Microsoft: Bing generates images with artificial intelligence – 03/22/2023 – Tech

Microsoft announced two releases with artificial intelligence (AI) this Wednesday (22). The Bing search engine, which already responded to requests from users in chats similar to ChatGPT, will start creating images. The idea is to transform the search engine into a virtual assistant capable of creating images to be used in spaces such as posts on social networks and newsletters.

In addition, GitHub, a kind of programming code library purchased by Microsoft in 2018, announced that it will add GPT-4 technology —a turbocharged version of ChatGPT— to its Copilot programming assistant. With requests made by programmers in plain language, this tool creates program code snippets, provides explanations for bugs, and details computing concepts.

The new features available in Microsoft’s browser are aimed at turning Bing into a creative assistant, according to Yusuf Mehdi, the company’s director of consumer marketing.

“Just type something like ‘draw an image’ or ‘create an image’ into the chat, followed by the description of the image you want, like a look for a newsletter or inspiration for redecorating your living room,” he said.

Users will receive this material with a watermark, to attest that the content was made with AI.

Microsoft says it has integrated a more powerful version of Dall-E into its browser, a platform from OpenAI, the same creator of ChatGPT, which generates images from text descriptions.

Bing search will also start delivering cards [postagens que combinam imagens e textos] and stories [vídeos curtos na vertical] created with OpenAI’s artificial intelligence in response to surveys. Bing already allowed asking questions to the chatbot, based on search results.

To access Bing’s new functionality, you need to access Microsoft’s search engine through Edge or download the Bing app (Play Store or App Store). Then you need to sign up for a chatbot waitlist. Bing’s image generator is in the chatbot itself

According to a statement signed by the CEO of Github, Thomas Dohmke, the updates to the platform not only improve the option of automatically completing programming codes, but also increase the interactivity between the user and Copilot in the chat format.

The new programming assistant can, for example, comment on a computing project at the request of the responsible professional. This practice facilitates collective work, by making each stage of development clearer.

Copilot was developed before Microsoft invested $10 billion in OpenAI, in a partnership between the artificial intelligence startup and Github itself. Microsoft bought GitHub for $7.5 billion in 2018.

Since Copilot was announced in June 2022, the platform has served millions of developers and more than 5,000 companies, according to GitHub. The new version of the platform is available for testing by subscribers of the Microsoft 365 corporate package, which costs from BRL 359 per year).

Microsoft has doubled its bet on artificial intelligence that generates content, after the success of ChatGPT with the public, which already has more than 100 million accounts.

In January, the company disclosed a billionaire contribution to OpenAI to gain access to its AI products, and said it had adapted the new platforms to ethical standards.

On March 14, Google announced the deployment of AI tools across its office platforms: Docs, Sheets, and Slides. The new resources will be able to generate content such as image, audio, text, programming codes and video, according to the search giant.

This Tuesday (21), Google launched its own artificial intelligence chatbot, Bard. Users in the United States and England can now chat with the new technology.