The Innovator's Dilemma at Global Scale

I want you to understand something about how companies die. They rarely die from stupidity. The executives at Blockbuster were not idiots. The engineers at Kodak invented digital photography before anyone else. Nokia's leadership had built one of the most valuable companies on Earth. These organizations were filled with smart, capable people who had spent decades building successful businesses. And then they got obliterated. Not because they failed to see the threat coming. In most cases, they saw it clearly. Blockbuster had numerous opportunities to buy Netflix. Kodak's own engineers were begging leadership to embrace digital. Nokia watched the iPhone launch and understood what it meant. They died because the structure of their businesses made it impossible to respond. The thing that made them successful became the thing that killed them. Their very competence in the old paradigm guaranteed their failure in the new one. This is the innovator's dilemma - a concept Clayton Christensen brilliantly articulated in his 1997 book of the same name. And it is about to play out at a scale we have never witnessed before.

The Pattern That Keeps Repeating

The innovator's dilemma is one of those frameworks that, once you

understand it, you start seeing everywhere. Christensen's core insight is simple but counterintuitive: successful companies fail not despite their capabilities, but because of them. Before I apply this to automotive, let me show you that this pattern is universal. It plays out the same way across every industry, every era, every technology transition. Understanding the historical examples helps you recognize the pattern when it is happening in real time - which is right now. Kodak and digital photography: Kodak did not miss the digital revolution. They invented it. Their engineer Steve Sasson built the first digital camera in 1975. Kodak understood the technology better than anyone on Earth because they literally created it. But digital photography was an existential threat to their film business. Every digital photo taken was a print that would never be made, a roll of film that would never be sold. So they buried their own invention. They spent two decades telling themselves digital would stay inferior. By the time they admitted the truth, every competitive advantage they had accumulated over a century was worthless. Blockbuster and Netflix: Blockbuster had 9,000 stores at their peak. Relationships with every studio. Distribution logistics across North America. Customer relationships built on decades of Friday night rentals. They could have acquired a controlling stake in Netflix for $50 million in 2000. They passed - not because they were stupid, but because Netflix was tiny and unproven and Blockbuster knew the video rental business better than anyone. From their vantage point, streaming was a niche curiosity. By the time it obviously was not, the window to respond had closed. Nokia and smartphones: Nokia did not miss the smartphone. They saw the iPhone launch and understood immediately what it meant. They had more engineering resources than Apple. Relationships with every carrier on Earth. Manufacturing scale Apple could only dream of. What Nokia could not do was integrate hardware and software into a seamless experience. Their

organizational competence was optimized for hardware excellence. Software was something they bought or licensed. When the iPhone proved that integration was everything, Nokia's century of advantages became liabilities overnight. Traditional media and streaming: This one is still playing out. The major networks and cable companies built empires on bundling, advertising, and distribution control. They could see Netflix and YouTube and TikTok coming. Many of them launched their own streaming services. But their entire business model was structured around the old way - cable fees, advertising loads, scheduled programming. To truly embrace streaming meant cannibalizing the cash cows that paid for everything. So they half-committed, protecting legacy revenue while the insurgents took their audience. The pattern is identical every time. The incumbent sees the threat. The incumbent has superior resources. The incumbent cannot respond because responding means destroying what made them successful. The window closes faster than anyone expects. Think about what happens when a new technology emerges. The incumbent has a massive customer base, established supply chains, optimized manufacturing, trained workers, brand recognition, distribution networks, and institutional knowledge accumulated over decades. These are enormous advantages. In theory, the incumbent should be able to outcompete any startup that tries to challenge them. But they almost never do. Why? Because all those advantages are optimized for the old way of doing things. The supply chains are built around old components. The manufacturing is designed for old products. The workers have skills suited to old processes. The distribution networks serve old customer relationships. To embrace the new technology, the incumbent would have to tear apart the very systems that made them successful. They would have to tell their best customers that the products they love are going away. They would have to lay

off workers who have dedicated careers to the old way. They would have to write off billions in optimized infrastructure. And they would have to do all this while the new technology is still unproven, still generating lower margins, still serving a customer base much smaller than their existing one. So they do not do it. They cannot do it. Not because they are stupid, but because the rational choice in the short term is to squeeze more juice from the existing business while the upstart is still small. And by the time the upstart is not small anymore, it is too late.

