OpenAI announced Intelligent UI in ChatGPT on October 7: answers can contain interactive elements. Rollout is gradual, with Free and Go scheduled to start on October 8. That does not establish activation on every account. This announcement does not change Work or Codex.

The change is more than cosmetic. Readers can try another input and see whether an explanation still holds. That guarantees neither correct mathematics nor trustworthy data. A chart can look more authoritative than prose, making content checks at least as important.

Our realistic application scenario is a small learning exercise: a student changes one parameter, predicts the result, then reveals the explanation. Comparing two budget scenarios or exploring a simple decision step by step might work similarly. These are possible applications, not features we tested on a particular account.

Practical use needs tests of edge inputs, units, mobile controls and accessibility. For financial or health decisions, an interactive card does not replace a verified calculation or qualified professional. A useful application must expose its assumptions rather than hide uncertainty behind an elegant button.

Our optimistic editorial estimate is 1–3 weeks for a small, manually checked learning exercise once the relevant account receives access. We have no basis for dating reliable general deployment. The interesting shift would be from reading a finished answer to experimenting with one's own question.