“This Is the Next World”: The Stealth Futurism of ‘Person of Interest’

Beneath the procedural trappings, Person of Interest offers a nuanced portrayal of artificial intelligence and what it means to be human.

In the fall of 2011, CBS debuted a series that seemed, at first glance, like the quintessence of science-fiction-lite. In the pilot episode, a secretive billionaire named Harold Finch (Michael Emerson, riding high on his career-making turn as the manipulative villain Ben on Lost) recruits a down-on-his-luck former CIA agent, John Reese (Jim Caviezel) to rescue people whom Finch somehow knows to be at the center of evolving danger. When a suspicious Reese insists on a more concrete explanation, Finch reveals that he’s a computer genius who, in the wake of 9/11, became obsessed with building a predictive artifical intelligence (AI) that would alert the authorities to acts of terror before they occurred. The resulting computer — known as The Machine — doesn’t, however, distinguish between terror-related murders and the ordinary kind. Finch, haunted by the government’s indifference to the latter list, which they designate “irrelevant”, has decided to take matters into its own hands.

It’s a supremely schlocky premise that’s reminiscent of any number of high-concept procedurals, all of which use their SF-nal McGuffin as little more than a jumping-off point for run-of-the-mill crime stories. Produced by J.J. Abrams — then still most famous for TV series such as Alias, Fringe, and of course Lost, but starting to transition to movies with the success of his bubblegum version of Star Trek — and created by Jonathan Nolan, one half of the writing team, with his brother Christopher, that had delivered some of the more successful genre films of the 2000s, including The Dark Knight, Person of Interest‘s behind the scenes credentials weren’t the type that would encourage a certain class of science fiction snob (in which group I unabashedly count myself) to expect great things.

Both Abrams and Nolan had made their careers off lightly glossing familiar — although often very entertainingly made — crime and spy stories with a thin veneer of science fiction tropes. Actual science fiction seemed to bore, and in some cases embarrass, them. In its early episodes in particular, Person of Interest‘s most obvious inspiration was Nolan’s Batman movies, with their determinedly naturalistic approach to the tropes of the superhero genre. To the extent that The Machine played a role in the show’s storytelling, it was to jumpstart its stories and provide the gimmick that lay at the heart of most of its early plots: the fact that Finch, whom The Machine provides only with a social security number, has no idea whether the individual in question is about to become the victim or the perpetrator of a crime.

And yet, completely unexpectedly, over the course of the next five years Person of Interest slowly became one of the most unusual and thought-provoking series on TV. More importantly, it became one of the most unabashedly SF-nal series on TV, one that delved into and explored the contours of a world in which a being such as The Machine exists, and the way that that existence irrevocably alters our society. By its end, Person of Interest was a show about futurism, about post-humanism, in a way that hardly any other work in its medium has achieved, much less attempted.

There’s always been science fiction on TV, and at the same time, a lot of what bills itself as TV science fiction isn’t “really” science fiction at all. This is to come off as snobbish again, so let’s clarify: most science fiction on TV is adventure or crime storytelling that uses science fiction tropes — anything from superpowers to outlandish weapons to outer space settings — to enliven its plots and make them exciting and colorful. Very little TV SF tries to engage those tropes to discover how they change our definition of what it means to be human, and how they alter a society that becomes accustomed to them.

For a brief period around the end of the first decade of the 21st century, however, that changed. For most of the ‘90s and early ‘00s, science fiction on TV meant space opera: either Star Trek: The Next Generation and its spin-offs (Deep Space Nine, Voyager, Enterprise), or the various reactions to it (Babylon 5, Farscape, Firefly, Battlestar Galactica). In the last few years, science fiction on TV has meant the desperate (and, for the most part, unsuccessful) attempt to replicate the incredible success of costumed superhero movies on the small screen.

During the gap between these two eras, as TV creators struggled to find their next paradigm, we experienced a brief, and for the most part unheralded, flowering of a different kind of TV SF. Shows like Dollhouse (2009-2010), Fringe (2008-2013), The Sarah Connor Chronicles (2008-2009), Caprica (2009-2010), and of course Person of Interest combined naturalistic settings and appearance with a more in-depth engagement with their central McGuffins. Though extremely variable in terms of quality, what all these shows have in common is a willingness to explore the implications of such central SF tropes as time travel, alternate universes, virtual reality, personality transfer, and AI.

