ChatGPT now chooses the shape of the answer too

When I ask an AI a question, getting an answer matters, but so does what I can do with it. A paragraph is sometimes enough. When comparing two options, I would rather see the same information side by side. I think the next test for conversational interfaces is knowing the difference.

Announced on October 7, GPT-6 and Intelligent UI bring text, visuals and interactive elements together in ChatGPT responses. I see this as giving the model responsibility for the presentation of an answer as well as its content.

Does a polished answer make things clearer?

For me, a useful interface should make the information easier to question. When comparing two subscription plans, for example, I care more about the assumptions behind the total cost than the color of a card. Being able to change an input and follow the result would help.

An attractive chart can also make incorrect information convincing. I would therefore be careful about treating visual polish as evidence of reliability. I would want calculations, estimates and interpretations to remain distinguishable. These are my product expectations, rather than a promised accuracy guarantee.

Every question does not need another screen

I would not want to tap several buttons to discover the answer to a simple question. Interaction has value when it improves understanding. If it multiplies choices and makes a decision harder, presentation has overtaken the task.

To assess this experience, I would start with a small job: understanding two options, editing a plan, or changing the inputs of a calculation. If I need fewer follow-up explanations, the interface is doing useful work.

What interests me is the possibility of each answer finding a form that helps with the immediate task. The number of things moving on the screen would tell me much less.

The cover is an AI-generated illustrative scene. It does not show the actual ChatGPT interface.