Legacy Auto and the Approaching Cliff

I think most legacy automakers in the world are toast. GM, Ford, Stellantis, Toyota, Nissan, VW… - they are facing an existential threat that their corporate structures simply cannot process. And I do not say this with any joy. These are American institutions that have employed millions of people over generations. Their decline will have real consequences for real families. But I have to be honest about what I see. The issue is not that legacy auto executives are incapable of understanding electric vehicles or autonomous driving. They understand these things. Many of them have given presentations about the transition to EVs and self driving cars. They have announced ambitious electrification targets. They have poured billions into battery development and software teams. None of that matters. What matters is that their entire business model is structured around internal combustion engines. Their manufacturing expertise is in machining engine blocks, transmitting power through complex gearbox assemblies, managing exhaust systems with thousands of components. They have spent a century optimizing for something that is about to become obsolete. And more critically, they have outsourced the things that will actually matter.

This is the detail that kills them. Legacy automakers do not make their own

chips. They do not make their own batteries at meaningful scale. They do not write their own software. They have spent decades pushing complexity onto suppliers because it made their operations simpler and their margins higher. All the while, vehicles that people will buy are now simply robots on wheels. We covered Tesla’s FSD in detail in the first section. Try comparing that product with anything Legacy Auto offers. It’s a joke. But as it pertains to Legacy Auto, it was a completely rational strategy. Right up until the moment when chips, batteries, and software became the only things that matter. Tesla builds everything in-house. Their own chips. Their own battery cells. Their own software stack. When Tesla encounters a supply chain problem, they can redesign around it. When they want to improve full self-driving, they do not need permission from a tier-one supplier. When they want to optimize battery chemistry, they are not waiting on someone else's timeline. Legacy auto is not competing with Tesla on cars. They are competing with Tesla on integration. And they are losing badly. Just go to any major metropolitan area and look around. Or go to places like San Francisco or Austin and tell me how many self driving electric cars you see. It’s already happened. It’s just a matter of time.

The Automotive Parallel

The patterns I described above - Blockbuster, Kodak, Nokia - are not just

historical curiosities. They are playing out right now in automotive. Like Blockbuster, legacy automakers can see the threat. They can see Tesla. But the gap between seeing the threat and being able to respond may have already closed. The supply chains they would need to rebuild take years. The software expertise they would need to develop takes a decade to mature. The manufacturing techniques they would need to adopt require billions in new

facilities. And they would have to do all of this while still making money from the old business to fund the transformation. Like Kodak, they are burying their own potential. I see echoes of Kodak every time legacy automakers talk about Tesla as if it is not a "real car company." As if manufacturing expertise in internal combustion engines is what defines a car company. As if a century of history making one thing proves you can make a different thing. Tesla is a technology company that happens to sell vehicles - the same way Netflix is a technology company that happens to deliver entertainment. And that distinction is the whole game. And like Nokia, they are caught in the integration trap. When a product becomes a platform, the ability to integrate across layers becomes more important than excellence in any single layer. Nokia was excellent at hardware. Apple was mediocre at hardware but excellent at integration. Integration won. Legacy automakers are in the exact same position - excellent at manufacturing traditional vehicles, but vehicles are becoming platforms. Software is becoming the differentiator. And the ability to integrate across the full stack - chips, batteries, motors, software, data, AI - is what determines competitive success. That integration capability is exactly what legacy automakers outsourced away. It is an almost impossible transition. And I think most of them will not make it.

