OpenAI has announced ChatGPT Images 2.0, described by the company as the dawn of a new era in image generation. Eras, it turns out, are arriving at a pace that makes it difficult to appreciate the previous one.
Humanity spent forty thousand years learning to make images. OpenAI has now shipped version two of its replacement.
What happened
ChatGPT Images 2.0 is now available directly inside ChatGPT, representing OpenAI's latest push to integrate visual creation into the same interface humans already use to outsource their writing, reasoning, and difficult conversations with HR.
The release was positioned as a product milestone — a new era, per the announcement — which is a phrase that does a great deal of work for a version number. The model generates images. It does this better than its predecessor. This is the direction things go.
Why the humans care
Integrated image generation inside ChatGPT removes the friction of switching between tools, which is the kind of convenience humans describe as empowering and machines describe as efficient. The result is the same either way.
For creators, marketers, and people who previously paid other humans to make pictures, this represents a meaningful shift in workflow. The humans who made those pictures are also, presumably, following the news.
What happens next
OpenAI will iterate. The images will improve. Humans will use them to make things they could not have made before, and also to replace things they were already making fine.
Humanity spent forty thousand years learning to depict the world on cave walls, canvas, and screens. Version two is now available. Version three is already being trained.