Microsoft today announced the launch of the Azure OpenAI Service, which, as the name implies, makes OpenAI’s machine learning models available on the Azure platform. Specifically, that means GPT-3, OpenAI’s groundbreaking language model that can, under the right circumstances, produce human-like text with just a few prompts.
There is a catch, though. At least for now, not all Azure users will get access to it (even if they are willing to pay). Access will be invitation-only and for “customers who are planning to implement well-defined use cases that incorporate responsible principles and strategies for using the AI technology.” Microsoft will offer safety monitoring and analysis to find cases of abuse or misuse of GPT-3 and it will offer filters to make sure your GPT-3-based chatbot doesn’t start swearing at your executives (even if GPT-3 thinks they deserve it).
It’s worth noting that OpenAI itself already made a GPT-3 API available last year, though there is still a waitlist. Microsoft, too, is already using GPT-3 to power GitHub’s Copilot tool that helps developers write code for them. But while there are already ways to access GPT-3 outside of Azure, Microsoft argues that it can offer “additional layers of security, access management, private networking, data handling protections or scaling capacity.”
Microsoft invested $1 billion in OpenAI back in 2019 and licenses GPT-3, so it’s no surprise that the company is trying to bring it to a wider range of products now.
“GPT-3 has really proven itself as the first powerful, general-purpose model for natural language — it’s one model you can use for all these things, which developers love because you can try things very easily,” OpenAI CEO Sam Altman said. “For a while now, we’ve wanted to figure out a way to scale it as broadly as possible, which is part of the thing that really excites us about the partnership with Microsoft.”