My personal journey into the vast ocean of AI and Creativity: “Surfing human creativity with AI — A user’s guide” book is Here!
I’m proud to announce that my book, “Surfing Human Creativity with AI — A User’s Guide,” is now available on my Bandcamp page!
(2024.2.10 It is now available on Amazon.com and other major online bookstores as well.)
This book is a translated and revised version of my 2021 Japanese book on AI and human creativity. However, it goes beyond that and delves into my personal history and statement as an artist, musician, and founder of AI creative and music companies. It explores my journey of augmenting my limited musical creativity with AI capabilities and my aim to help others with the tools my team and I are currently developing.
In 1997, I encountered an artwork by CG pioneer Karl Sims at a digital art museum in Tokyo. Sims’s installation was based on the concept of evolutionary computing and artificial life. In the accompanying description, Sims stated his lack of understanding regarding the evolution of the virtual life forms in his simulated world: “I had no idea how these virtual creatures evolved, and I cannot design them manually myself.” This one sentence intrigued me a lot. As a junior computer science student at the time, I struggled daily with debugging and managing software behavior. Sims, on the other hand, viewed unexpected behavior as a feature rather than a bug. This realization made me understand that a computer system can be creative and surprise its designer.
Since then, my focus has been on creating AI/computer systems that surprise me and others.
Since then, my focus has been on creating AI/computer systems that surprise me and others.
Later, I discovered that this is the exact question that AI pioneers like Alan Turing and Ada Lovelace contemplated. While mentioning these historical figures, including arguably the world-first “AI Artist,” Harold Cohen, in the book, I also delve into various topics such as:
- How Jorge Luis Borges, an Argentine novelist/poet, provides new perspectives on creativity and AI
- TR-808 and Hatsune Miku on how machine imitation of artistic endeavors failed and yet flourished.
- The intersection of imagination, Japanese tea ceremony, and AI hallucination
- The risk of AI-based optimization and tales of peacock tail
- Why the world's best Flamenco dancer decided to perform with his AI simulacra
- Personal experiences of DJ-ing alongside AI in front of over 15000 people
- What is the lesson surfing gave me on the future of AI and creativity
After the release of the original Japanese book, I have given numerous talks on the subjects covered. In this blog post, I will use a few slides from those presentations to introduce some key takeaways from the new book.
Making “Mistakes” with AI
Here is a fact: AI is an excellent tool for making interesting and meaningful “mistakes.”
As an artist, I always strive to create something new and original. This often requires deviating from conventional ideas or “optimal solutions” and daring to make “mistakes” from a conventional perspective.
AI has no inherent understanding of common sense or norms (unless explicitly taught), and it operates on a logic that differs greatly from ours. This can lead to interesting deviations from common sense or what we may consider “mistakes.”
It is our responsibility as humans to identify interesting mistakes from mere random errors and maximize their unique aspects. This is where art truly lies.
Here is a quote I particularly like from Scott Adams, the author of Dilbert:
In this sense, I’m sure AI can be creative.
Surfing with AI
To maximize deviation, one must be in a fluid state. It is essential to actively seek out interesting new ideas, be open to unconventional thinking, and not cling rigidly to original ideas or conventions. One must be actively passive.
I have found this state of mind between activeness and passiveness in the art of surfing. When you surf, you proactively choose which waves to ride. Once you catch a wave, your body is propelled and carried by its power. You surrender to the wave’s force, which may take you in unexpected directions. However, at some point, you consciously decide to make turns and maneuvers.
This is precisely how I envision the future of AI and human co-creation. We must let loose, surrender, and follow AI’s suggestions, but we must always be the ones to make the final decisions. We should not surrender too much. There is a clear distinction between surfing and drowning.
When you saw the title of the book, you may have wondered why surfing. Now you know why.
(I found a quote from a guy called John McCarthy on surfing in an old Japanese surfing magazine. It goes as the following:
“Surfing is the most blissful experience you can have on this planet, a taste of heaven.”
In the magazine, he was introduced as a researcher who coined the term Artificial Intelligence, but I couldn’t confirm that the John McCarthy was a surfer indeed. If you know the truth, please let me know!)
The Importance of “Misusable” AI Tools
Throughout history, music and technology have often intertwined, with new technologies being misused by artists.
Consider turntables (record players) and samplers. Vinyl records were not meant to be touched on their surfaces, and samplers were originally intended to reproduce the sounds of expensive physical instruments, not repurpose someone else’s recorded materials. However, thanks to pioneers who found ways to “misuse” these devices and technologies in ways the original inventors couldn’t foresee, we now have Hip Hop and other music genres.
However, AI is often more challenging to misuse because it lacks a physical entity that can be manipulated and tends to operate as a black box. Artists can feel constrained by the technology and tools provided by large AI tech companies. This is why I started the Neutone project: to provide AI tools that artists can “misuse” in their unique ways. The premise of the Neutone project includes the right to train their own models with their own datasets. AI models used for creative purposes do not necessarily have to be well generalized. Artists do not need or want to create artwork in all styles. Their goal is to produce art or music in their own unique styles with AI.
(Note: I am not advocating the misuse of AI technology for malicious purposes, such as misinformation, discrimination, or military applications. I am referring to misuses in creative domains.)
This is why I started the Neutone project: to provide AI tools that artists can “misuse” in their own unique ways.
The Intertwined Dance Between Imitation and Creation
I have always been fascinated by the story of the legendary Roland drum machine, the TR-808. Initially, the TR-808 was introduced as a replacement for human drummers, allowing musicians to record demo songs even without a drummer in their band. However, the distinctive bass drum sound of the 808, partly due to the low-cost, somewhat defective transistors used for cost-saving, was nothing like human drumming. The sales initially flopped, leading to the model’s discontinuation. However, musicians found the unique sound intriguing and began using it in their productions. This eventually led to the birth of Hip Hop, techno, drum’n’bass, and other electronic dance music genres.
New creation often begins where machine imitation of human creativity fails. Let’s try hard to imitate what we do with AI, then see where it fails. There must be a hint for new creative potential for AI.
The book was expected to be available on Amazon and other online bookstores by the end of January 2024, but the review process is taking longer than usual. (I suspect the Amazon Kindle store is now flooded with substandard books filled with AI-generated texts.)
I am currently accepting orders for a limited run of 50 copies at a discounted price on my Bandcamp page until it becomes available on Amazon and other online stores.
I hope you will enjoy the book as much as I enjoyed writing it. Join me on the journey to discover new waves to ride with AI.
For more details on the book
Interview on the Gradient Podcast
I talked about the book and my projects in the last couple of years.
PS. I appreciate David d’Heilly for his tremendous effort in translating and editing the original Japanese text. When we started this project in 2021, I had no clue how challenging it would be. I learned that translating a book into other languages is not just about converting text from one language to another. It also requires considering cultural backgrounds and common senses where each language is spoken.
The book’s original designer, Hata Yurie, once again delivered an amazing book design. I am aware that dealing with my repeated requests was challenging. Thank you.
I thank BNN, the publisher of the original book, who kindly approved and encouraged the release of the translation.
I am incredibly grateful for the following individuals. Without you, I would not be able to complete this project:
Project Management: Sakiko Yasue
Cover Graphics: Shoya Dozono
Proofreading: Shizu Yuasa, Jeff Kim, Marcos Alonso
