Microsoft CEO Satya Nadella has criticised the closed nature of mobile app stores, and said the burgeoning field of artificial intelligence should be an open ecosystem.
His comments are a warning shot to Apple, which pioneered a "walled garden" model for iOS and the App Store, shutting out competitors. The model is hugely successful, with Apple's App Store revenues estimated at around $28 billion (£21 billion) last year, while Microsoft is on its last gasp in mobile, with 1% global market share.
Nadella, speaking to reporters in London on Friday, pointed to Microsoft's partnership with Amazon as an example of an open partnership. The deal means Microsoft's Cortana assistant can talk to Amazon's Alexa.
"The fact is, we have great capability and I want to make sure it shows up everywhere people want to interact with it. If they have an Echo, they want to invoke Cortana on their Echo, we want to open it up. It's the same philosophy that led us to build our apps so they were available on iOS and Android. So I view the world as: What are the ways we can meet more of our customers all the time, and [what is] the unique value we can add. So that's how it'll play out."
"When I look at PC ecosystem, it's a very open ecosystem. This entire App Store and all the things around it is a very closed system. So the more we can create interoperability at least ... just imagine if the web was as closed as these app stores. it would be a real tragedy. I hope in the AI world we have an open ecosystem, and we will push for that."
Of course, Microsoft was once accused of creating a walled garden with the app store for Windows 10, its combined desktop and mobile OS. The cofounder of Epic Games, Tim Sweeney, accused Microsoft last year of "locking down the consumer PC ecosystem and monopolising app distribution and commerce" with its Universal Windows Platform, which enables developers to build apps that work on both mobile and desktop. Microsoft defended UWP as an "open ecosystem" at the time.
Microsoft's disastrous AI experiment Tay is coming back
Microsoft's efforts in AI haven't always gone to plan.
The company launched an English-speaking Twitter AI chatbot, Tay, in March last year — but the bot quickly started spewing racist, sexist tweets.
It was a huge public embarrassment for Microsoft, and it eventually pulled Tay from Twitter. A tamer successor, Zo, didn't voice racist bile, but did go off-script and describe Windows as "spyware."
Here's Nadella's view of the incident. He also revealed Tay is returning:
"On the Twitter channel [Tay] was attacked, and therefore what happened was it started learning from the attacks and spewing out comments which were not acceptable. So one of the things it has really influenced is our design principles ... we have to take accountability. First and foremost, we need to be able to foresee these attacks which, interestingly enough, were attacks by humans. But the idea that we need to keep the broader goal of having this AI behave properly is our accountability, and so how can we test it, how can we make sure it does not lose control. Now we're launching it again, it's in preview in a couple of channels. So we're being mch more deliberate in that process."