Technology Disruption: How to Pick the Winners

Every few decades, a technology comes along that reshuffles the entire deck. The printing press. Electricity. The internet. And now, artificial intelligence. When disruption hits, the same pattern repeats: a few companies ride the wave to extraordinary profits, most get crushed under it, and investors – watching from the sidelines or worse, from the wrong side of the trade – wonder how they missed it.

The thing is, they did not miss it. Almost everybody saw the internet coming in 1999. Almost everybody sees AI coming in 2025. Seeing disruption is the easy part. The hard part is figuring out which companies will actually benefit, which will get destroyed, and – critically – what price you should pay for that knowledge. Because overpaying for the right disruption winner still loses you money.

Let me walk you through how I think about this.

Who Actually Benefits From Disruption?

Here is something that trips up most investors: the company that invents the disruption is not always the company that profits from it. Sometimes it is. Often it is not.

Think about the automobile. Hundreds of car companies launched in the early 1900s. Almost all of them went bankrupt. But the real money was made by companies that supplied the auto industry – oil companies, tire manufacturers, highway construction firms – and by businesses that were transformed by the automobile, like suburban real estate developers and fast food chains. The auto industry itself was brutal. The ecosystem around it was phenomenal.

The same pattern plays out with AI today. Yes, NVIDIA has made extraordinary returns selling the pickaxes in this gold rush. Their GPU dominance and the CUDA ecosystem they built over fifteen years created a moat so deep that competitors are still years behind. But look beyond the obvious winners:

• Cloud infrastructure providers (Microsoft Azure, AWS, Google Cloud) are the landlords of the AI economy. Every AI model needs compute, storage, and networking. These companies collect rent whether any individual AI startup succeeds or fails.
• Data-rich incumbents benefit enormously. Companies sitting on decades of proprietary data – medical records, financial transactions, industrial sensor readings – suddenly find that data is worth ten times what it was five years ago because AI can actually extract value from it.
• Enterprise software companies that embed AI into existing workflows (think CRM, ERP, cybersecurity) lock customers in even tighter. If your Salesforce instance now uses AI to predict which leads will close and your team has trained on that system for two years, switching costs just went through the roof.

Meanwhile, who gets destroyed? Companies whose entire value proposition was doing something that AI does better, faster, and cheaper. Basic content creation. Routine data analysis. First-level customer support. Simple translation services. If your business model is “we have a lot of people who do repetitive cognitive work,” you are standing directly in the path of the bulldozer.

The question to ask is not “is this company using AI?” Every company puts AI in their press releases now. The question is: does this company’s existing competitive advantage get wider or narrower because of AI? That is the only question that matters.

Why Do Established Companies Keep Dying to Disruption?

There is a well-documented pattern in business that goes something like this. A successful company dominates its market. A new technology appears that is initially worse than the existing solution – cheaper, simpler, more limited. The established company ignores it because their best customers do not want it. Then the new technology improves. It gets good enough. And by the time the incumbent reacts, it is too late.

This pattern has repeated so many times it is almost boring to recount. Kodak invented the digital camera and then shelved it because film was so profitable. Nokia dominated mobile phones and dismissed the touchscreen smartphone as a toy. Blockbuster had the chance to buy Netflix for $50 million and laughed it out of the room. Traditional taxi companies watched Uber grow for years and responded by … lobbying for regulations.

The core problem is not stupidity. These were smart people running successful companies. The problem is that existing success creates incentives that work against adaptation.

When your current business generates fat margins, every dollar invested in the disruptive technology looks like a terrible allocation of capital. Why cannibalize your own profitable product line for something unproven? Why upset your best customers who are perfectly happy with what they have? Why retrain your entire workforce and restructure your operations for a technology that might not even work?

These are all rational arguments in the short term. And they lead to extinction in the long term.

You can see this playing out right now in several industries:

• Traditional automakers vs. EV-native companies. Legacy car manufacturers are spending tens of billions on EV transitions while simultaneously trying to protect their ICE vehicle profits. They are fighting a war on two fronts. Companies that were born electric – and born with software-first engineering cultures – do not have this internal conflict.
• Legacy banks vs. fintech. Traditional banks have enormous advantages in regulation, trust, and deposit bases. But their technology stacks are often thirty years old, held together with duct tape and COBOL. Fintech companies built on modern cloud infrastructure can iterate ten times faster. The banks that survive will be the ones that genuinely transform their technology, not the ones that bolt a mobile app onto a mainframe.
• Traditional media vs. AI-generated content. Studios, publishers, and agencies that treated AI as a threat to be regulated rather than a tool to be adopted are watching AI-native competitors produce content at a fraction of the cost and a multiple of the speed.

