Prompts vs. Netflix: How I Whipped Up Angry Jets with Grok in One Evening

So picture this. It is a random evening, I have about four hours of free time, and instead of melting into the couch with Netflix, I decide to build a game using nothing but Grok and Photoshop. No Cursor. No fancy dev tools. Just prompts and pixels.

The Big Idea: Evolving My AI Game

I have been messing around with AI for a while now, borderline obsessed with figuring out how these machines tick. How do they reason. How do they save information. And how can I bend them to my will using nothing but clever prompts.

The game concept was simple but fun. Angry Jets. Think Angry Birds vibes, but instead of slingshotting birds, you are piloting a jet using the up and down arrow keys while dodging a swarm of enemy aircraft. Classic arcade madness.

The Process: Prompts, Pixels, and Patience

I fired up Grok and started typing. Step one was nailing the game logic. Grok spit out a rough framework — basic mechanics like jet movement, enemy spawns, and collision detection. From there, it was all about progressive prompting. "Make the jet move smoother." "How do I spawn enemies at random intervals?" It felt like training a super smart intern who occasionally overthinks everything.

Then came the visuals in Photoshop. A sleek jet for the player. Some goofy enemy planes. A basic sky background. Nothing award-winning, but it had charm.

The Build: Four Hours of Chaos and Coffee

The whole thing took about four hours. I used Grok to simulate the game logic step by step. By hour three, I had a working concept. By hour four, I stitched together a rough but fun prototype. The result was an HTML-ready, drop-in game that uses no external code or files.

The Payoff: Office Chaos

The next day at FlyUSA, within an hour, half the staff was huddled around my desk hammering keys, trash-talking, and trying to beat my high score. Productivity absolutely suffered. Sorry, team.

The Real Win

Angry Jets is not winning Game of the Year. That was never the point. The real win was leveling up how I work with AI. I learned how to chain prompts into something complex, how Grok reasons through iterative steps, and how to wrestle its output into something usable. You do not need a full dev suite to build something fun. You need a sharp mind and a sharper AI. What should I build next?