Microsoft Builds on AI with Team Copilot, New Partnerships, and More

Microsoft’s Build Developer Conference served up a veritable banquet of conversational AI products, flanked by tables bearing visions of AI for every developer. Here are the menu items:

Team Copilot Takes the Spotlight: Microsoft’s AI assistant is going beyond the individual with Team Copilot, a new feature within its communications and collaboration app, Microsoft Teams, that serves as an intelligent, virtual teammate that can coordinate agendas, take notes, and answer questions within a meeting in real-time, improving collaboration and productivity.

Copilot Gets Extensions: Extensions make it possible to extend Copilot’s capabilities, building on top of the existing versions to combine its strength with external resources such as other tools, repositories, or applications. The first extensions are available for GitHub, providing developers with more options to customize their coding journey.

And of these collaborations continues: In a bid to strengthen its position in the rapidly expanding AI ecosystem, Microsoft announced a slew of expanded partnerships with big names such as chip-maker Nvidia, text-generating AI lab OpenAI (whose GPT-4o language model has made waves for its capabilities), NLP and ML hub Hugging Face, and the vast dataset-compile Cognition AI, with the stated goal to accelerate the development, adoption and implementation of AI across various sectors.

The new model unveiled: Microsoft releases Phi-3-vision, a small language model that excels at audio and visual tasks. A more affordable, task-specific model, Phi-3 invites more developers to tap AI, as the software is lighter to run than bulkier large language models that are harder to train.

Real-Time Intelligence with Fabric can give organizations real-time responses, allowing them to make vital decisions faster by using AI to identify problems before they arise.

Custom chips and infrastructure improvements: Microsoft is serious about building an AI-optimised foundation, and in 2022 it introduced a preview of its Cobalt 100 custom silicon processor offering up to 6x performance improvement for general and cloud-native workloads. AI additions to Microsoft’s infrastructure also included becoming the first cloud provider to offer AMD’s upcoming leading MI300X AI accelerator.

Democratization of AI as a Key Focus: A key message across the conference was to make AI more accessible, from the collaborative features of Team Copilot to the availability of Phi-3.

This emphasis on collaboration, personalization, and accessibility speaks to Microsoft’s vision of AI: a technology that enhances human potential at the individual and organizational levels.