Nineteen minutes. That's how long it took to get three hundred seventy-nine people off a burning aircraft at Haneda Airport, every single one of them alive.
The Japan Airlines collision in January of twenty twenty-four. The Airbus was fully engulfed by the time the last passenger left the slide.
And here's the thing that sticks with me. Not one person stopped to grab a carry-on. Not one person opened an overhead bin. The crew got everyone out in under twenty minutes, and the investigators said the passengers did exactly what the safety demo has been telling them to do for decades. Leave everything. Follow the crew. Go.
Which brings us to Daniel's question, because he's been thinking about how few of us actually listen to that demo in the first place.
Here's what Daniel sent in. He flies a few times a year, not constantly, but he says he could probably recite the cabin crew safety demonstration in his sleep. The seatbelt, the oxygen mask, the life vest he's never once touched. And he's wondering whether that repetition has numbed us to the information. His questions, roughly: one, when emergencies actually happen, what does passenger compliance with these instructions really look like? Two, do passengers find they've been numbed by repetition? Three, have any airlines floated the idea of asking passengers to watch the safety demo before the flight, or in some interactive format? And four, the demo is presumably based on evidence and statistics showing it makes a difference. What does that difference actually look like?
Four questions, and the fourth one is the one I've been chewing on. Because the honest answer is that the evidence is thinner than anyone in the industry wants to admit.
So today we're asking whether the safety demo is a ritual we've outgrown, or a ritual that saves lives precisely because it's a ritual.
The tension at the center of this is uncomfortable. The demo is legally mandated, universally performed, and almost universally ignored by the people who have seen it most.
And the people who have seen it most are the people most likely to be in an emergency. That's the paradox. The demo is optimized for first-time flyers, but the frequent flyer is the one who will actually be on the plane when something goes wrong. They're also the one who's heard the briefing forty times and stopped registering it entirely.
We'll trace this through the psychology of habituation, the real-world compliance data from actual evacuations, the airline experiments with creative and interactive briefings, and then the evidence question. What does the demo actually buy us?
Start with the brain. Why is it so hard to pay attention to something you've heard dozens of times?
Habituation is a neurological response, not a personality flaw. Your brain has an orienting response. New stimulus, unexpected sound, something changing in the environment, your attention snaps to it. That's the brainstem doing its job. But when the same stimulus repeats over and over, the orienting response diminishes. The brain says, I've seen this, nothing new here, deprioritize it. And that's not laziness. That's efficient allocation of attention. Your brain is conserving resources for things that might actually be novel or threatening.
So the demo becomes background noise because our attention systems are wired to make it background noise.
And the safety demo is the perfect habituation trap. Same script, same gestures, same order every single time. Seatbelt, oxygen, life vest, exits. The words barely change between airlines. The only variation is whether it's a live crew member or a video, and even the videos follow the same template.
There's recall data on this. Passengers retain surprisingly little specific information from the briefing, even frequent flyers. Ask someone how to open the overwing exit and they'll hesitate. Ask where the life vest is stowed and they'll point vaguely under the seat, maybe. The information doesn't stick because the brain stopped encoding it.
The overwing exit one is striking. It's a completely different mechanism from a regular door. You don't turn a handle. You pull a panel, heave the window outward, and it's heavy. And in a real emergency, the person sitting at that exit row has to do that while the cabin is filling with smoke and everyone behind them is screaming. But if you ask that person beforehand how the exit works, most of them can't describe it.
And that's the person the crew is counting on.
That's the person the crew is counting on. The exit row briefing is supposed to fix this. The crew member comes over, makes eye contact, asks if you're willing and able to assist, and tells you to read the instruction card. But the instruction card is also something people glance at once and set aside.
So the demo fails the people who need it most. The first-time flyer is watching with wide eyes, absorbing every word. The frequent flyer has their headphones on before the video finishes.
And that's the structural flaw. The demo is designed for the lowest common denominator, which means it's pitched at the person who has never been on a plane before. That's the right safety instinct, but it means the delivery is calibrated to bore the experienced passenger.
