Gemini 3 feels like a big next step

Before vibe-coding had a name, I made a "game" to help my kiddo learn left from right, and then I was able to drastically improve it with better LLM models. And then today, Google reveals Gemini 3 which can do this.

Pretty slick demo! I decided to try and one-shot it with my Asteroid Blitz game by first asking Claude, "Write me a prompt that I can give to a new conversation that will create this game. Don't be too detailed because I want to leave room for interpretation and possibly even improvements. Capture the high level gameplay objectives and flow."

Pretty open-ended, which Gemini v3 then used to create this.

I kinda love it! It didn't nail the learning aspect, which is fine based on the prompt. This worked out of the box on the first try. Even that first "left right" game took me a couple hours back in the day. The pace of improvement with these models, and the UX to play with them so effectively, is astounding. 

Can you beat my high score of 160? Maybe this can be my "pelican riding a bicycle."