By Sheryl Finley, AI Engineer, 9/20/2023
OpenAI is previewing DALL-E 3, an improved version of its image generation tool that can create images from written prompts. They plan to integrate it into their ChatGPT chatbot, expanding its reach, despite increasing calls for restraint on such technology.
DALL-E 3 features AI technology that seems to have a better understanding of user commands, rendering text into images more coherently. It can also handle complex instructions and will be available to ChatGPT subscribers in October, potentially reaching a much wider audience.
This move comes as OpenAI faces competition and a slowdown in traffic to their AI tools. Integrating DALL-E 3 into ChatGPT aims to boost both.
Over the past few years, text-to-image AI tools like DALL-E have proliferated, raising concerns about their potential misuse, such as deepfake creations and copyright issues.
OpenAI is working to address these concerns, inviting outside experts to test worst-case scenarios and publishing system cards to outline safety measures. They are also experimenting with mechanisms to identify AI-generated content and its provenance.
While the integration of DALL-E 3 into ChatGPT has potential benefits, it also highlights the ongoing challenges of regulating and managing the ever growing avent of AI-generated content.
