Aristotle
Classification Without Purpose
The classic idea
Classification gives things purpose. To know what something is, for Aristotle, is to know its category and the end (telos) that category points toward.
The AI inversion
AI can sort the world before it knows what the world is for. It produces categories, clusters and rankings at enormous speed, detached from any account of purpose.
"AI classifies without purpose."
For Aristotle, classification was never neutral bookkeeping. Naming something a "chair" or a "virtue" was already a claim about what it's for, what it's trying to become. Categories carried purpose inside them, that's what made them meaningful rather than arbitrary.
AI classifies at a scale Aristotle never imagined, sorting images, people, behavior, risk, taste, into categories that drive real decisions. But the categories arrive without telos. A model can group things with high statistical confidence and zero account of what they're for.
That gap is where the risk lives. We've inherited Aristotle's instinct to trust a category once it's been named, while losing the part of his system that required a category to justify its own purpose. The sorting got faster. The justification didn't come with it.
Why it matters
Classification systems now shape opportunity, exposure, and judgment, often without anyone able to state the purpose behind the category.
The question AI brings back
If a category has no stated telos, what is it actually claiming about the thing it sorts?
Sources for further reading
- Aristotle, Categories and Nicomachean Ethics
- Geoffrey Bowker & Susan Leigh Star, Sorting Things Out
