That depends on the nature of the problem. Solutions we've seen so far to problems of discoverability in massive amounts of content have been to use an algorithm of some sort to provide results either through searching, recommendations or both.
It seems like Mastadon or a 3rd party would have to come up with a novel, effective solution to the issue of content discover that no one else has found yet. It's not just about providing search results, but in general showing users what they might be interested. Right now that can be done via lists and directories that are still relatively small enough to browse a lot of it and get a general sense of a lot of what is there to choose from.
But maybe it doesn't have to solve that problem better than others have to avoid their pitfalls. The second link up in this thread is at least an open source search engine so the algorithm can be scrutinized. Even if they converge on similar content discovery solutions, as long as the search engines that pop up are open source then people can at least know how the results are weighted.
> It seems like Mastadon or a 3rd party would have to come up with a novel, effective solution to the issue of content discover that no one else has found yet
No ty!
Very happy with my home timeline and 'local' timeline, don't need anyone trying to get me addicted to something else.
While I certainly see the utility in truly global search / discovery systems (they're great!), all the big ones turn deeply abusive and manipulative in time, because they are all ranking systems of some kind, which can be gamed[1]. Nothing has ever fixed that issue, and I'm not sure it's possible to. Small communities can more reasonably deal with it by hand when necessary, so they can get closer to what they meant rather than what they said, which is what I think most people want/need.
[1]: "Closest text match" ranks in favor of both precise words and phrases, and commonly used ones (you can game it by being predictably-niche or by using the most used things). "Most recent" ranks heavily on frequency (game it by doing things a million times per day). "Page rank" is literally just a popularity contest, easily swayed by money and connections like all popularity. It doesn't matter what you're doing, you're promoting something and thus providing a way to control your promotion to determined parties.
It seems like Mastadon or a 3rd party would have to come up with a novel, effective solution to the issue of content discover that no one else has found yet. It's not just about providing search results, but in general showing users what they might be interested. Right now that can be done via lists and directories that are still relatively small enough to browse a lot of it and get a general sense of a lot of what is there to choose from.
But maybe it doesn't have to solve that problem better than others have to avoid their pitfalls. The second link up in this thread is at least an open source search engine so the algorithm can be scrutinized. Even if they converge on similar content discovery solutions, as long as the search engines that pop up are open source then people can at least know how the results are weighted.