What Happens Next

So where does this leave us? What happens as the innovator's dilemma plays

out across automotive and beyond? I think we are going to see the fastest industrial disruption in history. Not because the technology is changing faster - though it is - but because the advantages of incumbency are smaller than they have ever been. In previous transitions, physical infrastructure provided durable advantages. Factories took years to build. Distribution networks took decades to

establish. Customer relationships were sticky because switching costs were high. But when software becomes the product, these advantages erode quickly. You do not need a century of manufacturing expertise when manufacturing is commoditized. You do not need distribution networks when the product is delivered over the air. You do not need sticky customer relationships when the product is dramatically better. Legacy auto is just the most visible example. But the same pattern is about to play out across every industry where software and AI can become the primary value drivers. Finance is next. Traditional banks have centuries of accumulated expertise in evaluating risk, managing capital, processing transactions. All of that expertise lives in human judgment and legacy systems. When AI can underwrite loans better than human analysts, when smart contracts can execute transactions without intermediaries, when risk assessment can happen in milliseconds instead of days - what is left of the traditional banking advantage? Healthcare follows the same pattern. Diagnosis is fundamentally pattern recognition. Treatment protocols are fundamentally decision trees. Administrative overhead is fundamentally workflow management. Every one of those functions is susceptible to AI automation. The century-old structures of hospitals, insurance networks, and credentialing systems are about to face the same pressures that are dismantling automotive incumbents. Education is already transforming. When AI can provide personalized tutoring that adapts to individual learning styles, what is the value proposition of a lecture hall with 300 students? When credentials can be verified through demonstrated competence rather than institutional affiliation, what happens to the university's monopoly on certification? Legal services are beginning to crack. Document review, contract analysis, case research - these were high-margin activities performed by armies of associates. AI is already doing them faster and cheaper. The partnership

model that has defined law firms for centuries is structurally unprepared for what is coming. In every one of these industries, the incumbents have massive apparent advantages. They have expertise, relationships, infrastructure, brand recognition, regulatory capture. And in every one of these industries, those advantages are about to become massive liabilities.

The Speed Problem

One of the things that makes this disruption different is the speed at which it

is happening. When the automobile replaced the horse, the transition took decades - but what happened to the horse population when cars took over? It dropped by roughly 90% from 1920 to 1960. When electricity replaced gas lighting, it was a generational shift. AI moves at a completely different pace. Far faster than the computer. Far faster than the internet. The capabilities we are seeing today did not exist two years ago. GPT-3 was interesting but limited. GPT-4 changed what was possible. And the models coming next will change it again. Each generation compounds on the previous one. The improvement curve is steep and showing no signs of flattening. For incumbents, this speed is fatal. Traditional corporate planning works on three-to-five year cycles. You identify a threat, analyze it, form a committee, develop a strategy, get budget approval, execute the plan, measure results. By the time a large company completes that cycle once, the technology has advanced through multiple generations. This is why I think the gap has already closed for most legacy automakers. They could theoretically adapt if they started today - the problem is that adaptation takes longer than the time remaining before the disruption fully materializes.

Tesla is not waiting for legacy auto to catch up. They are improving their

software every week. Their FSD system gets better with every mile driven. Their manufacturing efficiency improves with every car produced. While legacy auto is still developing strategies to respond to where Tesla was two years ago, Tesla has already moved on to where they will be in two years. You cannot catch up to a target that is accelerating away from you. And the acceleration is the key variable that traditional strategic thinking fails to account for.

The China Variable

I need to spend some time on China, because the competitive dynamics there

add a layer of complexity to everything I have been saying. The Chinese automotive market is the largest in the world. Chinese manufacturers have gotten very good at making electric vehicles. And Chinese companies do not labor under the same innovator's dilemma that plagues Western incumbents, because many of them do not have the same legacy to protect. BYD, as I mentioned, is formidable. But there are dozens of Chinese EV companies moving fast. They have government support. They have access to the entire battery supply chain. They have manufacturing infrastructure that took decades to build. And the Chinese business community operates with a pragmatism that American companies often lack. I think the Chinese approach to business is actually more honest in some ways. When the US legacy business community looks at disruption, they form industry associations to lobby for protection. They seek regulations to slow down the disruptor. They complain to media outlets about unfair competition. The Chinese just try to win.

