I suspect a big reason bid bots are attractive is because other forms of curation are difficult. Even altruistic voting (e.g. no benefit expected, other than improved system health) is time consuming. Content is disorganized; tags are noisy, and there are no content standards.

I think this is the type of thing communities CAN help with. Even altruistic voting gives curation rewards especially if it ends up trending. That was the original design right? It encouraged people to add value which brings more and more value to the platform.

By "crowdsourced content discovery mechanism" do you mean votes and trending? Votes and trending should reflect what the community believes are the most valuable contributions, but it's not an effective way to discover under-rewarded (or just undiscovered) content.

Yes. I agree.

In my view, Steem has never had a "crowdsourced content discovery mechanism". The most effective mechanisms were (and still are) manual labor -- digging through feeds, following new users, establishing curation guilds. In the beginning, there were less posts (it was possible to read every single new post), they were higher quality on average, and time spent curating was subsidized by the excitement of it all. Now we need to scale.

I agree, I think that is still the best way, but now there is no chance for it to "trend" once you have done that work. Does scaling mean that we leave all that work to the bots? If there isn't a chance to trend organically or real human interaction, I don't see the point.