Nanowrimo 2022 Project

Before we wrap up this chapter and move on to the next segment of this book, I wanted to discuss how this book project evolved. I have mentioned previously that the idea of writing this book got formalized around October 2022. A few days later, the month of November arrived. During November, the Annual Nanowrimo (National Novel Writing Month) competition takes place globally. The aim of this competition is to encourage authors to write one novel of atleast 50,000 words (or about 1,800 words per day) in November.

With that goal in mind, I almost put off the idea of writing this book. My main goal was to complete a science fiction book revolving around Climate Change. The idea of this book has been on my mind since 2016, and every other week I used to pester my wife (my biggest supporter and my greatest critic) about Shatabdi, the working title of this fiction book. Over the past few years, the plot of the book was ready in my head, but that had resulted in zero words being written. After an ultimatum form my wife, I decided to use a rather creative approach. (To put it mildly, she said," If you do not complete writing your book in November, don't you ever mention that book to me again.")

I booked a recording studio for the entire month- with the aim that I would record the book from Monday to Friday, and use speech to text to transcribe the novel. With much apprehension, the fiction project began promptly on November 1st. By the middle of the month, I had recorded and transcribed nearly 80,000 words! Which meant that I still had about 20 hours of recording studio time that could be put to use for other projects. Which led to the creation of this book.

This book is a the first of the 4 part series of books I am writing about images and content. The themes for the other books include Image compression and optimization, Content Delivery Networks, and Content Management Systems.

#Back to this book

Once I started recording the chapters for this book, in a couple of weeks' time, I was able to make some significant headway. Around the first week of December, I was able to transcribe and structure the book, and by the middle of December, a test site was created.

I was quite excited by the pace at which things were taking shape, this inspired me to do a soft launch of this book, and et some early feedback. Based on that feedback, I decided to update some sections, and moved from brick.do to typemill, a PHP based Content Management Systems that used Markdown.

I hope that this site will server as a beneficial resource for those looking to take a deep dive into the world of AI Imaging.

Because this field is evolving so rapidly, the idea of a "book website" makes more sense to me at this point in time, so that the content can be updated as and when new developments keep taking place. I might release this book as an eBook, particularly for the non-English versions.

My main intention behind the print or eBook is to reach out to the non English language speakers, particularly Hindi and Marathi, and empower them with a resource that they could benefit from. More on that in a subsequent update.

#Timeline: How the idea for this book evolved

I mentioned at the beginning of this section that the idea for writing a book on AI Imaging took shape around end of October 2022. Over the past six or seven weeks, the following steps led to the creation of this book and site.

1. During #nanowrimo2022, I voice recorded the chapters in a recording studio, and then used otter.ai speech to text to transcribe. After editing the script, the first draft of this book took its current form. This may seem overly simple, but speaking into the microphone was the easiest part in my experience. Editing was tedious and rather painful- imagine having to listen to your own voice multiple times, and reading your own words multiple times too!
2. During the months of October to December, I generated hundreds of images to understand what commands or prompts work for my intended purpose. My goal was to create cover art for podcast episodes for Baalgatha and Fairytales of India podcasts. I also wanted a distinct "India" theme, because our stories have a distinct India connect. Therefore, the sights, people, seasons, food, buildings... all had to reflect Indian-ness in some way or form.

Image of a castle and a temple on a hill, created using art.elbo.ai

3. One of the very first images I generated was for an episodes of Fairytales of India podcast. I wanted an image to complement a story involving a Merchant who lived in a palatial home in a valley, and the Royal Palace in the background. i was unable to get the right look and feel. Maybe the tool was still in a learning phase, as indeed I was myself. That's when I decided to change the prompt that resulted in the above image. This represented another element of the story quite well.
4. Next, I curated a list of sites and apps where one could generate AI driven images (Nearly 100 such sites and counting at the time of writing). This list includes services where once can generate AI images for free, and you will also find paid and freemium sites. I have included this list in Chapter V.
5. During my process of experimentation, which I consider a tremendous learning experience, I identified many use cases for AI imaging. Some of these include architecture, book cover designs, and so on. Some of these use cases have been listed in Chapter III.

6. I also spent a lot of time understanding different painting styles and art forms across different regions of India. This includes Pahadi, Pichwai, Madhubani, Patachitra and many such art forms. Also tried to understand some technicalities and processes used in photography, such as focus and lighting, as well as different materials such as glass and terracota, colours and inks, and textures. I also learnt about painting styles from Japan. Below is one such example.

Much of my exploration around art form was a self motivated and self learnt deep dive. I would not recommend to others to go down that path- a structured course may offer a far better experience and value for time invested.

7. Initially, it was difficult for me to understand many of the technical terms used in art, or even art styles. But over the weeks, I could see more and more times words like Volumetric Lighting, or design styles such as Art Deco or Artstation. And artists like Greg Rutkowski and Artgem. This prompted me to create a Glossary /definition of terms.

Initially, I was hand curating the definitions of terms from sites like dictionary.com and wikipedia. But in December 2022, I decided to replaced the definitions with replies from ChatGPT.

You will find these terms in Chapter IX of this book. I might add more sources in a future update.

Chapter VIII is the only chapter in this book where I have used AI tools as a writing aid. The majority of the content in that chapter was generated using ChatGPT, but the first section of that chapter remains written by yours truly.

8. Finally, I recorded the summary of this book in Marathi मराठी and Hindi हिंदी language. My focus was to prepare short summaries at this time. Time and resources permitting, my aim is to translate this book in these languages. Hopefully that will happen in the near future. The prompt for the below image as given in Marathi language.

As an idea, since audio for this book was recorded in a studio, the voice quality is pretty good. This makes me wonder if I should create short videos with voice overs. This will enable me to create a "visual book" for a "visual medium" , i.e. visual arts. This idea could be developed at a later time.