My fascination with how artificial intelligence (AI) is reshaping our world led me to a unique project: observing how AI art evolves. I wanted to capture AI's creative abilities in a visual format. To start, I crafted a prompt that I can use periodically to visually observe how AI is changing. The prompt that I came up with was:

Draw a representation of yourself showcasing your creative abilities. Prominently display your name, software version, and the current date. Use maximum resolution in an aspect 3:1 ratio.

Choosing the Right Tool

Selecting the right AI tool for this task was crucial. I considered Midjourney for its ability to create wide images and upscale them for large prints. However, I felt Midjourey wouldn't quite showcase the evolution of AI. Don't get me wrong. Midjourney is amazing at what it does, but I don't expect it to become "self-aware" anytime soon.

I also thought about using Google's Bard, but it currently lacks direct image generation capabilities. So, I turned to the ChatGPT 4 interface to DALL-E 3, despite some of its limitations. This combination, while not perfect, promised a more intriguing approach to my project. And the result didn’t disappoint. Check out the image that ChatGPT generated:

Overcoming Resolution Limits

ChatGPT's resolution limits were a hurdle. To tackle this, I used the Topaz Gigapixel AI upscaler, a tool that promised better results than Photoshop. After buying Topaz for $100, I experimented with various settings, adjusting the height, resolution, and AI model to suit my print needs. I made sure to back up the original image before making any changes. Here are the Topaz settings that I used:

Aspect Ratio Challenges

Another issue with ChatGPT was its limited aspect ratios. I tried several methods to get the desired 3:1 aspect ratio, like cropping the image, but these attempts either cut too much of the original image or added unwanted space.

Reluctantly, I decided to create an additional image to fill the space in the wide print format. For the accompanying image, I used the following prompt, which generated the below image:

Generate an accompanying image that aligns to the right of the previous image. Use a 1:1 aspect ratio.

Preparing for Print

To combine the images for printing, I created a new Adobe Illustrator document with the following settings.

I used the following export settings. Note that I had to specifically set the resolution to 300 ppi to get the proper image resolution.

This is how the combined image turned out (downsized for this post):

I chose Mpix.com for printing. I ordered a giclee 12x36 on Fine Art Photographic paper; Double Weight Matboard.

Reflecting on the Journey

This project was more than just creating an AI-generated image; it was an exploration of the capabilities and limitations of current AI technology in art. From selecting the right tools to tweaking settings for the perfect print, each step was a learning experience.

Future Endeavors

Looking forward, I'm excited to continue this journey. I plan to create new AI-assisted images periodically, perhaps monthly, to see how the technology evolves and what new creative possibilities emerge. Each creation will be a testament to the blending of technology and art and a snapshot of AI's growing capabilities.

Conclusion

My adventure into AI-generated art has been a fascinating blend of technology, creativity, and learning. This project not only resulted in a unique piece of artwork but also deepened my understanding of AI's potential in creative fields. I look forward to sharing future creations and insights on this blog. Stay tuned for more updates on my AI art journey!

---

Through this blog, I hope to inspire others to explore the potential of AI in creative endeavors. Whether you're an artist, a technologist, or simply someone curious about the future of AI, there's a whole world of possibilities waiting to be discovered.