Can non-developer build commercial products with AI
I honestly think that even with AGI or ACI; without deep understanding of technical architectures and fundamentals , the last 10 miles will be tough! When I tried to build products using Claude Code, i had to interject , either to correct the code or suggest alternative SW designs, for the last few miles. And then operations/maintenance definitely required deeper intervention. Thankfully, I am faily technical - Wondering, if anyone else was able to accomplish this, being totally hands-off?
Can you put together a working product that you can put in front of someone and it will accomplish their goals? Sure, absolutely.
Will it be secure, reasonably free of bugs, scale well, and compliant with any regulatory requirements in your industry (if any)? Not a freaking chance.
Yes, non developers can build applications with AI just like non engineers can build cars with 3D printers.
I've been tinkering with some models and I'm currently progressing through a few personal projects with Gemma in Antigravity. I'm not an engineer, but I have a very good technical understanding, I'm competent enough to build something by myself.
I've been going though my personal projects feature by feature. So far I've had good success, and as I'm doing it step by step I'm checking what's being created. 90% of the time it's correct and when bugs occur I can work through them and identify the issue, and then explain it to the agent to fix.
I don't think you could ever just set an agent off to create something by itself, unless you have a very detailed comprehensive technical document for it to follow along outlining the big picture and all details within - even then I think the context window wouldn't be enough and it may start tripping up.
The projects I've tried to date: - A love2D game (success) - Buildroot linux for an SBC with above game embedded (success, but with several issues related to the framebuffer, other drivers etc. Fixing this took about an hour of my time and burnt through all of the available thinking model tokens in two sessions. - A few offline web projects (ongoing, success when going feature by feature) - A micro controller project (ongoing)
With a web based project, you'll need to know how to setup a server and all of that jazz too, I don't know if an agent can help you with this.
Even changing the technologies can be really hard. I usually use Java and GoLang. I decided to do an Android app with Kotlin. I hit all the Android/Kotlin specifics. Because of my technical skills, I managed to overcome most of the problems, but I did not catch them on time. If I knew that they existed beforehand, I would have saved a lot of tokens, time.. And probably, I would have approach the whole problem differently.
Having that said, non-developer can produce commercial products, but the learning curve will be so steep, and the expense so much, that probably does not make sense to do it.
I think you can be non technical, but that does not mean just blindly trusting AI. And it will probably take you WAY longer to get it correct. You still have to know (or figure out through iteration) what questions to ask, what edge cases might exists, etc. But I also think building things with AI, if you really lean into it, can be a great way for people without a technical background to learn.
People sometimes over-estimate the capability of software engineers behind a lot of products they use.
I'd say that there are risks, a bumpy road ahead, but as long as the non-developer accepts that he will have to actually learn something about coding along the way, it is possible to have a commercial product.
[dead]
[dead]