I've stayed with Hacker news because a lot of the content is of high quality, which appeals to me. However, I use hckrnews.com/ as an interface since I prefer seeing submissions in temporal order with the score clearly visible.

Regarding other content aggregator options: - I stopped looking up content on Facebook because of doomscrolling and, generally, bad content. - Twitter is all right, but the content is usually not deep. - I generally think it's a waste of time to visit news sites since important news will become visible to me through aggregators, peers, et cetera. - To find new research papers, I use Arxiv, recommendations from peers, and I participate in peer-reviews. Still, it can be time-consuming to explore too much.

Ideally, I want to dedicate some amount of time to explore. However, I also want a way to prioritize the content I'm receiving. Big bonus if the recommendations are interpretable.

You can think of it as a priority queue where my personal priorities are learned. Time should have some influence since, at some point, old news is not interesting anymore.

Is anyone building a product like this? Essentially, a smart filter where the user controls the input into the filter (it can be anything consumable, really). Furthermore, the user should be able to give some feedback on whether he is happy with the recommendations.

Is there any hope that we will ever build something like that which provides good recommendations while also having the right balance between exploration and exploitation?

What is your strategy today? Are you aware of its pros and cons?