What The Rehearsal Gets Wrong About Pilots–and Right About Leadership Training
Season 2 of The Rehearsal opens in a cockpit. Two pilots are on final approach. The first officer thinks they're lined up wrong and says so, mildly. The captain brushes it off. The first officer tries again, a little more nervous this time, and the captain keeps descending. Then the alarms, a go-around attempted too late, and fire filling the screen. The camera pans left and Nathan Fielder is standing there on the set, looking at the lens.
It's a re-creation of a real accident, and if you've spent any time in Lean or patient safety, you have seen this play out before. Somebody knew. Somebody said something, sort of. It wasn't enough.
Fielder spends the rest of the season arguing that first officers don't speak up to captains, that this causes crashes, and that nobody is doing anything about it. I watched all six episodes. I think he's probably wrong about the size of the aviation problem and right about the part that matters more to the leaders I work with: training people to speak up is not the same as changing what happens when they do.
The FAA said the training exists
After the finale, Fielder went on CNN. The FAA had given the network a statement saying it requires all airline crewmembers and dispatchers to complete Crew Resource Management training, before they start and on a recurring basis afterward, and that it isn't seeing data supporting the show's claim.
Fielder's response, quoted in coverage of the segment:
“The training is someone shows you a PowerPoint slide saying, ‘If you are a co-pilot and the captain does something wrong, you need to speak up about it.' That's all.”
He also called the FAA dumb, which is not how I'd have played it. But look at what each side actually said. The FAA answered the question “does the training exist?” Fielder was asking, “Does the training change what people do at 200 feet with a captain who outranks them?” Those are different questions, and Fielder's question is the right one.
I've been in the room for the leadership version of this more than once. Imagine a hotel conference space, air conditioning set too low, forty managers at round tables, a facilitator with good slides. Ninety minutes on psychological safety. Amy Edmondson's definition, two case studies, a discussion prompt at each table about barriers to speaking up.
At one of those tables is a nurse manager who has been up since four. She writes “make it safe to raise concerns” on her handout.
She means it.
Then she goes back to a unit with a vacancy she can't fill and a length-of-stay number she's not hitting, and on Thursday a tech tells her the new supply cart puts two look-alike vials at eye level. She says she'll look into it. Six weeks later, the engagement survey comes back; her unit's score hasn't moved, and someone above her asks what she's doing about it.
Nobody practiced anything in that room. Her director wasn't in it. Nobody went to the unit that week to see what happened when the tech spoke up. The attendance sheet went in the file and the requirement was met.
That's the parallel worth thinking about, and it doesn't depend on Fielder being right about plane crashes.
Was Fielder rehashing a solved problem?
Partly, yes. And the pushback is fair.
Slate interviewed a cargo airline first officer, identified only as James, who liked the show and still said
Fielder “is on the right track, but he's drawn some weird conclusions.”
Well, Fielder is a weird guy, in an endearing way.
The first officer's most useful observation: the accidents Fielder dramatizes are the same accidents shown as video re-creations in annual recurrent training.
“Those accidents that he thinks we need to figure out how to avoid in the future have already been incorporated into our training.”
He also said communication is “a huge, absolutely huge” part of airline training, and that he wouldn't call it solved, but the industry is further along than Fielder gives it credit for.
He's right that the case selection is circular. Tenerife, Air Florida, United 173, Korean Air 801 – those are the accidents that created CRM in the first place. Using them to argue CRM doesn't exist is a bit like citing the Ford Pinto to argue nobody does crash testing or that today's auto safety is abysmal.
And US scheduled passenger service went from the Colgan Air crash in February 2009 to the DCA midair in January 2025 without a fatal crash. That is an extraordinary record. That doesn't mean everything is as safe as it could be. Fielder's implication that cockpit miscommunication is still killing people with any frequency doesn't seem supported these days in the U.S.
So: rehash, or never-ending?
On September 6, a Boeing 767 operating as Amazon Prime Air Flight 7598 overran a runway at Miami International, went through a perimeter fence, struck a van and a car, and killed five members of a ground cleaning crew. The NTSB released an investigative update on September 10.
I want to be careful with that, and not only because it lands in the middle of a post about cockpit communication. An investigative update is not a finding of cause. The NTSB hasn't said what happened or why. The update doesn't identify which pilot was flying or which one raised the concern, so nothing in it establishes a captain-and-first-officer problem.
