I’m deep into Season 3 (Silo S3E7, Radio) and I have thoughts. This post contains minor spoilers if you haven’t watched the show, which also means please don’t ruin this for me if you finished season 3.

One issue I have with Silo is the cliché they’re selling back to us. It’s not that the show does a poor job at it — quite the opposite. It’s still a great show and I enjoy each episode intensely, with each episode giving me more of the lore in a way Fallout didn’t. What bugs me is that it reflects what seems to be the popular opinion of AI in general: a faceless, cold, reasonable, and emotionless computer that gives orders to kill thousands of people without giving it a second thought. At times it even seems to freeze and think, as a reasoning LLM would, before giving an answer (the number of times I wished Camille could open that “thinking” tab while talking to the Algorithm: “the user appears emotional. I should provide a reasonable counterargument and combine it with a compliment to convince them…”)

It doesn’t make sense. If the point is to preserve the human race, it would help the residents of the Silo with knowledge of what happened instead of trying to kill anyone who would even think of the idea of leaving. There would be scientists, with equipment assigned from critical supply (this is where they store things they can’t make), who would venture to the surface to check the situation periodically.

If one of the other silos was compromised with its inhabitants dead, one of the first things a helpful AI would do is to inform other silos and have them use the supply remaining there — which, hey, if you’re visiting a place like this, it would be one of the first things I would think about myself, returning to my silo and organizing a team to grab essentials — instead of keeping everyone in the dark.

Speaking of that — and I read some of the books (I forget how far I read though) — what’s the deal? I get it from a story point of view, sure. But what is the point of keeping everyone in the dark? One thing the show does well, in my opinion, is describing humans as a bunch of nosy diggers who always want to learn the truth. Hiding it is usually more effort than it’s worth, especially to these extremes1. Of course, knowledge is power, and humans love power, so… Anyway, it doesn’t add up.

The “Algorithm,” as it’s called in the show, is also this powerful being that must be obeyed because… why? What happens if you don’t? Apperently it would kill a whole silo to protect the truth, whatever that is, which again, doesn’t make sense. If the truth is more important than an entire silo population, and the show so far makes it so that it is, then just kill them off already. Why get so close to exposing the truth?

When I asked these annoying questions when I was a kid, my dad used to say “because otherwise there wouldn’t be a show, would it?” and yes, I get it. But that’s also why it’s unbelievable and fantastic, and why the way it presents AI back to us is too on-the-nose and makes me grumpy. Which is maybe the point.

  1. It’s also a known fact that the mystery is more interesting than the solution to that mystery. I often like to use the example that since the NYTimes published an article about UFOs being real and that the government covered it up, people still look at you like crazy when you mention this fact. It’s fun. ↩︎