Microsoft to add ChatGPT to Azure cloud services ‘soon’

17 Jan 2023

Image: © Peter/Stock.adobe.com

The two companies have expanded their partnership recently, while there are rumors that Microsoft plans to give more funding to Open AI.

Microsoft has expanded access to OpenAI’s software, making it generally available on its Azure cloud service.

The company said more businesses will be able to utilise OpenAI’s various large language models such as DALL-E 2 and Codex. Microsoft also plans to add OpenAI’s advanced chatbot – ChatGPT – to the service “soon”.

Microsoft plans to give access to a “fine-tuned version” of ChatGPT that has been trained to run on Azure AI infrastructure.

The two companies have been in partnership since 2019, when Microsoft invested $1bn into OpenAI to help bring “secure, trustworthy and ethical” AI to serve the public, while focusing on constructing Azure AI supercomputing technologies.

Eric Boyd, Microsoft’s AI platform corporate VP, said the Azure OpenAI service was first launched in 2021 to let customers “tap into the power of large-scale generative AI models”.

“Since then, one of the most exciting things we’ve seen is the breadth of use cases Azure OpenAI Service has enabled [for] our customers,” Boyd said in a blog post. “Customers of all sizes across industries are using Azure OpenAI Service to do more with less, improve experiences for end-users, and streamline operational efficiencies internally.”

The two companies have been expanding their partnership recently, as Microsoft plans to upgrade its Bing search engine using OpenAI’s ChatGPT.

Last October, the Wall Street Journal (WSJ) reported that Microsoft is in talks to give a new round of funding to OpenAI. The AI company was also reportedly in talks this month to raise capital at a valuation of almost $30bn, according to the WSJ.

ChatGPT is one of OpenAI’s latest creations and was designed to respond to questions in a human-like way. The experimental chatbot is able to answer follow-up questions, admit its mistakes and challenge incorrect premises, according to OpenAI.

public demo of the AI was released towards the end of 2022 and rapidly gathered attention, getting more than 1m users by 5 December.

However, while many users praised the experimental AI for its ability to respond to questions, some posted examples of the chatbot creating biased or racist content. Other examples showed the chatbot providing incorrect answers to questions.

Around the time ChatGPT crossed 1m users, Stack Overflow, a Q&A site for programmers, issued a temporary ban on answers created by the chatbot.

OpenAI said on Twitter today (17 January) that it “learned a lot” from the demo and that important updates are being made based on user feedback.

10 things you need to know direct to your inbox every weekday. Sign up for the Daily Brief, Silicon Republic’s digest of essential sci-tech news.

Leigh Mc Gowran is a journalist with Silicon Republic

editorial@siliconrepublic.com