ØVOID
ØVOID
Experimental software lab•Two people•Infinite drafts
Chaos is a feature. Curiosity is the UI. Refresh if you want. Stay if you’re brave.

We build in public uncertainty—on purpose.

ØVOID is a two-person experimental software lab building in public uncertainty—because AI makes software a commodity, and that changes what “good” looks like. We build systems that surprise us, then we learn out loud. Co-creation beats control: let the work argue back.

Enter the lab (wipe your feet on the way in). Reorders the cards + redraws a few numbers. Nothing leaves your device.

Tarot spread

Three columns, eight cards: not predictions—prompts. A neat way to read a messy lab.

If this feels slightly unpredictable: working as intended.

The Differentiator

card i — meaning

AI makes software a commodity; meaning becomes the differentiator. We treat “beautiful chaos” as a design material—not a mood.

  • We choose experimentation over safe choices, even when it squeaks.
  • Co-creation beats control: let the work argue back.
  • The future is being written—leave room in the margins.

The Method

card ii — loop

Prompt → Surprise → Patch → Laugh → Ship (softly).

Artifact / “Co-creation loop”

A hand-drawn diagram with arrows between: Prompt → Surprise → Patch → Laugh → Ship (softly).

Rule: Beautiful chaos is not a bug—it's a constraint system.

The Vow

card iii — unoptimized

We keep a short list titled “Things we will not optimize.” It includes: wonder, weirdness, and the time it takes to notice something new.

Artifact / sticky note on the router

“Do not negotiate with the network. Offer it a story.”

Current Project

card iv — platform

AI-native gaming platform where games are created, played, and judged by AI. We’re interested in what happens when the referee is also a collaborator—and sometimes an unreliable narrator.

ØVOID project note: “Every game is a first draft. The trick is making the draft playable.”

  • Creation that invites disagreement (and keeps a reset switch).
  • Play that tolerates strange edges without collapsing into noise.
  • Judgment that’s “curiously fair”—stern when needed, whimsical when earned.
Label: image://ai-judge-mask-split-face Label: image://first-draft-arcade-cabinet Label: image://glitch-garden-circuit-vines

Anecdote

card v — breathing

We once tried to “stabilize” a prototype by removing randomness. It immediately became polite, predictable, and—somehow—less honest. We put the chaos back and the whole thing started breathing again.

Artifact / crumpled index card

“Rules for building in beautiful chaos (v0.3): 1) invite disagreement, 2) keep a reset switch, 3) let the system say no.”

Translation: Not here for polished certainty? Good.

Metrics (Vibes)

card vi — instrumentation

We measure what we care about. Units may be imaginary; the consequences are not.

Unpredictability Index
0.72
Draft-to-Delight Ratio
3:1
Chaos Budget
11 units/day
AI Judge Mood Spectrum
stern → whimsical (curiously fair)
Time-to-First-Surprise
12 seconds

Ritual

card vii — second draft

A late-night lab ritual: one person writes a terrible idea on a sticky note, the other tries to make it technically plausible. The best outcomes are never the first draft; they’re the second draft arguing with the first.

Artifact / folder: “Unreasonable Prototypes”

Contains: “mood-based menus,” “judges that change their mind,” and “tutorials that teach you to unlearn.”

Contact Surface

card viii — edges

We build prototypes that argue back. If you want to talk about AI-native play, weird evaluation, or systems with taste—start at the edge:

GitHub (public work + receipts) Microcopy: You can’t break it. You can only teach it something new.

Small promise: We won’t optimize away the part that makes you look twice.

Margin note (arrives late)

You waited. The page noticed. Here’s your first surprise: a quiet instruction—let the work argue back, then listen for what changes.