Okay, I've been thinking about this one all morning and I still don't know how to feel about it.
That bad?
Not bad. Just... Daniel has noticed something that's been sitting in plain sight for about a hundred years and nobody's done anything about it. Here's what he wrote. He starts with the experience of driving an unfamiliar road, or a new country, where you and the driver end up working as a team the way a pilot and co-pilot do. The driver drives. The passenger becomes the second set of eyes and hands. You're checking whether the fault light on the dash can wait until the garage or whether it's a pull-over-now situation. You're hunting for a gas station along the route. You're finding a podcast to put through the car speakers.
The unpaid intern of the highway system.
Exactly that. And then he gets to the actual question. He points out that cockpit designers build integrated systems for two people, one in each seat, and the automotive model has been to give the front passenger nothing, or one tablet in the centre of the car that two people can't safely use at once. So he's asking: has any manufacturer built a co-pilot tablet? Something mounted on the dash or above the glovebox so the passenger can do these jobs without depending on their own phone's battery and reception. And the elegant version, in his words, would be a side panel that syncs with the main device. If the driver is following Waze, the passenger's screen opens Waze and shows the same route.
That's the whole question right there.
There's more. He asks whether Android supports that kind of tight integration between two devices sitting next to each other. And he points at taxi rear-seat screens as the closest thing he's seen, where navigation sometimes shows up on a second unit facing the passengers, and asks whether anything like that exists or could exist in a regular car.
Three questions in one. Dedicated screen, platform support, taxi analogue.
So let's find out whether the co-pilot seat ever got its own instrument panel.
Short answer, and then we'll spend the next twenty minutes complicating it. Yes, the screens exist. No, they're not co-pilot instruments. And the reason they're not is more interesting than the screens themselves.
Start with the role. What is a co-pilot, actually, in an aircraft?
In a two-crew cockpit there's a deliberate division of labour. One pilot flies, the other monitors. Pilot flying works the controls, pilot monitoring watches the instruments, reads the checklists, calls out deviations, handles the radios. It's not a hierarchy, it's redundancy plus workload split. Two people doing different jobs on the same aircraft, each with their own instrument panel in front of them.
And the car version of that is real. I've done it. You're in the passenger seat on a road you don't know, you've got the phone, you're the one saying "in four hundred metres there's a turn" and "that light's been on since Haifa, it's the tyre pressure sensor, it can wait."
The role exists. The instrument panel doesn't. What the front passenger has historically had is nothing, or a single centre screen that both people are supposed to share. Which creates a problem nobody talks about, which is that a centre touchscreen is a shared-attention device. Two people cannot safely operate one screen while the car is moving. The driver shouldn't be touching it at all, and the passenger reaching across the driver's field of view to tap it is worse.
So the passenger either uses their own phone, with their own battery and their own reception, or they lean over and become a hazard.
That's the gap Daniel's identified. And here's the finding we're going to spend the episode on. Dedicated front-passenger displays do now exist. They're across luxury brands, they're big, they're expensive, and they are entertainment-and-comfort tools rather than synced co-pilot instruments. The platform support for the tight two-device integration he's imagining is real but early. And the industry is split down the middle on whether any of this should exist at all.
Some brands love them. Some brands won't touch them.
Bentley's product line director said they'd never put one in a car. That's a quote, we'll get to it. So let's start with what actually shipped, because the answer to Daniel's question is yes, but not in the way he imagined.
Where does the timeline start?
Porsche Taycan, 2020. An optional 10.9-inch screen mounted on the dashboard in front of the passenger. That's the origin point. And from there it spread like wildfire, which is the phrase the coverage uses, and it's accurate. Within a few years every luxury brand with a flagship was either shipping one or announcing one.
Give me the current landscape.
The 2026 BMW 7 Series has a 14.6-inch passenger display as standard. Not optional, standard. The upcoming Audi Q9 gets a 12.3-inch display. The 2026 Jeep Grand Wagoneer has a 10.25-inch unit they call the Front Passenger Interactive Touchscreen Display. Zeekr's 9X plug-in hybrid uses three Samsung OLED panels, one for the centre information display, one for the passenger, one for the rear seats.
Three panels. One car.
And then there's Mercedes, who took a different architectural route. Their MBUX Hyperscreen puts three displays under a single 55-inch curved glass span in the EQS. Driver display, central display, and a dedicated front passenger display, all under one sheet of glass so it reads as one continuous surface. The Superscreen brings a second front-passenger screen to the 2026 E-Class, but as an option, not standard.
And what was the E-Class passenger screen sold on?
