In any introductory class to game theory, the prisoner’s dilemma is one of the thought experiments discussed. This game involves two rational agents, each of whom can cooperate for mutual benefit or defect for individual gain. Defection is rational, but cooperation yields a higher payoff. If we only play one game, rationality prevails, and we defect. It turns out that if we play a fixed number of games, logic says that we would defect on the last game, which then, by induction, we would defect on the game before the last game, and therefore, every game. There is one clear strategy. These are examples of finite games.
By changing one variable, if the number of games to be played is unknown or infinite, we now have an infinite game. Various strategies are mathematically provable as Nash equilibriums, where no player can gain by changing their approach. These include cooperating every time, defecting every time, and cooperating until you observe a defection and then defect from there. This revelation is a reminder not to apply finite game approaches to infinite games.
Since joining Sequoia, I’ve heard many examples of using finite game thinking on infinite games. They usually involve statements such as Company A will crush Company B because Company A has more capital and resources, a better product, a much higher valuation, or the market leader position, and the game is over. The statement shows finite game thinking. These advantages are not just temporary, they are fleeting. If that’s all it takes, first movers would win every time, and venture capitalists would be out of a job.
Pundits have made proclamations, such as Microsoft will kill Apple, Intel will kill Nvidia, Google will kill Microsoft, etc., but all of these companies are larger than they were many decades ago and have switched leadership positions multiple times.
By 1995, the internet reached critical mass due to the invention of the browser. Netscape Navigator was the most widely used web browser, and Netscape became a public company on August 9, 1995. Microsoft assembled an elite team and licensed Mosaic to create Internet Explorer 1.0, released with Microsoft Windows 95 Plus! on August 24, 1995. Over the next few years, the browser wars ensued, and Internet Explorer became the new dominant browser, attaining a peak of 96% browser share in 2001. This was a significant win, but the bundling strategy employed by Microsoft resulted in an antitrust trial and settlement that would slow the company down for over a decade.
With this opening, Google would emerge to win internet search and the second browser wars. Google enjoyed high growth, high profitability, and unlimited resources to invest in its future. Yet, they couldn’t win in social (Meta, X, Snap), web services (Amazon), or mobile (Apple). Despite four decades of chip dominance by Intel, with transformer architecture and large-language model technology, NVDA found a new revival after decades of irrelevance.
Peter Thiel famously stated, “competition is for losers” in his book Zero to One. The reason is simple: perfect competition leads to zero economic profits. We should focus on our customers, innovate with unique insights, and differentiate from what exists. Create a category where we can set prices, maintain high margins, and focus on long-term innovation, which will be a winning approach for some time (which could be a short or a long time). However, with more and more success and when the prize is enormous, competition will eventually follow. Amazon successfully went after commercial search despite Google’s search dominance. With the success of AWS, Google and Microsoft reconsidered web services as a low-margin business. So, while competition is for losers, in business, competition is inevitable.
Rather than being an inevitable loser, we should remember that business is an infinite game, and we shouldn’t use finite game approaches. As with the infinite prisoner’s dilemma, a combination of competition and collaboration could lead to the optimal strategy in business. The goal would be to create larger profit pools for all participants by focusing on where we are or want to become uniquely strong and compete in those core areas. In non-core areas, we should collaborate.
When we return from Thanksgiving Break and start laying the foundation for our 2025 plans, remember that we play a combination of finite and infinite games on any given day. Apply the appropriate tools to each.