Microsoft’s AI Copilot makeover lets it see and speak

Click here to visit Original posting

Microsoft's Copilot AI assistant has a sleek new look and a slew of new and enhanced features. The tech giant is clearly keen to get more people to use its AI tools and to make it as much a part of people's lives as possible. 

It's a strategy pursued by Google and Microsoft's other AI rivals as well. To stand out, Microsoft gave Copilot a makeover regarding how you engage with the AI and its capabilities when you do so. Copilot is noticeably quicker to adapt to your preferences and does so in more ways than it could before. You can test out the new Copilot experience on the web, a mobile device, or through the Windows app. 

"Copilot will be there for you, in your corner, by your side and always strongly aligned with your interests." explained Microsoft AI CEO Mustafa Suleyman in a blog post. "We are not creating a static tool so much as establishing a dynamic, emergent and evolving interaction. It will provide you with unwavering support to help you show up the way you really want in your everyday life, a new means of facilitating human connections and accomplishments alike."

We've highlighted some of the most notable and intriguing changes and new features below.

Voice

The loudest upgrade is the new Copilot Voice feature that allows you to converse vocally with Copilot. It's a bit like OpenAI's new AdvancedVoice Mode in that you can speak to the AI like you would a human, even interrupting mid-word, and hear what it has to say back. 

You can select four voices for the experience, with multiple accents, though only in English for now. Copilot Voice can also be your newscaster with the Copilot Daily feature. You can get audio summaries of the latest news and weather personalized to your location and interests. The AI relies on Microsft's partners, including Reuters, Axel Springer, Hearst, and the Financial Times, to source and summarize the news. 

Vision and Visuals

Not only can Copilot now talk, it can 'see' too. The new Copilot Vision feature lets the AI look at what's on your screen, whether it's a website or a personal document. The AI can suggest ideas for related information or analysis, even bring up products that might better suit your interests and budget if you ask, simply using the vision on the screen. It won't work universally yet, but it can handle most of the more popular websites. This is an opt-in feature, so none of the data is stored or used for training unless you agree. Notably, Copilot Vision isn't limited to printed text; it can even read handwriting.

For your own visual creations, Copilot assists with enhancing both photos and other visiuals. The AI-powered Super Resolution feature in the Photos app can enhance old and low-resolution images to 4K without getting grainy or pixelated. And if you make an image in Paint, you can now use the AI-fueled Generative Fill and Generative Erase tools to add or remove bits from an image using a text prompt or brush strokes.

Personalized Companion

Microsoft's main goal for Copilot is for people to think of it as a useful helper to turn to both online and when working on a PC. The new Copilot's look is no longer the usual chatbot but now works as cards that are supposed to make you feel more involved in running the AI. There's also the new main page that adapts to your interests called Copilot Discover. Your conversations with the AI and stated preferences will change the Discover page to provide searches and suggestions that are relevant to you, encouraging long-term use.

A subtle but potentially very useful upgrade is how Copilot helps you find and do things on your PC. For instance, Click to Do is an AI feature that puts interactive commands on your screen, offering suggestions relevant to what's on the screen, such as summarizing text, changing an image, or looking something up online. If you're working on a presentation, Copilot might offer to help expand the text in one section, remove the background of an image, or even email someone if you have their address in the presentation.

There's also the new Recall tool, which helps you find things you were looking at recently on your computer, document, webpage, video, or anything else. The AI can remember and recognize what it was and pulls it back up again for you, even if you don't remember the location or file name. Windows Search on Copilot+ PCs will even find files and settings that you haven't opened yet by using a description you write in.

You might also like