Skip to main content
Read
Watch
Listen
Read the essays

Three problems AI didn't solve

Author · Creator · Founder

Joshua Seymour

GnosisGenesisGenerative

Husband, father, American. I build with Claude Code and ship on Vercel, same as you. The question I keep working on: why do builders with the best tools in history still hit the same walls at $50K, $500K, and $5M? The answer I keep landing on has four phases.

Who is Joshua Seymour?

Most builders I talk to share the same frustration. They're using AI ten hours a day, shipping faster than ever, and still stuck at the same revenue, the same isolation, the same feeling that all this output isn't adding up to anything. 84% of us use AI daily. Only 29% trust what it produces. That gap — between capability and confidence — is the actual problem. That's what I've been working on.

The map I've been building is called Genius — not genius as a compliment, but as a process. Four phases: see where you actually are (Current), be precise about what you want (Desired), build the bridge (Actions), know the score (Results). Each completed loop produces one of three things:

Author

Gnosis

The power to see what's actually happening and articulate it clearly.

Creator

Genesis

The power to bring something real into existence from nothing.

Founder

Generative

The power to build systems that serve others at scale.

I write at joshuaseymour.com, build at avolve.io, and organize at supercivilization.xyz. I work for the Supercivilization and its stakeholders.


How did I get here?

The pattern I kept noticing was this: some people create real things that outlast them, and most don't — and the difference wasn't talent or luck. It was something structural. A way of processing experience into output.

I didn't have a name for it, so I started mapping it. That mapping became Genius. Not all at once — in pieces, over years, tested against my own life and against the people I watched closely.

The four phases (Current, Desired, Actions, Results) crystallized from watching where people actually got stuck, not from a theory I had in advance.

Then came the question of what to do with it. Writing it down wasn't enough. I needed to build something that made it usable — not just readable. That's where Avolve came from: a game where the loop is the gameplay, and progress in the game is real-world progress, not virtual points.

I'm still running these loops myself, against the same problems everyone runs them against: how to do meaningful work, how to be present with the people I love, how to build something that doesn't depend on me being in the room.

The organizing principle behind all of it is the shift from zero-sum to positive-sum. From systems that extract to systems that generate. I call that the Supercivilization. I'm still figuring out what it means in practice. But the direction is clear enough to work toward.


What am I working on?

Writing

Long-form thinking about what happens after AI solves the building problem — the knowing, creating, and sustaining problems that remain. Most posts start as something I needed to work out for myself.

Read the blog

Building

Avolve is a real-world video game — a web app where the Genius loop is the core mechanic. Players journal actual goals, log actual actions, and get XP for actual results, not for time-in-app. The first version is the individual experience: one person, their own puzzle, no leaderboards yet.

See avolve.io

Organizing

Supercivilization is the organization behind Avolve — and the larger bet I'm making. The argument is simple: when enough individuals solve their own problems in ways that help others solve theirs, the center of gravity shifts. Each positive-sum move makes the next one easier. I work for its stakeholders: the individual player and the collective they're becoming.

See supercivilization.xyz

Where to find me


Start here

If you're building with AI and hitting walls the tools don't explain, the blog is the best place to start. Three essays, one loop, no platitudes.

Read the latest