A Microsoft spokesperson said it did this without requiring deeper data permissions and while inheriting all of an organization’s existing security, compliance and privacy policies.
Charging
“Through these new products and tools, users will be able to access and reference content they already have access to, from emails and chats to presentations and documents. Existing rights controls are respected,” the spokesperson said.
“Copilot’s big language models are not trained on their tenant data or their ads. And most importantly, with Copilot your data, including your directions, stays within the compliance limit.”
Microsoft also has an AI companion for Teams, called Business Chat, that sits in on meetings and can be called in to make summaries or generate updates based on emails, documents, notes, or other data.
But the AI features aren’t strictly for business, as Microsoft’s new announcements for image-generating AI target consumers squarely as the company continues to press its temporary lead on search market leader Google. .
Bing’s latest features include a revised version of “Knowledge Cards” in search, which appear for certain topics and include images, summaries, charts, timelines, and more that look like an AI-generated school report. There’s also a new feature called “Stories,” which looks similar to the feature of the same name on Instagram or Snapchat, but has an AI voice that reads out AI-generated summaries of certain topics with images.
New AI-powered elements in Bing search results, including a story at the top and knowledge cards on the right-hand side.
Potentially more impressive is the integration of Bing Image Creator, which is based on a version of OpenAI’s powerful DALL-E 2 text-to-image system, into Bing Chat.
That means that in the same chat where a user asks the Bing chatbot for kitchen renovation suggestions, they could ask what that would look like, and Bing could create a believable image from scratch.
Yusuf Mehdi, Microsoft’s vice president and chief consumer marketing officer, said the apps ranged from “time travel” asking to see certain places at different times, to applying specific parameters to a creative project to get an idea of where to start.
Images generated by Bing Image Creator using various prompts.Credit:Tim Biggs / Bing Image Creator
Charging
“It is useful for inspiration. If someone wants to create a charcuterie board but doesn’t know what a good one looks like… or is planning a children’s party and wants ideas for, say, astronauts in a field of sunflowers, what would that look like? saying.
“It’s a quicker way to get to exactly what I want compared to what I’ve done in the past, which is I had to go through Pinterest or I tried to search and didn’t get exactly what I wanted.”
Simply explained, DALL-E is a system that can understand a text message and link it to visual representations, thanks to a machine learning process trained on a large number of photos and artwork found online. It’s far from perfect, however, with requests for photorealistic images in particular frequently creating impossible or disturbing scenes.
Create a banner for Bing Image Creator
While something like Google Search requires a short, precise query to be most effective, generative AI works best with lots of detail and specificity. To get the images above, we use:
- Close up painted portrait of a man’s face. Freckles, green eyes, trimmed beard.
- 90s heavy metal album cover featuring a chameleon.
- A portrait of a demon. Cubism.
- Mid-Century Modern Living Room with Lots of Plants.
- Sydney Harbor in 1952.
- A steaming cup of coffee but the steam is turning into a thought bubble image.
Currently, anyone can access Bing Image Builder at bing.com/images/create/, but you have to craft your message carefully on your own. The integration with Bing Chat means that the AI can potentially use your plain language explanations to create an ad that will result in images closer to what you want. Or you can easily repeat but modify a prompt by saying “that’s cool, but make it rounder and bluer.”
The AI-powered Bing Chat itself was only introduced in February and raised concerns when it started acting erratically or inappropriately in certain situations, largely when pushed or abused by users. Partly because of that, the integration with Bing Image Creator is gradually rolling out to users.
The system has built-in precautions to prevent abuse, but generative AI can be hard to predict.
“Obviously we are working a lot with responsible AI and putting security measures in place to prevent misuse of violent images, harmful images,” Mehdi said.
“We are playing with the watermark of a version of the Bing icon to make it very clear [these are AI-generated images] and we will continue to review and receive feedback.”
Charging
AI image generation has also come under fire from some artists, who say it plagiarizes works found on the web and discourages people from commissioning human artists.
Mehdi said that DALL-E trained only on publicly available and licensed content, but that there would be a mechanism for concerned artists to prohibit Bing from complying with requests to copy their style. Microsoft also hopes to reach commercial agreements with artists that compensate them when their work is used for these types of requests.
The new features come as Google has made an older version of its Bard chatbot available for the first time, to a limited number of users in the US and UK. The search giant also has generative AI features ready for its Workspace, to counter Microsoft’s Copilot. Built with Google’s internal language models, they will also be able to redact or refine entire documents based on natural language cues.
Get news and reviews on tech, gadgets and games in our tech newsletter every Friday. Sign up here.