Devs Finale Fumbles Free Will, Determinism, and Time Paradoxes

Original article was published by Not A Doctor on Artificial Intelligence on Medium


So what’s so wrong with the way this played out? In order to understand that, we first have to break down what “free will” actually means. First, let’s recall that literally every single character in Devs who watches the future via simulation has the exact same type of input that Lily does, which is a visual and auditory representation of the ostensible, DEVS-predicted future. Forrest and Katie watch it like a freaking soap opera. They’ve watched the future so many times they know it line-by-line (and again, there’s our time paradox: where did the lines come from if they’ve only come from watching the simulation?).

In one scene, a group of techs, gathered in the simulation screening room for a demo, all watch their milliseconds-into-the-future selves and copy them perfectly, thus fulfilling the prophesies made by DEVS. Since the characters in the simulation have their own simulation-inside-a-simulation, this introduces a recursive time loop, collapsing on a spacetime singularity. So here we go again — if the watchers behave as they do only because they’ve seen it happen, where is the causal root of the action? But I’m going to argue that the way this plays out is probabilistically untenable. Humans, upon watching the prediction of their own actions, would naturally rebel against the prediction or introduce some natural variation in copying it. The self-fulfilling prophesy version of a time paradox just doesn’t fly.

We see another paradox introduced near the end of the series, when Katie takes Lyndon (poor, sweet Lyndon) out the dam to test his belief in the multiverse theory. If Katie does this because she’s seen it happen already in the simulation, where does the action originate from?

So now we get to Lily’s time paradox: what is different about Lily and her knowledge of the future, that enables her to “break” DEVS? In the final episode, Katie and Forrest play the ending of the simulation for her, and, just like the techs in the previous episode, she’s able to view the predicted actions of her moments-into-the-future self. When she uses what she sees in the simulation to make a “choice,” why is she alone able to circumvent the laws of the universe — the laws that Katie programmed into DEVS? Is Lily the genesis of a universal law, a factor on the material plane, that somehow couldn’t be accounted for by the laws of physics that were able to perfectly predict everything else?

Bad new for Devs: That’s how they awkwardly play it off, with Forrest chuckling, “talk about a Savior!” as though Lily = Jesus? And Jesus = Free Will? Jesus doesn’t fall into recursive time loops. Only silly mortals do. Okay, let’s go with that: Lily is the fundamental genesis of some novel force in the universe, that’s how darn independent and spunky this gal is. True free will has never existed before Lily because she is Jesus, and Jesus breaks the shit out of time paradoxes.

Trying to reconcile free will with a material decision-making machine (the brain) goes all the way back to Descartes, who posited that the mind and the brain meet in the pineal gland (as though location somehow explains mechanism), and the soul is thereby able to influence our behavior. At Descartes’ time, we had a general understanding that the brain controlled behavior but we weren’t ready to give up our religious convictions. So we needed a device that resolved the mind-brain problem, though at this time, it was really a soul-brain problem, because, again: Jesus.

Funny thing is, we haven’t come very far since then. Even non-religious people cling to the concept of free will without realizing that in doing so, the onus is on them to explain what it is, and how this non-material, soul-like force interfaces with the brain. That’s a hell of a question even for scientists to attempt. And Garland definitely doesn’t answer it.