Some words on the criteria of judging movies:
In praising a meritorious film, there are a lot of words that we can bandy about, and while ultimately all of them mean something similar, i.e. "Watching this movie was an experience that made me feel like I had spent time well and was better for for it", they aren't all the same. I can and have referred to movies I found tremendously worthwhile as "important", "interesting" or "the best/good", without much thought. But what exactly is a "good" movie? There are a lot of good movies. Most (not all!) of the movies fêted with Oscar nominations are good movies, in that they are built according to an agreed-upon set of rules which make the establishment of narrative and character easy for the viewer to understand, and that according to these rules they are entirely functional and unproblematic. There are other ways we could call a movie "good", of course: it breaks the rules in a way that is coherent and driven towards an end which is achieved successfully; or it furthers a theme which is desirable and satisfying, and it does so without specifically violating any standards of aesthetic.
Now, like I said, a film can also be "interesting", which I personally use in the broadest sense to mean that it does something which is compelling and worthy of further consideration, because it seems like what has happened within the confines of the film is not easily classified or ignored (it is probably a more subjective thing to judge than "goodness"). An interesting film doesn't have to be "good", and a lot of them aren't. But interesting, I think, is generally a more valuable quality to possess, for any thinking viewer: it demands more of the mind and generally leads to more fulfilling thoughts later on, once the initial flush of seeing the movie is gone. Michael Clayton, by any yardstick, is a good movie, that in the main is not very interesting. From the same year, Across the Universe is fantastically interesting, although I'm not uncomfortable with the claim that it's nowhere near as good as the Best Picture nominee.
"Important" would simply mean that a film demands attention and must be grappled with, because understanding what it does promises to unlock not just itself but significant truths about the art form as a whole (in an historical context, "important" can also mean "influential", but that's not what concerns us here).
All of this is a roundabout way of apologising for a lie I've been perpetrating for a while now: I've put together a list that I called "The 100 Best Films of the Decade", and any normal person would take that to mean "The 100 Most-Good Films of the Decade"- because that's exactly what it does mean. In fact, I have really put together a list of "My 100 Favorite Films of the Decade", or better yet, "The 100 Films from the Decade That Had the Greatest Effect on Me, Ranked from Most to Least Affecting". It's a combination of interesting and good and simply how much I liked watching the movie, combined more or less arbitrarily. Which makes the list more personal for a certainty, and hopefully more stimulating.
But, don't think for a moment that I don't know this to be the truth: the 2001 science-fiction fantasy A.I. Artificial Intelligence, which I am brazenly calling one of the best films of the 2000s, isn't. It is a massively broken film; a film that walks up to the edge of outright fiasco, leans its head over, and waves. And by the way, there is absolutely nothing in this life that I adore as much as movies that almost completely fail, but don't actually fail in any one way - for they are by far the most interesting. And A.I. is so particularly interesting given some particulars of its creation, that I shall happily call it "important", maybe even the single most important film of the last ten years. But good, I just don't know, and I just don't care; because a "good" A.I. would be of infinitely less value than the A.I that we have.
(Though I think the claims of the movie's badness are readily overstated; especially the claim that the last act of the film somehow invalidates the rest, or comes after a "false ending", or worst of all, that it's sickly Spielberg sentimentality, a view that is based on on a widespread and totally indefensible misreading of the last scene and nothing else).
Let us revisit the details of this film's origin, because it has been a long time and though the story was everywhere in 2001, not everybody was paying attention: the great Stanley Kubrick had an idea to make a film adapted from Brian Aldiss's story "Supertoys Last All Summer Long", with story adaptation done by a number of people, of whom only Ian Watson was credited. There were significant practical reasons why Kubrick wasn't entirely prepared to take on this project, though, and here enters a filmmaker who had at some point met and befriended Kubrick, Steven Spielberg. The collaboration of one of cinema's great cynics and Hollywood's most eager-to-please showman is hard to believe, all these years later, but it obviously worked for the two men: Kubrick eventually suggested that Spielberg himself take over the Aldiss adaptation, since his rapid shooting schedules, rapport with child actors, and understanding of gooey emotion all seemed necessary to the project's successful execution. When Kubrick died unexpectedly in 1999, Spielberg immediately put the film, which had (I think) by now picked up the name A.I. (Artificial Intelligence came later), as his tribute to the late master. And that gives you enough to go on.
