The Genesis of the 100-Day Challenge

Shortly after, multiple iterations of DALL-E and Midjourney emerged, alongside a host of new platforms unveiling their own AI image-generation models. The concept of ‘Prompt Engineering’ gained substantial popularity, a trend that continues to thrive.

Since joining X in 2012, I have relied on it as my main source of news, diverging from traditional media. My network comprises mostly software developers, tech aficionados, and writers.

Eager to experiment with prompting, I decided to embark on a ‘100 days, 100 images’ challenge. The goal was simple: generate an image daily using an AI tool and share it on X with the hashtag #100DaysOfAI.

What Exactly is Prompt Engineering?

In simple terms, Prompt Engineering involves crafting precise queries or instructions that guide AI models, especially LLMs, toward producing the desired results. It’s akin to programming but more accessible. Instead of learning complex languages like Java or Python, here the medium is plain English (or any supported natural language).

Many of us are already acquainted with such technology in our daily lives. Voice assistants like Siri and Google Assistant utilize natural language processing; you've probably employed your language to command them to play music or manage your shopping list.

Andrej Karpathy, former AI director at Tesla, even referred to English as the latest programming language.

Prompt engineering has emerged as an essential skill recently, acting as the link that ensures effective human-AI interactions. The new LLMs have shown significant advancements in grasping context and nuances, meaning your prompts can greatly affect the quality of the AI's responses.

Illustrating the Power of Prompting for Image Creation

This topic can branch into various areas, but my focus here is on how prompting functions, particularly in image generation.

Prompts yield different results depending on the specific AI tool and the finely-tuned model it employs. Even within the same platform, prompts can produce varied outcomes each time, which is crucial for creating unique images; otherwise, we would all end up with identical creations.

Let me provide some examples using the same prompt:

Example 1: The Double Exposure Art

Using Leonardo AI, a popular text-to-image tool, I prompted:

"A mesmerizing double exposure art piece in the style of Alex Stoddard on a worn, cracked paper, capturing a dynamic scene that evokes a vintage aesthetic, combining two subjects, revealing layers of complexity and intrigue."

The generated image was stunning. It depicted a woman envisioning a house, complete with a garden and a lake, showcasing the power of double exposure art.

Example 2: Refining the Vision

Dissatisfied with the previous output, I adjusted my prompt and generated a different image.

This rendition resembled a professional photographer's work, aligning more closely with the double exposure style.

Example 3: Small Adjustments Yield Big Changes

After another attempt with the same prompt, the artwork slightly varied again.

While visually appealing, this piece appeared more distressed than I intended.

Example 4: The Final Iteration

My last attempt produced an incredible image, which became my favorite.

This artwork adhered closely to my original prompt, effectively incorporating the elements of double exposure in a style reminiscent of Alex Stoddard. It's something I would proudly display as a poster in my home.

The essence of this exercise highlights how prompts can yield varying results even when using the same AI tool. Slight changes in phrasing can lead to completely different outputs. Therefore, to create images with AI tools, multiple iterations with the same or different prompts are often necessary to achieve the desired result.

The Journey of the #100DaysOfAI Challenge

The rules for this challenge were straightforward: utilize any leading AI text-to-image tools, where you input your prompt, and the tool generates images based on that context.

I successfully generated 100 images, which turned out to be a remarkable learning experience. The AI platforms I explored included LeonardoAI, Midjourney, Bing AI Creator, and NightCafe, among others.

I also dabbled with various text-to-video tools like Runway, Pika, and Luma Labs, but we’ll delve into those another time.

Throughout this process, I gained insights into diverse art styles and learned about artists spanning different eras and cultures, their photography techniques, famous paintings, and the art of prompt writing.

Here are some of my favorite creations along with their prompts:

Candid Black & White Capture of a Young Girl in a Field

Prompt: A young girl with flowing brown hair races against the wind, in a field of wildflowers, her laughter echoes, captured in a candid photograph, black & white image, raw emotion in her joyful expressions, enhancing the contrast of light & shadows.

Hyperrealistic Portrait of the Goat-man in the Mist

Prompt: A hyperrealistic close-up of a Goatman’s face in the mist, emphasizing his frightening features, the eyes glowing red, the face a mix of human and goat, with large pointy horns.

Vincent van Gogh Reimagined as Harry Potter

Prompt: “Self-Portrait with a Straw Hat” by Vincent van Gogh, reimagined as the “Harry Potter” movies.

Photorealistic Light and Shadow Capture

Prompt: Strong light and shadows, insanely realistic unclean skin with slight wrinkles and freckles, gray streaks in hair, vividly detailed eyes, photorealistic.

Neon Dreams Illustration

Prompt: Bright neon colors, cartoon-style illustration of a woman experiencing hallucinations, stoned, splash art, splattered neon colors, iridescent glowing smoke, motion effects.

8K Photorealistic Digital Portrait

Prompt: A meticulously crafted digital artwork, rendered with remarkable precision and attention to detail, portraying a man wearing spectacles, resembling a photograph of the highest resolution.

Sketchbook Portrait of a Nordic Woman Warrior

Prompt: Sketchbook style, hand-drawn sketch, realistic depiction of a stunningly beautiful Nordic Woman Warrior, complete with a long blade and Nordic symbols.

The Punching Master vs. The Rookie

Prompt: 50-year-old Mike Tyson and 25-year-old Jake Paul in a boxing ring, the crowd roaring with excitement as they fight for a championship belt, hyperrealistic.

Hyperrealistic Woman Selfie at a Party

Prompt: Selfie, self-portrait of a woman at a party, candid, ultra-wide-angle, distinct facial features illuminated by neon lights.

The Lego Mona Lisa

Prompt: Lego blocks artwork of “Mona Lisa” by Leonardo da Vinci.

If you found this article engaging, you may also enjoy the following pieces:

• How One Small Bug Took Down Computer Systems Globally, Causing “The Largest IT Outage” In History
• I Compared The Best 2 “Free” AI Image Generators & The Results Are Incredible

For more insightful stories that enhance your learning and living experiences, consider subscribing. If you found value in this article, feel free to support my work with a coffee—only if you can! Connect with me on X. Thank you!