Interviews with key Microsoft staff shed light on its AI strategy, utilizing Cortana and chatbots, and the firm's plans to compete with Facebook and Google
Yoshua Bengio has never been one to take sides. As one of the three intellects who shaped the deep learning that now dominates … Tweets: @msftresearch , @harryshum , @nigewillson , @wired , and @mfordfuture . Thanks: @reckel Tweets: @msftresearch : In #AI, how Microsoft won't just “realize the spoils of #science, but push the field forward.” @jessiwrites @wired https://www.wired.com/... pic.twitter.com/dqQ4y8ivuC @harryshum : .@jessiwrites Exactly — A company with the resources, data, talent, vision & culture to push the field forward. https://www.wired.com/... pic.twitter.com/Xi2XVVMwr5 Nige Willson / @nigewillson : Comeback? Did we go somewhere Inside Microsoft's Artificial Intelligence Comeback http://bit.ly/2sBDrp8 #ai #ITRTGpic.twitter.com/a91YJzvhCJ @wired : “We don't want one or two companies, which I will not name, to be the only big players in town for AI.” http://www.wired.com/... Martin Ford / @mfordfuture : Microsoft is racing to catch up to Google and Facebook in #deeplearning . Its future depends on it. #AI http://www.wired.com/... via @WIRED Thanks: @reckel
Context & Ripple Effects
This 2017 Wired piece sits early in a decade-long arc of Microsoft explaining its AI ambitions in public. It follows [[a:871255|Satya Nadella and other executives laying out the company's AI plans and claimed competitive advantage in 2016]], and it lands at the moment when Cortana and chatbots were Microsoft's visible consumer AI products, with Yoshua Bengio framing the firm as an heir to the deep-learning research tradition rather than just a commercializer of it.
The interviews matter because they show Microsoft defining itself against Facebook and Google while insisting it will 'push the field forward,' not merely harvest the spoils of science — a posture Harry Shum amplified on Twitter by citing the company's resources, data, talent, vision, and culture. That framing is the throughline to everything the related coverage documents afterward.
First-order effects
- Microsoft's named AI leadership — Harry Shum's research organization and product teams behind Cortana and chatbots — commits publicly to competing head-on with Facebook and Google for both talent and consumer mindshare, making research credibility part of the pitch.
- The Bengio framing positions Microsoft Research as a first-tier lab alongside the deep-learning founders, raising the stakes for its hiring and publication output relative to rival labs.
Second-order effects
- Facebook and Google face a competitor arguing that data-plus-distribution beats pure research prestige, forcing the comparison onto who can ship assistants and bots into products people already use — the distribution argument Microsoft has repeated ever since.
- The consumer-assistant bet sets up the later pivot documented in the coverage: after the OpenAI board dispute, Microsoft diversified its AI investments and built its own models, showing how fragile single-partner strategies proved once the chatbot era gave way to foundation models.
Third-order effects
- The pattern across the corpus — 2016 strategy talk, 2017 Cortana-era interviews, then 2023 executive interviews on Copilots and Bing and Kevin Scott's 2025 agenda of agents, MCP, and NLWeb — suggests Microsoft's durable structure is a research arm continuously re-attached to whatever interface wins next, from assistants to copilots to agents.
- Brad Smith's regulatory playbook profiled in the coverage extends the same posture into policy: a company that frames itself as advancing the field also seeks to shape the agency-level rules around it, a template other large labs now follow.
The trend: Microsoft's AI identity has evolved from assistant-and-chatbot competitor in 2017 to a platform built on copilots and agents, with each era reusing the same claim that research depth plus product distribution is the winning combination.