Microsoft announced more than 100 new features and products designed to help companies that use or intend to use AI in current and future projects.
During Microsoft Ignite, the company also announced some interesting results from a study it conducted on the use of the Copilot tool it released earlier this year. 70% of respondents said Copilot helped them improve their productivity. 64% said it helped save time when working with emails, 87% said it was useful when creating a draft, and 75% said it was useful when finding information in files.
“What everyone wants to know now is: Will Copilot really change work, and how? Our research, using a combination of surveys and experiments, shows the productivity gains are real,” Frank X. Shaw, chief communications officer at Microsoft, wrote in a blog post.
Copilot updates across Microsoft’s portfolio
For Microsoft 365 users, Copilot now offers added features such as personalization options, a comprehensive whiteboarding and note-taking experience for meetings, and the option to utilize Copilot’s transcription services during meetings without the transcript being saved.
Microsoft is introducing a new tool called Microsoft Copilot Studio, a low-code solution that empowers users to customize Microsoft Copilot. This tool enables the integration of business data and the creation of custom copilots tailored for specific purposes.
Moreover, Microsoft is extending Copilot into mixed reality on the HoloLens 2, ensuring seamless integration into workers’ workflows. This mixed reality version utilizes natural language and human gestures to provide interactive guidance.
In addition, Microsoft is introducing specialized versions such as Microsoft Copilot for Service to assist contact centers and Microsoft Copilot for Azure to support IT teams.
As part of this update, the rebranding of Bing Chat and Bing Chat Enterprise to Copilot has also been announced. This move aims to streamline and unify the experience across Microsoft’s services.
Azure AI receives several enhancements
The new Model-as-a-Service feature allows developers to seamlessly integrate and customize AI models using their own data without the need to manage the underlying GPU infrastructure.
The Vector Search feature, facilitating the storage, indexing, querying, filtering, and retrieval of AI data, is now generally available.
In the upcoming public preview in Azure OpenAI Service, the GPT-4 Turbo model will be introduced by the end of this month. Furthermore, GPT-4 Turbo with Vision is set to be in preview soon, while DALL·E 3 is already available in preview.
The company has also announced the general availability of Azure AI Content Safety, aiding organizations in detecting harmful content in AI experiences. Additionally, new documentation has been released, offering guidance on how to avoid potential content infringement.
New chips optimized for AI
The company has unveiled two new chips, namely Microsoft Azure Maia and Cobalt. Microsoft Azure Maia serves as an AI accelerator designed for executing cloud-based training and inferencing for AI workloads. On the other hand, Microsoft Azure Cobalt is an Arm architecture chip tailored for general-purpose workloads, emphasizing performance, power efficiency, and cost-effectiveness.
Furthermore, the company has made Azure Boost generally available. This system optimizes performance by offloading storage and networking processes from host servers to hardware and software specifically optimized for these tasks.
Microsoft has also emphasized collaborations with AMD and NVIDIA. Azure will soon incorporate AMD MI300X accelerated virtual machines, and Microsoft’s NC H100 v5 Virtual Machine Series is configured to leverage NVIDIA H100 Tensor Core GPUs.