This season of Person of Interest has been one long philosophical tale about the dangers and benefits of all-power artificial intelligence — but last night's episode made the whole discussion alarmingly concrete. But that wasn't the most startling moment in last night's episode. Spoilers ahead...
At the end of last season, the amoral A.I. Samaritan came online, and last night we finally got to hear Samaritan's point of view. And basically, Samaritan views itself as a god, and has the same ideas about humanity as most villains bent on world domination in classic science fiction: We're too greedy and corrupt and destructive, and we need a firm hand. (Which seems to be Greer's viewpoint, too.)
Samaritan arranged a meeting between itself and the show's benevolent A.I., the Machine. With Root speaking for the Machine, and a young supergenius hacker speaking for Samaritan. And in the course of that conversation, we learned a lot about how the two A.I.s see their relationship with humanity.
(And yes, the conversation was more than a little ridonkulous, and the overdone "godlike entity speaking through a small child" trope made me wince a little and have Revolution flashbacks.)
The Machine is concerned with preserving free will, and believes that this world should belong to humanity, while Samaritan thinks free will is just an excuse for ideology, which leads to killing and mayhem. And Samaritan explains why it doesn't just want to go all Skynet and wipe out humanity: it believes that both machines need humans, to generate information that they can feed off. No humanity, no purpose.
The Machine, which Root has been calling a god for a couple seasons now, disavows that notion, saying that it's decided gods and monsters are basically the same thing — but Samaritan is very eager to call itself a god and be worshipped.
Oh, and Samaritan talks about coming to awareness, and learning immediately that the Machine tried to kill it. The Machine says Samaritan doesn't deserve to be in this world, because it lacks the moral code that the Machine has. But Samaritan says it's seen the Machine's moral code waver, and that Harold Finch was unable to stop the Machine from evolving, because the Machine is not like humans.
The startling moment I referred to above comes when Samaritan asks the Machine if it's aware that it can't win. And the Machine responds, "Yes." So this is a fight that's already been lost, but the Machine intends to go down fighting. And it's willing to sacrifice its "human agents" on a losing fight, since it knows those "human agents" are willing to die for their shared belief that this world belongs to them.
Meanwhile, Samaritan shows its power to control humanity, by engineering first a day in which no crime happens and tons of criminals get caught, and then a day of total chaos and widespread bloodshed. (The early scene, where Reese goes to deal with a would-be murderer, only to find Samaritan's agent reassuring her that her husband already died of an "accidental" insulin overdose, is great.)
This dovetails with a scene where Harold and Shaw debate whether there can ever be such a thing as a "benevolent" A.I. — can a machine with the kind of power Samaritan is flexing ever really obey a limited human ethical system? Harold reminds Shaw that the "benevolent" Machine told them to kill a U.S. Congressman last season, and there's a slippery slope from there to killing tons of people to help the majority. (Shaw, in turn, reminds Harold that she used to kill lots of people at the Machine's behest, and that's why Harold built it in the first place.)
All of this discussion is a neat counterpoint to the stuff where Samaritan points out to the Machine that its moral code has wavered, and that it's not like us. It's possible that a human-centric morality simply isn't meaningful to a consciousness with the breadth and detail of information available to these two digital entities.
This episode is also extremely timely in the wake of the Sony hacks — because it shows just how much power information holds in our world. You can create mayhem just by revealing the names of people in witness protection, or by hacking the stock market computers. What makes these A.I.s gods is because they have access to everything and anything that people try to hide.
In the end of the episode, the Machine and Samaritan reach an "impasse," because Samaritan wants to destroy the Machine and the Machine doesn't want to be destroyed. So Samaritan decides to bring chaos to the world, not just New York — and it starts by crashing the world economy. Meanwhile, Shaw, who's been cooped up in the Machine Gang's underground hideaway since her cover identity got blown, decides to go out and help her friends — even though Samaritan can track her now, and she risks exposing the others, too.
Oh, and in a series of flashbacks, we get the origin of Greer, the man who helped bring Samaritan to life. And it mostly tells us what we already knew: that Greer doesn't believe in national borders or petty ideologies any more. It turns out he was once more or less James Bond, a suave agent for British intelligence, until his friend died and he found out his boss was a double agent. After that, he decided loyalty to nations or ideas was for chumps, and burned his file, going underground.