This is part of a series. The Gutenberg/Semantic model describes physical versus logical layers. The Def-Use split describes how authors and users inhabit different semantic models. MDI versus tabs showed the pattern in window management. This post shows the same cycle in your car and your pocket.

1. Native Car Entertainment: The OEM Tribe

Mercedes COMAND. BMW iDrive. Audi MMI. Each a proprietary infotainment system with its own menu hierarchy, its own interaction model, its own update cycle — or lack thereof.

These systems were designed by engineers and UX teams inside the OEM, validated against internal standards, certified for safety compliance, and delivered to customers as a finished Def. Complete before a single real user touched it. Updates arrived every 3-5 years with a new model, not in response to Use signals. The tribe was the OEM, the Tier 1 supplier (Bosch, Continental, Harman), and the certification body — a layered institutional structure whose collective identity was bound up in owning the in-car experience.

The Gutenberg layer — the screen, the speakers, the steering wheel controls, the CAN bus — is legitimately the OEM’s domain. Cars have safety constraints that phones do not. The physical integration is real engineering. But the OEM tribe conflated owning the Gutenberg layer with owning the semantic layer: the maps, the music, the calls, the messages. Those are not car problems. They are phone problems. The tribe defended the Def across both layers simultaneously.

2. The Use Signal Was Brutal and Consistent

JD Power surveys showed infotainment as the top source of owner dissatisfaction for years running. The feedback was not ambiguous:

  • Deep menu trees to change the radio station
  • Rotary controllers that felt clever in the design studio and confusing at 100km/h
  • Maps years out of date the day you collected the car
  • Interfaces that required the manual to operate
  • Software that could not be updated without a dealer visit

The tribe filtered it out. Safety certification made change expensive. Supplier contracts locked in hardware years before production. The Def was the product differentiation — each OEM believed their interface was a competitive advantage, not a liability. The Use signal said otherwise for a decade before anything changed.

3. CarPlay and Android Auto: The Phone Is the Computer

Apple announced CarPlay in 2014. Google followed with Android Auto. The insight was the same one that produced tabs and VS Code: stop defending the old Def and read the Use signal directly.

What do people actually do in cars? They use their phone. So make the phone the semantic layer and the car screen the Gutenberg display. The split is clean:

  • Gutenberg layer (stays with the car): screen, speakers, steering wheel controls, microphone, physical integration
  • Semantic layer (lives on the phone): maps, music, calls, messages, apps — updated continuously, not every model cycle

This is the same Gutenberg/Semantic separation as containers and cloud zones. The semantic layer is portable. The Gutenberg layer is the physical substrate it runs on. You can upgrade one without touching the other.

The OEMs resisted. CarPlay integration was slow, incomplete, or absent on many models for years after the announcement. Some manufacturers treated it as a threat — it handed the user relationship to Apple or Google, made their own infotainment investment look overpriced, and undermined the tribal claim that the in-car experience was theirs to define.

GM announced in 2023 that future electric vehicles would remove CarPlay and Android Auto entirely, pushing users to a GM-developed system. The Use signal came back immediately and loudly — it was among the most complained-about automotive decisions in recent memory. GM partially reversed course. The tribe blinked.

4. Nokia and BlackBerry: The Earlier Chapter

The car infotainment story is the mobile phone story from a decade earlier, with the same structure and the same outcome.

Nokia owned the mobile phone market with Symbian — a complete, mature, carrier-certified operating system. Nokia had the distribution, the manufacturing scale, the carrier relationships, and the tribal Def of what a phone should be: a communication device that happened to have some smart features. The tribe included Nokia’s own engineering culture, the carrier certification bodies, and the IT departments that managed corporate fleets.

BlackBerry owned enterprise mobile with a tighter tribe still: BBM, the physical keyboard, push email, and IT department certification as the single most important procurement criterion. The BlackBerry tribe was the enterprise IT tribe, the security compliance tribe, and the carrier enterprise sales tribe simultaneously. Three certifications defending one Def.

Both had years of Use signal. Touch interfaces were demonstrably faster for browsing. App ecosystems created compounding value. Open platforms attracted developers. Users were buying iPods and Palms and wanting the same experience on their phone. The signal was consistent and growing.

