If your workplace uses the Google Workspace productivity suite, you may soon have a new teammate: an AI teammate.
In its mission to improve our real-life collaboration, Google has developed a tool to bring together shared documents, conversations, comments, chats, emails and more into a single virtual generative AI chatbot: the AI Teammate.
Powered by Google’s proprietary Gemini generative AI model, AI Teammate is designed to help you focus more on your role within your organization and leave the tracking and tackling of collective assignments and tasks to the AI tool.
This virtual colleague is given its own identity, its own Workspace account and a specifically defined role and objective to fulfill.
Once AI Teammate is set up, it can be given a custom name and also undergo other changes, including its function, a description of how it should help your team, and specific tasks it should perform.
In a demonstration of an example of an AI teammate at I/O 2024, Google showed off a virtual teammate named “Chip” who had access to a group chat of those involved in presenting the I/O 2024 demo. The host, Tony Vincent, explained that Chip was aware of a large number of chat rooms that had been set up as part of the preparation for the big event.
Vincent then asks Chip if I/O storyboards are approved – the kind of question you might ask colleagues – and Chip was able to answer because it can analyze all these conversations in which it was keyed.
As AI Teammate is added to more discussions, files, chats, emails, and other shared items, it builds a collective memory of the work being shared across your organization.
In a second example, Vincent shows another chatroom for an upcoming product release and asks the chatroom if the team is on track for the product launch. In response, AI Teammate searches everything it has access to, such as Drive, instant messages, and Gmail, and synthesizes all relevant information it finds to form its response.
When it’s done (which seems like about a second or a little less), AI Teammate delivers a digestible summary of its findings. It highlighted a potential problem to alert the team to, and then provided a summary of the timeline, showing the stages of the product’s development.
Because the demo takes place in a group room, Vincent said anyone can follow along and jump in at any time, for example by asking a question about the summary or having AI Teammate transfer its findings to a Doc file, which it does as soon as the Doc file is ready.
AI Teammate will be as useful as it is customized, and Google promises it can make your collaboration seamless as it integrates into Google’s suite of existing products that many of us are already used to.