OpenAI Introduces Dall-E 3, Latest Text-to-Image Tool With ChatGPT Assistance

An image created by Dall-E 3, the latest version of the text-to-image tool (photo: dock.openai)

JAKARTA - OpenAI on Wednesday 21 September introduced the latest version of their text-to-image tool that uses a very popular AI chatbot, ChatGPT, to help fill in demand.

Dall-E 3 will be available to ChatGPT Plus and Enterprise subscribers in October via API. Users can type in requests for images and adjust the request via chat with ChatGPT.

"DALL-E 3 can translate requests whose nuances are very detailed and accurate," the company said in a statement.

OpenAI says the latest version of this tool will have more security measures such as limiting its ability to produce violent, adult, or hateful content.

This tool also has mitigation to reject requests asking for pictures from public figures with names, or who ask for pictures in the style of an artist who is still alive.

OpenAI says creators can choose not to use some or all of their work used to train future text-to-image tools.

The OpenAI race to create accurate AI text-to-image tools has several competitors, including Alibaba's Tongyi Wanxiang, Midjourney, and AI Stabilities, which continue to fine-tune their image-making models.

However, there are some concerns surrounding images generated by AI. A court in Washington DC in August ruled that artwork created by AI without human contributions cannot becopied under US law.

OpenAI is also facing several lawsuits. A trade group for US writers recently sued the artificial intelligence leader on behalf of the author, including John Grisham and novelist George RR Martin's "Game of Thrones", on charges that the company violated their ChatGPT chatbots with their work.