If you’re thinking of Gemini 3 as a “Google ChatGPT,” you’re not going to be tuning your prompts to get the most out of the model.
The truth is these two models are both strong, but they have different specialties:
Gemini 3 is great at long messy context with video and images and text all thrown in, while ChatGPT excels as a “brain in the jar” that processes complex tasks but prefers a very clean workspace.
This means OUR job as the prompter is very different—we need to be better at naming the mess with Gemini, and we need to be better at defining the task with ChatGPT.
Curious to learn more? I did a full breakdown here, plus I wrote a prompt configurator that converts your prompt into a Gemini 3 OR ChatGPT friendly format (your choice). This is a SUPER practical post, so hop in and have fun!