Posts for: #Software

Application Paradigms enabled by AI

Before AI, most apps chased the largest possible market, the only scale where the economics worked. AI removed that constraint. Software that took months now takes hours, and what was intractable before will start to show up.

Personal apps

Something only you care about, software built for yourself by yourself.

A script that texts you when your favorite coffee shop drops its prices. A dashboard that tracks your running routes against the weather.

[Read more]

The Cambrian Explosion of Apps

Something is happening to software that happened to life 540 million years ago.

The original explosion

Before the Cambrian Explosion, life was simple. Single cells, some basic multicellular stuff, not much else. Then, in a geologically short window, nearly every major animal body plan that exists today appeared.

The organisms were already there, what changed was the environment. Oxygen rose enough to support complex metabolisms, and predation emerged, creating pressure that drove innovation at a pace the planet had never seen.

[Read more]

Allow Spontaneity

A conversation that cannot surprise you is not a conversation, it is a script. The more structure you add, the less room there is for anything unexpected.

Three times structure got in the way instead of opening things up:

A. Philosophy meetup

I saw a poster for a philosophy meetup about a philosopher I liked, so I went. I expected people comparing interpretations, recommending books, following ideas wherever they went.

[Read more]

Cosmetic Changes shouldn’t block Software Development

The most valuable thing a software team can do is ship. Users care about functionality, about experience, not about your SDLC, your agile rituals, or your naming conventions.

SDLC exists because collaboration needs structure, and structure is overhead even when necessary. Every round of the cycle takes time. Write, review, test, merge, deploy. SDLC slows you down by definition.

A healthy ratio is 80% real work, 20% overhead from the SDLC. I have seen it inverted. Developers stuck in the procedure, not the product. Five times the capex for the same result, and developer time is already the biggest line item in most companies. That is the spend, but the real cost is the release you never made.

[Read more]