In the discourse about social media, the term "algorithm" is exclusively used to refer to purposefully-maligned algorithms engineered to addict and abuse people. Nothing about any of the Fediverse services is designed this way because they're not chasing money or engagement, they're made to help people converse in a human way.
But I feel like it misses the point. What about a service where you can design and use your own "algorithms", and it's built into the platform?
Such a platform would have thousands of algorithms, but none of them designed for chasing money or engagement, just different preferences. But Mastodon could still claim "We don't use The Algorithm and is therefore better than other places" while a platform with custom user-owned algorithms could get the best of both worlds.
A very quick peruse of the Mastodon issue tracker came up with information about this on the ActivityPub level (albeit in an old toot): https://mastodon.social/@reiver/113668493283013849 and someone kindly rounded up similar feature requests here: https://github.com/mastodon/mastodon/issues/33098#issuecomme...
I like the idea of it not being related to a platform or implementer, but baked into the spec ala a feature of ActivityPub.
So it's literally just "bad algorithms" (the ones other platforms make) and "good algorithms" (the good algorithms good platforms make, like us). Which is kind of literally how it is, there are good ones and bad ones, except both of these kinds of platforms employ "bad" engagement driving discovery algorithms, so it's really just 'us vs them'. The trending and news algorithms are literally just driving engagement and discovery, and top hashtags feed is proudly clamoring how much engagement there is. Doesn't seem like they're not "chasing" it.
I think this is the part of Mastodon's code that calculates the Trending page: https://github.com/mastodon/mastodon/tree/main/app/models/tr...
These sorts of algorithms tend to promote posts or people that have recently been popular for the purpose of being useful to folk. On the likes of tiktok, facebook and twitter they are the culmination of very large sums of money and an ocean of professional psychological collaborators with the aim to purposefully harm and addict people, e.g. to manipulate public opinion and democracy, incite the suicide of transgender people and the perpetration of genocide. For money. I find it difficult to believe that you're arguing, in good faith, that the two types of "algorithm" have much in common.
I am not sure how it is "evil" showing recently-popular posts on a social media server's home page to logged-out people, and how that's pushing anything. It's not an agenda, it's not a series of posts that are picked because they are likely to addict and enrage people. I do suspect that there's some ragebait that shows up, because some people are still having to unlearn the indoctrination they're suffering from.
It's not "evil" to be showing an algo feed per se. But mastodon and a bunch of other platforms refer to algo feeds as "bad/evil" or something of the sort, market themselves as not having them, and yet thoroughly employ multiple algo feeds. Is that not just hypocritical? It looks glaringly dishonest. They could at least have some integrity to say "we don't like the yucky algorithms, but here we only have good™ algorithms", when that's literally what it is.
I feel like the wording needs a bit of rewording/rework. I agree chronological order facilitates better discussions, but just saying that "Mastodon lacks algorithms" doesn't really help people understand things better.