Codemonkey talked about vibecoding in his last Game Dev Report and I think this is an excellent example to be at least careful when using AI. It is not "DO NOT EVER", it is "USE WITH CARE". I really recommend everybody in here to read it. One developer shared how they spent about 40 hours over a couple of months trying to "vibe code" a metroidvania in Godot, hoping AI could handle the technical side while they acted as the creative director. At first it looked promising, they got some basic movement and systems working quickly, AI nowadays is absolutely excellent at making good small code snippets, but once the project started getting more complex and systems needed to be connected with one another then everything began falling apart. The AI struggled with enemies, combat, refactors, and broader architecture, and every fix seemed to create more problems. Eventually they gave up and concluded that full-project vibe coding is basically a myth, at least for something as complex as a real game. And yes, that is what I am constantly saying. AI is excellent if you use it as a tutor to help you understand something. If you ask it to explain some piece of code line by line so you can understand then that's awesome! But the problem is when people ask it to generate something and just blindly copy paste code. If you do that then you don't understand what the code is doing. And if you don't understand it then when something breaks (inevitable) you will not know how to fix it. If you're making something simple like Flappy Bird then it's possible you can build that whole game using AI without anything breaking, but for making any game even slightly more complex than that then it will break guaranteed. When you're vibe coding you're not actually building the game, you are just stacking mystery boxes on top of each other until the whole thing collapses. The developer asked the AI to generate a bunch of stuff and it successfully generated a working character controller that could run, jump, climb ladders, read tablets, use a scanner and more. All that done in just a few hours, impressive. But then when they ran into issues they tried solving them with more AI which didn't work, for example trying to refactor some code into a state machine just made a massive mess of global data. Since the developer did not understand the code they had no other choice but keep asking the AI to fix things and hope that it worked, and over time it worked less and less.

Anyone participating in the Cursor Vibe Jam this year? https://vibej.am/2026/ Rules: - Anyone can join and submit a game - You must include the required JavaScript widget snippet in your game - At least 90% of the code must be created by AI - The game must be newly created during the jam (after April 1, 2026) - The game must be playable on the web, free, and without login/signup - Multiplayer is preferred, but not required - You can use any engine (Three.js is recommended) - No loading screens or large downloads — the game should start almost instantly - Only one game submission per person - Focus on building one high-quality game instead of multiple entries - Deadline is May 1, 2026 at 13:37 UTC - You can submit early and continue improving your game until the deadline Found out someone already submitted Tower Survivors (https://ts.pliete.it/?ref=vibejam), just days (hopefully) before I release the android version of my Tower Survivor game 🤣 I was on vacation the last 2 weeks so two weeks left for me. Will post my first idea in a seperate post. So who will submit their own game? Post a link to your submitted game for support 👍

Background: I've been making games and apps with Unity for 16+ years. I had someone ask a question about how to use Claude Code to make a game and after writing a short novel, I thought, "hey, this could be helpful to others" -- but I'll let y'all be the judge. (note: a lot of this would also apply to making an app or any piece of software with Claude Code) Someone asked me about making something based off an existing game, but also how would he attract programmers to help him / work for him. Below is my response: --- To start... you need Unity (and some basic familiarity with it) and the Unity MCP and Claude Code (the CLI) -- and claude opus. Opus is extremely good at letting you drop the whole vision at once (including screenshots) and tell it what you're wanting to do. In this case, make a game like XYZ. And it'll tell you where to start and how to progress from there. The one thing that AI cannot do is make a game that's fun. Not intentionally, anyway. It's pretty much always going to need tweaks. In your case, wanting to make a game with big monster trucks... The key to ANY game is usually going to be one core piece or mechanic. I would tackle that hardest piece first. You need to get those vehicles rolling and feeling good to play. AI can build something, and it can build something you can tweak, but it cannot "make a fun game" any more than AI can "make a song that moves me" -- it's too subjective to the person interacting. Talk to Claude Opus, tell it the grand plans, ask it where to start, ask it to build a tech spec for the v0.1 -- again, I'd absolutely start with the car. Tell it you're going to use claude code and the unity mcp. It can build a markdown file very specifically for claude code, along with instructions. Then go to a directory with that file in it (probably needs to be an existing but empty Unity project if you're going to use Unity), go into claude's CLI and reference that file with @ saying something like, "I want you to plan out and then build to the specification in the @car_game.md file" (and the @ will bring up a list of files, fyi)

Describe your weekend so far with a GIF (or 2) and maybe a word if you're feeling daring! (Bonus points if it's the name of a game!)

🛠️ Question: ❓ Which AI tools do you use on a daily basis?