John Rogers

about me | github | linkedin

2 June 2020

Review: Humans In Boxes

Simulated human life is a popular topic lately.

Devs’ take on this topic is the most interesting in my opinion. Upload uses human simulation for satire, Black mirror uses it for distopia-ish imaginings and Westworld uses it to discuss free choice. Like Westworld, Devs’ central topic is free choice, but of the four shows only Devs actually tackles the mechanics of human simulation. Spoilers ahead!

Devs supposes a quantum computer of unlimited power. This computer, through sci-fi wizardry, can perfectly simulate the universe. The supposed mechanism for getting the universe into the box is worth talking about.

Here’s how you get the universe into a box. First, the computer is given all possible information about some objects on a plate. With that information, the computer can infer perfectly all of the other information in the universe. It can do this because the universe is 100% deterministic. There is only possible one way for those objects to get on that plate. From the big bang onwards: stars forming & exploding, life evolving, humans making plates, building computers, putting objects on plates in front of computers. There is only one initial condition that results in that exact arrangement of objects on that plate. This is a suspect premise, but it is a cool idea nonetheless.

A person holding a box that contains the universe can peek at any moment in time, past or present, anywhere. Many characters look back into the past. They look at cavemen, dinosaurs, Jesus, famous people doing it, and other things. Only three characters look into the future. And by looking at the future they create at least one paradox. Something akin to the halting machine. The paradox is this: a prediction changes you, and that change can invalidate the prediction. For example, if a genie tells you: you’ll jump in 30 seconds. You can easily not jump in 30 seconds and disprove the prediction. The same problem holds for the computer. The computer can predict that you will jump in 30 seconds, but by showing you the prediction, the prediction is invalidated and the simulation is no longer perfect.

(There is also another paradox here: how does the computer perfectly simulate itself, simulating itself, simulating itself, …?)

Impressively, this conumdrum is addressed directly in the show twice. Two characters are watching the simulated universe. One says to the other, ‘suppose we looked at ourselves 30 seconds into the future and did the opposite of whatever was predicted?’. In response the other rolls her eyes. Then the scene ends.

Initially this scene is extremely frustrating. However, when you come to understand how the machine actually works, and the motivations of the players, you will arrive at a very satisfying answer.

And the second time the paradox is addressed the show delivers a very satisfying answer to the question of free will.

To top it all off there’s a (mostly) happy ending.

Well worth a watch!

tags: TV