Aviation attorney Pablo Rojas told CBS News that without the actual recording it's hard to know what the NTSB means by an inconsistent verbal response, and I think that's the right caution. A pilot hand-flying a fast, unstable approach may not be answering because he has run out of attention, which is a different problem with different countermeasures.
I'd say neither, cleanly. The honest answer is probably closer to what we say about safety culture generally.
Speaking up in the cockpit is not a problem that got solved and stayed solved.
Patient safety hasn't been solved. In fact, it's far less solved, even with the effort to learn from aviation.
Safety is a condition that has to be maintained, because every year brings new captains, new first officers, new hierarchies, and new pressures.
The part Fielder found that still looks unresolved came from John Goglia, the former NTSB member who appears throughout the season. After the 2009 Colgan crash, the NTSB recommended that the FAA require role-playing or simulator-based training to teach first officers to assertively voice concerns. The FAA didn't adopt it. [VERIFY: recommendation number and current status]
That's not a solved problem. That's a specific gap somebody documented and nobody closed.
Role playing is less ridiculous than it sounds
Fielder's proposed fix might seem absurd to some. Before each flight, both pilots get a one-page character sheet, like they were actors (and he plays this out in the final episode of the season). The first officer becomes “First Officer Blunt,” who loves speaking up and is honest almost to a fault. The captain becomes “Captain Allears,” who loves feedback and isn't afraid to admit when he's wrong. When Fielder pitched this to a member of Congress, it went about as well as you'd expect.
But strip away the HBO of it. What Fielder is trying to do is lower the social cost of speaking up by giving somebody a role to hide behind. If the challenge comes from the character, the person doesn't have to carry it.
He also tried hard (if not awkwardly) to demonstrate this by inviting feedback from his first officer during his flight as a captain, in the final episode again.
We already do a version of this and we don't call it acting. TeamSTEPPS teaches the CUS words – I'm Concerned, I'm Uncomfortable, this is a Safety issue – and the two-challenge rule, which says that if your first concern gets brushed off, you are obligated to raise it again. That's a script. The whole point of a script is that you don't have to invent the words, and gauge the relationship, and manage your own nerves, all in the two seconds you have.
My retired pilot friend Steve Montague teaches even more specific language for how to speak up to those in power through these three podcast episodes:
Crew Resource Management in Healthcare: Lessons from the Cockpit, with Steve Montague
Aviation Lessons for Patient Safety with Steve Montague
A Pilot's Least Proud Moment Was Saying Nothing — with Captain Steve Montague
An andon cord works the same way. Pulling it is a defined act with a defined response. Nobody has to compose a sentence.
Where I'd push back on Fielder is the emphasis. He puts too much of the problem on the first officer's assertiveness. Most of what determines whether people speak up is what happened the last few times somebody did. Goglia made the point on the show that a captain can get a first officer fired, and pretty quickly. No character sheet or dry reading of a PowerPoint deck fixes that.
He went to gemba, which is the part I keep thinking about
Fielder spent two years getting his license and a 737 type rating so he could fly the finale himself, with 150 people in back. All of them were actors, and that is the only reason it was legal. As Columbia Journalism Review pointed out, he could fly a passenger jet without the experience an airline would require of him because nobody on board had bought a ticket. He was short of the hours for an airline transport pilot certificate. The loophole is that the rules turn on whether the passengers are paying, not on whether they are people.
Whatever else you want to say about the season, though, that is a serious attempt to understand the work by doing it.
And the detail that got me was a training detail, not a comedy one. A pilot can be type rated on a full-flight simulator, pass the checkride in the simulator, and then fly the actual airplane for the first time on a real revenue flight with passengers in back.
AOPA's flight training magazine puts it flatly:
“Yes, you fly the real McCoy for the first time with paying passengers.”
It works because the simulator is a high-fidelity replica of the actual work, and because a line check airman is sitting right there, watching, with authority to intervene and to decide you're not ready.
That's simulation plus supervised real work, with a person accountable for judging competence in the actual conditions. Fielder found his way around that system. The system itself is real, and it's the good version.
Now think about the leadership version. We put managers in a class on Lean or psychological safety. There's no simulator. There's no supervised first attempt. There's no check airman (or airwoman) deciding whether they're ready to be released to the line. There's a certificate, perhaps.
I don't think the answer is more training hours. When a supervisor comes out of a psychological safety class and someone on their team raises a problem that week, who sees what happens next? Who coaches them? Who models the ideal behaviors?