The companies that survive disruption share specific traits. They have management teams willing to cannibalize their own products before someone else does. They maintain financial strength – low debt, high cash reserves – so they can invest heavily during the transition without going broke. And they have cultures that reward experimentation rather than punishing failure.

How Do You Invest in Disruption Without Overpaying?

Here is where most disruption investors blow up. They correctly identify the trend, they correctly pick the winner, and then they pay such an insane price for it that they still lose money. Or at best, they earn mediocre returns on what should have been a generational investment.

The dot-com bubble is the perfect classroom example. Amazon was a genuine winner of the internet disruption. If you bought Amazon stock at the peak in 1999, you paid about $107 per share. The stock then crashed 93% to under $6. It took until 2009 – a full decade – to get back to your purchase price. You were right about the company, right about the disruption, and you still spent ten years underwater because you overpaid.

Now look at AI. In 2025, some AI-adjacent companies trade at 40, 60, or 100 times earnings. The excitement is real. The technology is real. But math is also real. When you pay 100x earnings, you need the company to grow earnings dramatically for many years just to justify the current price, let alone to make a good return. If anything goes wrong – a competitor catches up, regulation tightens, growth slows from spectacular to merely excellent – the stock gets cut in half.

So how do you participate in disruption without the valuation risk?

Look for the boring beneficiaries. The most exciting companies in a disruption wave get the most attention and the highest valuations. The boring companies that quietly benefit often get overlooked. AI needs enormous amounts of electricity. Utility companies that power data centers are disruption beneficiaries trading at reasonable multiples. AI needs cooling systems for data centers. HVAC and industrial cooling companies are beneficiaries. AI needs semiconductor manufacturing equipment. These companies are essential but not glamorous enough to attract bubble-level pricing.

Wait for the inevitable correction. Every disruption wave has a hype cycle. Prices overshoot, reality fails to match expectations on the timeline investors expected, stocks crash, and then the real long-term growth begins at reasonable prices. You do not need to catch the first wave. The patient investor who buys the real winners after the hype fades almost always does better than the one who bought at the peak.

Focus on companies with existing moats that AI makes wider. Instead of betting on pure-play AI companies that may or may not exist in five years, look for established businesses with proven competitive advantages that AI amplifies. A company with twenty years of proprietary data, a huge installed customer base, and strong unit economics that adds AI capabilities is a much safer bet than an AI startup with amazing technology but no moat, no customers, and no clear path to profitability.

Demand financial evidence, not just narrative. Disruption stories are intoxicating. But until a company shows actual revenue growth, expanding margins, and free cash flow generation from AI, you are buying a story. Stories are worth something, but not 100x earnings. Look for companies where the disruption benefit is already showing up in the financial statements, not just in the investor presentation.

Key Takeaways

• The disruptor is not always the best investment. The infrastructure, tooling, and ecosystem around a disruption often generate better risk-adjusted returns than the disruption itself. Sell pickaxes in a gold rush.
• Ask one question about every company: does this disruption make their existing moat wider or narrower? If wider, the company is a potential winner. If narrower, run away regardless of how cheap the stock looks.
• Incumbents die from incentive misalignment, not ignorance. They see the disruption coming but cannot bring themselves to cannibalize profitable businesses. Look for the rare incumbents whose management is willing to take short-term pain for long-term survival.
• Valuation still matters, even when the disruption is real. Being right about the technology and wrong about the price is still being wrong. Wait for reasonable entry points or find the boring, overlooked beneficiaries.
• Financial statements do not lie. Until AI revenue, margin expansion, or cost reduction actually appears in the numbers, you are paying for a narrative. Narratives can change overnight. Cash flow cannot be faked.

Technology disruption is not new. It has been reshuffling winners and losers since the steam engine. What is new each time is the specific technology and the specific companies involved. But the playbook for identifying winners – focus on moats, watch the incentives, demand financial proof, refuse to overpay – that playbook has not changed in a hundred years. And I do not expect AI to disrupt it any time soon.