Personally, I've started seeing algorithmic discoverability as an anti-feature that mainly serves the interests of the attention economy.
For my part, I want more long-form, thoughtful articles that offer an enriching read. Not only have we got a good decade or two's worth of experience showing that no algorithmic discovery system ever favors that kind of thing over content that's morally equivalent to Five Minute Crafts, but I've got a couple decades' worth of experience telling me that, since long-form bloggers tend to link each other quite a lot, never needed an algorithm to help me with discovery in the first place.
If I want anything, it's algorithmic filtering: Take the feeds I'm already subscribed to, and filter out the stuff that I tend to skip over without reading. Because my blog feed already delivers me a couple hours' worth of reading a day, so I can afford to be choosy.
Yeah, this is very true. I may have overstated my initial idea. When I think more about it, maybe what I'm looking for is handcrafted curation, but that let's anyone share their curation speccs and discover relevant things to curate from. To me, that suggests some kind of algorithmic component, maybe in the form of search, voting, etc., but I wouldn't necessarily want the whole thing driven by either machine or mob.
Like stitchfix, but for magazine articles, blog posts and books. I might be cynical, but I think humans do this much better than any algorithm I've ever seen. :)
For myself, I usually read in two "phases". I collect material from sources that are usually high quality for me, and do some superficial skimming to trim out the content I don't particularly care for.
Later, when I'm more "in the mood" or have a long period of time, I pop() and read(). The stack gets pretty large at times, but it only takes a couple days of vacation to blow through most of it. If I pass over something in the stack enough times, it gets free()'d.
I don't know if I'd trust an algorithm to do either of these stages for me. I'd trust an algorithm provide input to phase (1), but never to replace phase (1).
> If I want anything, it's algorithmic filtering: Take the feeds I'm already subscribed to, and filter out the stuff that I tend to skip over without reading. Because my blog feed already delivers me a couple hours' worth of reading a day, so I can afford to be choosy.
For my part, I want more long-form, thoughtful articles that offer an enriching read. Not only have we got a good decade or two's worth of experience showing that no algorithmic discovery system ever favors that kind of thing over content that's morally equivalent to Five Minute Crafts, but I've got a couple decades' worth of experience telling me that, since long-form bloggers tend to link each other quite a lot, never needed an algorithm to help me with discovery in the first place.
If I want anything, it's algorithmic filtering: Take the feeds I'm already subscribed to, and filter out the stuff that I tend to skip over without reading. Because my blog feed already delivers me a couple hours' worth of reading a day, so I can afford to be choosy.