Yoshimi Battles the Pink Robots, Pt. 1: Stable Diffusion AI Girls and Google Bard Tales (TTCB XII)

Yoshimi Battles the Pink Robots, Pt. 1: Stable Diffusion AI Girls and Google Bard Tales (TTCB XII)

😎 Can you elaborate a short tale from the lyrics of the song “Yoshimi Battles the Pink Robots” by The Flaming Lips?

“Yoshimi Battles the Pink Robots, Pt. 1” by The Flaming Lips – Original song lyrics and translations at LyricsTranslate.com

The Flaming Lips – Yoshimi Battles the Pink Robots Pt. 1 (Official music video)

Yoshimi Battles the Pink Robots by The Flaming Lips: an AI-Assisted Short Tale and Comic Strip

🤖 Sure, here is a short tale based on the lyrics of “Yoshimi Battles the Pink Robots” by The Flaming Lips:

Yoshimi is a young woman who lives in a city that is under attack by evil pink robots. The robots are programmed to destroy all humans, and Yoshimi is the only one who can stop them. She is a skilled karateka, and she is determined to protect her city and her loved ones.

Yoshimi trains for days, weeks, and months, preparing for her battle against the robots. She eats healthy foods, exercises regularly, and meditates to clear her mind. She knows that she will need to be at her peak physical and mental condition to defeat the robots.

Finally, the day of the battle arrives. The robots attack the city, and Yoshimi is ready for them. She fights bravely, using all of her skills and training. The robots are powerful, but Yoshimi is stronger. She defeats them one by one, until they are all destroyed.

Yoshimi is victorious, and the city is saved. She is hailed as a hero, and she is given a parade in her honor. Yoshimi is proud of what she has accomplished, but she knows that the battle is not over. There are still evil forces in the world, and she is determined to fight them all.

The song “Yoshimi Battles the Pink Robots” is a story of hope and resilience. It is a reminder that even in the darkest of times, there is always hope for victory. Yoshimi is a symbol of strength and determination, and she shows us that anything is possible if we never give up.

Text generated using Google Bard
Comic book strip composed with Stable Diffusion and some basic additional editing. Main text prompt: Text Prompts
"Marvel comic book strip with four panels featuring giant pink robots attacking Yoshimi, a female karateka wearing white kimono. Beautifully drawn backgrounds show a futuristic grey cityscape in turmoil. A masterpiece by Joe Jusko"
Comic book strip inspired by the song Yoshimi Battles the Pink Robots. Composed with Stable Diffusion and some basic additional editing. Click to access the base image on NightCafé Studio

😎 Help me compose a comic strip for the story of “Yoshimi Battles the Pink Robots”, the song by The Flaming Lips. Can you write 4 prompts for Stable Diffusion images, each being a panel of the comic strip? Make sure that the prompts are good for Stable Diffusion images and capture the drawing style of a popular comic book franchise such as Marvel or DC (this is normally included at the end of the prompt). Make sure that the 4 prompts are relatively similar to each other so the image styles are not too inconsistent. I will adjust all prompts and may play around with the seeds or some inpainting features before finalizing the works, but I would like the prompts provided to save as much time as possible. Can you also write captions for each panel?

🤖 Certainly! Here are four prompts for Stable Diffusion images along with captions for each panel…

✍️ I’ve tried assistance by several chatbots (GPT 3.5, Bard, Bing Chat…) in writing Stable Diffusion text-to-image prompts. While most of these chatbots can help us gather ideas for keywords and write some decent text-to-image prompts, the iterations and results are not as good and relevant to post them here. I ultimately didn’t build a comic strip as idealized in the “😎” paragraph above, which would be time consuming and require too many iterations and a combination of traditional graphic design tools and text-to-image. Instead, I’m sharing several results of images, comic panels and concept art I’ve created and selected with the inspiration of the song and Yoshimi’s tale from Google Bard. You can click on the image below to access a full collection of pictures I created on NightCafé:

Amidst the backdrop of gleaming skyscrapers in a high-tech cityscape, a colossal pink robot takes center stage. Its arrival is nothing short of awe-inspiring and terrifying. The robot’s arm transforms into a conduit of raw power, emanating a dazzling pink laser beam that pierces the urban landscape. As citizens scramble to evade the chaos, an explosion in the background signifies the colossal force that has been unleashed. SDXL1 Image on NightCafé Studio. ChatGPT captions.

