Native data models
Models that know the domain, not the words about the domain.
LMM Technologies builds native data models: models trained on a domain’s own data and prompted with it, not with words about it. Human movement is the first.
The bet is that movement carries structure the way text does, and that a model trained directly on it learns that structure because nothing less survives the objective. A model trained on descriptions of motion knows the language of motion. A model trained on motion has no notation to hide behind.
Today that means a forecaster that watches a few seconds of a person moving and streams what comes next, and a written form for movement, so that motion can be recorded, searched, and compared without keeping the video. A clip can now be written down and read back, and a model that reads the writing continues it.
The first place this matters is remote musculoskeletal care, where a clinician needs to see how someone moves without the footage ever leaving the room. Where the work stands.
Writing
- Essay Someone Turned It On A large language model is a file. It never starts. On where accountability lands when the tool can’t be fired, sued, or jailed.
- Essay Ron Hardy Doesn’t Prompt in Text The mix is the prompt. On native data, symbolic control planes, and why the most interesting models won’t be steered in words.
- Note Machine learning, or manufactured instinct? Why learning may be the wrong word, and what it means to build instinct first, deliberation second.
- Note The second half of the architecture If today’s models are manufactured instinct, the part still missing is learning, and the measure of it is how little data it takes.
- Aside Look Closer For sixty years we’ve searched for a technology and called it a search for life.
- Aside Too Fleeting and Too Far Organic molecules are everywhere, and that’s the bad news.
- Aside Snapshots of a Universe That No Longer Exists On quasars, obituaries, and why “now” doesn’t reach that far.
- Aside What If Life Only Happened Once? On why a universe full of ingredients can still be silent.
- Aside Preparing for the Data Apocalypse On building an offline knowledge archive for when “just Google it” stops working.