After Google I/O 2024, the tech giant Microsoft concluded its annual developer conference in Seattle. On Wednesday, the company held the developer keynote of Microsoft Build 2024, unveiling a string of new products and services and its plan of action to ensure its lead in AI. During the conference, the company also announced an upgrade to its various Copilot products, including support for OpenAI’s GPT-4o and a Team Copilot for groups. Also, Microsoft unveiled its broader Azure AI Studio, which includes many models and is now generally available, as well as major changes to GitHub.
In this article, we have made sure to include all the key highlights and the latest AI innovations announced and unveiled during Microsoft Build 2024.
Microsoft Build 2024: Key Highlights & Major Announcements
- Team Copilot goes beyond delivering a conversational experience. It proactively and automatically engages to improve collaboration and streamline project management. According to the Microsoft blog, “Team Copilot expands Copilot beyond a personal assistant to act as a valuable team member—participating and contributing along with the team. And of course, you’re always in control—assigning tasks or responsibilities to Copilot so the whole team can be more productive, collaborative, and creative together.”
Microsoft’s Early Access Program offers Team Copilot and custom agent-building features. An extensive rollout is scheduled for late 2024. - Microsoft’s Copilot+ PC was launched a day before Tuesday’s Microsoft Build 2024 keynote. Its new Windows-based hardware earns a spot on the list of Build 2024 highlights as the company seeks to make amends for the AI PC. Starting at $999, Copilot+ PCs are available to pre-order before being generally available on June 18. It will help creative professionals generate images in third-party apps, including Adobe Photoshop, DaVinci Resolve, CapCut, and more.
- Microsoft added Phi-3-vision to its family of small language models (SML). It is designed for low computing-intensive tasks, such as on mobile and IoT devices. The tool is available with multimodal capabilities, along with its 4.2 billion parameters and 28K context length, enabling it to transcribe text from images, making it an ideal tool for image, chart, table, or graph analysis.
- Microsoft claims Phi-3-vision outperforms Claude-3 Haiku, Llava-1.6 Vicuna 7B, and Gemini 1.0 Pro V on multiple benchmarks.
- Microsoft introduced the Snapdragon Dev Kit in collaboration with Qualcomm. This Snapdragon Dev Kit for Windows, powered by Snapdragon X Elite, has a 4.6 TFLOP Adreno GPU, a 45 TOPS Qualcomm Hexagon NPU, 512GB of NVMe storage, 32 GB of LPDDR5x RAM, five USB ports, an Ethernet port, and an HDMI port each, to enable building, testing, and porting new AI experiences and native applications.
- The ability to manage and track coding projects within the Windows native program is made convenient by the latest Microsoft File Explorer’s integration with Git. It allows monitoring file statuses, committing messages, and their branches from within File Explorer, which now also supports 7-zip and TAR compression.
- Microsoft and Meta are collaborating to enable Windows Volumetric Apps on Quest headsets. The goal is to “extend Windows apps into 3D space,” which will allow common Microsoft users to continue using their familiar tools on a brand-new platform for spatial computing.
- Apart from Meta, the company also expanded its partnership with Hugging Face to incorporate the latter’s extensive language models into the Azure AI studio.
- A real-time video translation feature powered by artificial intelligence (AI) is coming to Microsoft’s Edge browser. It can translate videos from websites like Coursera, LinkedIn, YouTube, and Reuters. The feature provides translation from Spanish to English or vice versa, as well as from English to German, Hindi, Italian, and Russian. It is compatible with a limited number of languages. According to Microsoft, more languages and video platforms will be added in the future, along with the feature.
- Microsoft is planning to launch a custom emoji feature for Microsoft Teams. With this feature, the company is adding the ability to add your own emoji to Microsoft’s Slack competitor. Like in Slack, admins can limit who is allowed to add emojis, and they won’t be visible outside of your organization’s domain. They’re coming in July.
- The PowerToys suite for Windows 11 now includes Microsoft’s new Advanced Paste feature, which allows you to convert the contents of your clipboard as you go. By pressing Windows Key + Shift + V, you can open the Advanced Paste menu and use additional keyboard shortcuts to convert your paste to other formats, such as plaintext, markdown, or JSON. Another way to convert is to type into the prompt box, which also allows you to change or summarize the text before pasting it. The catch is that the AI portion will require an OpenAI API key and credits in your OpenAI account.
- Microsoft provided an update on its chip-building endeavours as well. The software giant’s first internal processor, the Azure Cobalt 100 central processing unit, will be available for preview to Azure cloud computing service users on Tuesday, according to the company.
What is the Difference Between AI PC and Conventional PC?
Microsoft CEO Satya Nadella called the company’s partnership with OpenAI the “most strategic, most important.” To that end, Microsoft CTO Kevin Scott invited OpenAI CEO Sam Altman onstage to discuss how we got here and what the future holds.
Later, he also mentioned the registration of 200,000 participants and the huge number of people attending the event in Seattle.
AI Seoul Summit 2024: Program Schedule, Theme and Host Participants