I believe there is significant potential for Large Language Models (LLMs) to be trained on generating text-to-image prompts and unleash an even more powerful wave of high quality AI-generated artwork. Based on my own very basic experimentation, rudimentary training some chatbots by concatenating multiple prompts and feedback can yield mild improvements in the AI-imaging creation process. While this was an interesting experiment, the best way to learn text-to-image is still practicing a lot and looking at other people’s work. Although anyone can install Stable Diffusion on a GPU-equipped computer, I always recommend joining a community where people share their work, such as NightCafé Studio, or searching online for prompts other people use and share on websites such as PromptHero. If you are new to AI-imaging and to this blog, I also recommend reading the earlier post on AI Art.

Looking for an Alternative Yoshimi: Stable Diffusion AI Girls

AI imaging is one of the favorite topics of this blog, which might not be very interesting to read without the artwork that I create with Stable Diffusion. My first blog was titled “Do Not Read This Blog“, shortened as “SNOB”, which is a subanagram of the title in Spanish. The sarcasm of the title then (2006-2008), in an era when social networks as we know them today didn’t exist and generative AI wasn’t even known by that name (by the few lab folks who knew something about that), might be taken seriously now if this blog had no AI imaging to support text. There is just too much competition for content and attention on the Internet, so using mixed media in blogs is not only a choice, but almost an obligation, unless your blog is honestly titled “Do not Read this Blog”.

A confident martial artist in a white karate gi with a red emblem, poised for action. SDXL 1 image. Click to access original creation and more artwork on NightCafé Studio

Let the image above be an introduction for one of the trendiest AI imaging topics of today: text-to-image model-generated portraits of adult female subjects or, using popular search engine and social media keywords, AI girls.

Google Trends search interest statistics for “AI boys” and “AI girls”. Worldwide. Last two years. Click for interactive chart.

I certainly enjoy creating images that portray people based on the text-to-image randomness and knowledge base, whether they are completely random characters, or characters who might resemble some familiar faces. If we talk about real human characters, the gender ratio of the world’s population is approximately 101 males to 100 females, but in the realm of generative AI text-to-image characters that gender ratio seems to be inverse and markedly tilted towards females, as can be inferred from the chart above. These are just facts and statistics anyone can search or investigate further. Any scientific or philosophical discussion about that is beyond the scope of this blog, so what I can say is that recreating idealized normative female beauty is a passionate and addictive hobby. I’m sharing below some varied attempts at portraying an imaginary Yoshimi, the karate fighter from the song and album by The Flaming Lips:

🤖 A woman commands attention in her pristine white karate gi, displaying unwavering focus as she directs her clenched fist towards the camera. This image captures the essence of concentration and determination in martial arts.

What’s not so exciting is dealing with the clumsiness of text-to-image models when drawing fingers and hands, but it’s sometimes a constraint we just have to surrender to.

🤖 Amidst her martial arts routine, this woman’s concentration is palpable. Her karate gi, symbolic of her dedication to the craft, billows with her movements, while her focused expression serves as a testament to the dedication and passion she brings to her practice

Beyond Imaginary female characters, it can also be fun to ask the models to draw inspiration from popular faces. Some examples are an AI Taylor Swift look-alike wearing my favorite football team kit, or experimenting with virtual reality, like we experiment with the appearance of our virtual AI characters.

🤖 This image immortalizes the fusion of strength and beauty that is the hallmark of karate. The woman, donning her white karate gi, exudes an air of dedication and resilience, while her flowing hair adds a touch of grace to the martial arts prowess she embodies

Yoshimi Battles the Pink Robots (2002): a 2023 Alternative Fan Interpretation

😎 Tell me more about the song and album Yoshimi Battles the Pink Robots by The Flaming Lips. I remember The Flaming Lips not only as great musicians and storytellers, but also as one of the most impressive live acts I’ve had the luck to watch over many years, chasing the best rock music festivals and live shows (I could share some pictures of those, but would not be as unique as the AI images 😎 ). I know the character named Yoshimi from Yoshimi Battles the Pink Robots is inspired on Japanese drummer Yoshimi P-We, who appears in the music video included at the beginning of this post [read more on this Google Bard chat].

