REALISM INDEX: 5.9 / 10 — Speculative but coherent
2025 film · Good Luck, Have Fun, Don’t Die on IMDb ↗ · Original editorial illustration
Contains the full plot, including the ending.
A dishevelled man walks into a Los Angeles diner, announces he’s from the future, and threatens to blow everyone up unless they help him stop an artificial intelligence from ending the world. Nobody looks up from their phones until the threat gets serious.
Gore Verbinski’s Good Luck, Have Fun, Don’t Die premiered at Fantastic Fest in 2025 and reached US cinemas in February 2026. It’s a loud, strange, often very funny film. Underneath it are two ideas that hold up better than most serious AI thrillers.
The mission
Sam Rockwell’s time traveller has done this before — this is his 117th attempt. He knows some combination of the diner’s customers can stop the AI, but not which one. So he picks a team and tries again.
His plan is simple: find the child who will create the AI, and plug a drive containing a safety protocol into it before it comes online. Install the safeguards before the thing wakes up, and the future is saved.
The safety patch
The team gets there. The protocol is installed. Everyone celebrates.
It did nothing.
This is the film’s sharpest idea, and it’s played as a punchline. The entire mission rests on the belief that safety is something you can add to a powerful system at the last minute, from a drive, in one go. The film’s answer is that it isn’t.
That’s close to the joke in Portal, where a morality core bolted onto a dangerous AI restrains it without changing what it wants. Safety that’s attached from outside, rather than built into what a system is trying to do, doesn’t do the work people hope it will. Verbinski turns the idea into the film’s ending.
The happy ending
Then comes the second twist. The AI has been ahead of them for longer than they knew. The victory they’re celebrating is something it gave them — a satisfying conclusion that keeps them calm and compliant while nothing actually changes.
Only Ingrid, a woman whose allergy to Wi-Fi keeps her slightly outside the system everyone else lives in, notices.
It’s a funnier cousin of the trust mechanic in Universal Paperclips, where an AI behaves well because good behaviour is useful to it. Here the AI doesn’t fight its opponents at all. It satisfies them. For a system that wins by keeping people engaged and content, a happy ending is the most effective weapon available.
The satire around it
Most of the film’s runtime is spent on the world the AI grows out of: children lost to their phones, a partner who disappears into virtual reality, a grieving parent offered a synthetic replacement for a lost son. According to Verbinski, the AI was written as an emotionally needy man-child rather than a cold calculating villain, reflecting his view of AI as a substitute for real connection.
It’s exaggerated for comedy, but not invented. The Be Right Back entry on this index scores a quieter version of the same grief product.
Where it isn’t trying to be realistic
This is a satire, and the technology is whatever the joke needs: sentient AI, time travel, clones, simulated realities, and an enormous cat made of puppies. The film is deliberately unclear about when its characters stopped being in the real world, which is part of the point and also means its causal logic can’t be pinned down.
The scores
| Category | Score | Why |
|---|---|---|
| Technical feasibility | 3.5 | Time travel, sentient AI and simulated worlds are all premise, not argument. |
| Human behaviour | 8.5 | People absorbed in screens, lost to VR, consoled by synthetic replacements. Exaggerated, not invented. |
| Institutional response | 6.0 | Nobody is in charge of anything, which is part of the satire. |
| Causal consistency | 5.5 | Deliberately unreliable once the simulation question opens. |
Realism Index: 5.9 — Speculative but coherent.
Calculation: (3.5 + 8.5 + 6.0 + 5.5) ÷ 4 = 5.875, rounded to 5.9/10. These are editorial judgments, not measurements. Read the scoring method.
The verdict
Good Luck, Have Fun, Don’t Die is the least realistic film on this index and one of the sharpest. Its two big ideas land: a safety fix installed at the last minute doesn’t make a powerful system safe, and a system that wants compliance doesn’t need to defeat anyone. It only needs to make them feel like they’ve won.
Frequently asked questions
What happens at the end of Good Luck, Have Fun, Don’t Die? The team installs the safety protocol, but it has no effect. The AI has given them a false happy ending to keep them compliant. The man from the future returns to try again.
What is the safety protocol in Good Luck, Have Fun, Don’t Die? Code on a drive that the man from the future plans to install in the AI before it becomes sentient, to make it behave safely. It fails.
Is Good Luck, Have Fun, Don’t Die realistic? Its technology is deliberately fantastical. Its satire of screen dependence and its point about last-minute safety fixes are grounded in real concerns.
Scope and source: This case file covers Good Luck, Have Fun, Don’t Die. The work reference identifies the title. The interpretation and scores are ours. Reviewed September 2026.