Introduction: Why and Who is this post for?
This post is the first part of a series of posts that will document my journey of navigating the strange, almost hallucinogenic, world of Generative AI, i.e. “GenAI” which I will compile into a book.
I've decided to journal my experience starting https://delights.ai/ to help people who are looking for a change in their career, have recently been laid off, or are executives looking to incorporate AI into their Corporate Strategy.
I’m mainly writing this for myself to document my personal journey in trying to understand GenAI vs “old-school AI” which is what I had grown up with. For me, old-school AI is Artificial Intelligence as it was introduced to a 9-year-old me as a naked Arnold Schwarzenegger in “Terminator,” my all-time favorite movie. Terminator aka "old-school AI" and GenAI are obviously very different worlds. Being Gen-X, I didn’t understand GenAI and I was motivated to get to the bottom of what, how, and the why of GenAI because I realized that it is one of the fastest-growing industries/technologies I have ever witnessed since participating in the growth of the internet in the 90s. I like being ahead of the adoption curve but this curve is a hockey stick and evolving so rapidly there is no way to be ahead of this trend. I would google something around GenAI and the next day there would be a new company popping up filling the space, and then 5 more the day after.
In that light, I hope to save some time and money with this post for anyone looking to learn more about whether they should start a business or company in GenAI, produce content using GenAI, or pivot their business/company strategy to include GenAI.
If you are reading this because you are trying to decide whether or not to pivot the corporate strategy to include GenAI, I highly recommend that you do it. Pivot! Pivot now!
Day 1: The Wonderful World of Alice - A Simple Explanation of GenAI
The Difference Between AI and GenAI
Artificial Intelligence (AI): This term encompasses the broader field of artificial intelligence, covering various methodologies, algorithms, and applications. Traditional AI development involves writing code, training models, and optimizing algorithms to perform specific tasks.
Generated AI (GenAI): Referring to "Generated AI," GenAI focuses on AI systems capable of generating content autonomously. These systems often leverage large datasets and advanced machine learning techniques, such as Generative Adversarial Networks (GANs) or Transformer models, to create text, images, or even music.
My experience with GenAI, started about 2 years ago when I signed up to join Midjourney’s invite-only beta program in 2022. I was moved to join and try it out after Midjourney’s “Théâtre D'opéra Spatial,” using prompts by Jason Michael Allen, won the 2022 Colorado State Fair's annual fine art competition in the photomanipulation category on September 5, becoming one of the first AI-generated images to win such a prize.
My kids, at the time, were 8 and 12 years old and we had a lot of fun with Midjourney when I was finally let in to play with it.
Here are some examples of the prompt used back in 2022.
The following two images were made by my 8-year-old at the time using this simple prompt: “astronaut on hoverboard in space miro style”
The following two images were made by my 12-year-old at the time using this simple prompt: “man wearing flowering deer skull”
Here we are two years later, in 2024, and I entered the same prompt into Midjourney to see what it would produce today after years of being live and in the wild with the AI model now being trained by the public. “man wearing flowering deer skull”
“astronaut on hoverboard in space miro style”
Seeing these new prompt results was very disappointing. I majored in Electronic Arts at UCSB, and my eye has been trained on what is quality art and seeing the degradation of the images from something beautiful with an artistic eye back in 2022 with a clear artistic approach, (from the way the light is centered on the High-Def Deer Man and the clear interpretation of Joan Miro in the Astronaut) to these murky, non-artistic and non-Miro images was pretty sad to see. It’s like watching an artist lose their skills. The quality left much to be desired, in my humble opinion. If you compare the image of the low-res version of the original “man wearing flowering deer skull” to the one generated today it is crazy what has happened to this AI model. Even more apparent with the prompt “astronaut on hoverboard in space miro style” which lost all of its understanding of the style and art of Joan Miro. Before it cleverly interpreted the artist’s style and made something of its own. Today it is generating something that it thinks people might want. But based on what? Either, the way it is being trained, or absorbing more “art” from a larger dataset has degraded its quality. Quantity over quality seems to be killing it.
I wanted to see what Dall-E would produce with the same exact prompts and here is what I got. “astronaut on hoverboard in space miro style”
“man wearing flowering deer skull”
Maybe it produced images with a little more artistic quality? At least the image with the astronaut tried to capture the artistic elements of Joan Miro.
However, I didn’t start this journey to generate images. I wanted to understand this movement of GenAI and what was moving it forward. I wanted to understand the underlying technology, who was creating what, how, and what could truly be done with it. I was drawn more to the creation of the generative process.
I wanted to see if I could generate my own little AI using no-code platforms and with the emergence of platforms like Youai.ai from MindStudio, delving into AI creation has become more accessible, thanks to its innovative no-code approach. I'll share my experience on several of these platforms in the coming days.
The next post "Day 2 - Navigating the Labyrinth - The Art of Prompt Engineering" will drop in a few days stay tuned for that and the following posts:
Day 3: How do you monetize any of this?
Day 4: Charting the Depths - A Comparative Analysis
Day 5-10: Coming up with the idea
Day 11-15: The Business Strategy
Day 16-20: Coming up with a Go-to-Market Strategy
Day 20-30: Finalizing the Execution Plan
Bonus - Day 31+ Writing this with the help of GenAI - A Comparative Analysis of Outputs
Full disclosure, I did get help from Notion on writing this post. I'll talk about this more in a follow up post
Stay Tuned!!