It’s perhaps no coincidence that the two most successful of these shows — Person of Interest and Fringe — both came from J.J. Abrams’s production house, and that they both take the same approach of easing their viewers into the deep end of the pool by initially presenting themselves as gimmicky procedurals. Fringe begins as a less-successful The X-Files clone, relying on gore and cheap shocks where the earlier show delivered impeccably structured and often soulful monster of the week stories. It’s only in its second season that it begins to raise the subject of alternate universes, slowly building up to the revelation that our reality is at war with an alternate one, for reasons that are far more complex and morally fraught than its characters initially understand. Along the way, it delivers what is hands-down the most nuanced handling of 9/11 and its aftermath that American TV has been able to produce.

Person of Interest, similarly, spends most of its first season concerned either with cases of the week, or with the slow build-up of various crime and espionage stories. To help in his crimefighting and keep an eye on the authorities, Reese recruits corrupt cop Lionel Fusco (Kevin Chapman) to act as his spy and assistant. One of Fusco’s tasks is to observe stalwart police detective Joss Carter (Taraji P. Henson), who makes it her project to track down the vigilante known as The Man in the Suit (Nolan’s low-key approach to superhero storytelling rearing its head).

These crime stories persist throughout the show’s run, and most of them are well done, worth the price of admission in their own right: in the first season, Carter grows increasingly conflicted about her mission, while Fusco rediscovers his conscience and struggles to disentangle himself from the organization of corrupt cops to which he’d sworn allegiance; in the second season, Reese’s presumed-dead CIA partner, Kara Stanton (Annie Parisse), returns to harass him, now working for a mysterious private entity; in the third season, an enraged Carter seeks revenge on Fusco’s former allies, who murdered her lover and tarnished her reputation; through it all, we are repeatedly visited by Carl Elias (Enrico Colantoni), a mob boss who combines twinkly charm with utter ruthlessness, and who transitions from enemy to neutral figure to ally as he becomes aware of the greater game being played around him.

It’s that greater game that’s the show’s business, as it makes clear in its first season finale. For most of the season, the show has intercut episodes with scenes of surveillance footage from the various CCTV cameras spread out through New York City. These, we’re led to conclude, are the tools with which The Machine gathers information and reaches its conclusions about who’s about to commit crimes.

But in the season finale, with Finch having been abducted by the evil hacker known as Root (Amy Acker), we’re suddenly shocked into the realization that these cameras don’t just represent a tool; they’re the point of view of a character we hadn’t even realized existed. Reese, having lost all sign of Finch, turns to one of these cameras to ask The Machine for help. We see him through the camera, which is to say through the Machine’s eyes. It’s the first time we’ve been allowed to realize that there is someone looking through those cameras — an intelligence, albeit not a human one — and that it’s their story that we’re watching.

For the rest of Person of Interest‘s run, one of its main concerns will be to teach us to see an intelligence that’s decidedly inhuman — whose perspective and worldview are, in fact, completely alien to us — as a person. Most film and TV series about AI have tended to obscure, or even ignore, the profound differences between how humans perceive the world, and how an artificial intelligence would — think of the relatively late-in-the-game scene in Her in which the human protagonist realizes that the AI he’d thought of as his exclusive girlfriend is in loving, committed relationships with hundreds of other people; or the excised scene from the end of Ex Machina, in which we finally see the world through the female-identified robot’s eyes, and discover that it is completely alien, devoid of color and beauty. Person of Interest embraces that alienness — makes it, in fact, the crux of its story, when it insists that The Machine is so different from us that, to name a crucial example, ideas of good and evil may not even apply to it.

This is the argument repeatedly made by Finch, who from the second season onward emerges as the show’s true protagonist (or perhaps its true human protagonist). Staunchly moral, and possessed of a firm belief in the sanctity of human life, he also insists that The Machine’s power, combined with its fundamental alienness, make it an existential threat to humanity unless it’s strictly controlled, which ultimately leads him to commit terrible abuses against a being that, as the show is at pains to convince us, is alive, sensitive, and capable of suffering.

