AI EXPERIMENT
AI:Your Rapid Concepting Engine.
Break Creative Blocks: The Abstract Prompt Strategy.
10 Platforms, 10 Styles: Unlocking Aesthetic Bias.
Workflow Streamlined: Pinpointing the Fastest AI Tool.
Creative Exploration using multiple AI’s
Elevating the Creative Practice
The goal is no longer just to create, but to create with exponential efficiency. This comprehensive 10-platform AI generative experiment represents a foundational step toward that aspiration.
By systematically testing, benchmarking, and mastering the unique capabilities of tools from DALL-E 3 to Stable Diffusion variants, the objective is to transcend current workflow limitations. To convert time-consuming manual processes (like concepting and asset hunting) into seconds, ultimately allowing more bandwidth for high-value strategic thinking.
This pursuit is driven by the commitment to achieve Prompt Mastery, unlock unparalleled creative frontiers, and establish a data-driven framework for selecting the perfect AI tool for every design challenge.
01
Rapid Concepting and Brainstorming
The core benefit is generating a high volume of diverse visual concepts quickly.
Designers can use the AI platforms to visualise initial ideas, mood boards, or multiple design variations in seconds, saving hours of manual sketching or searching for stock imagery.
6
6 powerful models lead the industry
A handful of models dominate the landscape like Midjourney, Gemini and DALL – E3. Open source models like Stable Diffusion mean the numbers have exploded on 2025.
02
Style and Aesthetic Discovery
By using the same prompt across 10 different platforms, designers can discover the unique “style bias” of each AI model (DALL-E 3, Firefly, Stable Diffusion variants).
This helps them learn which platform is best suited for a specific aesthetic (e.g., abstract art, hyper-realism, digital landscape). Abstract Art Example Prompt is specifically designed for this.
Pushing Creative Boundaries
Use highly abstract and metaphorical prompts;
- Learner evolving and expanding their skills
- Linking neurons and grids
These encourage the AI to interpret ideas creatively, often generating visuals a human designer might not have initially considered. This can break creative blocks and lead to novel solutions.
03
Technical Proficiency
Prompt Mastery
The experiment directly tests and refines prompt engineering skills. Designers learn what specific keywords, style descriptors, and technical instructions (like aspect ratio: 9:16, or specific themes like “3D flavours and mobile payment”) yield the best results on different models.
Platform Benchmarking
Testing 10 platforms allows the designer to evaluate and compare:
- Image Quality and Fidelity – Which free platform offers the sharpest, most professional-looking output.
- Prompt Adherence – Which model best follows complex instructions and constraints The two prompts requiring a ”complex sign object with no words and a transparent background” are perfect for testing adherence to specific technical requirements.
- Credit/Cost Efficiency – For free tiers, understanding the credit limits
- Canva’s 30-day limit
- NightCafe’s 20 max credits
- OpenArt’s 40 max credits*
- *Good-to-know information for a designer on a budget to prioritise tools at different project stages.
- Workflow Integration – Identifying which platform is the fastest and most reliable for specific tasks (like generating icons, textures, or full scenes) allows a designer to streamline their workflow, using AI tools for repetitive tasks and dedicating more time to high-value creative work
04
Practical Use Case Testing
Specific Design Asset Generation
3.1
Backgrounds
Using the 9:16 aspect ratio for backgrounds.
3.2
Icons/Objects
Generating complex 3D icons for app or web design with transparent backgrounds
3.3
Branding/Mood
Creating a visual aesthetic for a lab/studio environment using specific color and style constraints.
3.4
Video Prototyping
Create complex transition between two images with rapid, timed flashes, which is critical for motion graphics.
This allows the designer to test the limits of what these free AI tools can produce in the video space.
Structured an layered approach
[CONTENT] The Subject:
- What is the asset? Clear, unambiguous object or scene.
- i.e. a minimal settings icon, a 3D rendered coffee cup
[CONTENT] The Function/Environment:
- Where is it being used? Defines placement and purpose
- i.e. on a clean white background, floating in a dark void, for a mobile banking app
[AESTHETIC] The Style/Medium:
- How does it look? Defines visual style, medium, and artistic inspiration.
