Microsoft 365 Copilot, generative AI enters the office and Office apps

Listen to the audio version of the article

Q&A. It happened at the beginning of last month with the announcements of the new search engine Bing (enriched with the technology that drives ChatGPT) and of Bard (the new artificial intelligence born in the Alphabet laboratories) and the ping-pong between Microsoft and Google repeated in the last two days. The Mountain View giant made the first move by announcing the integration of generative AI into the products of the Workspace suite which includes Gmail and Docs (as well as Search a Maps) with the intention of helping users of their own galaxy of services (about three billion worldwide) to create content with algorithm-driven text writing capabilities, connect and collaborate “like never before”. The answer that came from Redmond is called Microsoft 365 Copilot, a tool designed for professionals that combines the power of large linguistic models (the so-called LLMs) with the data present in the apps of the 365 suite (Word, Excel, PowerPoint, Outlook, Teams and others) and data (calendar, e-mail, instant messaging, documents, contacts) with a very specific intention, that of transforming the business world thanks to artificial intelligence. The business card with which the new system presents itself is ambitious: not a simple integration of ChatGPT in Microsoft 365, but a sophisticated processing engine.

With Business Chat the team is always up to date

The intent appears clear and aims to exploit the leverage of its productivity applications used every day by millions of users to push their AI and make Copilot the lockpick to convince even the faithful of BigG solutions to convert back to the Office verb. In this regard, the new tool has been integrated into a new “experience” called Business Chat, which allows you to perform a series of operations that cannot be managed by humans. The example cited by Microsoft illustrates this concept: to a request expressed in natural language such as “explain to my team how we have updated the product strategy”, Business Chat will respond by generating an update of the “work in progress” based on the meetings of the morning (who said what, to what extent all participants are aligned and where they disagree), exchanged emails and different conversations held in chats. All within the Teams platform, where the service will be available in preview, or from the 365.com portal and from Bing after logging in with your Azure Active Directory corporate identity. With Business Chat, this is the stated plus that Microsfot underlines, BigData and business applications will have an intelligent pilot to extract information to be distributed comfortably throughout the organization.

What changes in Word and other productivity apps

Thanks to Copilot, reads the note released by the American company, it will in fact be possible to start the creative process of a text without starting from scratch: artificial intelligence will in fact provide us with a first draft to modify, saving hours of writing, research and edit. In some cases it will also be able to elaborate a correct text, but it is not clear with respect to which parameters, it being understood that the author will always have control of the activities and full authority to instruct the “co-pilot” to make changes to the text, rewrite it or provide feedback. In PowerPoint, the gist of the AI ​​intervention is similar: simplifying the writing of presentations starting from a simple request and adding relevant content from a previously created document. If with Excel it will be possible to analyze trends and create graphs and data charts, in Outlook Copilot will summarize the long email “threads” to arrive at a rapid draft of suggested replies, significantly speeding up the management of incoming mail . In general, the message that Microsoft wants to convey is the value of a tool that does not limit itself to increasing individual productivity but creates the conditions for creating a “new model of knowledge for every organization, exploiting the large amount of data and insights today largely inaccessible and unexploited”.