I’m not sure but I think it’s probably because they are too big to care and using friendly URLs on a very large scale might be harder as they would probably end up being very long and difficult to organise.
However I think youtube does have “friendly” URLs as well. For example the one you mention is also found at https://youtu.be/qJcz5wX9BsM
This video offers some insight… Not saying this is 100% the reasoning behind it but offers a different perspective of “IDs” for video URLs
My Take away from it was “there is many hours of video uploaded every minute”… meaning I think it would be a nightmare, since a lot of people would use the same terms for video titles, they would end up “ugly” fast even if the first were “pretty/friendly”.
YouTube’s URL is not a “token”. It’s a perfectly “friendly” URL. It’s just not SEO’d.
A “token” is given to one person and won’t work for other people. That’s done as a form of rights-management / copyright protection.
URLs like YouTube’s go back to the early days of the internet - there would be one script that served something and it would take a certain number of parameters, so you just added the parameters onto the end ?param1=val1¶m2=val2. The idea of what you think of as “friendly” is simply an SEO construct done to make Google happy. Well, Google doesn’t feel the need to make itself happy, so…