Disclosure: Microsoft is a client of the author.
Microsoft’s decision to move away from mixed reality (where it was arguably the leader and market-maker for augmented reality with HoloLens) was a big surprise. And its massive investment in OpenAI and ChatGPT was equally surprising, given the company’s advanced internal AI efforts.
What we’re seeing, I suspect, is a company girding for war — and working to make sure what happened to it when the internet, iPhone, and Google arrived doesn’t repeat with the arrival of generative AI.
While Microsoft may not succeed, it won’t be for lack of trying. What will be fascinating is whether getting an early, aggressive jump on this technology will turn out better than its past efforts to catch a wave from behind. Its recent bold moves remind me of approach it took with the internet, when Microsoft took the browser market so aggressively from Netscape that Netscape didn’t survive.
Here’s how the coming generative AI war is likely to affect Microsoft.
A mixed reality failure?
In many ways, Microsoft approached mixed reality (MR) the right way, at least at first. It developed HoloLens as an industrial-level product that, while expensive, gained a foothold with customers like the Lawrence Livermore lab. The effort was marginally profitable, with sales successes in aerospace, microprocessor FABS, general manufacturing and even the military (though the military trial ran into serious problems).
Microsoft approached virtual reality (VR) differently; it had more of a consumer focus, but with technology that didn’t meet the minimum bar. Facebook’s attempts were arguably better, but Facebook poured more down than the money hole than Microsoft did. So, while VR was arguably a failure for Microsoft, it could have been much worse.
For a business to make sense to Microsoft, it needs to see sales in the millions. With AR, the company never seemed to get to those numbers. The issues included a heavy headset, a lack of occlusion, ghost-like virtual objects, poor field of view, less-than-stellar battery life, the lack of accurate hand indexing, and unfocused applications. Even so, it was arguably the best AR solution on the market. (Things might have turned out different if Microsoft had spent more on marketing to get a critical mass in sales.)
Remember, Xbox was initially a huge money loser, too. But Microsoft fought to make it successful. So why not ride out the AR and VR troubles, as well?
Simple: generative AI scared the hell out of the company.
Windows and MacOS — the GUI problem
Back when Microsoft was just getting started, one major technology change helped ensure the company’s success. Computers started moving away from command lines and embraced Xerox’s concept for a Graphical User Interface (GUI). Apple was convinced GUI was the future, and when the market started to pivot away from the command line, it scared the hell out of a lot of big players. IBM licensed DOS, Microsoft wrapped it with the Windows GUI, and the OS wars began.
After a few years, Microsoft prevailed with Windows 95; it was both an example of marketing excellence and a warning that the industry — particularly Microsoft — didn’t yet know how to handle the service requirements of an OS decoupled from hardware.
Microsoft came out of that effort significantly stronger, but did so by buying, not building, DOS. Microsoft moved early, but it was still chasing Apple.
The web browser fight
When Netscape launched and the internet became a reality, Microsoft was caught napping. It pivoted hard by buying a browser to compete, pivoted its MSN efforts away from CompuServe and AOL, and again successfully weathered the storm to become dominant.
It didn’t help that Netscape’s management made the same kind of errors Microsoft would later on with the Zune later and tried to chase Microsoft’s market dominance. Netscape failed. If Microsoft hadn’t pivoted fast and effectively, Netscape might not have made that mistake, and Microsoft would either be gone now or be far weaker.
Compared to obvious failures like Zune, the Microsoft Phone, Plays-For-Sure, and others what made the difference was seeing a threat early, responding to it quickly, and adequately funding the effort.
And now, generative AI
Generative AI has the potential to reshuffle the tech landscape because, like the GUI OS wars, it promises to change how we interact with computers, and, like the browser wars, it promises to change how we interact with remote services. In fact, it may well be more disruptive as it matures than either the OS and browser wars combined.
If Microsoft wants to come out ahead now, it needs to acquire a leading technology and then focus company resources on making it a defining competitive advantage across the Microsoft ecosystem.
The potential benefits of being able to talk to a computer on subjects ranging from office products to telesales should boost productivity for the former and be cost effective with higher close rates for the latter. Both are only a taste of the disruptive potential for generative AI technology.
As it moves to market, generative AI could well launch the next Google — and kill any tech firm that doesn’t adapt to what’s coming. Microsoft wants a shot at the former and hopes to avoid the latter. That’s why it’s shifting away from marginal, long-term opportunities toward generative AI.
A company and its CEO are often defined by the choices they make. Though he was amazing with operations (and brilliant as an individual), former Microsoft CEO Steve Ballmer will forever be defined by the failed Yahoo merger (a response to Google), the Nokia acquisition failure (in response to Apple) and the Zune.
Current CEO Satya Nadella has so far been defined by the massive success of Azure, but this is the first time he’s faced a major threat that requires a corporate pivot. Among the risks: ChatGPT might not be the leading technology; Nadella might underfund the shift; or, as with mixed reality, he may have moved too soon.
That last issue seems unlikely — it looks as if funding is adequate and, given the nature of generative AI and its gathering popularity, the market-making requirement is reduced.
We won’t know for a while whether Nadella’s embrace of ChatGPT is the right move, but it looks like a reasonable bet.
In the end, we are watching a company pivot to war before that fight really kicks off, and regardless of how it all turns out, Microsoft should be better for it. But the outcome is anything but certain.