,
10 tweets,
2 min read
Read on Twitter
1/ There’s a phenomenon I’m noticing more and more that I’d like to christen as “Algorithmic Regret.”
2/ It is “the feeling of regret one feels when something they have spent considerable time and effort on can now be done trivially by computers.” It’s FOHMO - Fear of Having Missed Out because you dedicated yourself to a difficult task instead of just...waiting
3/ First example: I spent YEARS organizing my iTunes library. Collecting songs from illicit hard drives. Manually correcting song names, album names, artist names. Meticulously compiling playlists. You know what’s coming...
4/ One day iTunes Match came out and offered to do all that effortlessly and for free. Then Spotify took off and even the playlists became obsolete. I wonder what would be possible if I’d dedicated those hundreds of hours to, say, learning an instrument instead
5/ Second example: I again spent years organizing my photos, meticulously organizing them into folders and tagging faces in iPhoto. Then Google Photos comes along and organizes them all chronologically, by event, auto-recognizing faces, & you don’t even need to hit import
6/ I wish I’d spent that time learning to take better photos, or learning to draw, or practicing graphic design. What distinguishes Algorithmic Regret is that it applies to purely clerical organizing work. It’s difficult to see “what you got out of the experience.” It’s zero sum
7/ Now what’s interesting is that, in the typical creative process of input > organize > output, Algorithmic Regret is starting in the middle. It can’t yet replace curation and collection on the input end, nor creation and expression on the output end
8/ The ends still require human judgment and decision making. But the organizational work in the middle, when you have all the data collected but still haven’t started synthesizing it, is where algorithms shine. They can detect and reorient around patterns effortlessly
9/ This is why I advocate for “just-in-time” productivity. You should leave things in raw form at input, and then bring them all the way to output in one fell swoop, only when you know the problem you’re solving or outcome you’re creating
10/ In other words, you want to minimize time spent in the uncanny valley in the middle, where the data-hungry ravenous algorithms roam
