World Labs: expert evals for AI-made games, by game developers who ship.
World Labs Community
The Panel
Less vibes. More verdicts.
Join The Panel and help shape the future of AI-made games. Play and judge real output from the top AI labs and game-tool companies, and get paid for the judgment you spent a career earning.
Every panelist has real titles behind them. If you've spent years making games good, this is the room for you.
Level Design
Systems Design
UX & Onboarding
QA & Playtest Leads
Game Feel
No exclusivity requirement, no middleman markup. Just a curated panel of shipped game developers who share a high bar for what counts as a good game.
How it works
Three steps in. Then you're in.
1
Apply
Send your shipped credits, your specialty, and a few sentences on the best and worst game you played last year, and why. Fifteen minutes, no portfolio required.
2
Calibrate
A short conversation with a current panelist, plus one sample eval on a real AI-generated game. We're checking judgment and writing, not trivia.
3
You're in
If you're in, you're in. Assignments start with the next round. No probation tiers, no ranking games, no fighting for gigs in a feed.
Assignments route by specialty: game-feel people judge game feel. Each session is a game, a timer, and a structured writeup. Paid per completed eval, invoiced monthly.
Community
A private room of people who ship
A members-only space to compare notes across rounds, argue about scores after they lock, and see where AI game tools are actually heading, months before the public does.
Early access
The tools, before the hype
Panelists play output from the newest prompt-to-game tools under blind conditions first. When identities unmask at the end of a round, you'll know which tools are real.
On your terms
Flexible, remote, no exclusivity
Take the rounds you want, skip the ones you don't. Everything runs in the browser, on your schedule. Keep your studio work, your freelance clients, your side project.
FAQ
Fair questions, straight answers
What does the work actually look like?
You're assigned a game generated by an AI tool (identity hidden), you play it in the browser for a minimum session, usually around 15 minutes, then score it on a seven-dimension rubric, write a critique, and rank the top three fixes. Some assignments are head-to-head A/B comparisons instead.
How does payment work?
Flat rate per completed eval, set per round before you accept any assignments. A rubric session pays more than an A/B comparison. No bidding, no rate negotiation per gig, no platform cut ambiguity. Invoices settle monthly.
What's the time commitment?
Whatever you take on. A typical round assignment is a handful of sessions over two to three weeks, each about 30 to 40 minutes including the writeup. You can decline any round with zero penalty. The only expectation is that accepted assignments get finished.
Do I need AI experience?
No. We're paying for the opposite: judgment about games, built by shipping them. If you can tell a coherent mechanic from a broken one and explain why in writing, you're qualified. The harness handles everything else.
Is my name attached to my scores?
Never. Published results and licensed data are attributed only to specialty and years of experience: "Level designer, 12 years," not your name. Your credits get you in the door; your identity stays out of the data.
How is this different from a playtesting marketplace or staffing agency?
Marketplaces sell volume; we sell calibrated expert judgment. Every panelist is vetted on shipped titles, every eval follows the same anchored rubric, and every game gets multiple independent raters with disagreements flagged and re-judged. Agencies place you with clients. We run the whole harness and you just judge.