I’m going to open this post with Walter Benjamin’s first thesis in his Theses on the Concept of History:
It is well-known that an automaton once existed, which was so constructed that it could counter any move of a chess-player with a counter-move, and thereby assure itself of victory in the match. A puppet in Turkish attire, water-pipe in mouth, sat before the chessboard, which rested on a broad table. Through a system of mirrors, the illusion was created that this table was transparent from all sides. In truth, a hunchbacked dwarf who was a master chess-player sat inside, controlling the hands of the puppet with strings. One can envision a corresponding object to this apparatus in philosophy. The puppet called “historical materialism” is always supposed to win. It can do this with no further ado against any opponent, so long as it employs the services of theology, which as everyone knows is small and ugly and must be kept out of sight.
In this passage, the automaton Benjamin is referring to is The Mechanical Turk, an elaborate prank throughout Europe, which involved a machine with the torso and head of a Turkish man attached, who could beat anyone at chess. The machine, however, was actually an expert chess player surrounded by machinery.
The second part of the passage is Benjamin suggesting that the Marxist view of history is that historical materialism is the machine. Benjamin then equates the man inside the machine, playing the chess, with theology. He suggests that the core of dialectical materialism is a theology, implying (and later in the text, explicitly saying) dialectical materialism has idealist features that aren’t accounted for.
I will explore Walter Benjamin’s contributions to Marxism another time on this blog, but that’s not what this post is about. The Turk represents a type of artificial intelligence “play” that can be a lens to analyze human relation with technology.
For example, The Turk immediately draws comparisons to Deep Blue. The Turk was an actual human inside machinery, but Deep Blue was the machinery itself. It was the first time a computer beat a world champion at chess.
The main way that the Mechanical Turk and Deep Blue are different is that the Turk was actually a human in a mechanical suit, and Deep Blue was a computer programmed by humans. Deep Blue supposedly made all the decisions itself (Kasparov argued some moves seemed too “creative” to be done by a machine), and the Mechanical Turk made none of the decisions, while simultaneously all of them, if you consider the person inside part of the machinery.
We can also recontextualize Benjamin’s historical materialism metaphor while considering Deep Blue, because the Mechanical Turk is emblematic of a type of feudal political economy and Deep Blue is emblematic of capitalism. We can then speculate what a future chess machine will be.
In the 2014 movie Ex Machina, an employee at a Google stand-in gets invited to the CEO’s fancy home. His job, while there, is to interact with a robot. The CEO claims the robot can already pass the Turing Test—the machine is already able to convince people that it’s behaviors are indistinguishable from a human. [spoilers ahead]
However, the CEO is interested in more. He never explicitly spells out what he’s looking for, but it seems as though he’s using our protagonist for a Lovelace Test, which is much like the Turing Test, but it’s a test to determine AI’s ability to make original ideas, rather than depending on trickery. He wants to know how “real” our protagonist determines the robot to be, based on interaction.
The movie frequently hints to the viewer that potentially our protagonist himself is a robot, and he’s being tested himself, to see how a humanoid robot interacts with a robot with a human face, but exposed robot body. This creates a lot of tension, and is highlighted by the fact that we have a slippery understanding of what, exactly, our protagonist’s goal is.
The movie never overtly gestures to the idea that the protagonist could be a robot, but you know, the movie knows you’re considering that. For example, the CEO actively encourages our protagonist to pursue a sexual relationship with the robot. This makes it seem like the real test is whether a robot who thinks he’s human would have sex with a robot who knows she’s a robot. The movie flags a lot of little moments like this, that make both the viewer, and protagonist, question the nature of these tests.
In a near climactic moment, the protagonist, after discovering that the CEO’s maid is a robot, and he has a closet of sexualized robot women, begins having the same anxieties the audience has been latently experiencing the whole movie: he wonders if he is actually a robot, and cuts his arm open to test it.
Eventually, the protagonist tries to hatch a plan to allow the robot to escape, knowing that her memories and character will be wiped out after he leaves. The CEO confronts him about this, and tells him that was the real test—to see if a robot could convince a human that they’re “human” enough to liberate.
This is a great movie, and is thought-provoking in two ways. It both makes the viewer question the criteria for determining humanity in robots (and consequently humanity in general), while also making the viewer question the humanity of humans. Where most media in this realm focuses on the criteria for determining humanity in AI, or the criteria for determining humanity in humans, this movie does both. This creates a disarming and off-putting effect where we are measuring one thing (AI) against a a paradigm (humans), while also being forced to question the nature of the paradigm along the way.
What can we learn about the Mechanical Turk from Ex Machina? By the movie exploring different criteria to determine humanity, it shows us different directions that the idea behind the Turk (namely Artificial Intelligence) can evolve.
The Turk is associated with the pre-capitalist, late feudal nature of its time. Deep Blue is associated with the capitalist nature of its time. Whether this association is purely by happenstance, or directly caused by the economic system, the association remains nonetheless. The Turk existed because it was using the technology of the time, and the technology was informed by the economic system.
What I’m suggesting is that, just as our conception of what humans are, the humanity of humans, the humanity of children, etc, change over time, in reaction to society, then our understanding of what makes AI “intelligent”, and how “artificial” that intelligence is, changes with society.
Just as the Turk was seen as artificial intelligence, so was Deep Blue. But now, not too far from removed from Deep Blue, we no longer see it as an example of artificial intelligence. It was simply a computer that was able to process the mathematics of numerous statistics quickly.
Now, we try to conceive of future ways to achieve AI: for example, not only can it have “intelligence” but it must be able to reason, AI should have intuition, AI should learn the way that humans learn etc. There’s many of different metrics for determining how AI could be in the future.
But, in conclusion, I want to tie this notion of AI back into Benjamin’s metaphor of historical materialism. I think, in the grand scope of human history. we won’t be able to understand an “elevated” form of AI, unless we radically start re-understanding society as a whole. There will someday be a paradigmatic change in the nature of what we consider AI, and I think that change will likely happen in concert with radical changes to economic systems, as that will radically change the lens through which we see the world.
To be more specific, every version of AI that I’ve discussed beyond Deep Blue are robots that emulate humans. But in reality, I think AI will end up being something more nebulous, collective, singular, etc.
The future of AI won’t be individualized, human-like figures. Of course, there will be those. But the true evolution of AI, will be more akin to a collective conscious, or something like a sentient internet.