👁👃👁 algorithmic curation, friction & Elton John
Thoughts on present platform capitalism & the role of friction
Thanks so much for your feedback on the first episode.
🤓 what we’re thinking about
Depending on where you look, taste-anticipating algorithms have been either the 'answer' or the blunt instrument for solving the problem of information superabundance. And with all the hype around GPT-3, this is especially potent…
Andrej Karpathy @karpathyBy posting GPT generated text we’re polluting the data for its future versions
Algorithms that churn experiences and recommendations based on consumption history leads to an echo chamber. The echo chamber issues are well documented (a favorite here). The more you participate on TikTok, for example, the narrower the range of voices that will be served up. In a hilarious example, a set of users were funneled into such a niche where they were convinced that TikTok was only for lesbians.
Beyond the echo chamber, many have spoken and written about the dangerous psycho-dynamics brought about by today’s ad-driven platform capitalism - including the wonderful Mat Dryhurst. In order to gather a following and perform well on today’s platforms, you have to play to satisfy platform logic. It’s creating new power structures with algorithms and editorial teams controlling what we consume to the point where creators are so obsessed with feed or playlist placement that it’s dictating what content they create. Working with creators and artists at YouTube, it was really sad to see creators calibrate creative output around the algorithm. We can do better.
Over the last two decades, we’ve made huge leaps with the consumer internet to give people a microphone. But as per the latest episode of means of creation, when you give everyone a microphone, it gets really loud. Hence, we’re seeing the exodus (or refuge) to better curated experiences today.
Extending Scott’s analogy, platforms today feel like Starbucks. But a lot of people want the independent third wave cafe. This connoisseurship movement is fascinating and, if you want to geek out, this paper goes into the ethnomethodologically-informed sociology of taste through the lens of coffee 🤓.
At its core, developing taste requires work, it is burdensome.
Ethan Diamond spoke of how he thinks of “friction” at Bandcamp:
“Is that song playing because they went to make a sandwich and the next track autoplayed? Or does this song mean something to the person? And the best way to figure out if this means something to you, is, well, you listened to it a few times and you decided you needed to save it. You listened to it a few times and you decided, man, I want this artist to keep making this music so I'm directly supporting them.”
The problem is that people don’t identify as much with automated choices as with what they’ve discovered or dug up. When there’s friction, you’re brought into full awareness and it means something.
The challenge then becomes balancing between “optimizing for convenience while appreciating the frictions that fuel ritual”. This is an ongoing discussion - subscribe for more.
🧐 trends on koodos
We’ve been seeing a lot of emojis that involve some face-like combination of emojis, such as 👁👅👁, 👁👄👁, 👁👃👁, 👁️🔺👁️.
Much like the ‘it is what it is’ movement, we’re seeing a contagious pattern of usage. These emoji combinations have become memes to poke fun at boomers. These emoji combinations are usually paired with throwbacks - older songs that boomers would like. From Rick Astley to Diana Ross & Elton John.