]

@zealous crane I'm not sure Block is sunsetting, where did you get that?

I believe harmony is rather a remplacement for Wix Classic Editor. I imagine support for block app will come later (hopefully)

@thorn ore Could you bring some clarification on this? 🙂

[thorn ore]

To clarify, around the Wix Editor... Wix Harmony and the Wix Editor are both available and live for you to use.

The launch of Wix Harmony does not affect any existing Wix Editor websites or the ability to create new ones. The Wix Editor sites remain live and you’ll be able to keep editing and maintaining them.

If in the future there will be any changes or updates to the Wix Editor, we’ll notify you in advance.

Will ping the team for clarity on the other questions 🙂

[fleet garnet]

@haughty gulch Blocks is not be supported on Harmony however will keep working on Wix Editor and Studio as mentioned.
If in the future there will be any changes or updates to the Wix Blocks, we’ll notify you in advance.

[zealous crane]

the documentation implied so. because at first they said harmony slowly will replace wix editor. and in documentation they said to use cli or self hosted. in one reply too, they are looking for way to convert some of the blocks apps only. but now it looks like wix editor is there to stay!

[haughty gulch]

@fleet garnet
So the only way to have a cross-editor app is to build it using CLI or as a self-hosted solution, right? Given how uncertain the future of Wix Blocks is, I’ll probably stop using it altogether in favor of the CLI.

@zealous crane
My prediction is that Harmony will eventually replace Wix’s Classic Editor (either by Wix’s decision or through adoption by users). Because of that, I wouldn’t place my bet on Blocks, especially since it already lags behind the other editors. But that’s just my opinion 🙂