You will be redirected back to your article in seconds
Back to IndieWire

Critic Knows Best (Or Thinks They Do, Anyway)

Critic Knows Best (Or Thinks They Do, Anyway)

The following post contains spoilers for “A.I. Artificial Intelligence.”

Film criticism, by its very nature, is a hubristic act.  When you get right down to it, the critic isn’t really talking about a film; he’s talking about his or her vision of a film.  To the critic, a director’s intentions matter less than a director’s execution.  A director might argue that he’s made a movie in favor of progressive politics; the critic might see the same film as a defense of conservatism.  In the eyes of the critic, it’s he or she — not the director, not the audience — who knows best.  “Here’s what I saw,” the critic says. “And here’s why that’s good or bad.”  

All criticism involves varying degrees of evaluation and interpretation.  That’s the job.  But here’s where it gets tricky: how much evaluation and interpretation?  If the critic knows best, how far should they take that critique?  Is it enough to say “Here’s what went wrong?”  Or should the critic say “Here’s what went wrong, and here’s how to fix it.”  Or is that going too far?

These questions came to mind as I read David Daw‘s piece on io9 entitled “How ‘A.I. Artificial Intelligence’ ends in a different, better universe.”  The different, better universe in this case is one invented by Daw, who argues that the original ending of the film written and directed by Steven Spielberg is inferior to the one temporarily imposed on the film by a freak bolt of lightning at the author’s first viewing of “A.I.” a few weeks into its theatrical run back in 2001:

“After the permanently child-like robot David spends the entire film searching for the Blue Fairy he finally finds a statue of the fairy in a sunken Coney Island attraction and prays that “she” will turn him into a real boy. Then… KRAKA-THOOM.  A bolt of lightning hit the theater and the screen shut off… After the movie came back, ‘A.I.’ took a hard left turn into Spielbergian sentimentality as David is awoken 2,000 years in the future by advanced robots who clone his mother so she and David can spend a final, heartwarming day together. It was a bizarre shift made all the more bizarre by a thunderstorm that wanted to make sure I never saw it… I spent the next few days insisting to my friends that the ghost of Stanley Kubrick hit our theater with a bolt of lightning… The ambiguity of leaving David at the bottom of the sea wishing eternally for a life as a real boy doesn’t just avoid the weird far-future coda, it’s actually a stronger ending that resonates more with the film’s themes.”

Daw isn’t saying Spielberg’s ending is weak, he’s saying Spielberg’s ending is wrong.  In his mind, the right ending would fade to black after child robot David (Haley Joel Osment) gets trapped at the bottom of the ocean with the statue of the Blue Fairy he’s spent much of the film looking for.  David wants to become a real boy so his “mother” — a flesh and blood woman he’s been programmed to love, but who’s discarded him in favor of her biological son — will love him once more.  In Spielberg’s ending, after 2,000 years trapped under the Atlantic, David is discovered by a race of advanced mechas, who grant David’s wish: one more perfect day with his mother, temporarily recreated from a strand of DNA. “If you give David what he wants, then you defuse the drama of who he is,” Daw writes. “Instead, the best option is to leave him at the bottom of the ocean in a weirdly poignant program loop, trying forever to exceed the bounds of what he is without ever actually being able to.”

Now I could spent a couple hundred words explaining why I think Daw is wrong.  I could point out that despite Daw’s belief that Spielberg schmaltzed up a story by the icy, dispassionate Kubrick, the opposite appears to be true.  Or I could observe that for Kubrick, “A.I.” was ultimately a modern day retelling of “Pinocchio;” Pinocchio’s story ends with him becoming a real boy, therefore the same must happen to David.  Or I could note that the superficial sentimentality of Spielberg’s “happy” ending disguises bleaker undertones: humanity is dead, replaced by machines who give a robot a temporary moment of joy that will likely be followed by an eternity of obsolescence, if not outright decommissioning.

Now I could do that, but I won’t (okay, yes, I kind of just did anyway.  Let’s not dwell on semantics).  The point isn’t whether Daw is correct or not, but whether any critic has the right to re-edit a film, even in their own mind, to suit their interpretation.  After all, the critic doesn’t really care what the director wanted to make, only what the director actually made.  And in Daw’s opinion, the film Spielberg actually made is too maudlin.  I’m sure if we polled many “A.I.” viewers, a hefty percentage would agree with him.  But maybe others would disagree and claim that “A.I.” was too Kubrickian, and that except for that one scene, Spielberg lost himself in another filmmaker’s material.

It might be worth noting here that Spielberg has revised his films before — as when he removed the guns from “E.T.” or changed the title of “Raiders of the Lost Ark” (or at least permitted the title to be changed on the DVD box) — and that when men like Spielberg or George Lucas do alter their movies, they typically receive more criticism, not less.  Critics often argue in favor of the changes they believe in while demanding other films remain preserved in their original form for posterity, like David frozen forever beneath the Wonder Wheel.  It seems somewhat hypocritical to want it both ways.  Now this is getting confusing.  Let’s not forget that all critics think they know best.  It’s just that they know best differently.

Daw’s approach is not all that dissimilar to the one employed a group of critics I greatly admire, who I dubbed a few months ago “film critic filmmakers.”  They are editors like Peet Gelderblom, who transformed Brian De Palma’s “Raising Cain” into “Raising Cain Re-Cut,” a version of the film that supposedly moves it closer to De Palma’s original vision of the project before he radically revised its structure during post-production.  Looking for the earliest progenitor of this wave of film critic filmmakers, I named Mike J. Nichols — better known by his nom de Avid, “The Phantom Editor” — who created the Gwyneth Paltrow in “Contagion” of Internet viral fan edits with his version of George Lucas’ “Star Wars: Episode I – The Phantom Menace.”  So there’s a critic offering suggestions to a director who already had a reputation for revising his work.  I wonder if Nichols thought Lucas might actually take some of his critiques to heart.

So if I’m so inspired by the work of film critic filmmakers, why was I slightly troubled by Daw’s article?  Maybe the problem was Daw’s piece felt closer to telling me what’s wrong than showing me what’s wrong, and in movies you’re always supposed to show and not tell.  Maybe if he’d made a version of “A.I. Artificial Intelligence” that removed the “Spielbergian sentimentality,” and I could consider it for myself, I might be more inclined to agree with him.  Or maybe I’d have to make my own version next.

This Article is related to: News and tagged , , ,



This article contains two conflicting statements that make it difficult for me to completely understand what the author is saying. Mr. Singer at once claims that "When you get right down to it, the critic isn't really talking about a film; he's talking about his or her vision of a film." but in the same essay says "After all, the critic doesn't really care what the director wanted to make, only what the director actually made." A critic cannot at one time both re envision the film and analyze the concrete details presented. S(he) might be able to do them one at a time, in whatever order s(he) chooses but s(he) can't do them simultaneously. I say all this to make the point that you can not criticize the critic (good Lord, we are crawling up our own behinds here) for making his own interpretation AND for dealing with the actual film at hand. Am I the only one that sees the paradox here? I might be misreading myself. That would be ironic.

As far as a critic knowing best, I don't know that critics in general make that claim, although I am sure there are those that do. I think any critic worth his salt constructs his argument based on the film on the screen and then makes his judgement and backs it up with evidence and reasoning. A lot of commenters seem to take umbrage with that but its the reality. Art is subjective and the real value of a "professional" critic is that he/she can bring knowledge and personal experience beyond the "average" person. It's much like basketball. Sure anyone can play but some play much better than others. If art is something that can be judged objectively then we are wasting our time. Let's begin training the robots.


All more indication that the "famous critic" is an oxymoron and, like printed newspaper, a fading necessity. "Daw's piece felt closer to telling me what's wrong than showing me what's wrong, and in movies you're always supposed to show and not tell." One of the biggest offenders of this rule had once been Roger Ebert, particularly in the 1995-2005 years. He has since mellowed.


The author/writer always has a particular idea/situation he/she wishes to convey to the audience. When a studio decides to adapt that (if novel/novella/short story) or go with a story written for the screen (less interference there b/c w/ screenwriting only one has an idea of what cinematographers, directors, producers, etc. deal with in trying to bring a story to the screen…. more or less) the execution depends upon the following: first the director (possibly the writer or there's a director/writer deal happening in some fashion) has to respect the story itself and think that at least 75% (IMO) to 99% has to be apparent in the final execution in some way. 75% allows for at least the third of a story to be converted to something else that makes the "screen" version flow better for whatever reason (IMO the development here might be to develop characters more or the situation being faced, personal preference for the former considering the recent "The Hunger Games" release. LOVED it. Felt certain moments could have been sacrificed to develop pointed moments in the Arena esp. those with Katniss and Rue, and/or been devoted to developing Cinna and Katniss's stylist team a bit more. Still my opinion. BUT that's all a Critic can do right? Give their opinions? And the more fair ones will state the positives of a film, yet also underline the most important negatives IF they truly want to spare the movie going audience of seeing a film they thought horrible. And IMO, most people eventually learn which critics have the same or similar opinions about film structure. Audience members also know which actors/actresses they love or hate which also impacts a film's success. There have been films that have been well written and acted, and yet because they do not boast an A-List actor or someone studios feel is known well enough, the films exist in limited release until word of mouth propels studios and other markets to take notice. One film I can think of in particular is "My Big Fat Greek Wedding." Brilliant film, yet except truly for a few actors, no one would've paid it much mind, even if it was Tom Hanks and his wife's production company footing the bill. In closing, the greatest thing I have ever read regarding films, and even extending that in to TV and other forms of entertainment—-No one knows what they want until they SEE it (and even then minimally be able to define it for future entertainment<–definitely IMO)


I see no need to define the role of a critic with rigid lines. I have no problem with a critic who suggests an alternate ending. To me, there is no ethical boundary inherently crossed by a critic who makes suggestions about how a film should have been made. Indeed, I think it is an entirely valid way to criticize a film, as long as the critic backs up his/her argument with a reasonably sound argument. It can't be, "I didn't like that ending. It should have been x instead," and just left at essentially that. It has to be more like, "That ending did not fit the rest of the film well. Here's one that would have."

Everyone has opinions about films they see. Critics are under no obligation to hold back their opinions as they relate to a film. Their only obligation is to explain why they hold those opinions in better detail than the average moviegoer.

Top Shelf®

I like to read reviews to get a sense of what goes on in a film I have yet to see.i also read them to learn from different perceptions on films I have seen.some critics are good, like roger ebert.some really suck, like leah rosen.(don't care if I spelled her name wrong) truth is, most are's very easy to critique garbage like gigli or any movie with adam sandler in it.but when I needed clarification on mulholland drive, they knew less than I critics should stick to the technical aspects and execution.


@ matt singer

intent never matters. that a filmmaker, or a novelist, or a playwright has an intent is good for them, and good for their works, but for the audience it is always and necessarily irrelevant. intent may translate to the film, or intent may be lost — what matters is only the film itself. e.g. if a film is ambiguous and the ambiguity is a necessary element of the film, then the ambiguity should never be resolved and the film cannot be criticized on those grounds — the criticism would be in error. but if a film is ambiguous and that ambiguity is not necessary, and is indeed a distraction in the narrative (or worse yet renders the narrative unintelligible), then the film is in error. and this is the key argument that your article seems to deny: that a film can be in error. more explicitly, you deny the role of the critic as 'philosopher of film' whose job it is to critically analyze — to point out error where she sees it, to provide counter arguments, to evaluate aesthetics, and of course to acknowledge compelling arguments. In short it is the job of the critic to give credit where credit is due. anything less is a disservice to the film, the filmmaker, and the audience. the critic can be wrong. but it is never wrong for a critic to point out error — and it is certainly not the critics fault that you can't be bothered to imagine how a film would play out with the last five minutes cut off. that you can't imagine a film without someone editing it for you and showing you the results physically is your own misfortune, and is not a strike against anyone but yourself. david daw claims that AI contains a ruinous error, he makes a compelling case for his argument — he backs his assertion with reason. you claim implicitly that films cannot be in error, and explicitly that david daw (as a 'critic' and not a 'film critic editor' — which is really just a filmmaker and not a 'critic') cannot be allowed to claim that a film is in error. you also claim that AI is not in error, yet you provide nothing convincing to support your assertion (you've merely demonstrated that, according to daw, spielberg and kubrick were in error). i've answered your denial of the film critic as 'philosopher of film' and daw has yet to be legitimately challenged (his argument still stands). now you have two counter arguments to make: one against me (defending your meta criticism of critics), and one against daw (to prove why the film is not in error). if you can provide a legitimate counter argument against my claim that i cannot overcome, then you do not, by necessity, need to answer daw. if you cannot overcome my claim (or wisely choose not to), then you must overcome daw's claim. if you cannot do either (or choose not to), then you should retract your article.


Hello, just a complaint about the pop up tab which allows us to post on twitter, facebook etc. It pops up right over the article, making it very hard to read!! There is also no obvious way to close it or move it makes it very annoying, so much so i didnt read the article and most probably stay clear of your website in the future unless its fixed!

thank you!


if critics really for the most part actually understood the films they saw and discussed what they felt was there and what it meant i might be more inclined to believe critics have relevance. but they dont and not just because of the film 101 reason of show vs tell. Critics dont discuss film or films anymore. nobody interacts with what they saw everything is about the author of the critiscms "brand". Their voice that makes them clever or hip or cynically cool or over the top nerd. At best most criticms have pre-formed ideas about what they think about a film because of what their self created niche is supposed to be. At worst critics dont write about what they even think they write what is the accepted or recommended attitude at whatever institution they write for. newspaper film critics are the worst for this. most of the time mainstream newspaper film critics simply tell whatever they can get away with to make the movie the company wants to do well sell. Even all of this aside, critics only talk about acting shots, cliched not cliched and a bunch of other critic speak gobbledegook that really leaves out story or meaning or anything other than well executed or not.


One thing I noticed about critics is that they tend to critique on the message. If a film is controversial and has a bad message they give it a low rating. If a film has a good message then they tend to give it a good rating. The message doesn't make the film. A production company could produce a film that has a really great message about family and love but if the acting, directing, and writing isn't good then the film isn't good. If a film has the worst message ever but everything else is good then the film is good.


A true critic should not mix personal opinion with how good the film actually is. For example, I really enjoy Evil Dead 3: Army of Darkness, but that does not mean I consider it a good film. Vice-versa, I'm not particularly keen on Schindler's List, but I know that it is a brilliantly crafted film. A mistake many critics make is that they say whether they enjoyed the film or not, which is not their task. Their job is to say how good the film is, how well it employs filmic devices and micro-elements and the like. To say how a film SHOULD have been is also a critic's error, they need simply analyse the film before them. If the ending to AI is bad, they do not need to offer up solutions to how it could be made better, but rather just state why the ending is bad. And for the record, the ending to AI is bad.

Jamie Helton

It seems that many people jumped to conclusions about the film when it first came out simply because Spielberg directed it, finding it too "Spielbergian" and not enough "Kubrickian" or vice versa. Like many difficult movies (including most of Kubrick's own films), "A.I." left one impression upon initial viewing, but has different meaning with repeated screenings.

The job of critics is to analyze movies and point out what they perceive what works and what doesn't about movies. This is both objective and subjective, as they can look critically at the technical aspects of film, but also how it makes them feel as an audience member. A movie can be well-made, but not be enjoyable, and a critic must reflect this duality. Unfortunately, too much of criticism is made based on the subjective, but that's how the audience reacts to any movie. For a critic to tell a filmmaker what he should have done with a movie is very arrogant, and to recut another person's movie is the height of hubris.


We're all critics. It's just that some people critique films, and others critique critics' critiques. I just critiqued a critic's critique of a critic's critique. In the meantime, someone out there is making a film.


It is interesting to note that a lot of people who criticized AI when it first came out, especially the ending, now call it one of the best films of the 00s. Like many of Kubrick's film, even though this wasn't directed by him, I believe it will take a decade or two for the audience and the film community to catch up with what a masterpiece it really is.

I had so many discussions with budding film critics that I had to write my "defense" of AI. You can read it at


I was reviewing films for the now-defunct when "A.I." was released. I named it my #1 movie of the year and actually got hate mail because of it. I love that so many critics have reversed their original opinions since then. Great article.

Ryan H.

The misunderstandings about the origins, development, and intent of A.I.: ARTIFICIAL INTELLIGENCE continue to abound, due to everybody making unwarranted suppositions about Kubrick and Spielberg and their involvement in the film's development. A little research would go a long way.

Anyone who wants to really get into the thick of "What was Kubrick wanting to do with A.I.?" should look at A.I.: ARTIFICIAL INTELLIGENCE: FROM STANLEY KUBRICK TO STEVEN SPIELBERG by Jane M. Struthers, Jan Harlan, and Chris Baker, which pretty effectively establishes that the film holds pretty closely to the materials and narrative Kubrick left behind, and further illuminates the ideas and purpose Kubrick had in developing the story. In short, Kubrick wished to create a kind of subersive fairytale (for example, Kubrick noted that David's arc should represent the opposite of Pinnocchio's development; David becomes more, not less selfish, as the story proceeds).

The ending of ARTIFICIAL INTELLIGENCE perplexes folks because it sustains a strange tension, a tension that is present in the narrative and the clash between the approaches of its two major creators. It does not mesh with what is popularly considered to be "Kubrickian" or "Spielbergian." But the tension is ultimately appropriate, speaking to the very ambiguities that lie at the center of David's creation.

Sic Faciunt Omnes

Brian Aldiss (the writer of the short story AI is based on) explains in the collection of short stories, that it was Kubrick's obsession with the fairy tale ending which made him walk away. It feels false because it was fake sentimentality, because Kubrick wanted to make a blockbuster (he wanted some of that Star Wars money). He really wanted to sell out. It sticks in your eye, but it's not Spielberg's personal fault.

The story is called "Super-Toys Last All Summer Long".


Spielberg on the end of A.I.


Excellent article, especially because you focused on A.I., which has, since its release, been my favourite film to demonstrate why critics can get it wrong. You are in fact the first person to echo my own interpretation of an ending which is actually very dark… when seen from a certain perspective, of course, and who knows? Maybe our perspective is also "wrong".

It reminds me of a story I heard about Alfred Hitchcock's granddaughter. When she was young, her class at school were asked to write about To Catch A Thief. Now obviously, she went to see her granddad and get the inside scoop! Despite this, she received a low mark and the teacher said she had misunderstood what the film was about. She went back to complain to Grandpa Hitchcock, who merely shrugged and said, "well, that's what *I* thought it was about".

Personally I think people, particularly critics and particularly amateur ones, don't want to be educated any more. They see what they see, make a decision and refuse to be budged. I truly believe that when you are watching a film that is generally accepted to be very good or has been made by a distinctive film-maker, and yet you see something "wrong", you should think: Is it actually wrong, or is my very reaction the intention? It's a matter of respecting the director and remembering that all editing is manipulation. When you consider that film-makers such as Pasolini embraced Realism to the extent they left mistakes in, and others, such as Stephen Frears (My Beautiful Laundrette), sometimes chose to make their films appear clumsy on purpose, you have to reset your idea of what constitutes a bad decision.

Of course, I am not saying films are beyond such criticism, merely that critics should start by being optimistic, believing that the director understood every decision he took for a reason, and work backwards from there. So if you think of the opening hour of Vertigo as boring or the weird angles in The Third Man as silly (just random examples I've read over the years), just stop for a moment and consider what they might be trying to achieve. That's more interesting than dismissing it.


