You can listen to the full conversation here: slate.com/podcasts/wha...

I think in the early days of Wikipedia there was an idea that it could be written like older encyclopedias. Going to a "everything must be cited" system in the English language version has been a large change.

I think a generation of students heard "don't cite Wikipedia as a source" and internalized "Wikipedia is unreliable" rather than "encyclopedias are secondary sources, go cite the actual primary source".

I used to love reading encyclopedias in elementary school. They were my own personal gifted and talented program.

To be fair, there likely has not been an encyclopedia written within the last two hundred years or so that has not been composed principally by stealing copy (to include, in many cases, whole articles) from competing extant encyclopedias. Wikipedia was originally built atop the 11th ed. Britannica.

Hopefully the new competitor will consider this in its future versions, either through hyperlinking or "See Also" suggestions embedded in each and every article. Otherwise, the pedagogical strengths of Wikipedia and even Britannica and Brockhaus alone will continue to vastly outmatch xAI's product.

In many ways what makes an encyclopedia recognizable as such in the modern West is its capacity for illustrating the interconnectedness of human knowledge -- regardless of the political bent of its content. Without this, it is merely a titanic compilation of topical essays.