With OpenAI’s unveiling of GPT-4 today, its partner Microsoft acknowledged something that had been largely assumed but never confirmed: its AI-powered Bing search engine has been using GPT-4 all along.
“If you’ve used the new Bing preview at any time in the last five weeks, you’ve already experienced an early version of this powerful model. As OpenAI makes updates to GPT-4 and beyond, Bing benefits from those improvements,” said Yusuf Mehdi, Microsoft corporate vice president, Modern Life, Search, and Devices, in a post this afternoon.
Microsoft cloud chief Scott Guthrie and CTO Kevin Scott make cameos in OpenAI’s GPT-4 teaser video. The new Bing can compose text, respond to messages, and sometimes get into disputes with users, a tendency that Microsoft has quelled by trying out different limits on the length and number of sessions that users can have with the AI chatbot.
In addition to an upgraded ability to process and respond to language, GPT-4 can also accept images as input. This feature is in preview, not yet publicly available, but you can see it in action in the OpenAI developer livestream here.
Earlier today, Google annouced new generative AI features coming to its Google Workspace suite, including the ability draft, reply, summarize, and prioritize email messages, and write documents, among other capabilities.
Microsoft is holding an event on Thursday where it’s planning to announce new AI productivity features.
Read this report for a technical deep dive on GPT-4.
Discussion about this post