Soon we’ll all be AI artists

Image generated using Canva with the prompt "reading “The Peak” email on a phone”

Just a few months ago, text-to-image AI seemed like super-futuristic tech reserved solely for hardcore researchers. Now, it’s available for free on one of the most popular design apps.

What happened: Canva has rolled out an AI-powered text-to-image generation tool that can generate four images at a time in various art styles based on individual user prompts. 

  • The feature runs on the open-source image-generator Stable Diffusion, the creator of which (Stability AI) just raised US$101 million at a $1 billion valuation.

Why it matters: With over 100 million users, Canva is bringing AI image generation to the biggest audience yet—with access to this technology spreading so rapidly, it could soon become as ubiquitous as the text box feature or WordArt… do people still use WordArt? 

Yes, but: There are fears over the tech being used to infringe on artists' work to create unlicensed products or to create content featuring violence or hate speech. 

  • When asked by The Verge about these issues, Canva said it was still figuring out copyright ethics, but stressed it had added extra filters to prevent NSFW content.

Bottom line: From algorithms to deepfakes, and now generated art, AI is slowly but surely creeping into the technology we use day-to-day.