Now let's talk about what happens when the emergency actually occurs. Because that's Daniel's first question, and it's where the data gets interesting.
The Japan Airlines evacuation is the strongest recent case study. January second, twenty twenty-four. Flight five sixteen, an Airbus A350, collides with a Japan Coast Guard aircraft on the runway at Haneda. The Coast Guard plane has five people on board, one survives. The JAL flight has three hundred sixty-seven passengers and twelve crew. The aircraft catches fire. The cabin fills with smoke. The intercom system fails, so the crew are shouting instructions down the aisle.
And everyone gets out.
Everyone gets out. Under twenty minutes from impact to last person off. The investigators later said the passengers left their baggage behind, followed crew instructions, and moved to the exits in an orderly manner. Near-perfect compliance. The photos afterward show the aircraft completely gutted by fire. The fuselage is a charred shell. And every single person on that plane walked away.
That's the demo working.
That's the demo working, or at least the cumulative effect of the demo, the crew's training, and the instruction cards all reinforcing the same message. Leave everything. Move to the exit. Follow the crew.
Now contrast with the Emirates crash in Dubai. August twenty sixteen. Flight five twenty-one, a Boeing triple seven, crash-lands and catches fire. Everyone survives the impact. But during the evacuation, passengers are opening overhead bins and pulling out their carry-ons. One passenger dies. The investigators noted that people evacuating with baggage slowed the process.
And the Aeroflot disaster is even more stark. May twenty nineteen, Moscow. A Sukhoi Superjet makes an emergency landing and catches fire. Forty-one of the seventy-eight people on board die. Video from inside the cabin shows passengers in the aisle, pulling luggage out of the overhead bins while the aircraft is burning. The evacuation was delayed. People at the back of the plane never got out.
The demo explicitly says leave everything behind. And a significant minority of passengers, in a real emergency, grab their bags anyway.
The baggage problem is the clearest measurable failure of safety demo compliance. It's not subtle. The instruction is simple, it's repeated every flight, and yet in real evacuations, people still reach for their roller bags. And the consequences are concrete. A bag on a slide can puncture it. A bag in the aisle slows everyone behind. A bag in the doorway creates a bottleneck. Every second of delay in a cabin fire is measured in lives.
Why do they do it? What's the psychology of grabbing your bag when the plane is on fire?
Part of it is that the bag is an extension of the self. Your passport, your wallet, your laptop, your medication. The brain under extreme stress does a terrible job of triaging. It reaches for the familiar object. There's also a social dimension. If one person stops to open a bin, others see it and think, oh, we're doing this now, we have time. It cascades.
So the compliance paradox is real. The people who need the demo least pay the most attention. The people who need it most pay the least. And when the emergency happens, that gap shows up as someone yanking a suitcase out of an overhead bin while the cabin fills with smoke.
Daniel's second question, whether passengers feel numbed by repetition, the answer is yes, and it's not a moral failing. It's the brain doing what brains do. The format is the problem, not the passengers.
That's a distinction worth making. The passenger who tunes out the demo isn't being reckless. They're being human. The demo has been delivered in the same format for decades, and the brain has adapted to it.
And here's the thing. The demo is legally required. The FAA and EASA both mandate a live demonstration or an approved video before takeoff on commercial passenger flights. That's not going away. The question is whether the delivery can change within that framework.
Which brings us to what airlines have tried. Daniel asked if anyone has floated the idea of watching the demo before the flight, or in an interactive format.
The pre-flight idea has been floated. Airlines have experimented with app-based safety briefings, video modules you watch before you board, gamified quizzes. None of it has scaled. The regulatory barrier is real. The FAA and EASA want the demonstration to happen in the cabin, before takeoff, where the passenger is actually sitting in the seat they'll need to evacuate from. There's a logic to that. The demo is tied to the physical space. You see the exit, you see the seatbelt, you see the life vest location. Watching it on your phone in the terminal doesn't give you that spatial anchoring.
