Skip to content
Article UX Collective Oct 2026

UX Collective: Tesler's Law and complexity that moved out of the chat field

Tesler’s Law states that every application has an inherent amount of complexity that cannot be eliminated, only moved — either into the product or into the user’s experience. This short essay from UX Collective, published October 9, 2026, applies that principle to AI chat interfaces and finds it explains something designers have been noticing without good language for it.

What the article is about

The piece examines what happens to complexity when a conventional UI is replaced by a chat field. On the surface, the interface becomes simpler: instead of menus, filters, and forms, there is one text box. The argument is that complexity doesn’t disappear — it shifts downstream, to the point where the user must evaluate what the AI returned, determine whether it is correct, and decide what to do if it isn’t.

In a conventional UI, users make their most consequential decisions before the system acts. They configure settings, apply filters, choose options, and see a predictable outcome. In a chat-driven interface, the most important decisions come after the system acts — by reading, verifying, and correcting the output. The article calls this a move from pre-action complexity to post-action complexity.

Why this matters for designers

The practical implication is that reducing complexity in the input phase — creating a simpler text field — can increase the burden placed on users at the output phase, if no corresponding design effort goes into making verification easier. Designers who treat the chat field as the end of the design problem are leaving the harder part unaddressed.

The article points to interface patterns that help distribute the verification burden more fairly: confidence indicators, provenance links, editable outputs, and diff views that show what changed from a previous state. These patterns acknowledge that checking AI output is itself a design problem, not a user responsibility.

Who it is useful for

Product designers and PMs building AI-assisted tools, or anyone considering replacing a structured form with a chat interface. The piece is a short analytical read that provides a principled frame for evaluating where AI simplifies a user’s life and where it just relocates the hard part.