Probably the smartest choice Bill Gates ever made came in 1980, when he decided not to hand over the copyright for Microsoft's first-ever operating system to IBM.
In 1980, IBM contracted a startup called Microsoft to deliver DOS, an operating system for its forthcoming IBM PC, on a tight deadline.
The IBM PC came out in 1981, and soon became a smash hit, surpassing the leading Apple II. A horde of competitors rushed to build their own "IBM Compatible" clones that could run all of the same software and use all of the same hardware upgrades.
But to build those IBM clones, they needed DOS. And if they wanted DOS, they needed to fork over cash to Microsoft. Microsoft kept the rights in lieu of royalties from IBM. DOS put Microsoft the very center of the PC revolution, even through the era of Windows, and even after IBM left the PC market, eventually selling off that business.
36 years later, it's been a long time since the IBM PC moment. And with the Apple iPhone and Google Android ruling the all-important mobile market, Microsoft missed its shot at the mobile operating system revolution.
That's why Microsoft, which keeps boasting about how much it loves selling cloud services and subscriptions, is suddenly investing so much in hardware like the HoloLens and the Surface. If no new IBM PC will come along like in 1981, Microsoft will just have to build it itself.
Jumpstarting the future
The iPhone and Android have a stranglehold on the mobile market. Apple has ridden the iPhone to becoming the most valuable company in the world, while Google's Android is now the most powerful operating system in the world.
Microsoft missed that boat. And Microsoft, going forward, has to decide if it wants to keep throwing good money after bad into its struggling Windows phone business while it tries to force the next big thing to happen.
Microsoft has decided to build the devices it wants to see in the world. And with PC sales shrinking, Microsoft is looking to more science-fictional concepts.
The tone was set in 2012, when Microsoft launched the Surface, its first tablet. That was followed up by the Surface Pro laptop/tablet hybrid, and eventually, the Surface Book, Microsoft's first full-fledged laptop.
Looking to the future, the Microsoft HoloLens looks to be the first commercially-available augmented reality headset, layering holograms on top of your field of vision.
The important thing that Surface and HoloLens have in common is that they're already sparking competitors to build their own clones. Samsung, HP, and lots more are building Surface-alikes, while Asus and perhaps Intel are working on HoloLens clones.
And in all cases, those cloned devices are running the Microsoft Windows 10 operating system.
So even if Microsoft isn't selling the hardware, it's increasing the reach of Windows 10 with every cloned device sold. HoloLens has the potential to set the standard for holographic computing, but it won't necessarily be the most popular holographic computer.
Margin of error
The problem is that building complex machinery is a low-margin business, across all industries.
Building a computer (or a hologram headset, or a car) is labor-intensive, requires a lot of specialized parts, and takes time to make each and every unit. Dell's margins hover around 3%; Ford's are around 7%.
Meanwhile, one of Microsoft's big advantages has always been that software is a much higher-margin business than hardware. In 1999, right at the height of its powers, Microsoft's operating margins were 51.7%.
Microsoft's smart move was to make profitable software, and let companies like IBM, Dell, HP, and Compaq build their low-margin, "IBM Compatible" PCs. After all, they all still needed buckets of pricey Windows licenses, no matter what they charged for their computers.
Therefore, it seems unlikely that Microsoft really sees all of this business of making hardware as something truly critical to its future, compared with its higher-margin Windows and Office businesses.
The Surface is popular, and even its Xbox One video game console is on the upswing in sales, but Microsoft has clearly signaled that in all corners of its business, the future is in Windows 10 and cloud services.
So while it's unlikely that Microsoft will stop selling these devices any time soon — at least, so long as they keep making money — it's equally obvious that Microsoft is mostly interested in using them to drive the future.
There's plenty of competition, especially in the holographic world. Magic Leap, for instance, is a well-funded startup with some truly impressive technology that has nothing to do with Microsoft or Windows 10.
But if Microsoft succeeds, it'll be the center of a new revolution, and the eighties will truly be back.