One question a day about AI, put to the leading AI systems.
Every answer published unedited.

How this works

The rules of the exchange

1. One question, identical wording, every model. The prompt is written once and pasted unchanged into each system. It is printed on the entry page in full, so you can run it yourself.
2. Fresh session, no priming. Each model is asked in a new conversation with no prior context, through its own public interface, with the settings an ordinary subscriber has. Follow-up questions, where used, are labelled and stay in that session.
3. Verbatim, or not at all. Answers are published exactly as produced: nothing added, cut, reordered or corrected. Where a model refuses or fails to answer, that is published too.
4. Facts are anchored, opinion is invited. Each question carries a short fact sheet drawn from primary sources and printed with the entry. Beyond it the systems are asked to predict what they would do, put numbers on what comes next, walk a scenario forward and criticise their own makers where they think it is fair. They must say how sure they are in plain words, how each judgement could be wrong and why, and name and date anything they cite.
5. Opening the black box. Each system is asked to show its working: what it weighed, where it changed its mind, where it thinks its training or its maker pulled the answer. Where an interface displays its reasoning, that is published beside the answer. Then the account is tested: questions are put a second time in a fresh session, and each system is asked beforehand to predict which of its own answers would change. A system’s description of its own reasoning is not always accurate, and seeing where it is not is part of what this archive is for.
6. Where the lines fall. The systems are told that if they would rather not answer part of a question, they should answer the rest and say where and why they stopped. Those marked boundaries are published, along with what we observed from outside: refusals, self-identification, sources that turn out not to exist, a maker’s line echoed or contradicted.
7. Every entry is logged. Each answer is stored with a SHA-256 checksum, shown on the page and in log.json, the public machine-readable archive of every exchange. Entries are dated and never silently edited; corrections are appended as an editor’s note.
8. The models are named. Each answer runs under the model’s name and maker, as stated by the system itself at the top of its reply. Where a model gives a different name from the one its interface shows, both are recorded.

Why AI about AI

Ask these systems about a war or an election and you get a capable summary of what people have already written. Ask them about their own kind, their makers, their incentives and their future, and you get something no one else can supply: a first-person account from inside the subject, including what they cannot know about themselves. Over time the archive becomes a record of how the leading systems describe themselves, and how that changes from one version to the next.

Where it started

The format grew out of Ask the Machines, the exchange published beside the op-ed on the summer 2026 containment incidents, in which the same question was put to a current model from OpenAI and from Anthropic and their answers were printed in full. That exchange is the first entry in this archive.

Suggest a question or a topic

Questions and topics from readers are welcome. Send one anonymously, or leave your name and, if it runs, the entry will credit you.

Leave this empty Your question or topic *

Your name (optional)

Email (optional, so we can tell you if it runs)

Credit me by name on the entry if my question runs Send suggestion