Major computing shifts: Will history rhyme?

I experienced the mobile computing shift in the trenches as a PM of Samsung’s mobile processors and later in the C-Suite as a consultant. Reflecting on those days, I see distinct similarities between the nascent stages of mobile technology and the current, early days of artificial intelligence. Some observations:

Operating System Wars vs. Foundational Model Wars

In 2007, the mobile OS war was in full swing and it was not at all clear who would win. Apple had the hype factor, while Windows Mobile seemed to be a sure thing due to its familiarity and domination of the Office productivity suite. Nokia, with its Symbian OS, had the largest market share, and Blackberry had the early cult following. In a few short years, however, the mobile OS space quickly became a duopoly of iOS (premium & closed) and Android (free! and open source). In the early days of mobile, the momentum of app developers created insurmountable moats. Today the moats in foundation models are created by CapEx and access to data.

I don’t necessarily think the AI foundational model wars will rationalize into a duopoly, but I do think that 1) a smaller number of players will emerge as the winners, 2) open source will have an important role and will force closed players to refine their strategies, and 3) developers will have to pay close attention to which models emerge victorious. Looking back to the mobile OS wars, startups that tied their fate to Windows Mobile or Symbian are probably not around today.

The entrepreneurial learning curve

Apple was not early to the smartphone game. Palm, Blackberry, and others had mobile devices in consumers’ hands. The proliferation of smartphones, which Apple masterfully accelerated, was enabled by inflections of existing technologies such as fast, cheap, and low power processors, fast mobile networks, software assisted GPS, capacitive multi-touch, and more. Following the launch of the iPhone in 2007, it took entrepreneurs a few years, and lots of experimentation in leveraging these technologies before companies like Uber (2009), Whatsapp (2009), and Instagram (2010) emerged. Before that, there were countless startups that saw early experimental adoption and then quickly fizzled out (e.g., Flashlight apps, iBeer, Bump).

Fast forward 17 years. It has now been 18 months since the public release of ChatGPT sparked the frenzy of today’s AI boom. The rate of model development and the corresponding boom in AI entrepreneurial activity have been dizzying. But much like the 2+ years that it took entrepreneurs to create truly iconic businesses in the mobile world, I believe that entrepreneurs are emerging from the experimentation phase and today are building AI native companies that will fundamentally change the way we do things for many years to come.