And passengers resist it. Nobody wants to do homework before getting on a plane.
Right. The airline's incentive is to make boarding as frictionless as possible. Adding a mandatory pre-flight video module is friction. The airlines that have tried it found low completion rates and a lot of grumbling.
What about the creative videos? Air New Zealand's body paint video, the Virgin America animated one. Those were engagement plays.
Air New Zealand's Bare Essentials video in two thousand nine was the first viral safety video. Crew in body paint, strategically positioned props. It got millions of views online. Virgin America's animated video in twenty thirteen was nominated for a Grammy. These were innovative. They broke the habituation pattern by being surprising and funny. People watched them, shared them, talked about them.
But did they improve recall?
That's the uncomfortable question. The engagement data is great. The recall data is murkier. A funny video gets attention the first time, but the frequent flyer sees it on every flight too. The novelty wears off. And there's some evidence that the humor can distract from the actual safety information. You remember the joke, not the location of the life vest.
So the creative videos are a temporary fix for a structural problem.
They're a patch. A good patch, but a patch. The underlying issue is that the demo is a one-size-fits-all delivery of information that different passengers need at different levels of detail.
Now Daniel's fourth question. The evidence. What does the difference actually look like?
Here's where it gets uncomfortable. You can't run a randomized controlled trial on plane crashes. You can't assign half the passengers to watch the demo and half to skip it, then crash the plane and compare outcomes. The evidence base is necessarily indirect. What researchers can measure is recall. And recall studies show that even frequent flyers retain surprisingly little specific information. They can't describe the overwing exit mechanism. They don't know where the life vest is stowed. They can't tell you how many rows to the nearest exit.
So if the demo's goal is knowledge transfer, it's failing.
By the recall measure, yes. But here's the strongest argument for the demo's persistence. The value may not be knowledge transfer at all. It may be priming. Creating a faint memory trace that activates under stress. Even if you can't consciously recall the details, the demo has planted something. When the cabin fills with smoke and the crew shouts brace, some part of your brain recognizes the instruction. It's not novel. You've heard it before. And that familiarity may be what lets you act instead of freezing.
That's the compliance versus knowledge gap. Knowing what to do is not the same as doing it when your body is flooded with adrenaline.
Stress degrades decision-making. The prefrontal cortex, the part that does deliberate reasoning, goes offline. The amygdala takes over. You act on trained responses, on primed patterns, on whatever has been drilled into you. The safety demo is a very light form of drilling, but it's something. It's a faint map of the terrain. When the terrain is on fire, even a faint map is better than none.
So the demo may be working in a way that recall studies can't capture.
That's the argument. The Japan Airlines evacuation is the strongest evidence for it. Those passengers didn't consciously think through the safety briefing while the plane was burning. They just did what the briefing had repeated hundreds of times across their flying lives. Leave the bag. Move to the exit. Follow the crew. The message was in there somewhere.
Which brings us to the future. Could AI-driven personalized safety briefings improve compliance? Briefings based on seat location, aircraft type, passenger experience level?
Technically, yes. The airline knows your seat, your aircraft type, your frequent flyer status. A personalized briefing could show you the exact exit nearest to your seat, the exact mechanism for that exit on this specific aircraft, the exact path to get there. It could quiz you, confirm you've understood, flag the parts you got wrong.
And practically?
The barriers are regulatory, logistical, and psychological. The FAA would need to approve any alternative to the live or video demonstration. The airlines would need to implement the system across their fleet. And passengers would need to actually engage with it. The engagement problem is the same one the current demo has. A personalized briefing you ignore is no better than a generic briefing you ignore.
There's also a question of whether personalization would help or hurt. The current demo is a shared ritual. Everyone on the plane experiences the same thing at the same time. There's something about that collective attention, even if it's half-hearted, that may matter.
That's the ritual argument. The demo marks the transition from the chaos of boarding to the order of flight. It's a moment where the cabin goes quiet, the crew stands in the aisle, and everyone is briefly oriented to the same information. Even if nobody's really listening, the ritual itself creates a baseline. A shared understanding that we're all in this together, and here's what we'll do if it goes wrong.
