Watch less, keep more: a system for remembering what you watch

Almost nobody has a shortage of things to watch. What people have is a shortage of anything left over afterwards. You watch a good talk, you feel like you learned something, and a week later the only thing you can reproduce is that it was good.

This isn't a memory problem. It's a system problem, and it has a mechanical fix.

Why watching leaves nothing behind

Three things work against you.

Video moves at the speaker's pace. Reading is self paced, so you slow down on the hard part and speed up on the filler. Video does the opposite: it dwells on the filler and rushes the hard part, because the speaker can't see your face.

There's nothing to mark. Reading gives you a highlighter. Video gives you a scrub bar. The act of marking something is most of what makes it stick, and video removes it.

The feeling of understanding isn't understanding. Watching someone explain something clearly produces a strong sense of comprehension that fades within days. This is well established and everyone ignores it, including people who know about it.

None of these get fixed by watching more carefully. They get fixed by changing what happens after.

The system, in four parts

This isn't elaborate. Elaborate systems get abandoned in week three.

One: convert everything you took seriously into text. Not everything you watched. Everything you took seriously, which is a much shorter list. Text is skimmable, searchable and quotable, and video is none of those.

Two: reduce to a short note the same day. Three to five claims in full sentences, the lines worth quoting, one line about why you cared. Two hundred words. If you can't write two hundred words about it, you didn't get anything out of it, which is useful information on its own.

Three: keep it all in one searchable place. The single most common failure is having four places. Notes in one app, transcripts in a folder, quotes in a phone note, links in a browser tab group that gets closed. One place, or it doesn't work.

Four: search it before you search the internet. This is the step that turns the archive from a graveyard into a tool. When a question comes up, look at what you already have first. If you never do this, everything above was filing for its own sake.

What each part costs

Converting is the only part with a real cost. Drop charges one credit per started minute, so a forty minute talk is forty credits and a ten minute video is ten. Credits don't expire. The first run on a new account is free.

The note is five minutes of your own time and can't be outsourced, because the reaction line is the part that makes it findable and only you can write it.

Keeping and searching is free once the first two are in place.

The part that compounds

A single note is worth very little. A hundred notes on adjacent subjects is worth a lot, because the value moves from recall to comparison. You start noticing that three different people made the same claim, or that two of them contradict each other and you now know which one turned out to be right.

That only works if you can ask across the whole collection rather than opening entries one at a time. In Drop, the library holds every saved drop with its transcript, ideas and quotes, and you can put a question to the whole thing and get an answer with the sources attached. That's the difference between an archive and a filing cabinet.

The rules that keep it alive

Don't save everything. The archive dies from volume. If everything is in it, searching it returns noise and you stop searching it. Save what you'd defend saving.

Don't save what you haven't reduced. A transcript with no note is a transcript you won't read. Reduce it or drop it.

Don't build categories. Folders, tags, hierarchies, all of it. You won't maintain them and search does the job. Spend the effort on writing better notes instead.

Do the note the same day. Tomorrow you've the transcript and not the reaction, and the reaction was the point.

What to do with the good ones

Some of what you keep is worth publishing rather than filing. The material is already in the right shape: claims, quotes, and your own angle. Turning it into something readable is a separate craft, covered in from transcript to post.

Some of what you keep is worth making public as a summary others can use, which is the idea behind the public knowledge ledger.

Start here

Take the last thing you watched that you'd recommend to somebody. Run it through, read the ideas, write two hundred words, save it. That's the entire system in one pass. If the note is still useful to you in a month, keep going. If it isn't, the note was bad, not the system, and the fix is in the note.