Whether you're a hobbyist investor, a tech enthusiast, or simply someone with a Wi-Fi connection, Nvidia has almost certainly captured your attention in recent years.
Their hardware underpins the most significant technological breakthroughs of our era, achieving a rare duality: They command the market with monolithic dominance, yet have the agility of a start-up. In the framework of The Innovator's Dilemma, a company this large, yet this nimble, simply shouldn't exist.
Tracing their arc from a 1993 founding at a San Jose Denny's all the way to the Big Bang of large language models (LLMs), it is clear that engineering prowess has been one of their primary differentiators, propelling them toward a $5 trillion valuation.
However, less intuitive in their story, but perhaps more important, are the efforts they take to master a wide range of industries to drive product development.
Stumbling Into Trillions
Graphics processing units (GPUs) are powering much of the AI revolution. But before these chips became a foundational product for AI development, their primary purpose was to transform video game graphics.
In the early 2000s, it came to Nvidia's attention that these GPUs were being repurposed by computer scientists for nongraphical applications, such as running simulations. The unanticipated use case prompted Nvidia scientist David Kirk to comment that "Really, the modern GPU, we had kind of stumbled onto."
Armed with insight into how GPUs can be used across many other industries outside of gaming, Nvidia then set out to grease the wheels for programming its chips with a model called CUDA. With that, a trillion-dollar moat was born.
Of course, it wasn't as linear as that. GPUs and CUDA have enabled the company's spectacular growth, but along the way Nvidia had to create a market for customers to optimize its use. From Tae Kim's The Nvidia Way:
Jensen [Huang] asked his team to figure out ways to create a market for CUDA—to solve the "entire problem", as he put it. This would require a systematic analysis of the needs of every industry, from entertainment to health care to energy, and to not only analyze potential demand but also figure out how to unleash it via special, GPU-centered applications in each field. If developers didn't yet know what to do with CUDA, Nvidia would teach them.
Comparing Semiconductors and… Insurance?
Nvidia's success isn't just a technology story – it's a reinforcement of "the more you know."
The alignment of Nvidia's technological expertise to the needs of various complex and specialized industries is something to be admired. And while the products and services differ entirely, insurance professionals have to engage in a similar process.
Whether someone is an insurance underwriter, broker or in another role, insurance expertise is beneficial to the extent that they understand what they are insuring. It is great to know all of the coverage forms, but what good is that without awareness of what the product is ultimately used for by the insured?
Historically, the range of these exposures wasn't overly specialized (hence the "general" in the general liability LOB!). But the world of risk has changed dramatically in the 21st century, leading to new insurance programs, a slew of novel coverage forms, and a steady shifting of accounts over to the E&S market.
There's no reason to think the pace of change will slow. Insurance professionals will be expected to continuously sharpen both insurance and domain knowledge so they can come up with solutions to effectively underwrite and protect businesses with complex and emerging needs (and risks).
Just like Nvidia, the most successful insurance carriers and brokers will be the ones who equip their staff to understand these different industry nuances and build/market the most sustainable risk solutions in a rapidly evolving risk environment.
