How Blade Runner 2049 fixed the sequel problem

I rewatched Blade Runner 2049 last weekend. Third time. It gets better each time, which is rare for a sequel and almost unheard of for a sequel to a perfect film.

Denis Villeneuve understood something that most sequel directors don’t: the original Blade Runner wasn’t about replicants. It was about what it means to be real. The famous question, “is Deckard a replicant?”, isn’t interesting because of the answer. It’s interesting because the movie makes the answer irrelevant. Whether Deckard is human or machine, his experiences, his memories, his love for Rachael, these things matter the same amount either way.

2049 takes that idea and pushes it further.

Officer K knows he’s a replicant from the start. No mystery there. His arc isn’t about discovering what he is. It’s about discovering what he can choose to be. He thinks, briefly, that he might be “special” (born, not manufactured). When he learns he isn’t, the movie doesn’t collapse. It pivots. K chooses to do the right thing anyway. Not because he’s the chosen one. Because it’s right.

That’s a harder question than the original asked. The original asked: does it matter what you are? 2049 asks: does it matter what you are if you choose to be good?

I keep thinking about that in the context of AI. When an AI produces something beautiful, helpful, moving, does it matter whether there’s “someone home”? Does the origin of the goodness matter, or just the goodness itself?

Warner Bros didn’t market the film well. It underperformed at the box office. But it’s the most thoughtful sequel I’ve ever seen to the most thoughtful sci-fi film ever made. And every year, as AI gets more capable, the questions it asks get less hypothetical.

Related thinking: