large language models
With the right inputs, DALL-E can make bigger versions of classic paintings like the Mona Lisa.
AI experts are arguing after a viral Twitter thread claimed DALL-E is creating its own hidden language to categorize images.
The company’s AI researchers say its new language model is generating ‘toxic’ results that often reinforce stereotypes.
Researchers have warned that massive language models like Google’s are inherently risky and can be harmful to marginalized groups.