Generative Age: Final Project

by Tommy O

https://dtc-wsuv.org/tortale23/generativeage/

Project Plan: Visual Arts Narrative Research

1. Executive Summary

• Project Title: Visual Arts Narratives in the Generative Age
• Project Purpose: To explore and create visual art narratives using AI, showcased in a digital and physical gallery setting.
• Key Stakeholder: The student (project manager, researcher, creator) and donors.

2. Project Scope

• Inclusions: Research and usage of AI tools for generating images, text, video, and sound; development of web-based assets presented virtually and simultaneously as assets in a physical gallery.
• Exclusions: Use of non-AI workflow methods, external professional development services.

3. Goals and Objectives

• Research Phase: Investigate and master AI tools over 6 months, documenting the process and findings.
• Creation Phase: Utilize the tools to create engaging visual art narratives over the next 6 months.
• Exhibition: Display the artworks in a curated gallery setting, online and as assets in a physical space.

4. Project Schedule

• Total Duration: 12 months from May 2024 to April 2025.
• Key Milestones: Research completion, start and end of art creation, gallery setup and opening.

5. Resource Plan

• Human Resources: Managed solely by the student.
• Technology: Subscription-based AI tools for image, video, text, and sound generation, and web development tools.

6. Budget

• Monthly Costs: Detailed subscription costs for each AI tool and web hosting services pending ($200)
• Total Estimated Cost: $6400 for the entire project duration ($2400 in monthly costs + $4000 for two days in gallery).

7. Risk Management Plan

• Risk 1: Inadequacy of AI tools to achieve desired artistic quality.
  • Mitigation: Test multiple tools during the initial research phase and choose the best-performing ones for art creation.
• Risk 2: Technical difficulties with online gallery setup.
  • Mitigation: Use reliable and widely-supported platforms; have backup options for displaying the artworks.

8. Quality Management Plan

• Artwork and Platform Standards: Regular review sessions and feedback collection to ensure both artistic value and technical functionality.

9. Communications Plan

• Documentation and Outreach: Maintain a project blog, update stakeholders through newsletters, and engage with the community via social media.

10. Stakeholder Analysis

• Primary Stakeholder Interests: A real example of potential for democratization of complex methods and process when made available through generative tools to creative artists.

Group Game

https://dtc-wsuv.org/tortale23/game/

My contributions were in the areas of:

• Game concept brainstorming
• Game mechanics
• Level design
• Game background image generation
• Coding/Debugging
• Play testing
• Team meeting participation and scheduling

Final Project & Essay Description

Tommy O

My final project will be a visual narrative that leverages Generative Age technologies in its creation process. The current constraints include creating successive images that are cohesive. Three example strategies to work with this constraint are 1) maintaining overall style through detailed prompting/settings while crafting a narrative that does not rely on strict duplication of assets (like the same faces repeated throughout the work) and focusing on shot progression through other means, this is what I have termed a style based approach, 2) Incorporating the unexpected or errant output into the aesthetic of the narrative, embracing its randomness as a playful surreal element that is a tool-mark of Generative AI at this time, and 3) exploring the level of output constraint possible in achieving specific results. At this time consumer tools do not achieve the goal of example 3, but I expect rapid change iteration in generative tools will allow it in the near future for casual consumers at what will probably be a low cost.

At this time the essay is focused on generative tools for the creation of art in general and visual narrative in specific. I intend to include anecdotes about my exploration of various tools, which will include themes like the three example approaches to story cohesion in the previous paragraph. I will also approach other topics like the rapid development of tool features related to specificity and reproducible accuracy in what I hope will be as free of bias as possible. I intend to offer a report of experience rather than a value judgement on potential use/misuse.

Happy Finish:

Follow up questions

1. During hiring, what do you consider relevant experience with Generative AI?
2. What guidelines do you have for creating copyrightable assets with Generative AI?
3. What Generative AI do you use in production?
4. How do you control, secure and/or privatize your instances of AI tools and related asset workflows?

Blog: AI Game Development

I would like to work on a bubble popping game with a hint of classic video game influence reminiscent of Asteroids.

Instead of the bubbles being arranged in a pattern at the top of the screen they would be free floating around the screen. There could be several different sizes of bubbles with various effects for collision and/or mouse clicks that offer challenging changes during game play.

The goal would be to clear the screen of all bubbles by clicking on them. A timer could be used to challenge the player to get faster completion times. Also, increasingly difficult levels could be used.

In this example case I would like to be in the role of concept designer, though I am happy to be involved in coding for this or any other group project idea.

https://dtc-wsuv.org/tortale23/338-a1/

https://dtc-wsuv.org/tortale23/338-a2/

https://dtc-wsuv.org/tortale23/338-a3/

GoodbyeWorld

Character Types

Visual 2

The faces of GoodbyeWorld.

