HyperAIHyperAI

Command Palette

Search for a command to run...

OpenAI Releases ChatGPT Image 2.0: Introduces "Reasoning Capability" for More Realistic Content Generation

Against the backdrop of continuous evolution in generative AI, OpenAI is strengthening its image generation capabilities. On Tuesday local time, the company officially launched ChatGPT Images 2.0, a next-generation image model described as possessing certain "reasoning abilities." Beyond generating images based on prompts, it can cross-check outputs using online information to enhance result plausibility and consistency. According to official demonstrations, this model further improves realism and expressive range. Some generated examples were deliberately designed as "screenshots within screenshots" or scenes simulating users flipping through physical magazines, blurring the boundary between AI-generated imagery and real photographs. In his launch livestream, Sam Altman remarked that "the team has done an excellent job on this model." Functionally, ChatGPT Images 2.0 supports multilingual prompting including Chinese, Japanese, Korean, Hindi, and Bengali. It can generate diverse formats such as advertising materials, magazine pages, and comic content, while enabling stylized reinterpretations of real individuals—for instance, transforming prominent investors into figures styled after the 1980s. Addressing external concerns over copyright issues, OpenAI stated that the model does not copy or reproduce specific works but generates content based on learned patterns. While the company restricts mimicking living artists' personal styles, it permits creation in broader studio-style contexts. Mitch Stoltz noted that unless generated content bears substantial similarity to specific works in training data, it typically does not constitute copyright infringement. However, as AI-generated content becomes increasingly realistic, related controversies persist. Currently, OpenAI faces multiple copyright lawsuits involving entities such as The New York Times and author George R.R. Martin. Industry observers widely believe that although the underlying technology resembles traditional image editing tools, advancements in efficiency and scale are amplifying associated social and legal impacts.

Related Links