Microsoft’s annual event for developers proved to be a treasure trove of AI-related news and features.
First off, announcements don’t get much bigger than a new line of Windows PCs designed specifically for AI applications. Microsoft claims that its new Copilot+ PCs will be the fastest and most intelligent Windows PCs ever built — which they’re pairing with all-day battery life.
With one of these devices on your desk, you’ll be able to use new features such as "Recall" where you can ask your PC anything about something you’ve done in the past which you need to (wait for it) recall. Can’t remember where you stored that one obscure file you thought you’d never need again? Just ask your Copilot+ PC.
A new AI era begins. Introducing the fastest, most intelligent Windows PCs ever. https://t.co/byVlr70zNH#CopilotPCs pic.twitter.com/UHloOYj4SdMay 20, 2024
Phi-3
Microsoft now has a new multimodal model, Phi-3-vision. It’s capable of general visual reasoning tasks and can also interpret charts, graphs, and tables.
Users can ask questions about a specific chart or input an image an ask open-ended questions about it. Phi-3-vision will then give a text response. This AI chatbot is currently available in preview.
Agents
Microsoft is also keen on delivering proper AI assistants which it refers to as agents or copilots.
These agents should be able to act independently and can be triggered by events, not just by your direct text prompts. For example, you could use a copilot to help with the onboarding of a new employee. Once the employee is registered, the copilot can greet them, answer their questions, introduces them to their buddy and sets up their first week of meetings.
Real-time translation
It also seems like a new wave of browser wars will be heading our way as Microsoft revealed real-time video translation is coming soon to Microsoft Edge. If you’re scrolling through videos on sites like YouTube and LinkedIn you’ll be able to listen to videos in a language of your choice either through dubbing or subtitles.
At first, translations will be available from Spanish to English and from English to German, Hindi, Italian, Russian, and Spanish. More languages are expected to be added later on.
Copy and paste
Copy and pasting is also getting an AI facelift. Microsoft’s PowerToys Advanced Paste is an AI-enhanced tool that you can use to paste text from your clipboard into different formats such as JSON or Markdown.
Love the new Advanced Paste tool in PowerToys! Paste as plain text, as markdown or as JSON are available for everyone, but you can hook up your OpenAI API to transform, translate or stylise the copied text on the go as well. Cool! #MicrosoftPowerToys pic.twitter.com/lKKN5YwgnRMay 22, 2024
Alternatively, you can paste text after it undergoes an AI transformation of your choice — be it summarizing, translating, or a changing it from a professional to a more casual tone.
All told, it seems like Windows users shopping for a new machine will have a lot more to consider as these features come to fruition. You can recap all the features discussed during the keynote in greater detail right here.
