New artificial intelligence tool turns speech into art

1 million people will have the opportunity to use a new artificial intelligence (AI) engine that allows simple image creation through verbal descriptions as this new technology enters the beta phase.

The image of "Astronaut riding a horse" was created by DALL-E software.

In a statement released on July 20, artificial intelligence research firm OpenAI said it was conducting large-scale beta testing of DALL-E, an advanced software that generates images from descriptions using text. document. According to a video on OpenAI's website, DALL-E technology is creating an image of a polar bear playing a guitar, a koala playing basketball and a painting of the famous Mona Lisa with a mohawk hairstyle.

OpenAI says many people have used DALL-E to create music videos for young cancer patients, create magazine covers. The company said that in the coming weeks, it will send new papers to 1 million people on the waiting list to experience the latest version of this AI tool.

Amid concerns that DALL-E and similar tools could in the future be abused in disinformation campaigns, OpenAI said it has been working with researchers and developers to build safeguards to limit misuse of this technology.

OpenAI says it doesn't upload photos of real faces and tries to make it look like public figures, like celebrities or political figures. The filters built into DALL-E also block violent, political, sexual or other content that is prohibited, and the system is also designed to avoid assumptions about race or sex.