The Secret AI Prompt Hack: Chrome Split View 🤫

If you’re prompting tools like ChatGPT or Gemini every day, I want to show you something simple that changes how you use them.

Instead of working with one model at a time, you can use Chrome’s Split View to run the exact same prompt in both, side by side, and compare the results instantly. If you want to watch the video tutorial on this, scroll to the bottom.

Let me explain how this works.

If you are in ChatGPT. To turn on Split View, go up to the tab, right-click, and select “Add tab to new split view.” Chrome opens a second pane next to it. In that new pane, I load Google Gemini.

Now I’ve got both tools open, side by side.

For this demo, I’m using a straightforward prompt. In practice, you’ll want yours to be detailed and specific. I’m asking both models for image concepts based on a Viking-style quote for an AI TikTok channel:

I paste the same prompt into both tools and hit submit.

You’ll notice I also asked for chain-of-thought reasoning. I want each model to outline how it interpreted the quote and how it arrived at its final ideas.

Right away, you will see differences.

ChatGPT responds quickly and gives a structured, step-by-step breakdown of its thinking. Gemini takes a slightly different approach, more direct and streamlined.

Then we get to the concepts.

ChatGPT gives three detailed image concepts, with extra context and even draft-style prompts for image generation. Gemini also gives three concepts, but the framing and interpretation differ.

Now I’m not looking at three ideas. I’m looking at six.

Six distinct creative directions, all from the same original prompt.

That’s the point.

Even when you use identical wording, each model interprets and responds in its own way. The structure, depth, tone, and suggestions vary. When you compare them side by side, patterns start to emerge. You see which model goes deeper. Which one is clearer? Which one fits your goal better?

Or you blend them.

Maybe ChatGPT gives a stronger structure, while Gemini offers sharper imagery. Maybe Claude or Copilot handles analysis better for a different task. The only way to know is to test them against each other.

If you normally stick to one tool, try this. Open Split View and run comparisons. Over time, you’ll build a sense of which model performs best for writing, strategy, ideation, analysis, or creative work.

Sometimes one stands out. Other times, the best answer comes from combining outputs.

That’s today’s Digital Spark.

Don’t limit yourself to one LLM. Test ChatGPT, Gemini, Claude, Perplexity. Compare them. Mix their strengths. See what fits your workflow.

Read all my Digital Sparks here.