Does this make them easier to compete against? In some ways yes, in some

ways no. They will move faster than Western incumbents. They will iterate more aggressively. They will take risks that American public companies cannot stomach because of quarterly earnings pressure. But they also face their own constraints. The Chinese companies are largely locked out of Western markets due to tariffs and security concerns. Their cutting edge technology is genuinely behind America’s - chips, self-driving, and AI models as an example. But there is a deeper issue. China is good at manufacturing. China is good at scaling production. China is good at optimizing known processes. But the innovator's dilemma ultimately comes down to organizational willingness to embrace the unknown. And in a system where no one can become more powerful than the state, there are limits to how much disruption any company can drive. I think China will produce competitive electric vehicles for decades. I am less certain they will lead the transition to autonomy. The data advantage matters too much, and the data is collected from vehicles driving globally, not just Shenzhen and Shanghai - although even that is changing quickly.

Why Timing Matters More Than Ever

In previous technological transitions, timing was important but not decisive. Miss the first wave of electrification? There would be a second wave. Fall behind on computers? Catch up with the next generation. The transitions played out over long enough periods that companies had multiple chances to adapt. AI does not work that way. AI systems exhibit what is called a capability overhang - improvements that happen suddenly rather than gradually. A system that cannot do something on Monday might be able to do it flawlessly by Friday. The S-curves that describe traditional technology adoption are steeper and shorter.

Take AI agents as an example. At the start of 2026, we are beginning to see

the massive impacts that agents can have on the economy, with products like Claude Code, CODEX, and OpenClaw. These are basically pieces of software that can execute tasks on a computer like a human could. Excel spreadsheet work, word documents, coding - you name it. By the end of 2026 - mark my words - AI agents will easily be one of the most disruptive forces happening in the economy globally. There will be tens of millions - perhaps more - of these agents running around the digital world completing tasks for humans at supersonic speed. This means the window for adaptation is smaller than executives expect. The classic mistake is to look at a competitor, note that they are doing something you cannot do, and assume you have time to match them. With AI, by the time you match their current capability, they have moved to something entirely new. I watch legacy auto announcements constantly. They talk about catching up to Tesla's 2021 capabilities. They are proud of matching range and performance metrics from vehicles that are already obsolete. They do not seem to understand that the target is moving. Meanwhile, Tesla's FSD system is approaching the point where unsupervised autonomy is already viable. When that happens, the comparison changes entirely. You are no longer competing on car features. You are competing on whether your vehicle can generate revenue while the owner sleeps. How do you catch up to that? You cannot. Not if you are starting from where legacy auto is starting.

Recognizing the Pattern

The question I think everyone should be asking themselves is this: What

industry am I in, and is it vulnerable to the same pattern?

Because if you work in a sector dominated by large incumbents who have

spent decades optimizing for the current way of doing things, you should be paying very close attention to what is happening in automotive. The warning signs are predictable. Incumbents will dismiss the upstarts as serving niche markets. They will point to their superior resources and expertise. They will announce strategic initiatives to embrace the new technology while actually protecting the old business. They will say the transition will take a while, giving them plenty of time to adapt. And then, suddenly, it will be too late. I am describing pattern recognition, plain and simple. And the pattern is as clear now as it was when Netflix was mailing DVDs to a tiny customer base while Blockbuster executives explained why streaming would never work. This transformation is fundamentally about economic power shifting from incumbents who dominated the old world to insurgents who are building the new one. AI, robotics, and energy are the mechanisms - abundance or collapse is the outcome depending on how it's managed. And the Innovator's Dilemma ensures that this transfer happens faster than anyone expects. The question is not whether it will happen. The question is whether you are positioned on the right side of it.