Love or hate either one of the men, it's fair to say that Kubrick and Spielberg are two of the most instantly identifiable auteurs in all the history of cinema; certainly, no collaboration between two filmmakers of such distinctive and antithetical concerns had ever been attempted in all my knowledge. The common idea that A.I. was a Kubrick project that Spielberg took over seems well-contradicted by the known history, and if Kubrick had lived, I think that the film would have ended up being a fairly unambiguous Spielberg movie, with some Kubrickian insights. But he didn't live, and Spielberg, in what was doubtlessly a fit of sentiment, decided not to make a Spielberg movie: he made a Kubrick movie instead, except that he was, at the end of the day, still Spielberg. The result is a profoundly fascinating movie that has, in a sense, no author, although it comes about this by being one of the most author-driven of all Spielberg's films (it is one of only three films for which he received a screenplay credit: the others are Close Encounters of the Third Kind and Poltergeist). It shares elements of both Kubrick's style and Spielberg's; however it is not at all the film that either of them would have, in any normal situation, created. And it's not "the average" of the two directors, either, but an unshapely hybrid that keeps trying to achieve a certain iconic Kubrick-ness that Kubrick himself would surely have avoided in such bald terms, while constantly undercutting itself with flashes of the actual director's panic that he was making something more mature, darker, inhumane, or whatever, than he was comfortable with. It very nearly tears itself apart; but the fact that it doesn't, that in fact it is a coherent and whole story with a thematic arc that can be tracked from first frame to last without a single skip, is in and of itself what makes it such a staggering creation - a staggering achievement, I nearly said, but who really achieved anything? There is my tale.
The fragmentary and highly episodic nature of A.I dictates that one can either give its situation in broad strokes, or in point by point detail, but little else; yet I'm going to try to split the difference: a narrator (Ben Kingsley) tells that a man-made climate catastrophe has left much of the world's coastlines drowned in water, and the governments of the world instituted strict population-control efforts to keep the much-reduced resources in check. As a result, many of society's basic functions are now performed by human-simulacrum robots, called mechas. The film opens with the president of Cybertronics, Professor Hobby (William Hurt), declaring that the need is now obvious for mechas with the ability to mimic familial love, which would of necessity also be mechas which can adapt, learn, and develop self-guided behavior, so far an impossible feat. Months later, his company has created a prototype named David (Haley Joel Osment), designed to look like an eleven-year-old boy. David is given to Cybertronics employee Henry Swinton (Sam Robards) and his wife Monica (Frances O'Connor), whose own child Martin (Jake Thomas) is in an irreversible coma. After much thought, Monica activates David's irreversible love program, shortly before Martin miraculously recovers. The subsequent trauma in the Swinton family leaves Monica to abandon David in the woods, knowing that if she brings him back to Cybertronics, he will be destroyed.
David's only goal is to prove himself worthy of Monica's love, and so, inspired by Collodi's The Adventures of Pinocchio, he embarks on a quest to find the Blue Fairy and become a real boy, aided by the prostitute mecha Gigolo Joe (Jude Law), framed for murder. David and Joe flee a Flesh Fair, where conservative humans destroy mechas as abominations, and eventually track down the possible location of the Blue Fairy, leading them to Cybertronics, where Hobby is ecstatic over David's development, and unwisely allows David to discover a roomful of perfect replicas of himself. He flees Cybertronics and ends up in the submerged ruins of Coney Island, where he finds himself face to face with a statue of the Blue Fairy in the park's Pinocchio garden. For 2000 years he stares at her, entreating, and eventually he is rescued by super-advanced mechas that have been looking through the arctic wastes of the Earth for any traces of the long-lost human civilization.
The simple meaning is that this is a film about love driving an individual to quixotic ends. That's the easy part. The exact nature of that story, including what it means specifically, is a real mess to figure out, because of the inherent tension between the Kubrick-meaning of the story and the Spielberg-meaning. Sentiment dictates that we read David as a tragic hero, beautifully pursuing the only thing that matters, love; and this is a reading borne out by several incidental details throughout the movie, but it also simply doesn't hold up under scrutiny. The darkest possible reading of A.I., and not for that reason the one that I tend to favor, is that it is a story about determinism, because as much as Spielberg apparently wants it to be otherwise, David is not a free agent. He does not love Monica and pursue her approval because it is his will to do so; he is compelled to love Monica for reasons that he cannot control, and by the same token he cannot ever allow another impulse to trump his love for Monica. He is, in effect, a purely obsessed being, and this is the one point, oddly enough, where Kubrick and Spielberg overlap; both of them show a healthy tradition of obsessed protagonists, though Spielberg seems to forgive or not even notice this trait, particularly in Close Encounters, in which the "hero" is so obsessed that he terrorises and then abandons his family in pursuit of a childish whim (to be fair, in later years Spielberg came to recognise this as the "accurate" reading of that film).
