I'm not sure that raw activity level is a good proxy for the amnount of genuine expertise in the topic that is present.
Some topics - of which AI is quite possibly one - are simply more niche than others, and therefore naturally will have less askers, less answerers and less experts.
Perhaps a better metric than raw activity level would be proportion of questions without an accepted answer? The tumbleweed sites that "don't work" would be characterised by many unanswered questions.
Some topics - of which AI is quite possibly one - are simply more niche than others, and therefore naturally will have less askers, less answerers and less experts.
Perhaps a better metric than raw activity level would be proportion of questions without an accepted answer? The tumbleweed sites that "don't work" would be characterised by many unanswered questions.