Canva launched text-to-image AI feature

Design platform Canva has launched a tool based on Artificial Intelligence (AI) using open source. It works with the Stable Diffusion model that allows you to switch from text to image. This will be available for free with the application of this Australian platform. The tests had begun in September and now it will be ready for its more than 100 million app users to use it openly.

This Stable Diffusion model will be open source and will have some additional security filters. In addition, there will be a customized user interface to help users achieve the results they are looking for. Canva will provide the tool both to those who use the application for free and to those who pay for additional features. All those who are registered on the platform will be able to generate about 100 images a day with the Artificial Intelligence.

The interface that users will have in this new tool will have a help that will guide them in the creation of specific images. In order to perform it, it will be necessary to load the function first. Then, the necessary indications for AI inspiration should be given. Afterwards, it is time to choose among the styles available there so that the tool generates a grid of four images to choose from. There will also be an option to report any image that contains violence, nudity, incitement to hatred or any type of biased or stereotyped content.

Canva and the search for new experiences

Cameron Adams, co-founder and chief product officer of Canva, when interviewed by The Verge made it clear that they are “treating this as a great learning experience for our community.” Adams also stated that they are “eager to introduce them to this technology because it is an emerging field, and exactly how it works and how customers will interact with it is still being developed.”

This tool, Adams made clear, has a variety of applications. “One of my favorites is for students to use it to visualize their stories, so they’ll write a story in English class and use text to image to generate an image that matches that story. We’ve also seen it used for images in presentations, brochures and T-shirts, which they can print through Canva,” said the company’s co-founder.

Road to the future

This feature is just one example of AI-enabled tools that move from text to image and reach increasingly large audiences. Stable Diffusion has accelerated access to these systems after their launch. The open source implementation allows companies to integrate it into their own products free of charge.

Creative platforms are increasingly adding this text-to-image option in the creation of images. Last month, Microsoft launched its own Microsoft Designer tool. It is powered by OpenAI’s DALL-E system and is part of its Office suite. Adams did not overlook the issue of royalties and believes that several issues in this regard are “pretty much up in the air.”