Tag Archives: #algorithm

ex machina 4

Ex Machina, Alex Garland, 2015.

A young male programmer employed by a Google-esque company wins a contest to spend a week in the paranoically secluded abode and laboratory of the company’s grandiose and manipulative CEO. The programmer is to assess an artificial intelligence in the form of a beautiful young female robot invented by the CEO, an A.I. generated through the culling of data from millions of company user searches (desire and knowledge) and images taken from masses of phone data (affect). Can it pass muster as human, with human defined as having complex intelligence, decision-making skills, and individual personality? This is the narrative schema of the new film, Ex Machina, a film that ventures into the under-examined territory of the algorithm with a degree of criticality not found today in most artworks or even discursive texts.

In a recent article about artificial intelligence in the Science section of the NYT, the journalist’s discussion of how unlikely is the possibility that we could create the kind of autonomous A.I. robots pictured in the film could not be more wrong about what is really at stake in the discourse and imagery of the film. The journalist reads the film concretely and reassures readers that we are far from achieving A.I. in robot form because the most advanced robots in the U.S. today are “glacially slow in accomplishing tasks such as opening doors and entering rooms, clearing debris, climbing ladders and driving through an obstacle course.” Robots, according to the article, are dependent on human operators to guide machines via wireless networks, though some are semi-autonomous. They are “largely helpless without human supervisors.”

Yet that is precisely what we have to fear.

A lot of the anxiety doesn’t come from any real situation that A.I.s are about to take us over or the world is about to change because of A.I.s in any fundamental kind of way — not at the moment at any rate. It’s got more to do with big tech companies and the Internet and search engines and social media and that kind of thing. I think there’s a sense in which we feel that we don’t understand how our cellphones and our laptops work … but those things seem to understand a lot about us. Now that’s not really about artificial intelligence, it’s about tech paranoia. So somewhere in this I think I’m trying to look at that, too.

Alex Garland, Director of Ex Machina, NPR, April 14, 2015

ex machina 3

Ex Machina, 2015.

What the algorithm achieves today is set in motion by humans, but once set in motion by humans, algorithms collect data on populations at a speed and with an accuracy unachievable by the human mind or hand. The A.I.s of today are not housed in attractive “fembots.” They are housed in microchips and abstract mathematical formulas that are unsexy…to most.

An Adaptive Data Collection Algorithm Based on a Bayesian Compressed Sensing Framework

An Adaptive Data Collection Algorithm Based on a Bayesian Compressed Sensing Framework

The other thing I was interested in was the way tech companies present themselves. So Oscar Isaac’s character Nathan talks in this very kind of familiar, pal-y way. He uses the word “dude” and “bro” a lot. And I felt that this was sometimes how tech companies present themselves to us. They’re kind of like our friends. They say, “Hey pal, hey dude,” like we’re kind of mates, you know, “I’m not really a big tech company, I’m actually your friend and we’re hanging out sort of at a bar or at the beach and we’re sort of part of each other’s lifestyle, but at the same time I’m going to take a lot of money off you and I’m going to take all of your data and rifle through your address book” and that kind of thing.

Alex Garland, Director of Ex Machina, NPR, April 14, 2015

The critical dimensions of Ex Machina depend on a synthesis of the visual, the visceral, the aural, and the scripted. But like a lot of great films, Ex Machina achieves visually what cannot be enacted through language. Yes, the storyline is creepy, and its arc is that of a moral tale about A.I.s taking over the world. But the visual and affective aspects of the film draw in the spectator in ways that go beyond that of a repressible moral tale. The film’s subtle visual effects, the decisions made about just how close to hew to the visual realism of human skin and features and affect, and just how far to veer from that realism and at which moments to do so- these create a spectatorial engagement semi-separate from that of the script. Just as an utterance or an exchange or facial expression in the psychoanalytic setting can have reverberating effects on consciousness and on the unconscious, so can the unscripted aspects of film, independent from a story line, independent from its narrative ending.

ex machina 5

Ex Machina, 2015.

Granted, it’s a common trope in science fiction or cyberpunk fiction to create a frisson of sameness/difference where the human and the robotic meet.

Screen frame from A.I. Artificial Intelligence, a film by Steven Spielberg, 2001.

 A.I. Artificial Intelligence, Steven Spielberg, 2001.

But Ex Machina deploys the visual trope through a deeply intelligent understanding of how to manipulate it to critical ends.

ex machina 1

Ex Machina, 2015.

I will admit to spending a good part of the film trying to figure out just what special effects they used to achieve the creepy elision of space between human and robot in the Ava character; the physical interface between human and robot in Ava is fascinating. I was able to attend to this throughout the film because of its pacing, and I think that musing on that intermittently was actually one way of experiencing the film. I was only able to find an explanation of the effects employed through the accretion of information from a few articles after the viewing. Garland’s script is allegorical, something the science writer didn’t pick up. It’s not about A.I. per se. It’s the throwaway lines that haunt – the references to the CEO’s access to the memories and desires of hundreds of millions of us; Ava as the encapsulation of that access.  Ex Machina‘s particular ways of blurring the space between human and A.I are what gives the story its significance and its affective pull. And this is best achieved visually, so as to leave the viewer with the question (a visually composed question) of just how large they’d like that space to be.