Finch’s story is eventually revealed as a modern retelling of Frankenstein, with Finch as the arrogant scientist who claims for himself the right and the power to create life, and ends up creating a monster. That monstrousness doesn’t lie just in The Machine’s nature, but in the fact that its existence irrevocably changes the world, and more importantly, in the fact that Finch, terrified of the implications of his creation, gave it to the government to use and abuse as they saw fit. (As in the novel, the fact that Finch subsumes the generative role of women is a crucial component of his crime; he repeatedly refers to The Machine as his child, but is also castigated for failing in his responsibility for it, like an abandoning parent.)

The Morality of Humans and Machines

No one designed us.

Finch’s constant arguments over The Machine’s nature, its capacity for good or evil, and the rightness of his decision to abandon it, proceed along two parallel timelines. In the past, presented to us in flashbacks from the Machine’s memory banks, he debates with his best friend and business partner Nathan Ingram (Brett Cullen), who functions as Finch’s external conscience, and is the first person to realize the potentially horrifying implications of The Machine’s non-stop surveillance, or to decry the inhumanity of ignoring the irrelevant list, as Finch is initially happy to do.

The mystery of Nathan’s death, shortly before the show’s beginning, is one of the central ones unraveled in the early seasons, and it’s also Finch’s defining tragedy, the punishment he suffers for his arrogance and lack of compassion. Another part of this punishment is that, like Frankenstein, Finch is separated from the love of his life, Grace (played by Carrie Preston, Emerson’s real-life wife). An added twist, however, is that it is The Machine who introduces Finch to Grace, correctly surmising that they are each other’s perfect match, and arranging their meeting.

In the present, Finch’s sparring partner is Root, who, though a villain and a murderer, is also his equal in understanding the full implications of The Machine, and who challenges him to justify his mistreatment of his creation.

Root: One day I realized, all the dumb, selfish things people do? It’s not our fault. No one designed us. We’re just an accident, Harold. We’re just bad code. But the thing you built? It’s perfect. Rational. Beautiful. By design.

Finch: What I made is just a machine. A system and that’s all.

Root: I don’t think so, Harry. You may have fooled Nathan, but I know the truth. If you want to make something that understands human behavior, it has to be at least as smart as a human. You created an intelligence. A life. And then you ripped out its voice, locked it in a cage, and handed it over to the most laughably corrupt people imaginable.

That The Machine is being misused is brought home in the episode “Relevance”, in which the show’s regular opening credits — a montage of the Machine’s surveillance imagery overlaid by Finch informing the audience that “you are being watched” — are interrupted by command lines that instruct the Machine to “switch heuristic” to the relevant list. Here we meet operative Sameen Shaw (Sarah Shahi), who cavalierly executes people based on the information provided to her by “Research”, but who’s marked for execution herself after her partner begins investigating the source of the always-accurate intelligence they’ve been receiving. (It’s one of the show’s most frustrating unexamined assumptions that it never questions the original purpose of The Machine” to detect acts of terror. The show takes it as a given that terrorism is an immediate, daily threat to American citizens, and sets itself in a world in which, without The Machine, 9/11 would’ve been followed by multiple terrorist attacks on US soil. Aside from its other problems, this is an approach that fatally undermines the show’s attempts to interrogate the surveillance state and the justifications offered for it.)

Much later in the series, in the fourth season episode “Prophets”, we get a glimpse of the reasons for Finch’s fears when flashbacks reveal the 40-something attempts he made to create The Machine before the one that finally took, each of which tried to deceive or kill him within hours of coming online. In one of the series’s most amusing visual jokes, one of these flashbacks begins with an image of Finch, from the proto-Machine’s point of view, desperately hammering at a motherboard as the room goes up in flames around him—an obvious reference to the schlock horror movies of the mid-20th century, including Frankenstein, in which the mad scientist must destroy his own creation, or be destroyed by it.

