Philosophy
Why Linse Does Not Show Model Selection
Scope AGI 2 min read
Published
Expanded
Users shouldn't have to act like system administrators for AI models. We explain our product philosophy of automatic intelligent routing and why hiding complexity behind a single simple interface leads to better outcomes.
Start with the task
Writing a thoughtful message, comparing options, and making sense of a document are different tasks. We believe you should be able to express those needs directly, without first deciding which model family or set of technical parameters to use.
A model selector puts a system-design decision in front of a person who may simply want to finish a piece of work. Hiding that decision is an interface choice: the product takes responsibility for organizing the capabilities behind the conversation.
Keep complexity behind the interface
Automatic routing is part of how Linse approaches that responsibility. The aim is to connect a request with an appropriate capability while keeping one coherent interaction for the user.
This does not mean every request is identical or that every answer is equally certain. The visible result still needs to be understandable, reviewable, and useful for the task you described. Simplicity should reduce setup, not make an answer harder to question.
Make the result easier to judge
The questions that matter at the surface are practical: does this answer use the right context, can I inspect its sources when available, and can I take the next step? Those are more useful design targets than exposing every infrastructure choice.
Our product philosophy is to put those questions first. We will keep refining the experience around the work a person wants to accomplish.