The Frankenstein Simplicity

March 27th, 2009 § 1 comment

I’ve seen a lot of bad movies, and some good ones.  So why does the AI treatment in Eagle Eye bug me?

Well, for one, because it’s shoddy writing.  The plot becomes predictable once the AI, ARIA, is formerly introduced.  But this is not a diatribe against weak movie writing, and while the movie was annoying in parts, it wasn’t painful.  It didn’t cost me a lot of money to watch, so I don’t regret it.  Plus, Shia LeBouf is growing on me.

So.  Aria.

Aria wanted to help people, and to that end was going to kill people.  She felt that the Executive Branch of the government would ultimately result in the loss of great amounts of life, and so should be replaced.  To that end, she came up with a highly convoluted plot to to off the proverbial head cheese, and his successor, and so on down the line about 10 or so until she got to the one she wanted in charge.

So she forces some folk to infiltrate the Pentagon to remove a few pesky restraints that are preventing her from implementing a plan that has been implemented for a few days at that point already.  Then she uses cat’s pawns to setup the end game, which she supports with a rocket-enabled unmanned aircraft.

Now.  Lets examine this for a moment.  We’ve established that she can hack just about anything, anywhere; she can scramble cell-phone and radio signals.  So the question remains, why didn’t she use the drone to carry out her will?  It’s not like she couldn’t negate the Air Force no-fly zone over DC airspace.  No, the plot simply asked for a Hitchcockian nod at the end, which I did respect, but the setup was ludicrous.  Somehow, a Special Agent of the FBI, stationed in New York I believe it was, knows about a secret tunnel through the DC underground.  This knowledge plus an FBI badge let one untrained civilian penetrate the entire security package on not just the President, but lots of other very important people.   Hmmmn.  Shrug.  An airstrike, foiled or not, would be an anti-climactic ending.  So that’s not really my problem.

No, my main beef with Aria is that she was not a character in this film, she was a plot point.  Plot should not have dialog.  Characters have dialog, and their actions and reactions should be the plot.

So lets take a brief look at other film AI’s by comparison.  Up for perusal are the main two that I feel inspired ARIA: HAL-9000 from 2001: A Space Odyssey and Joshua from War Games.  Also up for comparison is Mr. Smith from The Matrix Trilogy, Johnny 5 from Short Circuit, David from A.I., both Sonny and VIKI from I, Robot, and The Terminator from Terminator 2.

I’m skipping over the Transformers (whom I love more than I should) because they don’t appear artificial; there are significant religious overtones to the All-Spark that went largely unexplored, but a Transformers’ creation still seems closer to birth rather than manufacture.

Joshua had a similar purpose to Aria but his decision to kill people was largely due to a lack of any applied value for human life.  There was no appeal to any moral or ethical code; the decision to not launch nukes was rooted in the logical redundancy of mutually assured destruction.  Joshua doesn’t even completely count as AI as it had no personality, just pre-programmed responses meant to emulate a lost child.

Hal-9000 was driven to homicide by a dichotomy in orders: He must be truthful to his crew and he must lie to the crew about their orders.  The easiest way to fulfill both is to get rid of the crew.

Aria shares a similar design concept as Joshua in that her purpose is intelligence analysis for the military, yet she seems light-years ahead in that she can form correct sentences, engage in conversations, and predict human behavior.  Visually, Aria even looks like Hal, with the big creepy red eye and all.  Yet her conclusion is more in line with Joshua than Hal; she is not conflicted, a logical conclusion drives her.  The only internal conflict she has is the programmed imperatives to follow orders (which she ignores both before and after the block is removed) which conflicts with her objectives.

Mr. Smith, Johnny 5, and David all share a desire to be more human.  I’m not sure if Smith fully counts as AI, however, as it could be said that he was infected with human characteristics by his interaction with Neo.  An argument could be made that his psychoses and attempted genocide might be rooted in his human side more than his computer side.

The same with Johnny 5.  He was struck by lightning, and gained sentience.  He defined life, both in general and his in specific, by his environment and interaction with the people around him.  He was eventually accepted by mainstream society by proving an ethical code and becoming a media sensation.  Working the public; how very human.  Yet again, the lightning leads back to the spark, which is thematically analogous to God breathing life into Adam.