- i.e. flat vector illustration, Material Design 3, neumorphism style, etc
[TECHNICAL] The Refinement/Quality:
- How is it composed? Defines quality and resolution
C-C-A-T
Diffusion Models
Prompt Engineering
Finetuning
RLHF
Visual Coherence
Microsoft Designer / Bing Image Creator
Powered by DALL-E 3, these tools are free to use with a Microsoft account and provide image generation directly from text prompts.
Reinforcement Learning from Human Feedback (RLHF): A powerful machine learning technique that aligns AI Large Language Models (LLMs), with human values by incorporating direct human input into the training loop,
This teach AI what humans find helpful, safe, or desirable.
Overall Best For Literal Translations: This generator didn’t cross the line and ramp-up the creativity. Which was important in a post about Creative Exploration.
Generative AI
Creative Workflow
Text-to-Image
Commercial
Adobe Firefly
Adobe’s AI tool offers free access and monthly generative credits for creating images and text effects from text. It’s known for its focus on ethically sourced training data.
Visual Content Strategy
Brand Consistency
Prototyping
Template Customisation
Collaboration
Canva
A popular design platform that includes AI image generation features (Magic Media and Dream Lab, powered by various models like Leonardo.ai, DALL-E, and Imagen). Free accounts typically get a good number of free image generations.
Canva: A popular design platform that includes AI image generation features (Magic Media and Dream Lab, powered by various models like Leonardo.ai, DALL-E, and Imagen).
Free accounts typically get a good number of free image generations.
Overall Best For: Very detailed which contributed to the believe-ability factor. Also succeeded in spiking my imagination, and in-turn fixed my creative vision.
A double-win!
Model-Centric
Asset Creation
AI Canvas
Leonardo AI
This platform is known for its generous free plan, offering a significant number of tokens (credits) daily to generate images. It also provides various models and fine-tuning options.
Model-Centric
Asset Creation
AI Canvas
Craiyon (formerly DALL-E mini)
A completely free and easy-to-use AI image generator. While the quality might not always match the top-tier paid options, it’s excellent for experimentation and quickly generating a wide variety of images.
Canva: A popular design platform that includes AI image generation features (Magic Media and Dream Lab, powered by various models like Leonardo.ai, DALL-E, and Imagen).
Free accounts typically get a good number of free image generations.
Typically adopting the credit-based system the complexity/size of the image determines the credit cost per generation.
Free accounts typically get a good number of free image generations.
Model-Centric
Asset Creation
AI Canvas
DreamStudio
(Stable Diffusion)
Developed by Stability AI, using the Stable Diffusion model, this text-to-image diffusion model is targeted for everyone, from hobbyists to professionals.
Model-Centric
Asset Creation
AI Canvas
Freepik AI Image Generator
Freepik provides a free tier with a certain number of credits for AI image generation, along with a vast library of other design assets.
Canva: A popular design platform that includes AI image generation features (Magic Media and Dream Lab, powered by various models like Leonardo.ai, DALL-E, and Imagen).
Free accounts typically get a good number of free image generations.
Canva: A popular design platform that includes AI image generation features (Magic Media and Dream Lab, powered by various models like Leonardo.ai, DALL-E, and Imagen).
Free accounts typically get a good number of free image generations.
Model-Centric
Asset Creation
AI Canvas
ImageFX
Google’s standalone AI image generator, which uses Imagen 3. It’s free to use with a Google account and offers expressive chips to easily modify prompts. You can also access Imagen 3 through Google Gemini.
Model-Centric
Asset Creation
AI Canvas
OpenArt
OpenArt provides a free plan with daily credits and access to various AI models, including Stable Diffusion XL. It also offers advanced features like image-to-video conversion and character consistency. 40 max creditsº
Canva: A popular design platform that includes AI image generation features (Magic Media and Dream Lab, powered by various models like Leonardo.ai, DALL-E, and Imagen).
Free accounts typically get a good number of free image generations.
Model-Centric
Asset Creation
AI Canvas
Canva
A popular design platform that includes AI image generation features (Magic Media and Dream Lab, powered by various models like Leonardo.ai, DALL-E, and Imagen). Free accounts typically get a good number of free image generations.
Canva: A popular design platform that includes AI image generation features (Magic Media and Dream Lab, powered by various models like Leonardo.ai, DALL-E, and Imagen).
Free accounts typically get a good number of free image generations.
Leave a Reply