[SomewhatLikely]

An algorithm is desperately needed though. Sources have varying publishing rates and quality. Aggregating across them and culling is crucial. For the most part I've thrown up my hands and delegated this job to hacker news and Reddit subscriptions, but it's still not well personalized.

[gaius]

An algorithm is desperately needed though

An algorithm that weaponises jealousy, FOMO and outrage will always out-monetise a benign or no algorithm at all, over any financially viable timescale. The algo doesn’t “know” of course. It just knows that showing you X has historically made you spend longer on the site than Y.

The only way to win is not to play.

[aninhumer]

I don't think we need an "algorithm" just better RSS readers that let us filter things easily.

Or ideally more effort put into feeds by sources, so we get better curation of the feeds to start with.

[thescribe]

I do fine without an algorithm, everything appears at the rate it is published in chronological order, if I dislike a feed or discover I rarely read it I remove it.