Microsoft Corp. on Monday announced the general availability of its Azure OpenAI Service, which provides access to artificial intelligence models developed by OpenAI LLC.
The service is one of several Microsoft offerings that incorporate OpenAI technology. In the future, more products are expected to join the list. Microsoft Chief Executive Officer Satya Nadella told the Wall Street Journal today that “every product of Microsoft will have some of the same AI capabilities” as Azure in the future.
The Azure OpenAI Service provides access to a managed version of OpenAI’s GPT-3 model. The model, which originally debuted in 2020, can generate text, produce software code and perform other tasks in response to natural language prompts. Azure OpenAI Service customers have access to four different versions of GPT-3 that vary in performance, speed and cost.
The service also includes Codex, a neural network based on GPT-3 that is optimized for generating software code. It powers the Copilot programming assistant that Microsoft’s GitHub subsidiary introduced in 2021. Using Codex, developers can describe a task in natural language and automatically generate a code snippet capable of carrying it out.
Further down the road, Microsoft plans to add OpenAI’s latest ChatGPT model to the Azure OpenAI Service. ChatGPT is a more advanced version of GPT-3 that can generate text and code, as well as perform related tasks. ChatGPT can, for example, explain how a given snippet of code works and provide suggestions on how to improve it.
Azure OpenAI Service also includes several other neural networks from OpenAI. Among them is DALL-E 2, an AI system that can generate images in response to text prompts. Additionally, customers have access to the Embeddings series of neural networks, which OpenAI designed for use cases such as searching text and organizing documents by topic.
“Azure OpenAI Service provides businesses and developers with high-performance AI models at production scale with industry-leading uptime,” Eric Boyd, the corporate vice president of Microsoft’s AI platform group, wrote in a blog post. “This is the same production service that Microsoft uses to power its own products.”
The Azure OpenAI Service includes controls that are designed to ensure it’s used in accordance with Microsoft’s responsible AI policies. The controls will enable the company to detect abusive, hateful and offensive content. “In the event of a confirmed policy violation, we may ask the developer to take immediate action to prevent further abuse,” Boyd detailed.
Microsoft first partnered with OpenAI to commercialize the AI group’s technology in 2019. As part of the partnership, Microsoft invested $1 billion in the group and built a cloud-based supercomputer to support its research efforts. Recent reports suggest that the company could invest another $10 billion in OpenAI at a valuation of up to $29 billion.