We’ve been leaning into vibe-coding prototypes lately to capture more complex interactions.
The problem is that with static screens, it’s hard to communicate the full vision and “feeling” of a product/feature to a user, especially when there could be dozens of possible interactions that a user could take.
The other problem is that a lot of people are transitioning from a design-first mindset to a prototype-first mindset with the rise of AI tools that allow non-devs to pump out functional applications, fast.
This forces these prototype builders to be building blind. Everything is just a whim, a feeling of what’s right and wrong; pure intuition, which in some ways are good, but in a lot of ways can be bad without a good sense of smell.
Helio already does something very powerful, and that’s providing a huge audience base of 1 million+ users waiting for a survey.
Which means that the only thing that’s currently a gap, is a way to funnel users to a prototype, track their analytics, and have them answer a few questions.
My current sense is that this could start to become an interesting yet powerful way to collect feedback, fast, and at scale. You could even start to funnel the feedback straight back into your AI tool of choice with summarized syntheses, making the feedback loop move even faster.
Figured I’d share some of the idea/work here since it’s been fun building this out. Curious what other people are experimenting with feedback-wise.
