Quality Video Generation, Layer Split & JSON Prompting

June 18, 2026

Hello Friend,

Welcome back to the Generative Path, your weekly infusion of AI-powered creativity curated by ZenRobot to keep you ahead of the curve with the latest tools, trends, and techniques.

This week explores how AI is giving creators more control over every stage of the creative process. Grok Imagine 1.5 continues to blur the line between image generation and filmmaking by transforming still images into cinematic videos with motion, sound, and camera movement. We're also sharing a simple JSON prompting technique that produces more consistent results across image generations, spotlighting creative director Amandine Leblanc-Hamel and her AI-powered storytelling, and taking a look at Magnific's new Layer Split feature, which brings non-destructive editing workflows to AI design. Add the atmospheric sounds of Mechagnome Radio to your creative sessions, and you've got everything you need for a week of focused experimentation and visual exploration. Let's get into it.

Feature of the Week: Grok Imagine 1.5

Grok Imagine 1.5 is pushing AI creativity beyond static images. Built by xAI, it combines image generation and image-to-video capabilities, allowing creators to generate visuals from prompts and then bring them to life with motion, sound, and cinematic camera movement.

The latest update improves image quality, prompt accuracy, and generation speed, making it easier to turn ideas into polished visual content. For creatives, the real value lies in control. Start with a single image, then animate it with camera moves, character actions, ambient sound, and music.

Whether you're creating concept trailers, social content, storyboards, or visual experiments, Grok Imagine 1.5 helps bridge the gap between still imagery and video production.

It is another sign of how quickly AI tools are turning individual creators into full creative studios.

Creative Hack: Consistent Styles With JSON Prompting

Getting inconsistent results from Nano Banana or GPT Image? Try separating style from subject using a JSON prompt.

Here's the workflow:

  • Start with a structured JSON prompt
    Create separate sections for your style, subject, and constraints. Something like:
    "style": { "medium": "digital illustration", "art_style": "painterly", "inspiration": "mid-century editorial illustration", "lighting": "warm golden hour light", "color_palette": "soft peach, dusty blue, muted cream"
    "subject": "man surfing a wave",] }

  • Keep the style block identical

    Reuse the exact same style settings across every generation.

  • Only change the subject

    Update the subject while leaving the style block untouched to maintain a consistent look.

  • Add clear constraints

    Include instructions like "no text" and "no logos" to reduce unwanted changes.

What makes this workflow powerful is that you're separating creative direction from image content. Keeping the style block fixed while changing only the subject helps reduce visual drift and improves consistency across generations For designers, marketers, and content creators, this is a simple way to create cohesive image sets for campaigns, product collections, social media, and brand assets.

Sometimes the smallest change to your prompt makes the biggest difference in your results.

Community Spotlight: Amandine Leblanc-Hamel

Amandine Leblanc-Hamel is a creative director and visual storyteller who brings over 14 years of interior architecture experience into the world of generative AI.

Working with tools like Midjourney, Luma AI, ElevenLabs, and SketchUp, she transforms narrative briefs into rich visual worlds, helping brands develop distinct identities through AI-powered art direction and storytelling. As founder of Studio 13 Mai and a creative partner to leading AI platforms, her work sits at the intersection of design, emotion, and emerging technology.

From concept art and moodboards to brand storytelling and creative education, Amandine’s approach demonstrates how AI can be used not just to generate images, but to create meaningful visual narratives that leave a lasting impression.

This Week’s Flow State: Mechagnome Radio

Step into the steady pulse of Mechagnome Radio on Spotify. Blending atmospheric electronic textures, understated grooves, and forward-thinking production, this playlist explores the intersection of focus, creativity, and modern sound. From warm ambient layers to crisp downtempo rhythms, every track is selected to keep you moving without pulling your attention away. Whether you're designing, coding, writing, or simply unwinding, Mechagnome Radio provides a seamless soundtrack that keeps ideas flowing and distractions fading into the background. Press play and let the music do the heavy lifting.

Magnific has introduced Layer Split, a deceptively simple feature that could have a significant impact on AI-assisted design workflows.

Instead of regenerating or rebuilding an entire image to fix a small detail, creators can now isolate and edit individual elements layer by layer across Designer, Spaces, and Magnific MCP. What makes this notable is the shift toward more professional, non-destructive editing.

Whether it's correcting text, adjusting a product, refining a character, or updating a background, Layer Split gives creators greater control while preserving the rest of the composition.

As AI tools mature, features like this move the workflow closer to traditional creative software, reducing friction and making iteration faster, more precise, and far less disruptive.

AI vs. Reality: Can You Tell the Difference?

This week, we’ve got a brand-new pair of images to test your perception again. Think you can tell which is AI-made and which is the real deal? Take your guess and we’ll reveal the answer in the next issue!

Last week, we kicked off our AI vs. Reality face-off with some images of Albert Einstein, and now for the moment of truth, if you guessed the right one was the real deal, you got it right. The left image was created with ChatGPT showing off the program’s impressive ability to generate hyperrealistic images.

Join the Conversation! AI is constantly evolving, unlocking new creative possibilities every day. What caught your attention this week? Hit reply and let us know. We’d love to hear from you.

Follow ZenRobot for more AI insights, and if you found this interesting, share it with someone who would too.

Stay curious,
David and Ross

Every shared newsletter helps more people explore creative AI with confidence. If this edition was useful to you, we'd be grateful if you shared it with a friend or colleague. Thank you!