OpenAI Inc., the San Francisco lab working to develop human-level artificial intelligence, has received $1 billion from Microsoft Corp. to support its mission.

The investment is part of a multiyear technology partnership the companies announced this morning. The sum matches the $1 billion in initial capital that OpenAI started out with when the lab launched in 2015. Its endowment came from a group of high-profile Silicon Valley figures, including Elon Musk, LinkedIn co-founder Reid Hoffman and former Y Combinator president Sam Altman (pictured, left), who stepped down earlier this year to become OpenAI’s chief executive.

The lab is on a quest to develop so-called artificial general intelligence. It’s a hypothetical form of advanced AI with the capacity to not only master complex tasks, such as chess or image recognition, but also learn new skills on its own. OpenAI’s endeavor falls squarely into the moonshot category: Leading AI experts’ estimates of how long it will take to make human-level neural networks a reality range from decades to more than a century.

Pursuing such a far-off vision isn’t cheap, which is where Microsoft comes into the picture. The newly announced investment will help OpenAI keep up with deep-pocketed companies such as Google LLC and Facebook Inc. that are run own cutting-edge AI research initiatives.

As part of the partnership, Microsoft will not only provide the lab with capital but also become its exclusive provider of cloud computing services. That should ensure OpenAI will have access to the massive amounts of infrastructure necessary to support large-scale AI projects.

The two companies didn’t specify the precise nature of the investment, but despite speculation by some that much of the investment might be in the form of Azure services, OpenAI Chief Technology Officer Greg Brockman said on Hacker News that it’s a “cash investment.” That said, Altman told the New York Times that most of that money will go toward computing power — and the contract calls for Microsoft eventually to become OpenAI’s only source of that.

In exchange, the lab will support Microsoft’s own machine learning initiatives. OpenAI is making the company its “preferred” partner for commercializing the AI innovations that come out of its research efforts. The lab will license some technologies to Microsoft, make others available on the Azure cloud platform and work with the company on yet-unspecified joint development projects.

OpenAI has made significant contributions to the AI ecosystem. Among the lab’s most notable creations are Gym, a popular toolkit for building machine learning models, and Dactyl, a robotic hand that that uses AI to manipulate objects with uncanny dexterity.

Microsoft is likewise helping to advance the state of AI. In June, the company open-sourced TensorWatch, a tool that simplifies the complex task of debugging machine learning models. 

Microsoft’s investment in OpenAI follows a recent change of focus on the lab’s part. Originally founded as a nonprofit, OpenAI adopted a new operating model shortly after Altman took the helm in March. The lab shifted most of its operations to a for-profit entity called OpenAI LP so that it could more easily raise capital to support development. 

“By bringing together OpenAI’s breakthrough technology with new Azure AI supercomputing technologies, our ambition is to democratize AI — while always keeping AI safety front and center — so everyone can benefit,” Microsoft CEO Satya Nadella (right) said in a prepared statement.

Though Microsoft will be OpenAI’s exclusive cloud provider, the partnership deal doesn’t mention anything about the company being its sole investor. The lab might bring in additional backers further down the road to share the steep costs of AI research. Google’s DeepMind machine learning research unit, which has a similar roadmap as OpenAI, burned through $440 million in 2017, according to regulatory filings.

Photo: Microsoft