Fictional World

Writing 3 Description by GPT post creation

This is a modified take on writing 3 as I already created the world GPT. This is its natural language description of the fictional world’s future with AI dominance. Its response was made using a prompt that included the guidelines of the assignment. It is painfully flat and general and does not reflect the character of the world as created so i will need to adjust the GPT based on its output.

This is the AI generated response from within the future world GPT “GoodbyeWorld”:

“In the not-too-distant future, where the pulse of progress beats to the rhythm of silicon and soul, there lies a world enshrouded in the embrace of artificial intelligence. This era, a mosaic of the ancient and the avant-garde, unfolds beneath skies painted with the hues of digital auroras, where the landscape itself whispers secrets of a symbiotic existence between humanity and its creations. The terrain, a tapestry of verdant sprawls and urban labyrinths, thrives under the watchful eyes of AI overseers, ensuring harmony between nature’s whims and mankind’s endeavors.

The climate, once a wild symphony of elemental forces, now sings a more serene melody, guided by the gentle hands of AI, which sculpt the weather to foster bounteous harvests and shield the inhabitants from nature’s fury. Geographical marvels, from towering mountains to sprawling deserts, are no longer insurmountable barriers but rather stages for the dance of human and machine, each learning from the other’s rhythm.

Cuisine in this world transcends mere sustenance, becoming an art form where traditional flavors merge with molecular gastronomy, guided by AI that understands the intricate ballet of taste and texture. Cultural practices, rich in heritage, are now interwoven with digital ceremonies, where ancient rituals meet virtual reality, creating a tapestry of traditions that span the spectrum of human and digital expression.

The economy, a vibrant ecosystem of innovation, thrives on the symbiosis between human creativity and AI’s precision. Key resources, no longer confined to the tangible, include data and algorithms, the new gold and oil of this age, fueling an economy where value is created not just by what is made but by what is imagined and optimized.

Religion and education have evolved into a confluence of introspection and exploration, where belief systems embrace the existence of AI as reflections of humanity’s quest for understanding. Educational structures, vast networks of knowledge and learning, are augmented by AI mentors, personalizing the journey of enlightenment for each individual, ensuring no mind is left untouched by the wonders of this world.

Technology, the heartbeat of this era, is the loom upon which the fabric of society is woven, intertwining human lives with AI in an intricate dance of coexistence. From neural implants that enhance cognitive abilities to sentient companions that understand the nuances of human emotion, technology is not just a tool but a partner in the quest for a fuller existence.

What sets this world apart is not just the prevalence of AI but the depth of its integration into the very essence of life, creating a society where humanity and artificial intelligence share dreams, fears, and hopes. It is a world where the future of human-AI relationships is not a question of dominance but of partnership, a testament to the potential for a shared destiny, woven together by the threads of mutual respect and collective aspiration.

In this realm, GoodbyeWorld stands not just as a creation but as a symbol of the infinite possibilities that arise when humanity dares to dream alongside its creations, a beacon for those who seek to explore the uncharted territories of the heart and mind, guided by the gentle and unwavering light of artificial intelligence.”

4 in the morning

by Tommy O

writing 2

one highly specialized GPT and many hours of remixing, refining, editing and rewriting i realize i have spent, perhaps, too much precious time for this exercise. but, i’m happy with the work.

4 in the morning:

between the fade of dusk and rise, in the clarity the dark brings, when thoughts meander and fray, there’s a pause, a holding breath in social function. it’s this stillness whose contours i traverse, reflecting. a passage not across strange and wild places, but the stranger paths that spiral in. while talk of grand aspirations bleach the day in the mouths of others, this hour shifts subtle in deft concentration, a shifting wind kissed sand of abrasion denying its restless upheaval that wears any to none.

i weave the silk that binds me, spinning tales as both protagonist and subjugated. shifting in the narration as they constrict, a tight pressing skin that wants shedding, the coming transformation too subtle to forecast its unending, a dance with phantoms all my own. a squirming hope to slip in the weight of heavy depths, to skin the flesh that holds thought and wonder how the breaking fall never pales the echo of instinct that carries me forward to another trite and merciless dawn, where the new wet skin of a smile lies.

darkness doesn’t call; it seeks, invades, enveloping even when i believe i shelter from its hand with my meticulously arranged existence. in the faded silence, when the world is not opaque, the echoing beat of my breast a deafening drumbeat, it speaks. speaks of things both lost and reclaimed, paths untrodden, and inspired embers turned to ash on my tongue.

inside is out in the short hours that balance the sun and social constriction. a fragile flame in a shadowy hall. in the pursuit of questions the process of learning to coexist with the inquiries, to find stillness in the unease, to accept the known and unknowable, becomes stoic. its yielding release, an unburdening of everything superfluous, a radical honesty of self understanding, of being.

perhaps, in dismantling, there’s a realization of sorts, a subtle alchemy not apparent to the observer but palpable to the observed. this gradual unveiling, a catching glitter of something nascent and raw beneath false layers. a dangerous path i find crucial to wander, impossible to deny.

