Every time I open Splitwise I think about how I would build it differently. The ads. The currency settings that never remember. The flows that take five taps when they should take one. I know exactly what I want. I just can’t build it.

Everyone has a version of this. An app that almost works but not quite. A feature that should exist but doesn’t. A tool you’ve imagined for years that nobody has made because the market is too small or the problem is too specific.

What if you could just describe it and have it exist?

This is what I keep coming back to. We’re entering a world where describing software is the same as creating it. You tell the model what you want. It writes the code. The gap between imagining something and having it exist is collapsing.

What becomes possible

I’ve been thinking about all the apps that should exist but don’t.

A chemo patient who knows exactly what food recommendations would help during treatment. A high schooler who knows exactly what polling app her friend group needs. A fitness coach who knows exactly what tracking features his clients are missing. A parent who knows exactly what would make homework less painful.

These people understand their problems better than any product manager at a software company ever could. They just couldn’t build software. Now they can.

Creators become developers

Instagram created influencers. TikTok created short form video creators. I think this creates app creators.

A fitness influencer doesn’t just post workouts. They ship a personalized training app for their audience. A teacher doesn’t just lecture. They ship interactive learning experiences. An astrologer doesn’t just write horoscopes. They ship a tool that generates predictions based on your actual birth chart.

Apps as content. Software as expression.

The interesting thing is that software compounds in a way content doesn’t. A video entertains once. An app keeps working. A journaling tool I build gets more useful as I use it. A wardrobe suggester learns my style over time. Memory accumulates. Context deepens.

The organizational layer

I was watching Veo3 prompts go viral on Instagram. The loop was fascinating. Someone posts a video. People comment asking for the prompt. The creator DMs it. You copy paste into Gemini. Upload your image. Get output.

Prompts spreading through DMs and comment sections. No tracking, no social graph, no way for creators to know how many people are using their work.

Notion templates have marketplaces. Airtable templates have marketplaces. What if mini apps had the same thing?

A platform where you can see who created an app, how many people use it, fork it, remix it. Where security is handled by the platform so users don’t have to evaluate each creator’s data practices. Where your context travels with you across apps. Where memory persists.

This is what I mean by organizational layer. The infrastructure that turns scattered creations into a medium.

The verticals I’m excited about

Astrology. A billion dollar industry built on generic newspaper columns. Every prediction could be personalized to your exact birth time and location. Dasha matching. Varshphal. Rajayoga calculations. The depth is there. The personalization hasn’t been.

Healthcare. Medicine trackers that know your full list. Immunity trackers that adjust to treatment cycles. Symptom logs that surface patterns. Apps that would help so many people but never got built because the market was too small.

Personal finance. Your bank lets you set notifications for specific things. But imagine setting up anything. Spending anomalies. Subscriptions you forgot about. Patterns you want to track. The same data your bank has, working for you.

High school apps. Sarahaha. Poll.me. Slambooks. Teenagers understood exactly what their friends needed and those apps went viral. Same energy, but now anyone can build.

Multiplayer experiences. Invite your roommates to a Splitwise alternative you built because you were frustrated by the ads. Invite your partner to a date picker. Software that emerges from real relationships and real friction.

Deep personalization

I think the real unlock is context.

Today every app starts from zero. It doesn’t know who you are. It doesn’t know what you’ve done before. It doesn’t know your preferences or your history.

What if your context traveled with you? Your calendar, your health data, your preferences. Available to apps you trust. Accumulating over time.

Features that adapt to how you use them. Interfaces that match your aesthetic. Prompts tuned to your specific workflows. Not just faster software creation. Genuinely personal software.

What I believe

Every new medium creates new creators. Printing press created writers. Cameras created photographers. YouTube created video creators.

Software 3.0 creates app creators. Ordinary people shipping tools for problems they understand intimately.

The platform that provides the organizational layer for this is the next big thing. Where security is a given. Where social mechanics drive distribution. Where context is shared and memory persists.

We’re at the beginning. The tools exist. The behavior is emerging. The platform hasn’t been built yet.

I find this genuinely exciting.