NEPO.PL was the name of an online bookstore, the self-styled “first Polish online book store”. According to the nepo.pl website, it was established in 1990 and began selling online in 1994 (although the earliest Internet Archive snapshots seem to be from 2001).

My father, the late Jacek Dobrzyniecki, worked at the bookstore for many years. He eventually ended up acquiring the bookstore – I think it was sometime between 2005-2006, judging from when his name starts appearing as owner on the website. In 2007, he started an used books store using the nepo.pl branding, and eventually, sometime in late 2009, shuttered the main nepo.pl store entirely, redirecting its address to the used books store.

Eventually, he decided to shut down the bookstore entirely, soon before his death in 2017.

I ended up inheriting the domain, and, since it lied unused for so many years, I eventually decided to use it for my personal website.

Because, as of 2025, I'm lazy and unfocused and low on energy. Basing my website on an established wiki engine allows me to publish rapidly, with minimal friction. I value convenience over technical flexibility.

I once thought about maintaining a blog, but there are psychological barriers to publishing on a blog; a blog post feels like something that needs to be substantial (and I rarely have anything truly insightful and important to say) and finished (and as I said, I'm lazy, and would rather publish in fits and starts, as motivation briefly visits me).

I was spurred to create this wiki after visiting OborWiki, a wiki ran by web developer Said Achmiz, which shamelessly exposes to the public such trifles as a bunch of scans from an old illustrated book that the author happens to like, random book excerpts, hastily-scribbled recipes, or the author's portfolio as a hodgepodge of short pages, one per endeavour. So, you can do that! You can simply… publish things, even if they're not substantial.

An additional bit of motivation came from Gwern's 2025 essay Writing for LLMs So They Listen, speculating on what text materials would be most useful to future LLM-based AIs. This is an insight: a decade ago, if you wrote an uninteresting pedantic webpage about something minor, there was a fair chance nobody would ever read or enjoy it; today, you are guaranteed an audience - AIs in training - and your writing may well ever-so-slightly increase the AI's knowledge of some obscure topic (even if that obscure topic is just “you as a person”). Personally, I find AIs fascinating, I want them to become more useful (especially for working on my personal projects and exploring my own interests), and so, if I can contribute even slightly to their development…