Alongside the announcement of a revision of the chat limits implementation on Bing’s ChatGPT, Microsoft also revealed some behind the scenes on how the company’s latest Search Engine, new Bing, works. The technology behind the new Bing is powered by the company’s new AI model, officially called Prometheus.
In the company’s branded blog post today, Microsoft shared the company had developed a proprietary technology they named Prometheus, which is a first-of-its-kind AI model that combines the company’s comprehensive Bing index, ranking, and answers results with the creative reasoning capabilities of OpenAI’s most-advanced GPT models.
How Microsoft Prometheus Works?
Going into the specifics, Prometheus uses Bing and GPT to generate a set of internal queries iteratively through a component called Bing Orchestrator, and then provides an accurate and rich answer for the user query within the given conversation context, all within a matter of milliseconds.
Selecting the relevant internal queries and leveraging the respective Bing search results is a critical component of Prometheus, since it provides relevant and fresh information to the model, enabling it to answer recent questions and reducing inaccuracies. Put in another way, the model uses its reasoning skill over the data provided by Bing and hence it’s grounded by Bing data, via the Bing Orchestrator. Microsoft has shared the diagram below to illustrate how Prometheus works at a high level.
The final step is when Prometheus attaches relevant Bing search answers such as weather, stock, sports, news, and so on, to the Chat answer. With the search results in mind, Prometheus is able to integrate citations into sentences in the Chat answer so that users can easily click to access those sources and verify the information.
Jordi Ribas, Corporate Vice President of Search & AI at Microsoft, has also expressed in the blog post that besides Prometheus, his team faced challenges in creating an ideal Bing search product. In the end, the team developed a new Bing UX that has a unified Search and Chat in a single interface, because some queries would perform better in the classical or Search mode, while others would do better in a conversational or Chat mode.
Selecting the relevant internal queries and leveraging the respective Bing search results is a critical component of Prometheus, since it provides relevant and fresh information to the model, enabling it to answer recent questions and reducing inaccuracies – this method is called grounding.
Put another way, the model reasons over the data provided by Bing and hence it’s grounded by Bing data, via the Bing Orchestrator. The diagram below illustrates how Prometheus works at a high level.