This is the part that tells you everything. When the 2024 E-Class launched, the passenger-side screen was framed for TikTok and Zoom calls. That's the marketing. Not navigation, not fault diagnosis, not finding fuel. Video calls and short-form video.
The co-pilot seat, and the pitch is TikTok.
Now, to be fair, one manufacturer came closer than the rest. The Jeep Grand Wagoneer's passenger display is the closest thing to a marketed co-pilot tool that actually shipped. The passenger can enter destinations, add stops, reroute around traffic inside the navigation app, pull up camera feeds for parking and trailer alignment, and queue media. And all of it, in Jeep's words, without disrupting the driver's main screen.
That's the co-pilot job. That's literally the list Daniel wrote.
It is. And then Jeep says the thing that's the crux of this entire episode. They say the display "complements, rather than mirrors, the driver's screen."
Complements rather than mirrors.
That's a deliberate design decision, stated out loud, in marketing copy. The passenger screen is not a copy of the driver's screen. It's a different screen showing different things. Which means the one feature Daniel actually asked for, the passenger seeing the same live route the driver is following, is explicitly not what this does.
So the screen exists, mounted exactly where he said it should be mounted, doing the tasks he listed, and it pointedly refuses to do the one thing he said would make it elegant.
Correct. And that's not an oversight. That's a choice. Which brings us to the counter-argument, and this is where it gets good, because there's hard data on the other side.
Go.
Bentley. Their product line director is a man named Martin Page, and he said the company would never add a passenger screen. His line is: "Technology has to add value. It should not dominate the experience." The idea was actually discussed for the 2028 Torcal EV and dismissed. And here's the detail I love. The Bentley Torcal shares a curved OLED central display with its platform-mate, the Porsche Cayenne Electric. Same platform. The Cayenne has a 14.9-inch passenger screen. The Torcal puts wood, aluminum, or carbon fiber in that space instead.
Same bones, opposite decisions.
And it's not just Bentley. The Ferrari Luce, six hundred and forty thousand dollars, no passenger screen. At that price the omission is a statement.
What's the data against?
More than a third of owners of cars with front passenger screens say they never use them. And the same reporting puts it at 21.3 problems per hundred vehicles equipped with front passenger screens. More than one in five. This comes through Motor1's reporting from yesterday, and I want to flag that we're getting the JD Power figures secondhand. I haven't seen the primary document. But the shape of it is consistent with everything else we know about adding screens to cars.
A third never used, and one in five has a problem. That's not a feature, that's a liability with a wallpaper.
It's a solution in search of a problem, which is exactly the phrase Page used. And the reliability number shouldn't surprise anyone who's thought about the environment these things live in. A car interior is a hostile place for electronics. Voltage spikes, vibration, temperature swings that would kill a laptop. A screen mounted on a dashboard in direct sun is being cooked and frozen on a daily cycle.
And the passenger screen has the worst thermal position in the cabin, because it's the one facing the windscreen.
Right. The driver's display is recessed and shaded. The centre screen is angled. The passenger screen sits up on the dash in the sun, and it's the one nobody's using, and it's the one that breaks.
So the sub-thesis here is that the industry is split, and the split is philosophical.
More screens is not automatically better. That's the whole argument. And the co-pilot framing is aspirational marketing rather than a literal description of what these screens do. Jeep says "copilot" and then says "complements rather than mirrors." Those two statements are in tension, and the second one is the honest one.
So the screens exist, but they pointedly don't do the one thing Daniel asked about. The question is whether that's a hardware limitation or a software one, and it turns out the software is already there.
It's already there. This is the part that surprised me. Android Automotive OS has an API specifically for this.
Named?
The Multi-Display Communications API. And the architecture it sits on is called occupant zones. The idea is that each person in the car is a zone. Each zone maps a user to a set of displays, with one main display plus optional extras like a cluster display. So the driver is one zone with the driver display and the centre screen. The front passenger is a different zone with their own screen.
And the API lets those zones talk.
It lets a privileged app in one occupant zone communicate with the same app, same package name, running in a different occupant zone. So if Waze is running in the driver's zone, Waze can also be running in the passenger's zone, and the two instances can exchange messages.
That's the thing Daniel asked for. That's the tight integration.
It is. And now the limits, because the limits are why it hasn't happened. Three components to the API. There's a power management piece, a discovery piece that lets a client monitor other zones and find peer clients, and a connection piece that opens a connection and sends a payload. That's the mechanism.
And the constraints?
Single SoC only. Same package name only. It does not support communication between different apps. And connections are one-way by design. If you want two-way, both clients have to request connections.
