AI

What Hackathons and Failed Prototypes Teach Us About Trust

Two Doors. One glowing door labeled Winners with a man and a woma walking in, while another door labeled Other Applicants is left closed/dark.

Sometimes the way we communicate matters as much as the outcome itself.

When I heard about the SheBuilds hackathon, I was excited. Alongside 4,000+ other women, I threw my hat in the ring, hopeful for a chance to build something meaningful with a global community of makers.

This week, the results were announced, but not in the way you’d expect. I didn’t get an email. Neither did the thousands of others who weren’t selected. Instead, I found out from celebratory posts on LinkedIn by those who made it.

First, let me say this: congratulations to the 60 women selected! May your caffeine be strong, your code bug-free, and your ideas unstoppable. I’ll be cheering from the sidelines.

But here’s the thing: community trust isn’t built only by how you celebrate winners. It’s also built by how you treat everyone else. And in this case, Lovable missed a step.

They had the means to send a collective ‘not this time’ message. But there was silence. For thousands of applicants, that silence became the answer. And that’s a UX failure. Because even rejections deserve good UX.

My own failed prototype

At the same time, I was running an experiment of my own: building an AI Policy Decoder (not the same project for #SheBuilds). My idea was simple: upload policy documents and let users query them conversationally. In practice? Let’s just say the ingestion pipeline and I had disagreements.

No matter how I tested, the system repeatedly failed to chunk or retrieve correctly. Over and over, I ran into the same wall: “No basis found in the uploaded documents to answer this question.” Debug logs mocked me while my eyes grew dry from staring at the screen.

It was, in short, a failed prototype. Read Testing Lovable.

The parallel

And yet, failure here taught me the same thing Lovable’s silence did: rejection and breakdowns are part of the journey. But how you handle them determines whether trust is built or eroded.

For hackathon organizers, it’s about valuing every applicant enough to send closure. For builders, it’s about respecting your own process: documenting what didn’t work, reflecting, and deciding whether to iterate or move on.

Silence doesn’t build trust. Transparency does.

Because in the end, it’s not just about outcomes. It’s about how we close the loop.

Closing the loop

So a toast to the 60 #SheBuilds winners! And here’s to failed prototypes that teach us lessons along the way.

In design, whether it’s community design or product design, rejection and failure aren’t endpoints. They’re part of the user journey. And trust, at the end of the day, is built in how we close the loop.

Update: After I posted on LinkedIn about not receiving any notification and tagged Lovable, I received an email.

By then, the gesture felt delayed, more of an afterthought. It required me to ask for it. Nothing could undo the reality that thousands of applicants were left without communication. Transparency wasn’t prioritized from the start.

For me, this confirmed that Lovable isn’t the right tool at this stage of its development, and that our values don’t align. I want to invest my energy in spaces where transparency, respect, and thoughtful design are part of the culture.

The true test of character is in the small acts of respect that cost you nothing but mean everything.

For a more detailed hands-on account of me testing Lovable’s platform, see Testing Lovable: Lessons from Trying to Build an AI Agent.