Something strange happened yesterday. While perusing my morning reading, I stumbled across this headline: “Microsoft researchers say their newest deep learning system beats humans — and Google.”

For the first time in years, I believed it.

That is, I believed that Microsoft could actually do something better than Google. That it was not only relevant again, but a leader. That it was a company I could credibly claim enterprises should take seriously.

Bizarre, right?

The long road to relevance

Yes, I still snicker at those lugging around Windows laptops. And yes, I still don’t understand why anyone would bother to use a Windows smartphone, though they’re actually pretty good.

But when it comes to hardcore, enterprise infrastructure, Microsoft seems like a smart bet to me, and that’s not something I’ve been able to say out loud for many years.

This is Microsoft, after all — the company that scorned Linux, even as the open-source operating system emerged as the default standard for high performance computing, cloud computing, and every other significant enterprise infrastructure trend.

And this is Microsoft, the company that clung tenaciously to its staid desktop monopoly, while Apple and Google up-ended the market and pushed the world to mobile. Microsoft’s response was to try to sue Android into oblivion, even as it pushed a paltry excuse of a mobile ecosystem on the world.


Leading again

But now, Microsoft claims that it has built machines that can recognize and categorize images more accurately than humans can, delivering 26% better results than Google. That’s serious data science that few companies can hope to match.

Microsoft also has a cloud platform that is closing the gap on the leading Amazon Web Services (AWS), jumping three percentage points to 10% market share in 2014, compared to AWS’ 27% (and Google’s 5%). It’s a remarkable turnaround, powered in part by CIOs’ lingering affection for the enterprise giant.

Some of this growth is driven by Microsoft learning new tricks. This is a Microsoft that once fought open source and now embraces it. One big sign of this shift is that the Redmond giant (finally) open sourced .Net Core and more recently the execution engine for it.

Some chide this new Microsoft as much ado about practicality, not principle:

“In the final analysis, embracing open source has become the only option for Microsoft in the face of so many smaller, dynamic competitors. For sure, this is less Microsoft taking on board the principles of open development, sharing, and co-operation and more an attempt to buy its way into what has always been the dynamic world of the bazaar in order to harness the talents of developers outside the cathedral walls.”

Well, yes. But this has long been the reason the vast majority of companies — and frankly, individuals — get involved in open source. Very few can afford to save the world if it doesn’t also help them pay the rent.

A Microsoft we can believe in

Some credit Microsoft CEO Satya Nadella for all this change — and surely, much of the credit must go to him for making it safe to innovate in Redmond again. But former CEO Steve Ballmer deserves as much credit, if not more, for kicking off many of the projects that are now flowering under Nadella, as ZDNet’s Mary Branscombe writes.

Regardless of who gets credit, though, the reality is that Microsoft is credible again.

That’s good for Microsoft employees and shareholders, of course, but it’s also good for the industry. We need more than one company delivering a great cloud experience. We need more than a duopoly in mobile. We need competition again, and a revitalized Microsoft, oddly enough, is fantastic for competition.

These are words I never thought I’d write when battling Microsoft in the open-source trenches. But I believe them, and I’m grateful for the new Microsoft. You should be, too.