ImageLanguage delivers its judgment to whoever knows how to hear it…It is the realist’s imbecility, which does not pause to observe that nothing, however deep in the bowels of the earth a hand may seek to ensconce it, will ever be hidden there, since another hand can always retrieve it, and that what is hidden is never but what is missing from its place, as the call slip puts it when speaking of a volume lost in a library. And even if the book be on an adjacent shelf or in the next slot, it would be hidden there, however visibly it may appear. For it can literally be said that something is missing from its place only of what can change it: the symbolic. For the real, whatever upheaval we subject it to, is always in its place; it carries it glued to its heel, ignorant of what might exile it from it…a letter always arrives at its destination.

Jacques Lacan, Seminar on The Purloined Letter,  ÉCRITS

Rereading, for a project, Lacan’s Seminar on The Purloined Letter, and a few of the many critical texts that it has generated, I could not help but think of the role of the algorithm in the ways in which “letters” circulate today.

I won’t go into the details of Lacan’s argument, and those of his critics, and his critics’ critics, but suffice it to say that some of the salient points concern the circulation of signifiers. In the Poe story, a letter written to a Queen, containing what is alluded to as compromising language were the King to see it, is stolen under the Queen’s gaze and the unseeing eyes of the King by a Minister who replaces her letter with his own. She then hides that letter by crumpling it up. A Police Prefect who is offered a reward to retrieve the letter then offers to pay an amateur Detective to steal the letter from the Minister, as the police have not been able to locate it in spite of having searched the Minister’s accommodations thoroughly. The Detective surmises that the Minister has “hidden” the letter by leaving it in plain sight, proves himself right on first visit, and arranging for a second visit, manages to distract the Minister and replace the stolen Queen’s letter with one of his own. At no time is the substance of the letter described in the story. It is the signifier which circulates in relation to power and powerlessness, to femininity, the circulation of disclosure and concealment as they draw the characters into the established symbolic order.

Here are diagrams found online, from what looks like a students’ college course site. What can I say? It’s a complex story line:

chart_2chart_1Lacan, and some of his critics, analyze the story in relation to concepts of floating signifiers, the notion of subjectivity assumed through mis-recognition (i.e. a “letter” arrives at its destination as soon as someone receives it and presumes to be its recipient, correct or not), and the debunking of teleological illusionism (i.e. events that seem to have been pre-ordained are understood as such through a retrospective logic by the one who has received a “letter,” intended for them or not). The concept of mis-recognition, as rarefied as it sounds, actually has a great deal to say about the very popular process of how subjects today do or don’t assume political subjectivity.  Zizek’s quote from my previous post is apt. Those who recognize themselves in the Tea Party, for example, do so not necessarily because they assume an informed and educated identification with those anti-government politics. In fact, polls (for whatever they’re worth) indicate that many who identify with categoric anti-government politics (and I don’t mean anti-this administration, but anti-government overall) mis-identify which programs are state-funded at the same time as they identify with the opposing rhetoric. Somewhere in there lies resistance to assuming informed subjectivity, ironic in a discourse about self-determination.

In oblique relation to these analyses, the algorithm today can be viewed as functioning like one of those letters that always arrives at its destination. As adults, today we no longer receive information (and I use the term information here indiscriminately) in a collective sense, other than through those media that cross our paths without our choice – magazines and newspaper headlines in a subway kiosk or doctor’s office, billboards, crawling type in outdoor urban spaces, logos on building surfaces, etc.

zen_1The rest of the time we seem to make “choices” by selecting our customized news, entertainment, information, data, material items for consumption, etc. We create our entertainment through subscription processes, particularly as they are now distributed through the internet, rather than cable. Even radio has become exponentially expanded, non-regional, and customized through subscription satellite. And there now exist 50 sites similar to Kickstarter – “crowd-funding” sites that allow one to fund one’s personal future entertainment, reading material, and events. Through the “choices” we make – and including those of material consumption of all kinds – the algorithms cull information from the many sites we use to construct the “letters” that are sent back to us on virtually every page we search online. Customized advertising. When we click on those links, we recognize ourselves as the recipients of those “letters.” Hundreds or thousands of people reading a blog at the same time will be addressed by completely different ads, at the same time as the reading of that blog may partially determine the next advertisement you see, as the algorithm “monetizes” the intricate puzzle pieces of the shared desires, or the desires it presumes to exist. Now, there may be spaces of disjunction in the puzzle games played by the algorithm. The algorithm is a machine we humans have created, and it has its own fallibilities and quirks, as well as its overdeterminations. But the dynamics involved in mis-recognition are not inconsequential.

blog magIt has become commonplace for pop and media culture critics to extol the virtues of the digital in relation to political resistance. And some of those claims are legitimate. But the lack of a shared information realm – irrespective even of the question of quality – has its sinister consequences. The dispersion involved in technologized globalization often ends up producing isolation, as it creates pools of coherence. Perhaps that creates a more crucial opening for cultural production, since we are sharing less and less in the political realm.

It seems crucial to see the function of the algorithm as a floating signifier, in addition to its function as an applied instrument of capitalism today. We miss the symptomatic, and symbolic, aspects of the algorithm at our peril.

%d bloggers like this: