Algorithms Aren't Smart

When I was in college (and for some time after when I had my first office job), I helped tutor a class called developmental math for several years. Tom Ferguson was the professor of that class when I first started, and the guy was incredibly smart -- sometimes to the detriment of those poor students trying to get by in the bare bones of college-level mathematics.

Once, when explaining to students about how to reason about some math problem and when calculators were useful and when they weren't, he said simply, "Calculators are stupid. They don't know anything."

This was long before my journey into trying to understand the very basics of programming and computer science, so I didn't really know what he meant at the time. Of course, I do now. Calculators, computers -- anything digital really -- are stupid. They don't / can't know anything, and if you give them even the slightest wrong instruction, they will follow it with no regard for what you meant. (LLMs notwithstanding, though I think the argument can be made they're just stupid in a different way?)

This memory came back to me when reading this blog post ("If nothing is curated, how do we find things?"). Quote:

"And algorithms can only predict content that you've seen before. It'll never surprise you with something different. It keeps you in a little bubble. Oh, you like shoegaze? Well, that's all the algorithm is going to give you until you intentionally start listening to something else."

This is my experience exactly. The algorithms are stupid too. They only repeat, relentlessly. There's no magic in it, no surprise. Most often, it feels the the new, best experiences I've had with media or food or other kinds of consumption happened because I stumbled upon it by accident, through a friend recommendation (and no, not through a "social network"), or through making a connection to a work that I liked (à la Austin Kleon's note about "reading what your favorite authors read").

Algorithms aren't smart, but people sometimes are.

Tagged: technology, attention,