Where Dreams Shoot Their Shot
A launchpad for the internet’s wildest ‘what ifs.’
Challenge
Ever stumble across a comment online like, "What if Keke Palmer and Ziwe hosted a game show where celebrities have to defend their most problematic tweets" — and think, wait. That's actually brilliant?
But then you keep scrolling. Because what else are you gonna do? The idea disappears into the ether.
I started wondering: what if there were a way to capture those ideas before they vanished?
But then you keep scrolling. Because what else are you gonna do? The idea disappears into the ether.
I started wondering: what if there were a way to capture those ideas before they vanished?
My Goals
Coming into this project, I wanted to rapidly build something real around a frustration I'd personally experienced — that disconnect between fan creativity and collaboration reality.
- Build a functional MVP using AI-powered development tools
- Create intuitive UX for suggesting and discovering collaborations
- Ship something real that people could test
- Learn what it feels like to design with AI as a building partner
How I tackled It
The problem space
I dug into how people currently share these collaboration ideas:
- Twitter/TikTok comments: Can blow up and reach brands, but most suggestions get lost without systematic tracking
- Reddit communities: Passionate discussions with voting, but fragmented across different subreddits
The pattern was clear - while some suggestions do reach brands and create viral moments, there was no systematic validation process or reliable way to connect fan ideas to creators and brands.
My Target User
The Discovery Enthusiast — passionate fans who live for finding and sharing cool new things. What makes them special? Their suggestions come from genuine excitement, not content strategy. That's authentic demand you can't fake.
AI-Assisted Development Journey
Weekend build strategy that actually worked:
- Claude: Brainstorming and documentation support
- Bolt.new: AI-powered React development for lightning-fast prototyping
- Supabase: Backend setup without the usual database headaches
- Tailwind CSS: Professional styling without drowning in custom CSS
The secret? Crafting detailed prompts with database schema, UI requirements, and complete user flows. AI handled the heavy lifting while I focused on UX decisions.
Key Design Decisions
- Upvote-only system: Dropped downvoting after realizing it might discourage participation without adding clear value
- Collaboration format: Clean "X x Y" layout that works for any pairing
- Trending section: Popular content creates immediate engagement hook
- Authentication required: Account creation prevents spam while enabling user tracking
Learning on the Job
Working with AI development tools was honestly both intimidating and exciting. I got way better at writing prompts that actually worked and learned to iterate super quickly with Bolt — bouncing between button tweaks and entire feature flows while staying focused on user experience.
Core features that actually work:
- User auth and profile management
- Collaboration suggestion submission
- Browse feed and one-click voting
- Trending section for discovering popular suggestions
- Responsive design and search functionality
Takeaways
- Start with your data story — mapping out database tables upfront meant Bolt could build everything way more cohesively from the beginning.
- Weekend goals actually work — used Claude to create a build plan that kept me accountable and helped me know when to stop perfecting.
- Real-time everything — feedback loops went from weeks to literally minutes.
- Vision beats speed — development isn't the constraint anymore. It's how clearly you can articulate what you actually want to build.