Not a Photoshop killer — but dangerously close

A few years ago, I was stuck.

Our designer was out sick.

I needed perfectly polished graphics for a major client.

And I... was left alone with Photoshop.

What followed was hours of dragging, masking, redoing layers I didn't understand, coffee, countless YouTube videos, and tutorials.

If Gemini Flash Experimental had been around back then?

I’d have typed “Brighten the visual, sharpen the logo,” and gone home early.

That’s the shift happening now.

And it’s only been a few days.

In Case You Missed It

Launched on March 13, 2025, Gemini Flash 2.0 is Google’s new AI-powered image generation tool — experimental, fast, conversational. You can try it for free here.

Unlike Photoshop, it doesn’t need years of training.

You just… ask.

Like texting a friend who knows lighting, composition, design, storytelling — and delivers everything you ask for. No sleep required.

Why I Wrote This Letter

Honestly, I wasn’t sure whether to write about it. There’s already so much noise. If you’re reading this, you’ve probably seen it all over your feed —

Maybe even tried it yourself. But here’s what made me commit:

This is a shift.

And I’ve only felt that a few times before.

• ChatGPT (2022)
• Prototyping apps on my phone with Claude (2024)
• NotebookLM (2024)
• And now… Gemini Flash (2025)

So I figured: this deserves a letter.

Not to hype it. But to mark it.

To show what’s possible — even in its early, imperfect form.

I thought maybe I could give you something deep. A prompt formula. A secret technique. Something smart.

But the truth is — it’s too early for that.No secret sauce yet.

So what’s the best way to learn this tool? Play with it. Explore. Break things. Get weird. See what happens. I decided to give you some inspiration.

I’ve pulled together 19 examples of what people are already doing with Gemini Flash.

Things that sparked curiosity, ideas, or “Wait, you can do that?”

And beyond the examples —

I want to share why these things matter.

What they mean.

What are the limitations.

How they might change how we work, create, or just play.

Take a breath.

Let’s dive in.

Inspiration: Stuff People Are Already Making

Let’s start with the fun stuff.

1. Fantasy Worlds

“Make a gritty dungeon with flickering torches and rusty chains.”

2. Old Photo, New Life

Upload a scratched family photo. Say:

“Restore this.”

3. Posture Change: PS vs Gemini

Photoshop vs. Gemini?

A user changed a person’s posture in an image in 5 seconds using Gemini.

Photoshop took 15 minutes.

4. Storyboarding with Visual Consistency

“Show a pirate on a ship, then on an island fighting a tiger!”

5. Instant Room Redesigns

“Make the furniture go away, then decorate it with a modern chic aesthetic.”

6. From Anime Doodle to Masterpiece

A user sketched a rough anime figure. The creative potential is very high.

7. Put a Product on a Model

“Let the model wear this jacket, same pose.”

8. Motion Made Simple

Pair Gemini with ImageFX or Sora, and you’re making ads.

One small business claim to have made a full 15-second commercial in an afternoon.

Zero crew. Zero budget.

‍

9. Changing Angles Made Simple

“Make a side view of this photo” is enough.

Not bad with so little context.

10. Zoom Out Made Simple

“Make a zoomed out view of this photo to show his full body.”

(Doesn’t work without the “to show his full body” part).

Incredible. My wife said this one looks almost exactly like me.

‍

11. Colorizing Black & White Photos

“Color this image, realistic, HD quality.”

Interesting.

12. Removing Watermarks

This poses a significant threat to the business models of all image libraries.

13. Swapping People Like Magic

Powerful isn’t it.

14. Tell Illustrated Stories with Consistent Character

This is a killer prompt: “Generate a cartoon story about a group of wizards fighting in a alien planet. For each scene, generate an image.”

Honestly, this is by far my favorite use case. The ability to create comics is absolutely incredible.

15. Passport Photos at Home

This application is intriguing, though I have concerns about whether the results meet official passport photo requirements.

16. Quick Edits for Non-Designers

This is one of the few cases where I’m personally still using Photoshop extensively. I create images using Midjourney or Replicate, and while I'm happy with a specific result, they often have typical AI flaws—wrong number of fingers, mixed-up details, and so on.

I used to spend time fixing these issues in Photoshop.

Now it's just a prompt. Life-changing for me.

17. Everyday Retouching, Minus the Fuss

Blemishes, bad backgrounds, lighting fixes — all handled in plain English.

19. GIF Creation

Here's a powerful workflow for creating animations:

Midjourney + Gemini + Runway

Here's how it works:

• First, create your base image in Midjourney
• Then use Gemini to create different poses
• Finally, bring it to life with Runway's animation tools

What Gemini Really Brings to the Table

Let’s get real.

People are throwing around “Photoshop killer” like it's 2009 again.

It’s not.

But it's not nothing.

Here’s what’s actually impressive:

• Visual Consistency
• Scene Expansion & Iteration
• Prototyping and Product Use Cases

Prompting Tips

If you're building a story or trying to maintain style:

• Start with: “Generate same scene, but…”
• Upload products one at a time
• Keep your prompts specific and structured
• Don’t over-edit. Quality degrades after too many tweaks.

Limitations (Let’s Be Honest)

• Text in images? Still hit-or-miss.
• Complex edits? Falls apart fast. It handles the first prompt well, then starts drifting.
• Image memory? Weak. Context fades as the thread continues.
• Too many changes? Quality drops with each one.
• Bird’s-eye views? Inconsistent at best.
• Hands and faces? Often close, rarely perfect.

Bottom line — it’s experimental.

And it’s not replacing any serious image editor just yet.

Business vs. Casual Use

Everyday:

• Create social content
• Make flyers, posters, memes
• Tweak selfies
• Restore family photos

Professional:

• Visual storyboards
• Ad mockups
• Product shoots
• Style guides
• Visual scripting for AI videos

You don’t need a team to do this anymore.

One person. A good prompt. That’s it.

Try It Yourself

Here’s how:

1. Go to Google AI Studio
2. Sign in
3. On the left: Create Prompt
4. On the right: Pick Gemini 2.0 Flash Experimental
5. Output: Text + Images
6. Upload your image
7. Have fun

Bottom Line

Photoshop isn’t going anywhere.

But Gemini just made high-level visuals democratic.

Moreover, this is just the beginning of "Image Editing" AI features, xAI is catching up (and others will definitely do):

You don’t need the skills. Just ideas.

So — which example blew your mind?

Personally?

I can't stop playing with #14, generating comics. It's absolutely addicting!

Until next time,

Keep creating.

— Charafeddine