Inane Ramblings of a Cat Juggler

I put thirteen local decision models in front of two jobs, approving bash commands and moderating game chat, to find out which ones I can trust to say no.

I ran Qwen3.8-27B through every local inference engine I could build on an M5 Ultra, with and without draft models, to find out which ones are actually fast.

I'm building my own AI benchmark to test spatial awareness, color, web programming, and UI design, and keep a running record of model quality over time.

Sometimes the tech is lousy or misguided, the game design is garbage, or the art is simply bad. These are symptoms, not causes.

Every major shift in game economics was optimizing for something that was inefficient. Play to Earn is the next major shift, and many studios are trying to figure out how to do it right.

There has never been a time when you make players love your game--fans of your game are always going to love it, haters are always going to hate it--so being able to recognize your players is a crucial element of successful tailoring a game to maximize your reach.

What really matters is that you have answers for the following six fundamentals, and pack them into about 10 slides. Here's the list, in order.

Setting up your own secure private registry is easy, but there are a lot of options and blind alleys. Here's a quick walkthrough to get it done.

Let's discuss a few ways that you can crush the morale of your employees and guarantee disharmony, where work loses meaning and tempers flare quickly.

He said that all the technologies being put in play were unfamiliar, and that is what makes it complex. My immediate response was, "Just because something is unfamiliar does not make it complex."

Tools that work get used repeatedly, and sometimes you reach into your drawer full of tools and realize you don't have the one you're expecting to find at your fingertips.

There are many solutions out there, but here's the easiest to implement that gets you something for nothing.

Following curiosity instead of providing value is not virtue.

A dear friend asked me how to tune the interview and job posting to target the best person for the job. Bad jobs often start with bad job postings. Let's unpack that a bit.

Gambling has figured out that the secret to motivation is reducing the failure cases to weak successes. Education has not.

A conversation with a school teacher turns a mild-mannered game developer into an intragroup dynamics philosopher.

Pontificating about the art world having developed universal phrases and lingo to impart understanding, but software really hasn't.

I have found that not only is it hard to do many things proficiently, sometimes all you have to do is one thing really, really well.

Brief advice on how to help your chances of success running an independent game studio, from a veteran developer and studio head.

Actually, it was just the first chapter in the Game Engine Gems 1 book, but I think it was a rare, almost imperceptibly impactful article.