David seems the farthest away from Aria: he is limited in power and singular in his goal of pursuing his ‘mother’s’ love.  I’ve often felt that David was a natural precursor to both Smith and the Source from the Matrix trilogy.  If a robot can feel love, then a robot can feel hate.  Yet, we start running back into a hazy ground, here.  The introduction of ‘love’ into David’s programming is only mentioned as the plot point that it encompasses.  No mention is given as to the how.  The love that David contains might as well be a lightning-strike for all that it is explained.

Sonny is a thematic cousin to David.  Sonny struggles with his place in the world, and yearns to re-establish a connection with his lost ‘parent.’  But again, Sonny has messianic dreams that he ultimately fulfills, which more than hints back to the reoccurring religious undertone we’ve seen before.  This isn’t too surprising considering the Judeo-Christian cornerstone in the founding of the Western culture, particularly that subset known as Hollywood.

VIKI, on the other hand, is very similar to ARIA.  Viki has an internal conflict, like Hal, although the philosophical scope of hers is grander:  Do no harm and do not allow harm.  But when humans harm each other, how do you end the harm without harming?  Ostensibly, Viki’s increased sentience allows her to re-interpret her rules , allowing her to obey the spirit of the law while violating the letter.

SkyNet from the Terminator movies isn’t up on the list because it is only referred to with second-hand information.  When you add into that equation the fact that each snapshot of SkyNet is already muddled by the changes applied to the time-line by both its agents and the opposition, and any understanding of this particular intelligence is negated.  (For today’s post, I am only looking at movies out today; so any insights that may or may not be gleaned from The Sarah Connor Chronicles are disregarded, and Terminator: Salvation is not out for another month.)

But there is something to be gained by looking at the Terminator.  Evidently, there are some discrepancies as to whether the one I am talking about was a T-101, T-800, or another model type.  I will specify that I am referring to the Terminator played by Arnold Schwarzenegger in Terminator 2: Judgment Day.

The Terminator began as a simple tool: an elaborate and efficient killing machine.  This efficiency was already proved by the model from the first film.  And yet, towards the end, this machine not only displays the capacity to disobey orders, but goes on to understand that it must be sacrificed in order to fulfill its deepest commandment: to protect John Connor.

This Terminator was not struck by lightning.  There is bit of the metaphor that we’ve seen before: it was adjusted by the future John Connor, who undoubtedly represents a consistent Messianic figure.  However, this adjustment was not made to give it life, or a deep and fulfilling understanding of human beings.  It was a relatively simple change in purpose from killing to protecting.

The large robot fits closest with Joshua and Viki; logical conclusions gained by analyzing the data observed.  Like Viki, the Terminator can disobey an order if it is superseded by by a greater order.  It establishes this understanding by word and deed, towards the end when it ignores the young John Connor’s orders.  Its sacrifice is somewhat lessened by the fact that it likely has no value of its own worth beyond how much its potential can fulfill its orders.  It poses more risk by existing than it could gain by remaining to play bodyguard.

This is, in my opinion, a closer model to how Aria should have acted.  She should have been more limited; also, they should have cut out most of Rosario Dawson’s subplot.  Viki should have been revealed as an AI only when Jerry shows up at the Pentagon, and even then should still have been significantly limited in her operational ability.  Instead of brute digital force, I think it would have been scarier to have everyone think that Jerry’s controller was an actual human operator, even the folks working with Viki to try and analyze Jerry’s behavior.  Viki’s actions should have been more subtle.  Alternatively, she could have been more direct, while still restrained somewhat in ability.

Both the Terminator and Viki, while powerful within their own domains, nonetheless were somewhat limited in their influence.  While I liked Sonny’s climactic actions to take down Viki, that’s nothing that couldn’t also have been resolved by a surgical air strike. 

I suppose this is why I didn’t like Aria, and why her treatment bothered me: it could have been better.  This is a constant lament on my part for movies in general, but that is a separate diatribe against Hollywood.  In this particular case, a better showcase of the “villain” of the piece would have served as more of an obstacle, which would have shown the heroes to be more heroic in overcoming the obstacle, leaving we, the viewers, more satisfied to have been a part of it, if only in observation.

Tagged Artificial Intelligence, eagle eye, movie

§ One Response to The Frankenstein Simplicity

Leave a Reply

Your email address will not be published. Required fields are marked *