1
0
mirror of https://github.com/osmarks/website synced 2025-09-03 11:17:55 +00:00
This commit is contained in:
osmarks
2025-04-12 14:32:50 +01:00
parent 71e8e39b3b
commit 2cd0718a70
6 changed files with 170 additions and 11 deletions

View File

@@ -18,6 +18,6 @@ We can possibly rescue this with clever mechanism design: specifically, a [Harbe
However, unlike with most other land, I think "improvements" on a word or brand name are zero-sum, so disincentivizing them is fine. It seems that "improvements" consist of advertising to make of people think of your brand in association with some context, at the expense of whatever you might think of instead. This is also why I think we should not dismiss attaching popular brands to common/short names as merely a way for language to tend towards increased compression: every such attachment increases ambiguity and [computational overhead](https://en.wikipedia.org/wiki/Hick%27s_law).
It would be nice to extend this beyond trademarks - people often consume cool words and phrases with bad or rapidly obsoleted ideas - but transaction costs and operational difficulties don't make this practical yet. I think it would be valid to have a similar mechanism in domain names, though the cost of switching a domain name is higher because of security implications and remote configuration files. See also a [related gwern proposal](https://gwern.net/harberger) for copyright inheritance.
It would be nice to extend this beyond trademarks - people often consume cool words and phrases with bad or rapidly obsoleted ideas - but transaction costs and operational difficulties don't make this practical yet. I think it would be valid to have a similar mechanism in domain names, though the cost of switching a domain name is higher because of security implications and remote configuration files.
[^1]: Strictly, you can expand your character set, so to be precise we should be counting in terms of bits and not characters.