I mean if he’s genuinely interested in learning Unity and just using AI to help with learning that’s good, but yeah if he’s just making the AI do all the work and relying on it then yeah probably not gonna go good.
It’s a bit of both. I’ll try to help him if he asks for help if/when the AI leads him down the wrong path, but who knows how deep in the tangled weeds of spaghetti he’ll be in by then.
I mean spaghetti code is a part of learning a new tool, my first Unity projects I did back in highschool have some really bad spaghetti code. I’ve since learned from then and found better ways to organize things to avoid that. But yeah if you’re just letting AI write your spaghetti you’re not gonna understand why it’s bad and that prevents learning what you can do better next time.
As long as you can keep the vibe coded pieces tiny and modular you’d probably be fine. But that takes a robust knowledge of Unity and gamedev architecture, and at that point you should probably just write it yourself.
Complex, math-heavy stuff like gaming usually is too much for an AI. It’s better at like, basic Python scripts or writing a bunch of dirty CSS.
It can help out a lot in writing base systems. I wanted to learn rust and made gpt write a game with me, it really could handle more complex things than I thought. For example it wrote a diablo style rotatable inventory system (cli, no graphic) which impressed me a lot.
Another game prototype i had lying around in react i asked Jules to refactor to use event sourcing rather than mutating values and I only had to give it a few pointers on what it had missed. It even found an old bug I had missed.
I guess nothing of it was really math heavy but that inventory grid was really solid
I have a friend who is trying to vibe code a unity project. I don’t have high hopes for him, but I’m doing my best to support him and help him out.
I mean if he’s genuinely interested in learning Unity and just using AI to help with learning that’s good, but yeah if he’s just making the AI do all the work and relying on it then yeah probably not gonna go good.
It’s a bit of both. I’ll try to help him if he asks for help if/when the AI leads him down the wrong path, but who knows how deep in the tangled weeds of spaghetti he’ll be in by then.
I mean spaghetti code is a part of learning a new tool, my first Unity projects I did back in highschool have some really bad spaghetti code. I’ve since learned from then and found better ways to organize things to avoid that. But yeah if you’re just letting AI write your spaghetti you’re not gonna understand why it’s bad and that prevents learning what you can do better next time.
Agreed
Don’t. Let him drown. Bring the popcorn.
As long as you can keep the vibe coded pieces tiny and modular you’d probably be fine. But that takes a robust knowledge of Unity and gamedev architecture, and at that point you should probably just write it yourself.
Complex, math-heavy stuff like gaming usually is too much for an AI. It’s better at like, basic Python scripts or writing a bunch of dirty CSS.
It can help out a lot in writing base systems. I wanted to learn rust and made gpt write a game with me, it really could handle more complex things than I thought. For example it wrote a diablo style rotatable inventory system (cli, no graphic) which impressed me a lot.
Another game prototype i had lying around in react i asked Jules to refactor to use event sourcing rather than mutating values and I only had to give it a few pointers on what it had missed. It even found an old bug I had missed.
I guess nothing of it was really math heavy but that inventory grid was really solid
I don’t think he’s gonna need much math. He will need complicated editor things set up though, and I’m doubtful that the AI will understand much.