What looks like it ought to be a tribute to filial piety is absolutely nothing of the sort: it is a faintly horrifying tale of self-destruction as complete as anything in either director's canon, and I think the best way to sum up David's arc is that, for all the times I've watched the movie, I still don't know which of three possible moments is the one where he has his irrevocable psychotic break with reality - the possibility that he doesn't have a psychotic break is almost not worth mentioning. My tendency is to favor the scene where he discovers the only other functioning David in the movie, and goes positively apeshit, destroying the other mecha in a violent, shocking scene that is much the most disturbing moment in a film; it encapsulates the pure rage of a being whose sense of identity, based on his perfect uniqueness, has been shattered, and shows how his response that realisation is to attempt what amounts to a form of suicide (in this scene, it must be noted, "our" David keeps swinging his weapon in the air long after the other David has been decimated, attacking- what? The way the scene reads to me is that he's swinging so energetically and wickedly that it's as though he's hoping that he'll fly apart himself).
The film's insistence on the horrifying conundrum David faces, that he is a thinking, feeling entity that cannot ever get away from that single, unfulfillable emotional need, is a constant, and not even very subtle; and that brings me to the final scene, probably the most contentious moment in the film. The future mechas explain in some unfathomable technobabble that they can bring Monica back to life, with her memories and personality intact, for one day; at the end of that time, she will die again, permanently and the essence of her being lost to the cosmos. They propose this as a means to placate David, so that he will help them in their anthropological efforts; he just wants, badly, to see his Mommy again. They bring Monica to life, and she and David have the perfect day, that seems to give him the feeling of being a real boy, and this is often read as Spielberg's forced happy ending (it was, by all accounts, present in Kubrick's treatments). Despite the fact that: David is probably quite insane by now; Monica is going to die; the whole sequence pushes every last creepy Oedipal button possible it can, right up to the fact that when Monica dies, she's lying in bed next to David. There are no comforting overtones to this scene other than the narrator's puffy statement about "the place where dreams are born", which reads something like irony given the rest of his speech, which describes Monica's death in storybook terms and implies that David is either dead or completely shattered as a personality. I genuinely don't think Spielberg intended for this reading, but that doesn't mean it's not there, and there seems to be no possible way around the conclusion that we've just watch a man plunge finally into the abyss, with only his appearance as a child (a fake child, at that) to distract us from the nihilistic urgency of this fact. This ending, in which David's obsession finally obliterates his self, is the proper ending to the story that has been building all along, not the potentially trendy, rather more playfully nasty ending where he's trapped forever on the bottom of the ocean, that several people seem to think would be the best place to stop to movie.
Incidentally, about those future mechas and their plans for David; it's a fact that doesn't seem to get enough attention that every figure who interacts with the boy does so for their own ends. Henry and Monica want a son, and while Monica plainly feels a lot of guilt about what she does later, they don't seem to waste a lot of time discarding him when he becomes difficult, rather than spending time trying to understand his needs, like you would with an actual child. Gigolo Joe uses David as a ward to keep off the police. Hobby is so pleased by David's functional achievements that he completely fails to realise that as a direct result of those achievements, the mecha has become a sentient entity to be counseled, not a machine to be exploited. The future mechas pretty much just view him as a walking, interactive museum exhibit, and they're willing to grant him what is clearly the unhealthiest wish in the world in order to get what they want out of him. The whole universe is lined up to objectify and exploit David, in essence; that's a level of fatalist cynicism that even Kubrick didn't ordinarily wallow in, more the territory of Ingmar Bergman in his really dark moods.
