Hell is other algorithms.

Spoilers abound: Do not read further unless you’re up to date on The Good Place or have embraced nihilism and feel no remorse in ruining things for yourself.

Season two of The Good Place just wrapped its fall run, ending on a cliffhanger that likely spells damnation for our motley crew of “mostly good” souls. Despite the gang’s best efforts to ethically better themselves and circumvent eternal torment, the gig is up, with Shawn the “the all-knowing judge of all matters in the afterlife” (Marc Evan Jackson) having finally caught on to Michael’s (Ted Danson) scam. Given the show’s precedent for pulling the rug out from under its viewers, I genuinely have no forking clue what will happen next and am content to twiddle my thumbs and daydream about the Trolly Problem until the show returns at an unspecified date in early 2018.

I came to The Good Place with my own expectations about what a generically bright, colorful network comedy starring Kristen Bell entailed. I was not expecting to find myself at the mercy of deep cuts about Aristotelian virtue ethics, moral relativism, and Kant’s reputation as a routine-worshiping bore. And I was certainly not expecting this tonally delightful sitcom to engage in a frightfully relevant critique about the horrors of technological ubiquity.

Film and assorted media have a longstanding tradition of gleefully grounding the afterlife in man-made systems, typically some intentionally banal flavor of bureaucracy. You can see it in the DMV-styled limbo of Beetlejuice, staffed by suicide victims in a morbid display of contrapasso that would make Dante blush. Likewise, in Albert Brooks Defending Your Life we find a divine judiciousness in the all-too-familiar archetype of officious angels with quotas, corporate lingo, and three-piece suits.

The Good Place frequently aligns itself with this motif. Michael, the Good Place’s architect, and overseer, is a stickler for paperwork, frequently FaceTimes his boss, and dreams of climbing the corporate ladder. His head office looks like the accounting firm from The Producers; a drab no-place of coffee machines, tweed, and nondescript productivity. Like others before it, The Good Place employs an “afterlife express”; an efficient, reliable train to ferry souls between planes of eternal rest. Also, perhaps most explicitly, The Good Place shares a striking family resemblance with Heaven Can Wait, where Paradise itself, like any corporate structure, makes mistakes.

The Good Place

What modernizes The Good Place and distinguishes it from its predecessors is its use of something that looks a heck of a lot like ubiquitous computing; like the utopic, algorithm-supported proximate future we seem to be hurtling towards like that dog Michael kicked into the sun that one time. Every inch of the Good Place has been tailor-made for its residents; specific filter bubbles crafted via some inscrutable celestial black-box churn of “big data” analytics. Every aspect has been calibrated to individual preferences and histories; display screens appear out of thin air; and for everything else, there’s Janet (D’Arcy Carden), the system’s anthropomorphic supercharged Siri-like fetch-system.

For the majority of season 1, we are led to believe that an improbable glitch in the system had unhinged the finely tuned Swiss watch of Paradise; that every meticulously designed Jenga piece of Michael’s tranquil suburban heaven was problematized by Eleanor (who wasn’t exactly Gandhi on earth) being in it. However, during season 1’s finale, we learn that instead of Eleanor posing a threat to every precisely designed and calibrated detail in the Good Place, every detail in the Good Place had been precisely designed and calibrated to pose a threat to Eleanor. The gang has been in The Bad Place the whole time. In Eleanor’s words: “it looks like Paradise, but it’s actually a filthy dumpster full of our worst anxieties.”

Knowing this, the Good Place retroactively comes across as supremely and recognizably hellish. It’s like an uncanny valley approximation of what an Instagram algorithm might think “Good” people would expect heaven to look like: an unnervingly bland aspirational Paradise. Which is, it turns out, exactly what’s it is. Every fiber of “the Good Place” has been designed to evoke an unsettlingly generic but believable vision of Paradise: a pastel cruise ship fever dream of otters holding hands, cookie-cutter neighborliness, and frozen yogurt. In hindsight, the fro-yo is a good tip-off to understanding Michael’s stance on torture: that “there’s something so human about taking something great, and ruining it a little.” Pizza is good. Only being able to order Hawaiian is torture.

In this way, The Good Place grounds its suffering in our trust in computational systems; leveraging our ambivalent readiness to accept the higher power of algorithms and our subsequent distress when we’re unhappy with their outcomes. When you’re told the system has generated your dream house, it’s disheartening when the ceilings are too low, or the walls are full of evil clown paintings. If it’s a hyper-logical divine system, it’d be wrong to call it out, right?  There’s also the Good Place’s nightmare-mode version of Tinder that matches residents with their soul mates for all eternity (only, what if you’re not actually that into your algorithmically-determined soul mate?). The precipitous discomfort of feeling dissatisfied with a personally-tailored heaven is all par for the torture; it’s the spiritual equivalent of being served the completely wrong entree and feeling too torn up to send it back.

Eventually, triumphantly, the gang’s suffering alerts them to the fact that they’re in The Bad Place (no conceivable version of heaven would include a three-hour spoken-word jazz opera, Michael). This brings us to The Good Place’s most compelling, and hopeful suggestion: that while our blind trust in technologies can torment us, such systems are no match for the messiness of human behavior. No matter how many iterations he whips up, Michael’s design can never fully account for the emotional unpredictability of its subjects; for Jason falling in love with and “breaking” Janet; for Eleanor and Tahani to forge an unlikely supportive friendship, and for Eleanor to persistently try her best to become a better person.

The Good Place’s afterlife operates on a reductive essentialism gives us a sense of where the show may be headed. In The Good Place‘s version of the afterlife, there are good people and that there are bad people (…and Mindy St. Claire, a the Tantalus-inspired cocaine-loving lawyer from the 80s who improbably racked up an exactly equal amount good/bad points). Eleanor routinely calls out this blind spot: that the system has no way of accounting for the reality that most people aren’t good or bad, but somewhere in between. Michael banked on that ambiguity being the backbone of his nouveau Bad Place but as season 2 has shown us, it’s not that simple. People are tenaciously idiosyncratic; they break things, surprise themselves, and consistently muddy the waters in ways that no algorithm can fully account for. Like Shawn says, even getting humans to “do simple things like pulling out each other’s teeth is like…I can’t think of the right analogy.”

The way that “the Good Place” leverages our trust in algorithms is truly, recognizably,  horrifying. But if Hell is a misplaced trust in computational systems, perhaps salvation lies in our ability to break them; to be relentlessly unpredictable, messy, medium people.

More to Read: