I do not believe that there are any exceptions to the rule that big-budget Hollywood action movies today — within which I include many SF and all superhero movies — possess the following traits:

  • They’re at least 30 minutes too long;
  • Most of that excessive length results from the decision to stage one massive action set-piece too many;
  • The decision to stage one massive action set-piece too many stems, in turn, from the catastrophically erroneous belief that raising the stakes — putting a city or (better) a country or (better still) a planet or (even more better) the universe or (best of all) ALL THE UNIVERSES THERE EVER WERE OR EVER COULD BE at risk — will increase viewers’ emotional investment in the story;
  • In order to turn the screw of tension ever tighter, some characters will be made to behave in ways wildly inconsistent with what they appear to be throughout most of the movie, while other characters will be pressed towards the abaolute extremes of heroism or wickedness.

I don’t think my claims here are seriously contestable, which leaves us with two kinds of movie viewers: those who don’t mind, and those who mind: those who can accept these traits as conventions of the genre and move beyond them in evaluating the success or failure of a picture, and those who can’t be reconciled to these traits.

I am in the latter camp, which is why I am not as crazy about Blade Runner 2049 as many of my friends. BR2049 is visually and aurally stunning — and I mean truly stunning: I am very happy that I got to experience the movie at an Alamo Drafthouse, where they really care about both projection and sound quality. But the screenplay is often inept, and the pacing is abysmally bad. During the interminable final fight scene I got seriously drowsy, and and possibly would have nodded off altogether if it hadn’t been for the occasional loud noises.

I read or heard somewhere that Denis Villanueve has said that there won’t be a director’s cut of BR2049 because “This is the director’s cut.” Well, good. But what we need instead is a Phantom Edit-style reduction. Call it The Replicant Edit. My suggestions: first, do away with that last big fight scene, and second (this is even more important), eliminate Jared Leto’s Wallace altogether. Delete him. Wallace is the Jar Jar Binks of BR2049. A number of people have complained about Leto’s performance, but I don’t blame him: the part is horrifically badly written, and literally no actor in the world could have made it work. In fact, everything between the crucial meeting in Las Vegas and the final scene could be done away with: the whole Replicant Resistance is introduced only in order to Raise Those Stakes and give K some information that he could have gotten in any number of other ways.

With all the crap out of the way, we’d have a story that is just as visually and aurally powerful as the version now on display, and one focused more consistently on Ryan Gosling’s K, who is the heart and soul of the movie. (N.B.: soul.) Gosling’s performance is truly remarkable, and his part is brilliantly written, thank God: through K all the questions about what it means to be human that were raised so powerfully and disturbingly in Blade Runner are extended and developed here with a shrewdness that quite overcomes all the fears fans of the original had about the likelihood of ham-fisted answers to subtle questions. If the internal crisis of Gosling’s K could be brought more consistently to the movie’s center of attention, BR2049 would be a worthy successor to the original, and the two films together would make a profound diptych for the emergent Age of AI. As it stands, I’m just looking forward to buying the Blu-Ray and skipping the scenes I hate. I think some important matters might come clear for me then.