The State of Microsoft's AI

SEATTLE—Major tech firms—including Amazon, Apple, Google, IBM, and Microsoft—are all making big pushes into artificial intelligence these days. At the Build conference here, I spoke with Dave Forstram, Microsoft'south Senior Director of Communications for AI, for insight into what Redmond thinks sets it apart from the competition and where the company sees the field going over the side by side few years.

Two Teams

Microsoft'due south AI efforts fall into two different teams: the research division, headed by Harry Shum, and the Azure AI services for developers, headed upward past Scott Guthrie.

The first team'south work shows upwardly in things similar Bing, Cortana, ambient computing, and robotics. The Azure side makes AI services bachelor to developers, letting them build apps using customizable machine learning with speech, language, vision, and knowledge services. Tools offered by the partitioning include a Cognitive Services and Bot Framework, and deep-learning tools like Azure Machine Learning, Visual Studio Code Tools for AI, and Cognitive Toolkit.

AI Hardware

Microsoft is also investing in AI-powered hardware. At Build 2022, Microsoft announced Project Brainwave, which uses field programmable gate arrays (FPGA) to perform AI calculations. The setup offers latency 5 times lower than Google's TPU hardware, Microsoft CEO Satya Nadella said during his Build keynote. Brainwave hardware is also reconfigurable compared with TPU'south set blueprint. Brainwave even runs Google's Tensorflow AI code as well as Microsoft's Cognitive Toolkit and Facebook's Caffe2.

Brainwave Harward Card

AI on the Edge

Another key to Microsoft's strategy is edge AI—where the ciphering can happen on a local device rather than in the cloud. There are plenty of industries that tin can't afford to take everything terminate if cloud connectivity is interrupted. Azure IoT Edge offers a way to bring AI and auto learning away from the cloud and to places like the factory floor. Equally with a lot of Microsoft'southward AI tools, IoT Edge is open up source and available on GitHub.

Where Is AI, and Where Is it Going?

Forstram likened AI development to human development. "The starting time affair a child does is it sees the earth. Next it has to acquire speech. And the side by side thing is comprehension. AI tech has come up pretty far in categorizing the visual world, simply still need to brand more sense of it," he said.

This represents a whole level beyond the utilitarian, brand-focused chatbots, which are currently far ahead of personal assistants like Cortana and Alexa in terms of conversational AI. XiaoIce, a chatbot Microsoft launched in China, "has more than than 200 million users, has engaged in 30 billion conversations, and has an average conversation length of 23 turns, which averages out to about one-half an hour," Forstram said.

Smart home voice assistants are but now adding the power to support back-and-along conversations; Google appear it for the Assistant at I/O this week, while Amazon rolled out a similar feature for Alexa, dubbed Follow-Up Mode, in March. At this bespeak, Cortana can only ask you followup questions.

With XiaoIce (besides as Japan-based Rinna and the US-based Zo), "information technology's a completely dissimilar way of thinking most conversational AI. It simply interacts; it'due south natural and personal."

Azure AI

Forstram noted that XaioIce is working toward achieving human parity at translation from Chinese to English language; it's also considered "full-duplex," meaning you tin can interrupt and XaioIce can anticipate what yous're going to say. Information technology's so realistic that she'due south even hosted TV shows and written an acclaimed book of poetry, based on machine learning of centuries of verse.

"When I hear people talking near the AI vision, that's the closest I've seen," said Forstram.

This caste of language comprehension accuracy has business implications, too. Customer intendance, in particular, benefits from a bot that has some emotional intelligence and even empathy—"that'south the holy grail," Forstram said. It could even encourage people to engage with bots rather than furiously pressing 0 or screaming "REPRESENTATIVE!!" into their phones get a human on the line.

So how does Microsoft's AI differ from those of Amazon and Google? "Everybody started with the aforementioned building blocks," said Forstram, referring to machine learning, deep learning, neural networks and the like. "But nosotros use different algorithms, even though we often get similar output."

Google, Amazon, and IBM are edifice out cerebral services, just Microsoft is the first to let customers customize them. "In the past, if someone wanted to utilize our speech-recognition technology, we would simply give them libraries that they'd have to train their information to use. Google is just getting to that."

Finally, Forstram sees the Microsoft Graph equally one of the company'southward biggest differentiators in AI. If knowledge is broken downwardly into globe, work, and user data, work noesis piece is a distinguishing forcefulness of Microsoft, with Part, Windows, and LinkedIn information to draw on.

Well-nigh Michael Muchmore

Michael Muchmore

Michael Muchmore is PCMag's lead annotator for software and web applications, with an emphasis on photo editing, video editing, and Windows. A native New Yorker, he has at various times headed up coverage of web development, enterprise software (including databases and application servers), and display technologies (monitors and TVs). Michael cowrote 1 of the get-go overviews of spider web services for a full general audition. Before that he worked on PC Magazine's Solutions department, which educated readers nigh programming techniques like C+ and Visual Basic, as well as offering tips on using office productivity software. He previously covered services and software for ExtremeTech.com.