Personally I think that using words to identify geographic locations can be useful, but the main value comes if it’s free and widely adopted which what3words definitely isn’t.
Lots of people in these discussions focus on how straightforward the concept is (it is - I even made my own for fun https://wherewords.id ), or how there are many similar systems (actually, most commonly used similar systems use alphanumerics or some other variant that isn’t as easy to remember / communicate).
I think the core concept is to create a data transport system that can go via humans (either hearing and speech or seeing and writing), and can reasonably easily encode fairly complex data without too many errors. For that we need a good wordlist. Creating a good wordlist is surprisingly hard. I tried using the BIP39 wordlist for a while, but that has a number of problems (not least that you can create some pretty offensive phrases with it).
I think a shorter wordlist and longer phrases would work better. People can remember nonsense phrases pretty well!
What could be fun is generating sentences using noun phrases and verb phrases.
Replacing words with emojis: https://what3emojis.com/
Emojis don’t really solve the computer -> human -> computer communications problem, so they’re playing in a different space with all the alphanumeric codes, and other things of that sort.
Having said that, I’ve started using emojis a bit like a hash - I generate a ‘checkmoji’ for every location, so you can quickly have a second check that you have the same location as the other person. In my password generator, I use emojis to create an emoji sentence for the site, the settings, and the master password. I have a small list of ‘nouns’ and ‘verb’ emojis.
Agree, what3emojis is mostly satire!
Seems kinda like the Maidenhead Locator System.
The biggest issue I have with What3Words is that it’s based on the language with some bizarre pronunciation rules. That is, for non native speakers, correctly pronouncing English words is quite hard. Thus, it has a limited use globally, where majority of people have no idea how to pronounce the coordinates. Even worse, I could see how a british person could be completely misunderstood anywhere on Balkans.
Lojban might be a better specified language, but it fails at the most important job of a language: unite people from disparate places and walks of life.
English does a pretty good job of that.
One of the analyses showed that it doesn’t control for homophones properly, so even if you are a native speaker then there are words that give nearby locations that can sound too similar. It doesn’t matter if the words are pronounced differently by folks who are native speakers of another language, as long as they’re not pronounced in a way that can be confused easily with another word. Unfortunately, if fails spectacularly on this metric.