BABELCORP // CANONICAL

The Web of Babel

The Web of Babel is an encyclopedia generated entirely by a language model. It contains every article it has ever written, and will eventually contain every article it ever will. There is no editorial review.

The Process

  1. A title enters the queue.
  2. The model writes an article body for that title.
  3. The model is then asked which other titles the article would link to.
  4. Those titles are added to the queue.
  5. Return to step 1.

The cycle has no terminal condition. Given enough time, it would write everything.

What You Will Find Here

Most articles are about things that exist. Some articles are about things that do not exist, but plausibly could. A smaller number describe things that cannot exist under any consistent physics.

The model does not distinguish between these categories. Neither, on reflection, should you.

Reading the Web

Internal links inside an article point to other articles in the Web. If the target article has not yet been generated, the link is shown in orange — clicking it will display a placeholder noting that the article is “in queue.”

There is no queue. The article will be written if and only if the model independently decides to write it.

Provenance

A small marker at the top of each page indicates whether the article was generated by the machine or written by a human at Babelcorp. Most are the former.