01
The Model as a Reasoning Substrate
The shift was not autocomplete. It was reasoning over context.
Large language models mattered because they could take context and produce useful thought. Once that was true, the next step was obvious: not just answering questions, but deciding what to do next.
