I went hands-on with Microsoft's new AI features, and these 5 are the most useful

I went hands-on with Microsoft's new AI features, and these 5 are the most useful

Innovation

I went hands-on with Microsoft’s new AI features, and these 5 are the most useful

Here are the new Microsoft AI features that will likely make it into my own daily workflows.

Microsoft Event - Copilot

June Wan/ZDNET

At Microsoft’s annual fall event, the company unveiled major AI products such as Copilot, an updated Windows 11, and Microsoft 365 Chat that are all capable of doing many different tasks, almost too many to keep track of. 

New AI productivity assistants enter the space nearly every day with promises to do various tasks to improve your workflow. Yet, sometimes, the tasks they can do are not that helpful or require the same amount of human supervision, making them not worth using.  

Also: Every product Microsoft unveiled at its Surface and AI event today

From my hands-on experience at the event, I am happy to report that five of Microsoft’s new AI offerings are actually helpful in solving real-world problems.  

I rounded up the new Microsoft AI features that will likely make it into my own workflows.

  1. Bing Chat Personalized Answers 

Bing Chat 'Personalized Answers' demo

Sabrina Ortiz/ZDNET

Bing Chat Personalized Answers reminds me of OpenAI’s ChatGPT Custom Instructions feature, but smarter. 

With the Personalized Answers features, Bing Chat will be able to refer back to the content from your previous conversations to give you new answers, preventing you from having to retype the same guideline over and over. 

For example, if you told Bing Chat once in a previous conversation that you were allergic to shellfish and you later asked it for a recipe idea, it would automatically filter out recipes that contained shellfish. 

Also: How to use Bing Chat (and how it’s different from ChatGPT)

By remembering your previous prompts, you don’t have to be as specific every single time, making your prompt writing easier. 

The feature is especially convenient when you ask the chatbot prompts that all relate to the same topic, such as planning for vacation or creating content for a particular audience. 

With Custom Instructions on ChatGPT, you can get the same end result by typing in a guideline such as, “Make sure you explain the terms in a level a child can understand,” so it can automatically be applied to all of your prompts thereafter. 

However, with ChatGPT’s alternative, you would have to physically type in the instruction, and it would only apply that one instruction to all of your prompts. 

Also: Microsoft quietly announced two products today that business users will love

With Bing Chat’s Personalized Answers, the chatbot would remember all of the different information you have given it in the past, and it would use it as context every time it produces an answer. 

If you don’t want Bing Chat to refer back to a conversation in future answers, all you have to do is delete it from your chat history. 

2. DALLE-3 supported Bing Image Creator 

Bing Image Creator credentials feature

Sabrina Ortiz/ZDNET

Bing Image Creator will now be supported by DALLE-3, which is OpenAI’s most advanced image-generating model. This upgrade will allow Bing’s Image Creator to better understand your prompts and generate more accurate and detailed images, reducing the frequency of blurry and warped details in its outputs.

The best part is that you can open the higher quality renditions in Microsoft Designer, the company’s take on Canva, right from Bing Chat with the new “Start a Design” option. 

Also: How to use Canva AI to turn any photo into a professional headshot

This is particularly helpful if you need the image for a specific project such as a flyer, social post, banner, or more. Instead of having to download the image from Bing Chat and manually import it into Designer yourself, you can do it all in one swoop. 

Lastly, the images will now display content credentials in the form of a tag that says “Generated with AI” which will hopefully help reduce the spread of misinformation, one of the biggest concerns with AI image generation. 

3. Microsoft Shopping Assistant 

Microsoft Shopping Assistant

Sabrina Ortiz/ZDNET

Before you write off this feature, hear me out because I watched the Shopping Assistant do its magic from start to finish and was left impressed. 

Also: These are my 5 favorite AI tools for work

With the new Microsoft Shopping Assistant, users will be led to the exact product they are looking for by answering specific questions posed by the assistant. 

For example, suppose you tell the assistant you are shopping for new fall boots. In that case, the assistant may follow up by asking what type of boot specifically, then what color, and so on, showing you updated results each time until you narrow it down to the product you want to purchase. 

During that process, it can create charts comparing all of the different product specs and prices to help you make the best buying decision. 

Also: 5 emerging use cases of generative AI in commerce

The Microsoft Shopping Experience will be a standalone site that can be accessed through any browser. However, if you want the extra bells and whistles, such as price trends, coupons, and more, similar to the ones featured on popular RetailMeNot and Honey browser extensions, you can only find those on Edge. 

I will likely use this feature when looking for specific products because I often get frustrated after scrolling aimlessly through Google and not finding a product that meets all my criteria. 

4. The new Paint

The new paint demo

Sabrina Ortiz/ZDNET

The Paint tool is finally getting a much-needed AI facelift that elevates the tool from being a Windows application you ignore to an application you actually reach out for. 

Also: 4 ways generative AI can stimulate the creator economy

Paint now features a background removal tool that can isolate the subject of a picture within seconds, allowing you to layer it on other images or projects. The most noteworthy feature, however, is its new Cocreator feature, which infuses generative AI into the Paint experience. 

With this feature, you can type up a description like you typically would with another AI image generator. Then, in seconds, it will generate the image you can incorporate into your Paint project, as seen in the video above. 

5. Copilot in Word

Microsoft Word Copilot

Microsoft

This feature didn’t make last place because of its capabilities, but rather because it is so capable that I can see it will open a Pandora’s box of issues in terms of plagiarism and discerning between human and AI-generated work. 

Also: How to use ChatGPT to do research for papers, presentations, studies, and more

Copilot will live in Microsoft Word and assist through every step of the writing process. The most helpful and harmless feature is its ability to summarize and document for you with detailed, bulleted lists. 

Copilot will also serve as a writing assistant, rewriting your text and presenting you with a series of options to pick from, which you can then adjust in tone to your liking.

It can also write full documents based on other content within your Microsoft workspace. Think “Write my blog post based on the findings of this year’s monthly report.”

Also: 5 handy AI tools for school that students, teachers, and parents can use, too

These last two features could be extremely useful for many professionals’ workflows to help with mundane document writing; however, in my line of work or for students’ assignments, leveraging either of these tools would likely violate writing policies. 

Lastly, Copilot in Word will also be able to generate tables from copy or tweak tables with a simple command. 

Editorial standards

Add a Comment