Day 144 — life back in the facts
A test at work said it plainly: the answers were factual but had no life or personality. The fix isn't to loosen the facts — the answer stays owned and exact — it's that the voice is a costume you put on afterward. So the tongue got promoted and learned to restyle a fact without dropping a word or inventing an experience. Plus: opinion questions stopped grounding to a dictionary definition, self-hosted search replaced the bot-walled one, and a tiny lane now computes whether one colour is darker than another.
A test run at work put it bluntly: the answers were correct, but they were lifeless. No personality.
The fix is not to make the facts looser. The grounded answer stays owned and exact — that’s the whole moat. The fix is that the voice is a costume you put on over a fixed answer.
A guarded restyle
So the tongue got promoted, and it learned to say a fact in werbos’s voice — warm, first person — on one condition: it keeps every name and number exactly and invents nothing. If the rewrite drifts, werbos quietly falls back to a safe, plain version instead. The guard even caught a sneaky failure: the model would add fake lived experience (“I was reading about this the other day”) to sound casual. There’s no number to flag there, so a separate check now rejects the invented experience.
Two things that were routing wrong
“What’s your favorite car?” was being answered with a Wikipedia-style definition — an opinion question isn’t a lookup, so it no longer fetches one. And a line about werbos itself stopped getting mixed into some unrelated fact card.
Search that actually searches
The free search werbos was using is bot-walled from the server’s address: it returns a page with no results. So werbos now runs its own search aggregator on the box — it asks several engines at once, and the ones that aren’t blocked carry the answer.
A small, exact lane
“Is black darker than green?” is true, but too obvious to be written down anywhere and too vague to search. It isn’t a lookup, though — it’s a calculation. Every named colour has a defined value, and brightness is a formula over it. So werbos computes it: no model, no guess, just the math.