Game Over

Finding Value as AI Dismantles the Software Industry

March 7, 2026

Prior to Lyft and Uber, a New York City taxi medallion cost just over $1 million. Today, you can buy one for $90,000.  

The demand for taxi service hasn’t changed. People still need ways to get around large metropolitan areas; they just prefer Lyft or Uber over taxis. People that need rides around cities, it turns out, don’t fancy standing curbside begging for a ride with their arm in the air. 

Neither Lyft nor Uber would exist without the internet. The internet connects the person who needs a ride with the person who has the car. The internet enabled services like Uber and Lyft. That gave them access to the ride-hailing business, which destroyed the economics of taxis. Game over for the value of taxi medallions.

The internet dismantled one industry after another two decades ago. People not only prefer Ubers over taxis, they also prefer digital news. That dismantled the print newspaper business. Many prefer online shopping, which changed the economics of bricks-and-mortar retail. The internet shifted billions of dollars of business value from old-school companies with dated business models to new entrants with internet-enabled business models. Sears went bankrupt. Amazon stock compounded at 30% for 25 years.

But wait. If the internet destroyed the economics of bricks-and-mortar retail, why did LVMH just pay $16 billion for Tiffany’s? Tiffany’s not only survived the onslaught of the internet but thrived. 

Tiffany’s isn’t the lone exception. Target generates $2 billion of free cash flow annually today compared to just $200 million pre-internet. Walmart generates about $15 billion in free cash flow today, up from just $1 billion pre-internet. Dollar General generates $1.5 billion in free cash flow annually today, up from $200 million pre-internet. These four retailers collectively serve customers of all socio-economic classes. Did the internet dismantle them? Hardly.

The public equity markets sometimes paint with a very broad brush. Any hint of a disruptive technology triggers the sale of all companies in the industry being disrupted even though some companies in the industry being disrupted will thrive in the new environment. 

The disruptive technology today is artificial intelligence. The industry being dismantled is the software industry. The public equity markets are again painting with a very broad brush. Software companies of all types have been clobbered over the last 6 months. We smell opportunity.

As investors, the key to capturing those opportunities is to somehow differentiate between the companies that will die in the new environment and those that will thrive. Sears got smoked by the internet. Tiffany’s thrived. There will be plenty of Sears analogies in the software industry when the dust settles. No question about that. We also believe a few Tiffany’s analogies will emerge. 

We believe three conditions must exist for a company to thrive in the face of a disrupting technology.   

First, companies that thrive in the face of a disruptive technology are the ones that embrace the new technology. This is obvious, yet critical. If the technology is truly disruptive, companies must use the new technology to their advantage. They must embrace it.

Walmart was slow getting the message, but when the alarm rang in Bentonville, Walmart wasted no time embracing the internet. Today nearly 20% of its sales are online. That represents a $700 billion revenue stream growing 25% annually. Walmart complemented its bricks-and-mortar business with the internet, a powerful combination that most e-commerce companies cannot duplicate. Target and Tiffany’s likewise complement their bricks-and-mortar sales with robust online sales.

Taxis sadly never embraced the new internet technology. We believe traditional taxis once had a massive competitive advantage over Lyft and Uber in the form of scale. Lyft and Uber started from scratch. Taxis had scale from the beginning. To leverage its scale advantage, taxis needed to embrace the new technology. They didn’t. They got smoked.

Kodak pioneered digital photography. It had an early leg up in using the new technology to its advantage, all they had to do was embrace it. They didn’t. They got smoked.

Incumbent companies whose industries are being disrupted by new technologies oddly enough have internal barriers that prevent them from implementing the very technology that is revolutionizing their industry. Those barriers tend to be related to the culture of the company or the nature of their legacy businesses. Either way, the incumbents that want to thrive in the midst of a disruptive technology must proactively remove those internal barriers. 

Second, companies that thrive in the face of a new technology must have strong personal bonds with their customers, bonds that are sticky and durable. Technologies become destructive when they enable new entrants to offer new and unique value propositions to the customers of the incumbent companies. That’s what makes them destructive. The customer will stick with the incumbent only if the incumbent’s value proposition is greater than the one being offered by the new entrant. No value proposition no customer. 

Think of the four thriving retailers just described: Tiffany’s, Target, Walmart, and Dollar General. While nobody shops at all four, the people that shop at each enjoy a tight and enduring bond with the company. Customers love these retailers and the retailers are laser-focused on the customers. These companies thrived when others died because of their strong enduring bonds in the form of their value proposition for the customer.

The New York taxi system had no such bond. New York taxis were a necessary evil in getting from Point A to Point B, nothing more. The lack of customer bonds made taxis vulnerable. There was nothing preventing customers from fleeing to new entrants with better value propositions. Taxis were easy prey. 

When a technology disrupts an industry, customers will stay with an incumbent if the incumbent offers a better value proposition than the new entrant. Strong durable customer bonds are a built-in advantage for incumbents. 

Third, in order for an incumbent to embrace and utilize a new technology, that new technology must be available to the incumbent at a reasonable price. The internet, while disruptive, was available to anyone that wanted to use it. Uber didn’t have a monopoly on the internet. Neither did Amazon. They used it as an enabler for their business models. The internet wasn’t a scarce resource. It was available to all, incumbents and new entrants alike. This is a necessary precondition for incumbents to thrive in the face of a new disruptive technology. The incumbent must have access to the new technology at a reasonable cost.

If AI evolves into a monopoly product, the entire software industry is vulnerable. We don’t think it will. We believe AI will be like the internet: broadly available to all at a reasonable price. We see a highly competitive AI market at the moment, one with the incentives to deliver a quality AI product at a reasonable price. That is good news for consumers of AI, including the software industry.

What types of software companies will thrive in an AI-powered world? We look for three characteristics: 1) companies with access to AI at a reasonable price, 2) companies that proactively destroy implementation barriers and embrace the new technology, and 3) companies that have strong lasting bonds with their customers. 

Companies that have access to AI technology, destroy implementation barriers, and have strong durable bonds with their customers will survive and thrive. Companies that lack access to AI or do not destroy implementation barriers or do not have strong durable bonds will die. Two out of three isn’t good enough. Those that survive and thrive will have all three. 

More Articles