In contrast to all this bleak nihilism, Spielberg (I assume noticing that his material was getting away from him: nobody can be quite that unaware) keeps ducking back to his comfort zone, trying to find ways to add sentiment to a brutally unsentimental story. In keeping with his most characteristic work, the Swinton's home (I'd call it suburban, except that word seems to lack meaning in the A.I. universe) is a beacon of domestic peace that is quite literally equated with heaven, as an object of the pilgrim's progress and then at the end as the place where he ends up after he dies. The final act keeps placing odd notions about human beings as great dreamers and achievers into the mouths of the future mechas, sentiments that the film apparently agrees with, but which stands at stark odds to every last frame of the middle act (what we might call the "Gigolo Joe" portion of the story, since it is almost precisely defined by his entrance and exit). And in keeping with his treatment of child actors all the way back to the 1970s, Spielberg seems hellbent on making David seem a perfect innocent, framing him like a delicate, sympathetic, lovable being - thankfully, Osment seems much more aware of the needs of the material than his director and gives a performance that mostly scuttles Spielberg's sentimental attachment to childhood. Indeed, Osment's performance not only seems impossibly complex for even the most precocious of twelve-year-olds, I'd hesitate not one instant in calling it one of the finest performances, male or female, of the whole decade. It's physical flawless - there's not a single moment when he doesn't remind us in some subtle way that he's playing a robot - and pitched at exactly the right level, so that even David's sweetest, most affectionate moments have a somewhat indescribable aura of uncanniness and dread about them.
Now, this inherent tension between the way that Spielberg wants to tell the story (as Kubrick reincarnated) and the way that all his instincts as a commercial filmmaker demand that he tells the story (I didn't really delve into that enough, sorry - but Christ, this is a long review, don't you think?) could certainly mean that it's a failure and a wreck that doesn't know if it's coming or going, and if someone watched the film - really watched it, in good faith - and still felt that it was those things, I would concede that I can't really argue against matters of taste. But this tension could also mean that A.I. is the single most fascinating case study in 115 years of cinema on exactly what "authorship" means - a film made by one of the strongest proofs of auteur theory currently working that paradoxically seems to knock all the pins out from underneath the mere idea of auteur theory. Kubrick obviously didn't make A.I., but in a very real sense Spielberg didn't either; he shepherded it, and it clearly got away from him quite often.
Still, the film is not at all as dark as I've probably made it out: for all that Spielberg eschews the usual tricks of his trade, the warm glows and awe-filled onlookers that make his movies such friendly spectacles (there aren't but a dozen or so quintessentially Spielbergian shots, and the director's collaboration with the endlessly experimenting Janusz Kaminski here yielded some frames with enough objective chilliness that they might even fit in a Kubrick film - indeed, a goodly number of images seem to be deliberate quotes of Kubrick), he can't embrace the dark contours of this cruel utilitarian universe - not though Osment clearly jumps right in, and not though the reliably Romantic John Williams graced A.I. with the least typical score of his career, a jerky, borderline-atonal soundscape that leaves one feeling hollow far more often than comforted. He pulls back, constantly, and the one time he fully embraces the terrifying potential of the story, the result is the horror film-inflected Flesh Fair - by far the film's worst scene, jangling, loud and mostly pointless, with Holocaust overtones that are breathtakingly inappropriate, especially from this filmmaker of all filmmakers. Though, for what it's worth, I think without pushing himself into the darkness of this story, and try as he might to keep from allowing his sentiment to gush out, was the act of personal discipline necessary to permit Spielberg to make the run of such intriguingly mature films as he did throughout the rest of the decade.
The insistent Spielbergian overtones that keep butting in are what makes A.I. a mess, but its messiness is what makes it absolutely essential. Damning it for its failings is sufficient, I guess, if you just wanted to read it as a sci-fi narrative about the moral nature of artificial life, because superficially, that's exactly what A.I. is and how it has been read; but honestly, even as a lover of sci-fi, I can't call it particularly groundbreaking in that regard (David is plainly little more than HAL-9000's grandson). I don't agree that it's a failed sci-fi movie, mind you. But its generic concerns are secondary to that thing I mentioned up top - that "interesting" thing - precisely because it is such a patchwork quilt, A.I. almost can't be exhausted: its revelations about the intentional fallacy give way to its troubling narrative of free will which in turn raises questions about the interaction of script and the visual and from there we have to wonder whether lighting or framing or POV is the "key" element of a shot and thence...
Of all the films of this decade, and by all means there are plenty that I like watching a great deal more, there's probably not another one that I've found so perfectly rewarding to watch and re-watch. Ultimately, it's impossible to make sense of it, but it so woozy and rich with possibilities, both narrative and extra-narrative, that it always feels like that very next try is the one that's going to do it; it is the one film of the decade that completely, if accidentally, averts any possibility of a solution. It is itself; hypnotic and draining and frustrating and like nothing else ever made.