At OpenAI – one of the world’s leading companies in artificial intelligence, researchers have built technology that allows users to create digital images simply by describing what they want to see.
Alex Nichol, one of the researchers at OpenAI, demonstrated how Dall-E works.
Dall-E produces images with the keywords `avocado teapot`, `Shiba dog wearing beret and turtleneck`, `soup bowl that looks like a woolen monster`, `cat playing chess` and `room
Half a century ago, a group of seven OpenAI researchers built a tool that can identify subjects in digital photos and create their own images such as flowers, cars… Later, the company improved
`We can now use multiple intersecting data streams to create better and better technology,` said Oren Etzioni, CEO of the Allen Institute for Artificial Intelligence in Seattle.
However, this technology is currently not perfect as it misunderstands and creates unwanted results.
Not only creating photos, Dall-E can also quickly edit photos as required.
However, this feature also makes some experts express concern.
`You can use it for good things, but you can certainly also use it for other crazy, disturbing purposes, like creating misleading images and videos,` said Subbarao Kambhampati, professor of computer science.
Currently, OpenAI does not allow outsiders to use Dall-E.
However, a number of other companies around the world may soon create similar technology to Dall-E so that everyone can access it.
`People need to know that the images they see may not be real,` he said.
