Mind Matters Natural and Artificial Intelligence News and Analysis
abstract circle fractal background

Hulu’s Devs Series: Where They Get Determinism Wrong

Devs disposes of a key limitation of computers that can supposedly predict the future with psychobabble

Our science fiction film reviewer, Adam Nieri, recently reviewed the Hulu sci-fi series Devs for us at Mind Matters News. So I thought I’d take a look at it. Devs troubles me. After I explain my concerns as a computer nerd, I hope it troubles you too—at least, more than my nitpicking does.

David, a professor of ornithology, was a colleague of mine at the University of Washington. David studied birds. During an outdoor scene in the movie Out of Africa, where a bird could be heard tweeting in the background, his wife told me that David slumped into his seat with disgust. “Harrumph,” he grumbled to no one in particular,” That’s not an African bird. That’s a South American scarlet-banded Barbet.”1

Another friend, Lou, can be likewise irritating. Lou is a retired policeman. When watching cop movies, he always complains out loud about the specifics of police procedure: “Police would never bunch up like that when approaching a potentially hostile situation! They would spread out.” Or “Come on! A cop is trained to never place their finger on the trigger of a firearm unless ready to fire. Carrying the handgun like that is stupid and dangerous.”

I can be irritating company in the same way. I enjoyed the television sitcom The Big Bang Theory (2007–2019) until it trailed into tastelessness. Scenes often showed whiteboards with equations.

To their credit, The Big Bang Theory‘s developers hired a nerd consultant who made sure that all of the technology portrayed was correct. This made me happy. I would lightly elbow my wife and whisper “That’s Schrodinger’s equation.” Or “Oh look! It’s Coulomb’s law!” My wife was not impressed with my interrupting commentary nor with my obvious genius.

Which brings me to the television series Devs. The series is built around quantum computers that have the ability to travel back and forth in time computationally. In other words, there is no actual time travel. But the future can be computed and they are powerful enough to compute it.

This idea is not new. Early nineteenth century mathematician extraordinaire Pierre-Simon Laplace (1749–1827) believed that if all the properties of everything in the universe were known, the future could be predicted. It was possible in principle but would take a lot of calculation. Like the story concept underlying Devs, this “Laplacian determinism” posited that applying the laws of physics to current conditions could reveal the future.

Determinism, interestingly, does not imply knowability. Alan Turing first showed this. An arbitrary computer program will either stop or run forever. Halting can be determined by running the program long enough until the program stops. But one can never know if the program will never stop. Forever takes too long to find out. H. G. Rice generalized Turing’s result in 1953, showing that all non-trivial semantic properties of an arbitrary program are undecidable. Thus, besides halting, there is no way to determine whether an arbitrary program will never print the number 3.

All computer programs are deterministic.2 There they sit, on a hard drive. They carry out one deterministic step after another, as ordered by the programmer. Everything is fixed. Everything is deterministic. But in the general case, what the program will or won’t do is undecidable because it may take forever to find out.

Rice’s theorem applies to arbitrary deterministic computer programs. Laplace’s determinism applies to a specific program defined by the laws of physics and the current initial conditions of the universe. But what would happen in reality is moot. Laplace’s theory of determinism was offered before the current understanding of chaos, quantum mechanics, and even modern thermodynamics. No one today holds to Laplacian determinism.

We know that God does play dice with the universe (contrary to Albert Einstein’s view) and that the future can only be found probabilistically at best. Devs disposes of this problem with psychobabble, which includes repeating the quantum mechanical blasphemy that “everything is deterministic.”

As is characteristic of many modern streaming series, Devs contains filthy language and some religion-bashing. I see watching such shows as somewhat like eating chocolate cake even while knowing that a few mouse droppings fell into the cake batter. After a few bites, the taste became too offensive. Because of this and my nitpicks about computers and determinism, I did not finish the series.

No undecided problem here. I just halted.

1 I don’t recall the South American bird David identified, so picked “scarlet-banded Barbet” from a Google search.

2 Even random numbers generated on a computer are deterministic. That’s why they are called pseudo random number generators. For true randomness, a computer program must reach outside the code to other sources like a quantum random number generator.

Here’s our film reviewer Adam Nieri’s review of Devs:

Devs both grips and challenges Hulu viewers. I had fully expected Devs to be yet another series about sentient AI but it is something fresher. Alex Garland departs from conventional sci-fi themes to create a thought-provoking film, packed with action and based on a challenging underlying philosophy.

Also: But is determinism true? Does science show that we fated to want whatever we want? (Michael Egnor)

Robert J. Marks II

Director, Senior Fellow, Walter Bradley Center for Natural & Artificial Intelligence
Besides serving as Director, Robert J. Marks Ph.D. hosts the Mind Matters podcast for the Bradley Center. He is Distinguished Professor of Electrical and Computer Engineering at Baylor University. Marks is a Fellow of both the Institute of Electrical and Electronic Engineers (IEEE) and the Optical Society of America. He was Charter President of the IEEE Neural Networks Council and served as Editor-in-Chief of the IEEE Transactions on Neural Networks. He is coauthor of the books Neural Smithing: Supervised Learning in Feedforward Artificial Neural Networks (MIT Press) and Introduction to Evolutionary Informatics (World Scientific). For more information, see Dr. Marks’s expanded bio.

Hulu’s Devs Series: Where They Get Determinism Wrong