AI can now produce a working prototype, a whole app or a comparison of products in minutes. Three pieces argue that the result is worth trusting only when something was settled first or can be checked.
Jakob Nielsen (19 Sep · 01), a usability researcher who writes on his site UX Tigers, ranked the 100 design methods that teams use most, counting each method's cost twice against its benefits. AI prototyping, in which an AI model turns a written description into working software, came first. Nielsen warns that a demo looks finished while it leaves out error handling, permissions and performance. In his words, "a working demo argues exactly as hard for a bad idea as for a good one", so a demo is weak evidence. Cheap methods for deciding what to build, such as a one-sentence problem statement, fill most of the rest of his top ten.
Aha! (19 Sep · 02), a company that makes software for planning products, has built Aha! Builder, an AI tool with which product managers can create apps without engineers. Teresa Torres, who hosts the podcast Just Now Possible, interviewed its chief executive, its chief technology officer and the product's manager. The builder works in fixed stages, from a design system to a prototype to a working application, and uses ready-made parts for sign-in, databases and email. The team says the fixed stages give better results than one open-ended chat prompt. This is a company describing its own product, and only the show notes are open to read.
Jakob Nielsen (19 Sep · 03) also wrote a guide to comparison tables, which set products side by side so that shoppers can judge them without memorising every detail. Shoppers distrust the many tables made by sellers that work as advertising, with a checkmark in every cell of the seller's column. AI assistants now build comparison tables of their own, and Nielsen notes that "a wall of checkmarks gives them precious little to work with". He advises specific values with a source and a date, each cell making sense on its own.
In each piece, AI makes the visible result cheap: the prototype, the app or the table. What makes the result worth trusting is settled beforehand or can be checked afterwards: a written problem, parts that do not change between builds, and values with a source.
Nielsen's scores are his own judgements, and Aha! is describing a product it sells. None of the three pieces reports what happened when a team judged an AI prototype against a problem written in advance.
Jakob Nielsen ranked 100 design methods by value and put AI prototyping first, while warning that a working demo argues as hard for a bad idea as a good one.
Aha!, a maker of product-planning software, built an AI app builder for product managers that creates a design system first and uses ready-made parts for sign-in and data.
Jakob Nielsen argues that comparison tables exist to help people decide, and that AI assistants now build their own, so tables need specific, sourced values in place of checkmarks.