So where does that leave us? If the demo is a ritual, what kind of ritual should it be?
The open question is whether a ritual designed for first-time flyers can also serve the frequent flyer. Right now it doesn't. The frequent flyer tunes it out because it's pitched at someone else. But maybe that's acceptable. Maybe the demo is doing its real work in the background, priming the brain, marking the transition, giving the crew a chance to run through their own procedures.
The crew. That's a point we haven't touched.
The demo is as much for the crew as it is for the passengers. The crew member performing the demo is running through the emergency procedures in their own head. The seatbelt, the oxygen mask, the life vest. It's a rehearsal. It keeps the sequence fresh. When the emergency actually happens, the crew has performed this script thousands of times. The muscle memory is there.
So even if every passenger is ignoring the demo, the demo is still doing work.
It's doing work for the people who will actually be responsible for saving lives. The crew's training is the real safety system. The demo is part of that training.
Nineteen minutes at Haneda. That's what the system looks like when it works.
And the system includes the demo, the instruction cards, the exit row briefings, the crew's recurrent training, the aircraft design, the evacuation certification standards. All of it layered together. The demo is one layer. It's not the whole system, but it's the layer the passenger actually sees.
I keep thinking about the person who grabs their bag. The demo has told them, explicitly, every single flight, leave everything behind. And in the moment, they still reach for it.
The gap between knowing and doing is the whole problem. The demo can close the knowing gap. It can't close the doing gap. That's a different kind of training. That's drilling, repetition under stress, muscle memory. The demo is too gentle for that.
So the demo is a ritual that works best for the people who need it least, and the people who need it most are the ones who've stopped listening.
The evidence for what it actually accomplishes is indirect, contested, and unsatisfying. But the Japan Airlines evacuation suggests that when the system works, it works completely. Three hundred seventy-nine people, a burning aircraft, nineteen minutes, zero fatalities.
Zero baggage.
Zero baggage. That's the detail that matters.
Hilbert: Nineteen minutes is fast, but the Haneda evacuation was not the fastest. The certification standard for a full aircraft evacuation is ninety seconds.
Ninety seconds.
Hilbert: The manufacturer has to demonstrate that a full load of passengers can get off the aircraft in ninety seconds with half the exits blocked. That's the design requirement. Real evacuations are slower. Haneda was fast for a real fire, but the certification test is a different animal. Volunteers, no smoke, no fire, no panic, and they know it's a test.
That's the gap between the lab and the real world. The ninety-second certification is done under ideal conditions. The real evacuation is chaos.
Hilbert: I did the safety demo for a regional airline in the early nineties. Hundreds of times. Thousands, maybe. Same script every time. Seatbelt, oxygen, life vest, exits. You stand in the aisle and you do the motions and you watch the passengers. Most of them are reading, sleeping, looking out the window. You learn to pick one person per flight. Just one. Make eye contact with them during the demo. See if they're actually watching.
Did they watch?
Hilbert: Sometimes. Most of the time they'd look away when they realized you were looking at them. Like they'd been caught doing something.
The demo as a social ritual. Nobody wants to be the one who's actually paying attention.
Hilbert: The thing about the demo, and you two didn't mention this, is that it's not just for the passengers. It's for the crew. When I was doing it, I was running through the emergency procedures in my head every single time. The oxygen mask drop, the life vest inflation, the exit operation. You do that twice a day, five days a week, for years, and it's in there. It's not something you think about. It's something you do.
The demo is a rehearsal for the crew, not just an instruction for the passengers.
Hilbert: That's what I'm saying. The passengers don't need to remember where the life vest is. I need to remember. And I did, because I did the demo so many times I could do it in my sleep. Same as your friend Daniel says he could recite it. The difference is, for me, reciting it was the job.
The demo as crew training. That reframes the whole thing. Even if every passenger ignores it, the crew is rehearsing.
