← Back to posts

[Netflix] The Great Flood — My Thoughts After Watching

ScreenFilm

It’s been a long time since I picked up a pen to write anything — and I never expected the thing to get me writing again would be my thoughts after watching The Great Flood.

I only started because I saw someone say the film had a great concept and great actors, but that the story ultimately fell apart. Plenty of people online called it bad, said they couldn’t follow it, some even bailed before the end. But after watching it, I found none of that to be true. The subject it tackles may not be the freshest, but the premise behind it and the way it’s shot and told are excellent — as a science-and-engineering sci-fi fan, I was genuinely floored that a film like this got made.

If you want to watch it but would like a few hints or some mental preparation first, here’s my rundown:

The film splits into two halves — the first is real, the second is a simulation. Ja-in is an experimental subject (not an ordinary human, and quite the little terror). The film is shot using montage (think Nolan’s Tenet), so the timeline jumps around, and you have to piece it together yourself in the end.

What follows is my personal take, with the timeline reordered — and full spoilers. Read at your own discretion.

Anna is a researcher on an AI emotion-engine system. Five years before the Great Flood, she proposed using a virtual world to let the AI learn emotion, as a way to develop the emotion engine. The plan called for people of different ages to simulate different facets of emotion — but because the simulations might take too long and weren’t guaranteed to produce results, she talked it over with her supervisor Hyun-mo and pivoted the plan toward producing an AI-powered new-human experimental child, to be raised and used to perfect the emotion engine. That’s the origin of Anna’s son Ja-in, the little terror (a new human). It’s also why we notice that the kid’s behavior is so uncontrollable yet so patterned (he runs off, but can always be found in the closet; he loves playing with water and isn’t afraid of it), and why his body seems so fragile (because the human-printing technology isn’t fully mature yet).

On the day of the Great Flood, Anna realizes something is wrong and is preparing to evacuate when she gets a call from Hee-jo telling them to flee to the rooftop as fast as they can, where someone will come to pick them up. A lot happens in between, and a lot of foreshadowing gets planted — foreshadowing that all pays off in the second half of the film. I won’t go through it piece by piece; those who are interested can match them up one by one. This is also the source of the “wait, why does the second half feel so strange?” complaint a lot of people have. But the key point is that Hee-jo explains the cause of the Great Flood: three hours earlier, an asteroid struck the Antarctic continent; the meltwater, carrying seawater with it, is sweeping northward and overwhelming everything. Half of Japan has already sunk beneath the water, and at this rate this 30-story apartment building will soon be submerged too — today, humanity as we know it passes into history. He also explains that humanity actually detected the asteroid ten years ago, which is why the new-human program was launched, and why Anna is now working on the emotion engine. He mentions in passing that both the intelligence-printing and body-printing technologies are already complete — all that’s missing is the emotion engine. After they escape successfully, Hee-jo is to stay with Ja-in at a temporary shelter, while only Anna will head to the lab.

In any case, they finally reach the rooftop — only to discover there is no such thing as a temporary shelter. Hee-jo is shot dead by men in black, and Ja-in is there too, about to have his memory unit extracted. At Anna’s request, she gets to embrace Ja-in one last time and whispers to him: hide well in the closet behind me, and I will find you, no matter what. She then boards the helicopter. The men in black load Ja-in’s memory unit onto the aircraft, and everyone besides the essential personnel is left on the rooftop to accept their fate.

The next scene comes quickly: Anna and the other crew members board a space shuttle and, in the nick of time, make it into space. Director Lee lays out the current state of the mission, including the fact that a swarm of small meteorites will soon slam into Earth, and that most life on the planet is expected to be wiped out. Their objective is to produce new humans aboard the space station so they can eventually return to Earth — but for now, the emotion engine only has the “child” portion complete; the other emotional components still need to be finished. Anna says she wants to complete the “mother-and-child” emotional line first, using this method: “Set the experimental subject as the child’s mother, then make the child disappear. As the subject searches for the child, she’ll run into many obstacles — and if she fails, it starts over from the beginning.” What no one anticipated was that Shuttle No. 6 ahead of them was hit by a meteorite and exploded, and the blast caught them too. Anna is critically injured in the accident, and as her life is ending, she expresses a wish to become the experimental subject herself, to help complete the mother-and-child emotion engine. So the other crew members upload Anna’s and Ja-in’s memories and begin the simulation.

The first thing to note is that the extracted initial data — the raw data — may itself get amplified or turned strange by the AI simulation that follows. Those spots are exactly the ones people were discussing about the film. So let’s talk through the foreshadowing.

One: the number on Anna’s clothes keeps changing — this actually indicates which simulation run she’s on. And because simulations take a long time, not every run restarts completely from scratch; more likely a specific memory point is restarted, which is why she keeps getting stuck at certain moments.

Two: why does the staircase go from just one person praying to the gods, to everyone praying? Because if the AI is simulating the other characters, and an NPC only needs to follow the script in the key scenes, it’s entirely plausible for the other AI NPCs to over-panic and turn into something cult-like. But since the situation here doesn’t affect the later story, the simulation just carries on as-is.

Three: how does Anna suddenly become a crack shot, and so agile too? All I can say is that because this is a film, we feel like she changed instantly — but in reality, every simulation run has her constantly learning and evolving, so it’s reasonable that her marksmanship keeps improving. That same escalation is mirrored in how Anna starts out impatient and forceful with Ja-in and later erupts with genuine maternal love.

Four: why do they, on one occasion, enter a room on the 30th floor that looks like a bare, unfinished shell? This is actually consistent with scene simulation in games — for a scene that isn’t needed yet, we might not render it fully. Anna and the others accidentally wandered into that scene, which is why it feels so odd.

Five: why do the characters, or Anna, keep experiencing déjà vu? This is a lot like our AI today. Modern AI (LLMs) may seem to give great answers, but it’s quite possibly just an ultra-advanced tile-matching game — it matches to whatever it thinks will get the best reaction and feeds that back to you. All you need to do is give a good enough prompt, and you can very likely trigger and bypass the parts the system doesn’t allow — which is actually a lot like people. That’s why Anna can elicit different reactions from the other NPCs in the film, and why the NPCs in turn give Anna so much help. Of course, some human intervention along the way can’t be ruled out.

Six: why is run number 21499 the one that finally counts as a success? As stated in the original setup: “Set the experimental subject as the child’s mother, then make the child disappear. As the subject searches for the child, she’ll run into many obstacles — and if she fails, it starts over from the beginning.” The previous 21,498 runs all ran into obstacles and failed. We can also see that the last mile (30F → RF) took tens of thousands of simulations to get right, getting stuck on opening the closet door on the 30th floor and finding it didn’t hold what she wanted, then a crowd of men in black arriving to block her and force her to break through again — until she finally finds Ja-in in the rooftop closet. And rather than coddling him, she gives Ja-in the courage to swim alone toward the target spire. Only then, in run 21500 that breaks free of the data, do they meet at the spire together — and only then is the new human truly born.

At the very end of the film, we see many landing craft preparing to touch down on a post-Flood Earth, its geography drastically changed from before. But as new humans, Anna and Ja-in — I’d like to think they’ll manage to survive.

But here’s the interesting question: are new humans still humans? And if so, why wasn’t the original plan simply to keep as many of the old humans alive as possible? Was it because sustaining a population large enough to keep the species reproducing was just too difficult? These are intriguing questions well worth pondering.

References

https://www.beauty321.com/post/70835

https://youtu.be/Xdi65TUU0Nc

Signed in Taiwan · 2026