ChatGPT Images 2.0 is a Teaching Tool, Not a Toy

Today's topic: OpenAI released ChatGPT Images 2.0 on Monday, and it's the first image model with "thinking" built in.

What that means in plain English: the model plans the image before it draws. It can search the web for references, reason through layout, and double-check its own output. Text on images renders correctly. Characters stay consistent across multiple images from one prompt. Non-Latin scripts (Japanese, Korean, Chinese, Hindi, Bengali) actually work now.

Instant mode is on the free ChatGPT tier. Thinking mode comes with Plus.

For educators, this is the release that turns AI image generation from a novelty into a real teaching tool. Slides with readable labels. Consistent cartoon characters across a whole lesson. Language-classroom materials in the actual target language.

Here's my discussion question for today:

What's one thing you've tried to make with AI image generation that turned out faster to do by hand? That's the exact problem this new model is built to fix, and I want to collect real teacher workflows we can stress-test.

Drop them below. If you run one against ChatGPT Images 2.0 this week, come back and tell us whether it worked.

Plant ideas. Cultivate skills. Harvest results.