I save a pretty unreasonable number of links.

Most of them come from Twitter or HN, but sometimes an Instagram post or (on an off day) a LinkedIn post will sneak in. I’ll be scrolling around when I see a post that looks interesting, realize I don’t have time to give it the attention it deserves right now, and toss it into whatever read-it-later app I’m using. Then I keep scrolling, happy in the knowledge that I have successfully prevented this particular nugget of internet wisdom from disappearing forever.

Thing is, I’m most definitely not actually going to read it later. I’ve filled Pocket, Instapaper, Readwise Reader, Treechat, browser bookmark folders, and a few homegrown systems with things I had every intention of coming back to. For the most part I did not come back to them.

Saving the link still feels good though! There’s a little open loop created when you find something interesting at the wrong time: you don’t want to stop and read six thousand words on your phone, but if you keep scrolling there’s a good chance you’ll never find it again. Clipping it closes the loop. It’s safe, and future me can deal with it if he so chooses.

A few weeks ago I built something specifically for that gap between “this looks interesting” and deciding it’s actually worth my time. It’s currently (and hopefully temporarily) called FeedThing.

FeedThing is a sort of single-player social network. It has posts, feeds, quote-posting, search, and most of the other bits of interface you’d expect, except I’m the only user. It’s kind of like a social network where the poasters only poast my orphaned and temporarily homeless web clippings that past-me found interesting.

Some of the things that it does nicely: when I clip something, FeedThing saves the complete contents, not just the URL, so if the source disappears later my copy doesn’t go with it. It also gives each post a decent way to be consumed: articles get an Instapaper-style readability view, social media posts still look like social media posts, and videos play directly in the app. Everything, including the videos, gets a big fat “summarize” button, which can quickly create an AI-generated summary of pretty much any text or video (thanks to yt-dlp).

The inbox is the important bit. Most read-it-later apps are built around a pretty straightforward loop: save an article, then come back later and read it. Readwise Reader is my favorite of the current crop and does a much better job with highlights and resurfacing, but you can still feel that underlying assumption: it thinks you’re going to sit down and read something.

Thing is, when I hit save I haven’t actually decided if something is worth reading. I’ve decided that it looks interesting enough that I want to make that decision later. A headline and somebody else’s enthusiastic post are enough to get something onto the pile; whether it deserves my actual time is future me’s decision.

That pile increasingly isn’t even made of articles. It’s tweets, threads, Hacker News comments, videos, and various other forms of internet debris, so a text-heavy chronological list of article titles is not a great triage surface for it. A visual, interactive feed has a much better chance of competing with the infinite supply of new, shiny, animated engagement bait waiting for me back on Twitter; a list of old-ass articles does not. FeedThing treats the pile as an inbox, so I’m allowed to dump things in recklessly and sort them out once I have more information.

If the thing was a waste of time I archive it. If it’s good, I promote it into one of my categorized feeds. Sometimes it’s worth opening in the readability view and saving a few highlights, but reading the summary and deciding I’m done with it is also a perfectly successful outcome. I found something that looked interesting, FeedThing told me what was in it, and now I can let it go without adding another unread article to the guilt pile.

Archive doesn’t mean delete here. FeedThing keeps the full copy and search covers all of it, including things I dismissed from the inbox. That matters when I half-remember something months later but not its title, source, or where I put it. Twitter search is Twitter search (read: hot garbage), Google may or may not find the same thing again, and browser bookmarks leave me searching through a thousand links for one whose destination may no longer exist. FeedThing searches the copy I actually saved.

The categorized feeds are where the winners of the inbox go. I can scroll through the things I kept, add my own posts, and quote an old item into a new thought without having to maintain a graph or periodically reorganize a folder tree. It feels much closer to using a social network than maintaining a personal knowledge database, which is probably why I’ve kept using it.

After spending the last four years building Treechat, I was already pretty convinced that posts, feeds, and quote-posting are just a nicer interface for this stuff than folders and graphs; FeedThing is what happens when you keep that interface and remove all the other users.

I still use Twitter and HN to find things because they’re good at that. FeedThing picks up after I hit save: it stores the actual thing, summarizes it whether it’s a post, an essay, or a video, and gives me a much nicer place to read or watch it if it survives triage. If it matters six months later, the full copy is still there and searchable.

I’m still saving far more than I’ll ever read, so it hasn’t cured the pack-rat part of my brain. It has made the behavior considerably less pointless, though. Twitter can keep doing discovery; FeedThing is where I decide what deserves my attention and what gets to stick around.

Working well so far. Ask me again when I’ve filled it up with ten thousand things.