Founders, be honest: are you building an idea AI gave you?

Building analytics platform (https://spectry.io) that brings real value to customers. Now using AI to validate and enhance the features we can offer. Idea didn't come from AI, instead a long term dream coming into reality now.

Mine wasn't AI-generated. The idea came from a real problem I wanted to solve, though AI has definitely helped a lot during the building process.

To put it bluntly, I see Large Language Models as a rubber duck or a colleague who interviews me to develop my ideas. Instead of giving me brand new ideas, they are partners who are available 24/7 to brainstorm on any specialized field.

I treat ChatGPT like another co-founder during brainstorming sessions. It helps challenge assumptions and uncover blind spots, but I never rely on it to decide what to build. customer conversations still carry far more weight.

There's a good thing and a bad thing here. I built my first side hustle from an AI idea and it actually worked, I made money! This time I built flickapp.in, all vibe coded, but there aren't many users yet. It's honestly not about the idea, it's all about how you distribute it. The worst part about vibe coding is that it feels like you're making progress every day, but really it's just another way of hitting your dopamine.

Not at all! I actually have a huge backlog of my own ideas, some of which I've been sitting on for 10 years.

For me, AI isn't an idea generator; it's an execution multiplier. It finally allows me to bring those old ideas to life, especially using languages I only know on a surface level but are perfect for the job (like Rust or Go). It’s an absolute game-changer for quickly throwing together a Proof of Concept just to see if the idea actually works and feels right in practice.

Mine came from a friend, not a model. He runs a 12-person shop and got 230 applications for one junior role, most of them AI-written, and spent two full days reading them by hand with no HR team. The problem was so concrete I could not not build it. I use AI heavily to BUILD, but the idea was a real person's pain. I think that is the tell: if you can name the specific human who has the problem, it is probably yours; if you can only describe a market, it might be the model's.

Not necessarily. Mine was based on my own need. I built an app I needed after being laid off. You get in the mode where you're actively applying for jobs and your inbox just fills up that trying to track it all is pretty difficult (at least for me).