"Kakao AI Unveils Massive Model That Writes and Draws"

[Asia Economy Reporter Buaeri] Kakao has unveiled an artificial intelligence (AI) that can even draw pictures.

Kakao Brain announced on the 15th that it has released the ultra-large AI multimodal 'minDALL-E' on GitHub, the world's largest open-source community.

'minDALL-E' is an image generation model that creates desired images in real-time when users input commands in text. It is a smaller-sized model of Open AI's 'DALL-E' made accessible to everyone, pre-trained on a dataset of 14 million text-image pairs. For example, when commands like "Draw a chair made of banana peels," "Show a picture with a full moon and the Paris Eiffel Tower together," or "Draw in the style of painter Salvador Dali" are entered, the AI understands the context of the command and immediately generates the image.

Kakao Brain will hold an 'Open Mini Conference' online on the 17th to explain 'KoGPT' and the multimodal 'minDALL-E'.

Kim Ildu, CEO of Kakao Brain, said, "We are endlessly questioning what elements are necessary for AI to think and speak like humans, whether AI with memory, reasoning ability, and comprehension can communicate like people," adding, "Starting with the recently released language model KoGPT and the newly unveiled DALL-E, we will sequentially release various ultra-large AI models."

This content was produced with the assistance of AI translation services.