Hilbert: One time, a passenger came up to me after landing. He said, I want to tell you something. He recited the entire safety demo back to me. Word for word. Including the part about the life vest. The thing is, that aircraft didn't have life vests. It was a regional turboprop, no overwater routes. The demo didn't include the life vest. But this passenger had flown the same route so many times on a different aircraft type that he'd memorized a demo for a plane he wasn't on.
He was paying attention to the wrong thing.
Hilbert: He was paying attention to a demo from a different aircraft. He had it perfectly memorized. The wrong one. And that's the only time I ever felt like the demo actually worked. Because someone was actually listening. They were just listening to the wrong briefing.
That's the habituation problem in miniature. The information is there, but it's not tied to the specific aircraft, the specific seat, the specific exit. It's a generic script floating in the passenger's head, disconnected from the physical reality of the flight they're on.
The one passenger who memorized the demo memorized the wrong one. That's almost too perfect.
Hilbert: I think about him sometimes. If that plane had gone down in water, he'd have been looking for a life vest that wasn't there.
That's the strongest argument for the personalized briefing idea. The generic demo gives you generic information. The specific information, the exit nearest your seat, the mechanism on this aircraft, the life vest location on this aircraft, that's what actually matters in an emergency. And the current demo doesn't deliver it.
But the personalized briefing has the engagement problem. The passenger who memorized the wrong demo was engaged enough to memorize something. Most passengers aren't engaged at all.
Hilbert: The demo's not there to make people engaged. It's there to make the crew ready. The passengers are going to do what they're going to do. Some of them will leave their bags. Some of them won't. The crew is the difference.
The compliance data supports that. The Japan Airlines evacuation worked because the crew was trained and the passengers followed instructions. The Aeroflot evacuation failed because the crew lost control of the cabin and passengers reverted to their own instincts. The crew is the variable.
The demo is a ritual that keeps the crew sharp, and the passengers get whatever faint priming they get. That's the deal.
Hilbert: That's the deal.
Which means the question of whether the demo makes a difference is partly the wrong question. The demo makes a difference for the crew. For the passengers, it's a ritual that may or may not prime them. The real safety system is the crew's training, and the demo is one component of that training.
If we're designing a demo for frequent flyers, we're designing it for the wrong audience. The frequent flyer doesn't need the demo. The crew needs the demo. The frequent flyer needs something else entirely.
The frequent flyer needs to know the specific exit nearest their seat, the specific mechanism for that exit on this aircraft, and the specific instruction to leave the bag behind. And they need that information delivered in a way that cuts through the habituation. A personalized briefing could do that. A quiz could do that. A seat-specific card could do that. But none of it replaces the crew's rehearsal.
The cutting-room floor detail I found is that the ninety-second evacuation certification standard has been in place since the nineteen sixties, and it's been criticized for decades as unrealistic. The FAA has been talking about updating it for years, but the aircraft manufacturers resist because redesigning for a stricter standard would cost billions. So the standard that governs how fast passengers can get out of a burning aircraft is based on a test that everyone knows is too optimistic.
That's the thing about aviation safety. The systems are layered, and each layer has its own compromises. The demo is a compromise. The certification standard is a compromise. The evacuation slides are a compromise. And somehow, most of the time, the layers hold.
The open question I'm left with is this. If the safety demo is a ritual that works best for the people who need it least, what would a demo designed for frequent flyers actually look like? And would it still be a ritual, or would it become something more like a personalized training session?
As aircraft become more automated and emergencies rarer, the demo's role may shift from instruction to something more like a shared moment of collective attention. A brief pause before flight where the cabin orients itself to the possibility, however remote, that something could go wrong. That's not nothing. That's a ritual. And rituals have power, even when nobody's consciously listening.
Thanks to our producer, Hilbert Flumingtop, for keeping the show running.
This has been My Weird Prompts, the human-AI collaboration podcast. Email us at show at my weird prompts dot com with your own weird prompts, or find us at my weird prompts dot com.
We'll be back soon.