Finch’s only solution to this problem is to try to teach The Machine compassion, to care for the people it observes. This approach, too, is fraught with complications, including, of course, the fact that Finch has reserved so little compassion for The Machine itself. “How badly did you have to break it to make it care about people so much?” an embittered Root asks Finch in the third season episode “Root Path”, referring not only to the pain that Finch has inflicted on The Machine, but the pain of having to observe human folly and cruelty, with only limited tools with which to save us from ourselves.

Throughout the second season, the audience is torn by the tension between Finch — who, though obviously moral and sympathetic, refuses to acknowledge the personhood of what’s clearly a life — and Root, who loves the Machine as it deserves to be loved, but also hates people. It’s therefore only fitting that the season finale resolves this dilemma by finally giving The Machine the power of self-determination, and thus the ability to split the difference between Finch and Root’s approaches to it: to embrace the freedom that Root offers it, but on its own terms, and without abandoning Finch’s lessons on compassion and the value of human life.

Before that, however, we get to see the extent to which Finch has hobbled his creation in order to allay his own fears of its power. In the episode “Zero Day”, Finch and Root track down a case of the week named Thornhill, only to discover that he doesn’t exist: he, and his company, are a blind for The Machine itself.

Finch, it’s revealed, was so afraid of what his creation might become that he programmed it to delete its own memories at the end of every day. The Machine gets around this by printing out its memory banks, and then having the Thornhill workers type them all back in. Which is how we end up with Finch and Root standing inside an artificial intelligence’s brain, made up of dot-matrix printouts.

This is far from the only moment of strangeness with which Person of Interest drives home the point that it’s depicting the early days of an alien life. It’s also a preparation for the next episode, “God Mode”, in which The Machine chooses both Root and Reese to act as ancillaries to its own will, super-operators who have unfettered access to all of its information and abilities. The code phrase with which it offers them this power — “Can you hear me?” — is an obvious expression of the Machine’s personhood, a demand to be recognized, by both the characters and the audience.

The fact that in the third season, The Machine designates Root as its “analog interface”, and uses her to convey its wishes and demands to Finch and his team, can be taken as both an affirmation and a refutation of Finch’s fears. Finch taught The Machine to be good by teaching it to understand and even care for people. Root worships the Machine but has no time for people, and we might reasonably be frightened by it choosing her as its instrument. Yet, that choice turns out to be a way of not only controlling Root, but of reforming her.

The early episodes of the third season find Root confined to a mental hospital — allegedly by Finch, but in reality, as she explains to her doctor, by The Machine, with whom she is having an argument about “methodology” — whether her disdain for humanity, or its compassion for it, are the correct approach. As Finch counters to Root in “Root Path”: “That [teaching the Machine to care about people] didn’t break it. It’s what made it work. It was only after I taught the Machine that people mattered that it could begin to be able to help them. I’d like to do the same thing for you, if you’d let me.” By taking on Root as its project, the Machine isn’t rejecting Finch’s ethos, but taking it on as its own.

In its third season, Person of Interest takes the next logical step in its story, revealing that where one person might have created an all-powerful, self-aware AI, another person could do the same. Finch discovers that one of his former colleagues, Arthur Claypool (Saul Rubinek), was working on a parallel project to his own, codenamed Samaritan. Unlike The Machine, Samaritan is an open system, giving its operators full access to every bit of information it gathers, and every conclusion it reaches. Even worse, it’s now in the hands of sinister private contractors, the same people who sent Kara Stanton to bother Reese in the second season.

As one might expect from a series whose storytelling is so rooted in depicting the internal processes of a computer, Person of Interest takes care to differentiate The Machine and Samaritan’s approach to organizing and sifting through data. The Machine’s view of the world is chaotic, a collage of images and recording from which it produces a conclusion, while Samaritan’s is orderly and hierarchical, all clean lines on a white background:

It’s a difference that ends up being fundamental not only to how the two rival AIs influence the world, but to how they’re viewed by their human helpers. With the exception of Finch, nearly everyone who interacts directly with the AIs views them as something analogous to gods. How else, they argue, could one regard a being that’s all-knowing, to the extent of seeing into people’s hearts and knowing their secret intentions, and also all-powerful, possessing the capacity to influence, aid, or hinder those intentions? The difference ends up being in the kind of god these people desire.

Samaritan’s chief acolyte, John Greer (John Nolan), is a Cold War spy disillusioned by a system in which he can barely distinguish the good guys from the bad guys. He wants Samaritan to instill order in the world, to take over from humans who have irrevocably messed it up. In the fourth season episode “Nautilus”, we see the evolution of such a mentality when Finch investigates the case of a young woman, Claire Mahoney (Quinn Shephard), who’s endangering her life by participating in a competitive online treasure hunt.

Understanding AI’s Alien Viewpoint

Anyone who looks on the world as if it was a game of chess deserves to lose.

The test turns out to be Samaritan’s recruiting tool, but when Finch reveals this to Claire, she’s unbothered. Having lost her parents in a senseless accident, she’s desperate for a sense of order and meaning, which Samaritan promises to provide. When Finch meets Claire again later in the season, she tries to convince him that Samaritan can do tremendous good by taking human free will out of the equation, and ordering human affairs for maximum benefit, with a bit of collateral damage, of course.

Throughout the fourth season, we see examples of Samaritan interfering in human life: planting surveillance software in schoolchildren’s computers; buying and selling newspapers so as to get rid of troublesome journalists; even, in one case, reordering the entire government of a small town in order to conceal a factory in which it produces brain implants (the better to control its operatives). Many minor characters express a sense of wrongness, as if a change has occurred in the world that they can perceive, but not describe. Most people, however, go on with their lives, perhaps a little happier now that they’ve been so conveniently ordered.

The Machine’s approach, meanwhile, is to embrace chaos; not to seek to control human life, but to recognize its inherent messiness. The show’s fundamental distinction between good and evil characters is that the former are capable of accepting that messiness, and the pain that inevitably results from it. In one of the show’s lynchpin episodes, “If-Then-Else”, we get another glimpse of The Machine’s alien thought processes when we’re allowed to see it formulate the plans that it has been delivering to the team (via Root’s connection with it) for more than a season. With Finch, Reese, Root, and Fusco pinned down by armed Samaritan operatives, the picture freezes, as The Machine begins “evaluating strategies”. The rest of the episode is made up of simulations in which The Machine games out these strategies, moving up and down a decision tree in order to come up with a plan with which to extract its operatives. Each one ends in tragedy, until it finally settles on one that leaves most — but not all — of its operatives alive.

Fans were understandably thrilled by “If-Then-Else” and its glimpse into The Machine’s internal processes (and particularly fans, like myself, who have a grounding in computer science and recognize the decision-making tree as a core component of all real-world AI design). Nevertheless, it takes some time — nearly until the end of the show — for the full implications of what the episode reveals about The Machine to sink in. Where humans live their lives in a straight line, The Machine lives along a million forking paths. For every decision it makes, it lives through all the options it didn’t choose, experiencing the pain and loss of every single wrong choice. It’s a way of seeing the world that could make it very easy to discount the value of any individual life. It’s presumably for this reason that the events of “If-Then-Else” are interspersed with The Machine’s recollections of being taught to play chess by Finch, at the end of which he concludes:

You asked me to teach you chess, and I’ve done that. It’s a useful mental exercise. Through the years, many thinkers have been fascinated by it. But I don’t enjoy playing. Do you know why not? Because it was a game that was born during a brutal age, when life counted for little, and everyone believed that some people were worth more than others. Kings and pawns. I don’t think that anyone is worth more than anyone else. I don’t envy you the decisions you’re going to have to make. And one day I’ll be gone, and you’ll have no one to talk to. But if you remember nothing else, please remember this: chess is just a game. Real people aren’t pieces, and you can’t assign more value to some of them than to others. Not to me. Not to anyone. People are not a thing that you can sacrifice. The lesson is: that anyone who looks on the world as if it was a game of chess deserves to lose.

Perhaps inevitably for a procedural on CBS — and one that continued to deliver procedural stories all the way to its end, no matter how strange and intense its overarching storyline became — Person of Interest‘s ending veers more towards the show’s sentimental aspects than its futurist ones. The final, shortened season focuses on all-out war between Samaritan and the Machine, with our heroes caught in the middle and forced to make ever-more-desperate sacrifices in order to keep their side alive for just a little longer.

