Satu legenda tunggal yang menunggu GTA 6 diluncurkan telah memutuskan untuk membuat game dari awal sendiri: 'Tujuannya: mengalahkan GTA 6 yang sebenarnya untuk diluncurkan'

If you were waiting for GTA 6 to drop when it should have done on May 26 (well, on console at least), you were probably more than a little perturbed by its delay to *checks watch* no time soon, i.e. November. That's to allow Rockstar to "finish the game with the level of polish you have come to expect and deserve." Frustrating, but fair enough, we can wait.

But if you aren't content waiting, what's to do? 'Nothing at all' is what probably comes to most people's minds. But not Ziwen's. No, that would be far too simple for the AI start-up entrepreneur, who has decided to "beat the real GTA 6 to launch" (via X user Fashion Ruined My Life).

"Day 1 of building GTA 6. Still feels fake typing that out. Upgraded to Claude Max 20x just for this... No studio, no publisher. Just whoever shows up... The goal: beat the real GTA 6 to launch. Ambitious, probably stupid, doing it anyway. If you can model, code, build levels, or write music and lore, come join. Looking for a couple contributors to cook this."

This is it; this is what the brave new world of AI and vibe coding has brought us. The sheer brass, the gall, the bravery to stand up and say, 'I will make GTA 6, and I'll do it before November.' I love it. I hate that I love it, but I do.

I won't spoil the whole journey, which is of course still ongoing, but just to give some teasers for Ziwen's project:

After progressing from day one's Tic-Tac-on-a-field, day two takes a Roblox-esque character to a Roblox-esque world with jittery floors.

Day four involves a cinematic with alligators falling from the sky. And gameplay of a character doing his best impression of Edward Cullen in daylight.

Day six—the latest update as of the time of writing—is more of the same, but there's a "real main menu", a loading screen, and the "whole front-end wired end to end."

Things are already at a point where they're requiring development to get a little more serious, though:

"We've been building in Godot the whole way, it's starting to show its limits on the AAA stuff. So this week I'm going to go test Unity to see if we can get the actual game into AAA stuff as well. For a game built by a loop of agents, the only question that matters is which one they can read, edit, and push on their own. Godot's good at that. Unity might be better, with way more assets off the market. I'm going to find out... I'll report back which engine they build faster in."

Godspeed, Unity. May the AI agents treat you well.