Miami is full of selfish, entitled, distracted, irresponsible drivers. That’s why we all have to pay such high prices for car insurance. These are some of the misadventures of those drivers that annoy the rest of us.

Still not sure what the “right” balance is. Too much filtering hides things. Too little feels noisy. Experiment continues.

If you’re curious what the AI has been surfacing lately: tu.miami

Every time I tweak the prompts, the personality of the results changes. It’s weirdly fascinating.

Discovery in Miami is messy. That’s basically why I started building tu.miami.

Sometimes the AI pulls events I wouldn’t personally go to. That’s kind of the point. It’s not my taste. It’s broader than that.

One thing I’m noticing: AI is really good at finding *everything*. It’s harder to make “everything” feel useful.

I’ve been treating tu.miami like a small experiment. Some weeks the AI surfaces cool stuff. Some weeks it surprises me. Still figuring out the edges.

If you’re curious what I’ve been working on: tu.miami

This month has been useful.

Feedback beats assumptions.

For Miami specifically, I’m testing something small: tu.miami Still early.

I like projects that evolve in public.

Tools should help you decide, not scroll longer.

When it comes to Miami plans, what do you actually trust most?

Curation will always be subjective. That’s part of the appeal.

I’ve been experimenting with a small project around event discovery. Mostly learning.

Some cities are easier to explore than others.