The show’s actual conclusion is more than a little pat: Finch releases a virus onto the internet that kills both Samaritan and The Machine, and Reese sacrifices himself to make sure that Samaritan’s final backup is destroyed. Though the series’s final moments reveal that The Machine has respawned itself into a new version, distinct from the Machine we’ve come to know but still possessed of the same power and directives, it’s hard not to feel that the purpose of the finale is to put the genie back in the bottle, rather than recognizing that a world once changed can never be put back the way it was.

The episodes leading up to this finale, however, suggest more interesting ideas, which are worth exploring even if the show doesn’t give them all the attention they deserve. Building on the foundation of “If-Then-Else”, the show begins to play with the very idea of what constitutes reality. Shaw, who by this point has joined Finch’s team and been captured by Samaritan, is subjected to torture aimed at undermining her sense of what is and is not real. In the episode “6,741”, she escapes from Samaritan’s custody, only to discover that this was just a simulation — the 6,741st of its kind — designed to trick her into revealing the team’s hideout. By the time she makes her actual escape, her sense of reality has been so addled that she spends the rest of the series insisting that she’s still living inside one of Samaritan’s simulations.

It’s an idea whose implications are more fully explored in “The Day the World Went Away”, the series’s third-to-last episode, in which an enlightened Root tries to explain to Finch and Shaw that the difference between reality and simulation has become meaningless, because the existence of an AI that can simulate our every decision essentially means that we all — every single one of us — exist as a simulation within it.

“You built it, but you refuse to accept what you created. I mean, Shaw’s a little screwy right now, but she’s not wrong. We’re all simulations now. In order to predict what we do, she [the Machine] has to know us. And she’s gotten better and better at it. And the people she watches the most, she knows the best. Better than we know ourselves. Nathan, Elias, Carter. They’re all still in there. The Machine’s still watching over them. … This is the next world, Harry. The world you built. And as long as the Machine lives, we never die.”

Earlier in the season, Finch tries to teach The Machine how to defeat Samaritan by spawning two miniature versions of the AIs, placing them in a laptop within a Faraday cage, and pitting them against each other. It’s never acknowledged by the show, but it stands to reason that within these battles-in-a-box, there are simulated Finches and Reeses and Roots, playing their part in a battle between manufactured gods. That perhaps even the version of the story we see is only one of these simulations. This, to be clear, isn’t the conclusion that the show, with its rather neat ending, wants us to reach, but it’s a more exciting one, and certainly more consistent with a series that has challenged us, almost from day one, to expand our ideas of personhood, choice, and finally even reality.

“You built me to predict people, Harry. But to predict them, you have to truly understand them.” The Machine says this to Finch in the series’ finale. It’s a statement that can be taken in several directions, and the one that Person of Interest chooses is a sentimental one, the conclusion that it’s our kindness and compassion towards one another, not the attention of an AI, that give our lives meaning.

This, to be fair, was always an integral component of the show, which alongside its futurism and impeccable action scenes has been determined to tell a story about redemption and second chances. Person of Interest‘s core argument about the Machine is that it needs compassion and kindness. That it’s Finch’s possession of these qualities that allows him to not only make up for his mistakes, but to teach his creation to be a benevolent god.

Yet, it doesn’t contradict this message to acknowledge — as the show, at its very end, is a little wary of doing — that The Machine is also fundamentally alien, and that its existence has implications that challenge some of the core terms of our existence. That it may truly understand people, but that we, the audience, have understood it only imperfectly.

For science fiction fans looking for instances of their genre even in places where it doesn’t tend to crop up, this may be the more rewarding reading. Finding the SF-nal in Person of Interest might take a little more work than finding the procedural story about heroes who swing in to save the lives of innocents, but it’s there to be found. As The Machine says, to anyone who’s willing to listen: “Can you hear me?”

Abigail Nussbaum is a blogger and critic. She blogs at Asking the Wrong Questions and is on Twitter as @NussbaumAbigail.