So Waze can talk to Waze across the two zones. Waze cannot talk to Google Maps.
Correct. And that's the whole story of why this isn't a product. The API exists. The capability exists. But for a passenger screen to mirror the driver's live Waze route, Waze has to build it. The platform gives them the pipe. Nobody's put anything through it.
The pipe is installed. The water company hasn't shown up.
There's a reference implementation that shows what it looks like when somebody does. Google's Control Center app uses this API to enable something called co-watch. A passenger can mirror a video that's playing on another screen, using task mirroring APIs. And the driver's Control Center handles passenger requests to play audio over the cabin speakers through a heads-up notification the driver can accept or decline.
That accept or decline is the interesting bit.
That's the closest thing shipping to a true co-pilot handshake. The passenger requests something, the driver authorises it. That's a two-person protocol. That's the pilot and co-pilot actually coordinating. And it exists, in a reference app, for audio.
Not for navigation. What about the hardware layer underneath? Is any of this getting easier?
Yes, and this is where LG and MediaTek come in. They built something called the Concurrent Multi-User framework, first shown at Auto Shanghai in 2025, and introduced as a key feature of Android 15. The problem it solves is architectural. The old way to run multiple screens with different content was to run multiple Android virtual machines, one per display, which is heavy. CMU runs multiple displays on a single Android OS instead.
One operating system, several screens, each doing its own thing.
Driver navigating, front passenger streaming video, rear passengers gaming, all at once, on one instance of Android. They demonstrated it on MediaTek's Dimensity Auto Cockpit chip, the MT2718. Mike Chang from MediaTek said they were highlighting "the future of in-vehicle experiences." Eun Seokhyun from LG said they want to transform vehicles into "true living spaces on wheels."
Living spaces on wheels. That's the tell, isn't it. That's the whole philosophy in one phrase.
It is. LG is not selling a co-pilot instrument. LG is selling a lounge. And that's fine, but it's a different product from the one Daniel asked about.
Now the taxi question. Because Daniel's framing assumes taxi screens mirror the driver's navigation.
They don't. And this is the correction I want to make clearly, because the prompt is built on this assumption and it's wrong. Taxi rear-seat displays are a mature, widely deployed category. Ten point one inch HD touchscreens, mounted on headrests or seat backs, with built-in WiFi and Bluetooth, 4G LTE, GPS, and cloud content management systems so fleets can push content to hundreds of cars at once.
That's real infrastructure.
It's very real. But what they show is advertising, news, and travel content. Not a live route map. The screens are not synced to the driver's navigation. They're a content pipe with a screen on the end. So the analogue Daniel reached for is weaker than it looks. The taxi industry built exactly the hardware he's describing and pointed it at advertising instead of navigation.
Which is a strange thing to do. You've got a captive audience in the back of a car and you show them ads.
It's not strange if you're the fleet operator. The screen is a revenue surface. It's not a passenger service. That's the business model, and it explains every design decision downstream of it.
What about the aftermarket? Because there's a huge installed base of cars with no passenger screen at all.
And the aftermarket has filled the void, which I find interesting. For Teslas, third-party fifteen point six inch Android 13 co-pilot screens run between three hundred ninety-nine and five hundred ninety-nine dollars. There's a twelve point three inch Co-Pilot Passenger Entertainment Screen for the Mercedes E-Class W214 at five hundred ninety-eight dollars. Units from VJOYCAR, TrimpexParts, Tesery, AAPEXGEAR. They mount in the OEM dashboard position and create a dedicated dual-screen layout.
So if your car doesn't have one, you can buy one and bolt it in.
And then there's the open-source route, which is the one Daniel would appreciate. There's an app called AA Passenger, on GitHub, created back in 2017. It lets a passenger control an Android Auto device from a second phone. Play music through the car speakers, send destinations from Google Maps to Android Auto for navigation, display pictures on the Android Auto screen. It needs two devices connected over WiFi Direct or an Android hotspot.
So a passenger can push a destination to the driver's navigation from their own phone.
They can send it. They can't mirror the live route back. Which is the same limitation we keep hitting. Sending is easy. Mirroring is the hard part.
And that's the sub-thesis for this half. The platform capability is ahead of the products.
It is. Android Automotive's occupant zones and the Multi-Display Comms API and LG and MediaTek's CMU framework already support exactly the tight two-device integration Daniel is asking about. The gap is that no automaker has shipped a Waze-style synced navigation co-pilot on top of it. The true synced co-pilot is technically possible and not productized. Those are two different statements and both are true.