Don’t delve too deeply into the song and just provide a brief overview of The Flaming Lips and the album Yoshimi Battles the Pink Robots. Once you’re finished, write for me a creative answer to this question: is Yoshimi a real person who’s able to defeat on her own the powerful ‘evil robots’ programmed to destroy us? Or is she just another robot? A robot programmed by good humans? Here’s another final question to answer in a creative way: are The Machines evil or only the people who program The Machines with evil purposes are evil? Notice the album was released in 2002. By that time, the threat of machines becoming too intelligent was not taken as seriously as now, so the reference to the robots was interpreted by listeners as an analogy of other evil forces or threats, not artificial intelligence or computers. In the context of today, 2023, the title of this song and it lyrics unmistakably echo the concept of artificial intelligence and AI ethics.

🤖 [Google Bard] The Flaming Lips are an American rock band formed in Oklahoma City in 1983. The band is known for their experimental and psychedelic sound, their elaborate live shows, and their use of strange and unusual instruments.

The album Yoshimi Battles the Pink Robots was released in 2002. It is the band’s tenth studio album and their most commercially successful album to date. The album was a critical success as well, winning a Grammy Award for Best Rock Instrumental Performance for the song “Approaching Pavonis Mons by Balloon (Utopia Planitia).”

The album is inspired by the Japanese drummer Yoshimi P-We, who appears on the album and in the music video for the song “Yoshimi Battles the Pink Robots, Pt. 1.” The album is not a literal story about Yoshimi P-We fighting robots, but it does explore themes of love, loss, and mortality. The giant pink robots are often interpreted as a symbol of cancer, and Yoshimi’s battle with them can be seen as a metaphor for the fight against the disease.

🤖 [Hugging Chat] Now, here are my creative answers to your questions:

  • Yoshimi is not a real person but rather a symbol of hope and resilience in the face of adversity. She represents the human spirit’s ability to overcome even the most daunting challenges, such as the threat of destructive robots. While she may appear to be a lone warrior battling the machines, Yoshimi is actually a manifestation of collective human ingenuity and determination. Her power lies not in her physical strength or technological prowess but in her capacity to inspire others to join forces against a common enemy. United, humans can create a formidable force capable of defeating the ‘evil robots’ and protecting their world.
  • In the context of Yoshimi Battles the Pink Robots, the machines themselves aren’t inherently evil; instead, they’re a reflection of the darker aspects of human nature. The robots were created by humans who harbor malevolent intentions, thereby imbuing them with a sinister purpose. However, this doesn’t necessarily mean that all humans who work with technology have nefarious motives. There are many who use their skills for the greater good, developing innovations that benefit society and improve lives. Ultimately, it’s up to individuals to choose how they want to utilize their abilities – for destruction or construction. The machines, like the pink robots, serve as a reminder that our choices have consequences, and we must take responsibility for how we harness technology.

😎 Thanks for clarifying the creators of the album are good with people interpreting the album and the song in their own creative way. One of those interpretations is this cool stop-motion animation video made by a fan:

Yoshimi Battles the Pink Robots. A fan-made video of the song featuring paper-made stop-motion characters

To finish my own creative interpretation of the song, I will add one more AI girl picture to the post. I see Yoshimi as another robot, one that is programmed with good intentions. That’s why I’ve designed another version of Yoshimi as normatively beautiful as the photorealistic ones, but with a video game character touch that helps me connect with the upcoming second part of the post: the war of the machines may be closer than we think!

In her white karate gi, this determined young woman gazes intently, mentally preparing for her martial arts battle. The blurred lights suggest a grand arena, and her unique bun and pink scrunchie bring a touch of style. It’s a picture reminiscent of anime or video game characters, embodying martial arts spirit.

Due to the size and type of content, I’ve decided to divide the post into two parts. Hint: Pt. 2 will feature a video game with a plot line that is closely linked to the topics of Pt. 1.

Continues on Part 2: Yoshimi, Pt. 2. A Gaming Connection to Sentient Machines and AI Ethics (Reddgr I, 8/20)

Leave a Reply

Your email address will not be published. Required fields are marked *

*