Microsoft Research is a pretty laid-back place. It's where a lot of very smart people, almost all with PhDs, work on research projects without the pressure of a release date or productizing their idea on a specific timeline.
The problem was that they got too disconnected from the revenue stream, and product lines and great ideas went nowhere. A recent article by Bloomberg said that the researchers went too far in the other direction, and great ideas sat around in the labs for years, never being productized.
As the story goes, during a Microsoft executive retreat one month after he was elevated to the CEO position in 2014, Satya Nadella saw a research project that captured his attention. It was a speech recognition and artificial intelligence app that translated a live conversation into another language in real time.
Nadella told the team he wanted that research project to be integrated into Skype and ready to show off at his first public speech as CEO, which was three months later. The Skype team scrambled and got it working. Today you know it as Skype Translate, which does voice translations in five languages and 50 text languages.
That technology was hardly new. I attended a Microsoft Research open house in May 2010 where they showed off a bunch of technologies, one known as The Translating Telephone, which at the time did real-time translations between German and English.
Four years after I saw a demo, it was still a research project with no release plans when Satya Nadella saw it. That's not how research should work. Now, I know some of these experiments take time and often require waiting for the technology to catch up to the concept. And in the case of the translator, Microsoft didn't have Skype back then. So in this case, it was good timing.
To avoid a repeat of the translation product going nowhere, Nadella has reassigned about half of its more than 1,000 research staff to a new group called MSR NExT in late 2014. This group is focused on projects with greater impact to the company rather than pure research. The other half of Microsoft Research is getting pushed to find ways it can make more significant contributions to the company's product lines.
This is how Google and Facebook have operated for years, according to the Bloomberg article. Now, Research and products groups are in regular communication. Still, it makes me wonder how many more of the demos I saw in 2010 are still sitting around at Research with no direction or product plans. Hopefully that will change soon.