After spending time with Cursor/Junie/Copilot etc., game engines are starting to feel slow and behind the curve. I’d love to see faster adoption and deeper LLM integration (as native features, not just some random third-party plugins). Whoever builds the first “Cursor for games” is going to disrupt the market.
If Unity ever wanted to regain market cap, now’s the time. The solution seems obvious from my armchair.
I am rooting for Godot here because it has a human readable scene file format.
So when you have an AI agent modifying your project you can actually understand what it did, vs. the unity yaml based format that is mostly unreadable.