About

I build small interactive models of how systems work, and I publish the one condition that makes each of them fail.

Most explanations of systems are abstractions stacked on abstractions: feedback loops, queues, leverage points. You can nod along to all of it and still not be able to predict anything. A mechanism you can grab and push until it breaks is different. The breaking is the part that sticks.

So every piece here is the same three things: one claim, one model you can run in the browser, and one falsifiable line — this breaks if X. If the break condition is wrong, the whole piece is wrong, and I would rather find that out in the replies than not find it out at all.

The same rule usually shows up in more than one place. A queue in a coffee shop, a request queue on a server and a hiring pipeline are the same object wearing different clothes. When I can find a rule in science, in technology and in business at once, that is the one worth building.

What you get

One mechanism a week, on Friday, by email: the claim, the model, the break condition, and what I got wrong the week before. That last part is not decoration — it is where most of the learning happens.

— Gerald Jamisco, @GeraldJermisco