Both filtered it out:

  • Nokia’s response to the iPhone was that the touchscreen keyboard was impractical and that users needed physical buttons
  • BlackBerry’s response was that enterprise security requirements made the iPhone unsuitable for business — true in 2007, increasingly irrelevant by 2010

The iPhone launched in 2007 without MMS, without 3G, without copy-paste, without third-party apps. Technically inferior on the tribe’s own scorecard. But it had a semantic model that matched what users actually wanted — a computer in your pocket — rather than a phone that had been incrementally improved by people who still thought of it as a phone.

By 2013 Nokia’s phone business was sold to Microsoft. By 2016 BlackBerry stopped making phones. The tribe defended the Def until there was no market left to defend it in.

5. The Certification Body as the Tribe’s Weapon

The common thread in car infotainment, Nokia, BlackBerry, and the MDI story is that the tribe owned the certification process. This is the most effective institutional mechanism for slowing the Use signal down:

  • Carrier certification — Nokia and BlackBerry required carrier approval for devices and software. Carriers were part of the tribe. Apple famously forced AT&T to accept a new model: Apple controls the device, the carrier provides the pipe. A clean Gutenberg/Semantic split that the carrier tribe had never accepted before.
  • IT department certification — BlackBerry’s stronghold. Enterprise IT certified BlackBerry and made iPhone adoption a compliance question rather than a Use question. BYOD and MDM eventually dissolved this, but it bought years.
  • Safety certification — the automotive tribe’s equivalent. Infotainment systems require type approval, safety validation, and supplier qualification. Legitimate constraints that the tribe extended far beyond their actual scope to slow the Use signal.
  • Enterprise procurement — the general-purpose version. Long procurement cycles, multi-year contracts, and IT governance all create institutional distance between the Use signal and the Def.

Certification is not inherently tribal. Safety certification in cars is real. Security certification in enterprise phones is real. But the tribe uses legitimate certification requirements as a perimeter around the Def — making any change expensive enough that the Use signal cannot easily penetrate.

The tell is when the certification requirement expands to cover things it was never designed to cover. Car safety certification covering which music app is allowed. Enterprise security certification covering which keyboard design is permitted. At that point the certification has become tribal defence dressed as compliance.

6. The Pattern at Every Scale

Domain Tribal Def Use signal New entrant Tribe’s fate
Car infotainment OEM proprietary systems “I just want my phone” CarPlay / Android Auto Slowly capitulating
Mobile phones Symbian / BlackBerry OS Touch, apps, web iOS / Android Extinct
Browsers IE single-process Stability, speed, tabs Firefox, Chrome IE dead
IDEs Eclipse / Visual Studio MDI Lightweight, fast, tabs VS Code Losing share
Office productivity MDI, proprietary formats Cloud, collaboration, tabs Google Docs, web apps Adapting slowly

In every case the new entrant had no tribal investment in the existing Def. They could read the Use signal directly. The tribe arrived at the same answer eventually — after the market had already moved.

7. Usability as the Use Signal Made Legible

The Nokia and car infotainment stories are often told as usability failures — the interfaces were hard to use, the menus were deep, the learning curve was steep. That framing is correct but incomplete.

Usability is not a separate concern from the Def-Use split. It is the Use signal made legible. When users struggle with an interface, they are telling you that the author’s semantic model does not match theirs. The steep menu tree is not a design failure in isolation — it is the physical residue of a Def that was built from the inside out, organised around how the system works rather than what the user is trying to do.

Nokia’s menus were organised around phone features. Car infotainment menus are organised around system components. BlackBerry’s interface was organised around the IT administrator’s mental model. In each case the Def reflected the tribe’s internal structure rather than the user’s task.

The iPhone’s interface was organised around what the user wanted to do. CarPlay’s interface is organised around what you do while driving. The semantic model matched the Use, not the Def. That is the entire difference.

The weak link willing to learn reads usability failures not as user errors to be corrected with better documentation or a steeper learning curve, but as signals that the Def needs to move closer to the Use. Every confusing menu is a data point. Every workaround is a feature request. Every abandoned feature is a hypothesis that failed.

The tribe trains users. The weak link learns from them.