Something I hear people say often is something to the effect that they are currently pleased with what their algorithm is doing. It will be something like, “I’ve trained my TikTok so well this month. Everything I get is fire!”
People seem to feel like they are forming collaborative relationships with the algorithms behind their apps. And they are. We are all spending hours and hours telling different systems what we like and don’t like. It’s one giant data input experiment that has been happening for years.
Of course, teams are regularly experimenting with and making changes to how their algorithms work, and so sometimes you feel like ‘your algorithm’ isn’t working as well. You might get too many ads or sandwich videos for a few days. But then things even out again, and you are back in control.
It's fun and often shocking to watch people compare each other’s ‘Explore’ pages on Instagram. We quickly understand that, in fact, everyone sees very different things online. Seeing someone’s explore section actually feels to me a bit like looking through someone’s drawers at home. You get a feeling that you’re not really meant to be seeing that content, and that’s because we sort of understand or feel that algorithms are this private thing. It’s something that only the person who has the app sees.
Microsoft announced last week that it’s going to launch a feature that basically runs on your laptop and records everything it sees in an attempt to train a personal AI for you. Creepy.
There are lots of questions here about privacy and surveillance that are important. But I assume that the big players are all going to continue to package AI up as a friendly digital assistant, and I wonder to what degree AIs will be positioned as things we will have relationships with.
Some ideas and questions for the near future:
- Will we be able to name and give AIs identities?
- Is it still going to be called Siri?
- Can I ‘try on’ someone else’s algorithm in an app?
- Can I have ‘work’, ‘play’, ‘serious’, and ‘fun’ modes in apps/on phones?
- Could I get angry with an algorithm?
- What does it mean to ‘build relationships’ with things you can’t see?
- If I assume I’ll be, for example, an Apple customer for the next 40 years, how does that change how I’ll use my devices/services?
A final thought on this: It’s very likely at this point that I will be an Apple customer for decades to come. The hardware is and has been irrelevant for years now. All my information is in the cloud. If I knew that I had 10, 20, 30 years with a company and they had technology that could learn about me, what sort of product decisions appear?
Also, if my AI companion lives in the cloud, I assume I’ll be able to take this thing with me across devices? So then what? And what happens when I can’t access it? It all feels very Tamagotchi, and I liked my Tamagotchi.
Also, as always, the memes have it.