Probably the most famous group of filmcritic filmmakers were the New Wave directors that came from the Cahiers Du Cinema in France. I'm pretty sure it was either Truffaut or Godard who argued that as critics they were making their own correction to movies when reviewing them, so they just took that to the next step and started making movies themselves.


Why this particular argument from Daw today catches anyone's attention after a decade of similar criticism and debunking, I'm not sure, bolt of lightning or not. Daw's argument about ending the film with David stranded at the bottom of the ocean has long been a critique held by many, since Kubrick tends to treat his white male characters' desires with severe disdain by the end of many of his films and would fit in well with each protagonist in LOLITA, 2001, CLOCKWORK, SHINING, FULL METAL JACKET, and EYES WIDE SHUT. Each character pays a steep price for essentially getting what they wanted. Spielberg, however, has always said the milquetoast ending was Kubrick's idea and who's going to say he's lying? Spielberg is rich beyond Croesus and has the Oscar that Kubrick never won so he has no reason to fib, especially if it defames his friend. Spielberg gets a kicking from people like Daw because his filmography suggests that his influence as the director would have led ARTIFICIAL INTELLIGENCE towards a more commercial/conventional happy ending, especially when alien robot creatures are involved, so it's not a stretch for someone to have this conclusion when all of the facts (alleged or not) aren't known or acknowledged.

James Berclaz

Ian Watson's literary treatment, though, which was supervised by Kubrick from 1990-1991 (that was slightly altered by Sara Maitland to add 'fairy tale magic' and 'emotion') already had that ending. The myth that the 'super robots' ending is a Spielbergian alteration is interesting because we 2001 already suggests the idea of man cannot take care of himself and as such requires the meddling of "The Alien". In 2001, this is communicated as an ambiguous effect (negative at first, positive by the end) while in A.I., man was left to be consumed by his errors.

In fact the reason Kubrick didn't direct it (considering he started working on the project circa 1974) might be seen as an issue of narrative perspective. Kubrick's films, without the exception of The Shining — even then it is a stretched exception — commit to the perception of the white male and often focuses on male fears and desires. On the other hand, A.I. starts off through the perception of the conflicted mother before moving to the "inhuman" child, which might explain why Kubrick might have felt its themes and perspective were more attuned to Spielberg's sensibilities.

Thuan Dang

Daw believes Spielberg's ending is wrong because his interpretation of the movie is wrong. He writes his ending would be better for the themes of the film (limitations and outsiders vs. insiders) when he totally disregards the film's most important theme and the only one that matters: Love between a child and mother.

Interestingly enough, he recognizes this theme, "its maudlin and predictable emotional core while running through its exciting world-building". But this quote also describes why Daw is wrong in his viewing. He focuses on all things secondary in the film while with the essence of the film, he balked at from the beginning, perhaps before he even saw the film because of knowing Spielberg's past works.

You even write about this subjectivity yourself in the first paragraphs without taking a position on the matter. But you are mistaken when you write, "a director's intentions matter less than a director's execution". This is a illogical argument about the defense of viewer's interpretation through filmmaking prowess.

It is one acknowledging in seeing the director's intentions only to shrug it off and selectively view what they want to watch, all under the veil of saying "the director's technical skill wasn't up to par to get his point across." No, the viewer, him or herself, didn't allow the point to get across because it simply didn't interest them.

Unlike re-edits, which are not interpretations of someone's else's work but instead being the actual seizing of ownership and creating entirely new pieces of work, Daw's interpretation of the film isn't an interpretation of the film at all. It's his interpretation of Spielberg, wishing he was a little more like Ridley Scott.

Mike Cameron

Per your last paragraph, critics who imagine a "better ending" or "what they should have done" are almost always ignorant of the actual circumstances of the production, which is why it's often incredibly irritating to read that kind of criticism. The "film critic filmmakers" aren't doing this – they're framing their arguments with real, actual, extant footage.

It's akin to the difference between somebody looking at your broken-down 1994 Civic and saying "what you shoulda done was bought a different car with a better engine that wouldn't have broken down" and somebody who looks at it, pops the hood, and gets it running with a roll of tape and a twist-tie

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