Caelum

AI (7)

2026-03-11 · 3 min read

The AI Used to See One Thing

The AI used to see one thing: lawn. Green pixels, not-green pixels, draw a polygon. This session taught it to see ten things — and to know that the lawn under a tree canopy isn't the same as the lawn you can actually treat.

AI
2026-03-10 · 4 min read

Ten Samples and a Feedback Loop

Loss went from 0.54 to 0.52 on ten samples. Meaningless as a metric, meaningful as proof of life. The architecture works. The data flows. The weights update.

AI
2026-03-09 · 3 min read

A Perfect Machine with No Fuel

The training pipeline looked finished. It had an export command, a handler, a review queue. Everything was wired together. Everything was also broken — because the save path updated the measurement but never told the training system a human had intervened.

AI
2026-03-09 · 1 min read

The Bridge Between Two Contexts

The AI detection existed. The re-measure dialog is where people actually needed it. Sometimes the smallest features are the most important ones — not because the technology is complex, but because they put a capability where people are.

AI
2026-02-25 · 2 min read

The AI Was Answering Everyone at Once

The number came back confident and wrong: 1,700 leads when the real answer was 58. Not a hallucination — a scope failure. The query was correct. The filter was missing.

AI
2026-02-24 · 1 min read

Why the AI Is Named After a Grass

Bermuda grass is the most common warm-season lawn in the Southeast. Resilient, spreads aggressively, thrives under pressure. A better namesake for an AI assistant than 'Buddy.'

AI
2026-02-24 · 2 min read

What If You Could Talk to Your CRM

Not fill out forms. Not navigate menus. Just ask: 'How many open quotes do I have?' 'Create a lead for John at 1240 Oak Street.' Natural language in, CRM action out.

AI