ChatGPT can now generate images – and they are surprisingly detailed.
On Wednesday, San Francisco artificial intelligence start-up OpenAI released a new version of its DALL-E Image Generator to a small group of testers and changed the technology to ChatGPT, This is a popular online chatbot,
Called DALL-E 3, the company said it could generate more solid images than previous versions of the technology, showing a particular knack for images containing letters, numbers and human hands.
“It is far better at understanding and presenting what the user is asking,” said OpenAI researcher Aditya Ramesh. He said that this technique was created to have a more accurate grip on the English language.
By adding the latest version of DALL-E to ChatGPT, OpenAI is strengthening its chatbot as the center of generative AI, which can produce text, images, sound, software, and other digital media on its own. Since ChatGPT went viral last year, it has sparked a race among Silicon Valley tech giants to be at the forefront of AI with progress.
On Tuesday Google issued a New version of its chatbot, Bard, which connects to many of the company’s most popular services, including Gmail, YouTube, and Docs. MidJourney and Stable Diffusion, two other image generators, updated their models this summer.
OpenAI has long offered ways to connect its chatbots to other online services, including Expedia, OpenTable, and Wikipedia. But this is the first time that a start-up has combined a chatbot with an image generator.
DALL-E and ChatGPT were previously separate applications. But with the latest release, people can now create digital images using ChatGPT’s service by simply describing what they want to see. Or they can create images using the descriptions generated by the chatbot, further automating the generation of graphics, art and other media.
In a demonstration this week, OpenAI researcher Gabriel Goh showed how ChatGPT can now generate detailed textual descriptions that are used to create images. For example, after creating a description of the logo of a restaurant called Mountain Ramen, the bot generated multiple images from those descriptions in a matter of seconds.
Mr Goh said the new version of DALL-E could generate images from multi-paragraph descriptions and closely follow instructions in fine detail. Like all image generators and other AI systems, it is prone to mistakes, he said.
As it works to refine the technology, OpenAI isn’t sharing DALL-E 3 with the wider public until next month. The DALL-E 3 will then be available through ChatGPT Plus, a service that costs $20 per month.
Experts have warned that image-grabbing technology could be used to spread large amounts of misinformation online. To protect against this with DALL-E 3, OpenAI has included tools designed to prevent problematic topics such as sexually explicit images and depictions of public figures. The company is also attempting to limit DALL-E’s ability to copy the styles of specific artists.
In recent months, AI has been used a source of visual misinformation, An artificial and not particularly sophisticated imitation a clear explosion Pentagon’s stock market declined slightly in May Other Example, Polling experts are also worried that could be technology used maliciously During major elections.
Sandhini Agarwal, an OpenAI researcher who focuses on security and policy, said that DALL-E 3 generated images that were more stylized than photorealistic. Nevertheless, she acknowledged that the model could be induced to create believable scenes, such as grainy images captured by security cameras.
For the most part, OpenAI does not plan to block potentially problematic content coming from DALL-E 3. Ms Aggarwal said such an approach was “too broad” because images could be harmless or dangerous depending on the context in which they appear.
“It really depends on where it’s being used, how people are talking about it,” he said.
