ØVOID
ØVOID two-person experimental software lab
Current Lab Weather: Partly Glitchy, with a chance of elegant outcomes. We prefer prototypes with fingerprints.
Edition: The Unfinished Daily
Filed under: taste • risk • weirdness

ØVOID builds beautiful chaos on purpose.

ØVOID is a two-person experimental software lab building in public with a fondness for beautiful chaos. Our flagship is an AI-native gaming platform where games are created, played, and judged by AI—every game is a first draft, and that’s the point.

Start with the project AI-native gaming platform Enter the lab nothing here is finished GitHub see what we’re testing today
image://glitch-constellation-map image://ai-judge-scorecard-doodles image://chaos-dial-analog-instrument

Editorial: the manifesto

Make the chaos legible—then turn it back into music.

AI makes software a commodity; taste, risk, and weirdness are the remaining scarce resources.

  • Experimentation over safe choices. Even when the safe choice is wearing a nice blazer.
  • Beautiful chaos is not a bug. It’s a navigation system for the unknown.
  • We don’t “finish” products. We release them into the weather and learn their names later.
  • Unpredictability is a feature. When you’re co-creating the future, the future is being written.
“If it feels finished, we probably stopped one iteration too early.”
— Lab Note (found on a sticky note)

Lead story: current project

Proceed anyway.

An AI-native gaming platform: games are created, played, and judged by AI.

Core premise: every game is a first draft Judging: rubric-driven, weirdness-aware Mode: co-creation (gently)
  • Create: prompts become worlds; constraints become mechanics.
  • Play: AI agents explore the space like impatient critics with hands.
  • Judge: verdicts are explained—then argued with—then improved.
  • Learn: each round emits artifacts: rules, scorecards, and new questions.
Every game is a first draft: messy, earnest, and strangely confident.
On the flagship project
Filed report: “We tried to make it polite. It became opinionated.” open the door (it might move)
While prototyping the platform, we asked an AI judge to explain its verdicts. It produced a tiny manifesto about “fairness as a moving target,” and we taped it to the wall like a warning label.

Margin notes

Two voices, one sentence.

We’re here for the moment a prototype stops behaving and starts having opinions.

— ØVOID (two voices, one sentence)

Make the chaos legible—then turn it back into music.

— A margin scribble

If it feels finished, we probably stopped one iteration too early.

— Lab Note (found on a sticky note)

Evidence locker (selected artifacts)

Not nostalgia—proof of motion.

ARCHIVE

Anti-roadmap sheet

A crumpled “Anti-Roadmap”: three arrows pointing in different directions, labeled “maybe,” “probably not,” and “let’s see.”

UI FRAGMENT

Button specimen

Screenshot fragment: a button titled “Begin Anyway” with a sublabel “this might get weird.”

RUBRIC

AI judge scorecard

One-page rubric for AI-judged games: “surprise,” “internal logic,” “delight per second,” and “respect for nonsense.”

SKETCH

Chaos dial

A slider from “tidy demo” to “weather system.” (We keep nudging it right.)

NOTE

Rule #4

If the AI behaves too politely, give it a stranger prompt.

Nothing here is finished. Some things are deliberate.