I had a rough first pass at trying to use Opus 5.5 to make a fun game. It wasn’t fun. Maybe we’re past some kind of October inflection point, as Simon Willison puts it, but it’s still hit or miss.

Sand Nomad is not only not fun, it’s horribly un-fun. I have a few specific gripes:

• Zero onboarding. Most issues are downstream from this
• Poorly calibrated input information space (throws a whole bunch of info at you, all at once, with no guardrails)
• Very high choice volume where the choices don’t matter too terribly much, especially to a new player
• Lame graphics
• Mediocre sound design
• The core mechanic just isn’t fun to logic out.

This project was borne out of curiosity about why barter mechanics in games always suck, despite the fact that people keep adding them. So admittedly, I gave the model a hard problem to solve out the gate. It’s very much a follow-up to the space-trucking prototype in that way.

Maybe I’ll re-visit Sand Nomad one day. If I do, and it doesn’t suck, you’ll know what happened.