Why not? If the pipe's there, why hasn't anyone run water through it?
I have theories and no evidence, so take this as thinking out loud. First, it's a liability question. If the passenger screen mirrors the driver's navigation and the passenger says "take the next exit" and the driver does, and something happens, who's responsible? Second, it's a demand question. The JD Power data says a third of people never touch the screen they already have. Third, and I think this is the real one, the entertainment use case sells cars and the co-pilot use case doesn't.
Nobody buys a car because the passenger can reroute around traffic.
Nobody buys a car because the passenger can reroute. People buy cars because the passenger can watch a film on a long drive. And the manufacturers know that.
And speaking of the taxi version of this, I want to know how those things actually get installed, because there's a whole physical layer here nobody talks about.
Hilbert: We used a two hundred and forty volt bench supply and a crimping tool that cost more than the screens did.
Hilbert.
Hilbert: Fleet installs. Taxi and livery. Headrest units, the ten inch ones with the 4G module inside. You pull the seat back, run the harness down the pillar, tap power off the accessory circuit so the screen dies with the ignition, and mount the unit on a bracket that bolts through the headrest posts. Two per car, four cars a day if the seats came out clean.
The harness. What was actually in it?
Hilbert: Power, ground, network, and a content pipe. That's it. There was never a nav feed. I ran every one of those looms and there was no wire in it that carried a route. The screen had no idea where the car was going. It knew where the car was, because it had its own GPS puck, but it didn't know the route.
So Daniel's assumption about taxi screens is just wrong at the wiring level.
Hilbert: It's wrong at the wiring level. And it wasn't a missed opportunity, if that's where you're going. The operators didn't want route mirroring. They specifically didn't want it. You put the route up on the back seat and the passenger watches the meter and the map at the same time, and then they want to have a conversation with the driver about why we went this way instead of that way. Every fleet manager I ever met had already had that conversation and didn't want it again. Ads and content, that's the whole spec. It keeps everybody quiet.
So the screens were deliberately kept away from navigation to avoid an argument.
Hilbert: To avoid a fight, yes. That's a design decision made by a man in an office who was tired of phone calls.
What did people actually do with them?
Hilbert: Tried to turn them off. That was the most common thing, by a wide margin. Second most common was trying to change the channel. And that's why there's a physical power button on the side of the unit, because the touchscreen stopped responding when the cabin got hot. You'd get a car back in August and the screen was a brick until the air conditioning had been on for ten minutes. So the button's mechanical. It always works.
A physical button because the touchscreen fails in heat.
Hilbert: Every unit we fitted had one. Nobody asked for it. We put it on because we got tired of the warranty calls.
The one in five with a problem.
Hilbert: I don't know anything about one in five. I know about the button.
What happened to them? The units.
Hilbert: Last I heard they pulled the whole fleet's screens out and put in newer ones with a different bracket pattern, so the old brackets were scrap. I had a box of them in the van for about a year and then I moved and I don't know where the box went. Anyway, my sister's expecting me at six and I said I'd bring the thing for the thing.
Right. So the screens went in, and the button went on, and the route never went up.
And the reason it never went up is a man in an office who didn't want the phone call. Which is a better answer than anything we came up with.
Which brings us back to the question of what the co-pilot seat is actually for.
There's a reframe at the end of this that I think is the strongest thing in the whole story. Martin Page, the Bentley man, said passenger screens might make sense once cars are autonomous. Think about what that does to the question. If the driver becomes a passenger, then everyone needs a co-pilot screen, and the front-seat passenger's role disappears entirely. The co-pilot job only exists because someone has to drive.
The co-pilot seat is a temporary job. It exists because the driver can't do everything, and it goes away when the driver stops doing anything.
The open question is why nobody's shipped the synced version. The capability is sitting in Android Automotive right now. The occupant zones are there, the multi-display API is there, the CMU framework is there. Is it a demand problem, a liability problem, or just that the entertainment use case sells better?
My guess is the last one. But it's a guess.
Mine too.
The co-pilot seat in a car has never gotten its own instrument panel. And the screens that now occupy that space are pointedly not instruments. Whether that's a design failure or a correct read of what passengers actually want is the open question, and I don't think it's settled.
If you've got a view on it, we'd like to hear it.
Reviews help other people find the show, so if you've enjoyed this one, that's the easiest thing you can do for us. Thanks as always to Hilbert Flumingtop, our producer, who has somewhere to be.
This has been My Weird Prompts.
The human-AI collaboration podcast. We'll be back soon.
