HMM: Elon Musk’s chaotic business strategy for Tesla is actually brilliant.

The big challenge to understanding Tesla’s strategy is that most of us only look at it from one level of analysis. Namely, when we see Tesla, we see a company that produces cars. But when I teach executives how to invest in future technology, I encourage them to think at multiple levels of the technology stack: not just products but also components and systems. So let’s take a closer look at Tesla.

At the level of the product, although a Tesla looks the same as other vehicles, underneath the hood, the vehicle has a fundamentally different architecture—both in terms of hardware and software. This matters because a long research tradition underscores that when incumbents face a new technology architecture, they struggle to understand and adapt.

Even though they can see what the technology is, they struggle to adapt both because they are reluctant to give up the existing capabilities they have perfected over decades and to fully integrate the new ones. Although incumbents may imitate the new architecture, they have a hard time overcoming the way they have done things in the past and to match the superior performance of the new, purpose-built architecture.

You can see evidence of this playing out in the auto industry. Early electric vehicles produced by incumbents on internal combustion engine architectures paled in comparison to the Tesla, and even newer “blank slate” efforts sometimes don’t quite measure up.

I’ve been saying for a long time now that Tesla’s challenge is to move from boutique manufacturer to mass-manufacturer of electric vehicles before the mass-manufacturers figure out how to make electric vehicles. But it might be more on point to say that the race is between a tech company trying to become a car company, and various car companies trying to become tech companies.