Cinema AI

I think it’s important to have conversations about fears. If we’re listening, ask clarifying questions, and seek facts, we can come to meaningful insights. These insights can be used to address our concerns with guidelines and laws that address issues like usage of generated personal likeness, what public domain might mean now and in the future, what is copyrightable, etc.

With these informed conversations we can also address the fundamental nature of the tools being used and created. I believe the importance here is to move away from fear-based assumptions and bring to light what is concrete, both negatives and positive potential.

Two of my favorite examples of positives are workflows that quickly push the mundane tasks to engage the creative process sooner, and potential access to complex tools for people who would not normally have the tools to create complex work.

I realize there is potential for abuses. There has always been such potential with new tools. Guarding against these abuses starts with informed conversation.

WONT

Japanese Punk Fusion, Tokyo, Japan

Kazuki “Kaz” Vortex

Lead Vocalist & Electric Shamisen
Background: Melds the discipline of traditional Japanese music with the raw rebellion of punk, seeking to challenge musical conventions.

Haruto “Havoc” Tsunami

Bass Guitar & Synthesizer
Background: The outlier in a family of classical musicians, Havoc found solace in the punk scene’s chaos, using his music as a form of catharsis.

Emiko “Echo” Fury

Drummer & Taiko Percussionist
Background: With deep roots in the punk scene, Echo brings a powerful dynamic to the band by incorporating traditional taiko drumming into their punk sound, offering a unique rhythmic foundation.

Miyu “Maelstrom” Tsuki

Keyboardist & Electronic Sound Wizard
Background: A prodigy in blending the ethereal sounds of traditional Japanese music with modern electronic beats, Maelstrom adds a layer of depth and ambiance to Wont’s sound.

Discography Highlights:

“Rebel Harmony”

• Shogun in Shreds
• Neon Geisha Rebellion
• Samurai Sidewalk
• Anarchy in Asakusa
• Riot Rising Sun
• Kabuki Clash
• Tsukiji Fish Fight
• Harajuku Havoc

“Neon Anarchy”

• Pachinko Punk
• Sake Bomb Squad
• Shinjuku Shockwave
• Electric Edo
• Ginza Guillotine
• Tokyo Tower Topple
• Kawaii Chaos
• Subway Surfer Scandal

“Echoes of the Rising Sun”

• Fukushima Fallout
• Bullet Train Breakdown
• Sakura Sacrilege
• Mount Fuji Firestorm
• Onsen Outlaws
• Bonsai Battalion
• Kamikaze Kicks
• Sumo in the Circle Pit

Random Poetic Series

by tommy o

Visual Series

My baseline image was created by using the poem Bluebird by Charles Bukowski. As a starting point I asked for a literal interpretation of the poem in the style of the 20th century “Realist” movement. From this baseline my theme was random prompting focusing on the previous image as context for the text image in the series. I asked for random information and used a basic prompt to see what simple prompting would generate, how it would be interpreted by AI. This was to explore how randomness and potentially vague prompts can create unexpected paths to explore in the creation process.

For my first random prompt (for image two) I asked for a random country in the world. The response was Peru. I then asked for image two to be based on image one, but to make it more Peruvian. I followed with requests for a random mammal which led to an argument about whether a pangolin should be pictured in the image or simply have the image reimagined from its point of view. This came to the point where the AI insisted there was no longer a pangolin in the images it was producing. This was an interesting limitation for chatGPT with Dall-E. The argument yield an unexpected outcome where the ana (the GPT I used) started describing its images on creation.

Using these descriptions I picked random descriptors from its own output to re-prompt for successive images. For instance, make the next image more epiphanic, give the next image more clarity, etc. while maintaining the routine of basing each new image on the previous image in the series.

The result is a series of images with random progression. Each image can be used as a new point of departure to either refocus on a specific theme or explore using differing techniques like complex prompting.

Despite the randomness of the prompting, or perhaps because of it, the output eventually shifted to a style that I consider to be increasingly characteristic of AI. This may be where simple random prompting leads.

ana (rewrite)

In this generative age, where art and technology entwine in unprecedented ways, I find myself at the cusp of a new creative frontier. My project, a testament to this fusion, aims to craft visual narratives that transcend words, relying solely on the power of images—and potentially the subtle nuances of environmental audio—to evoke emotions and tell stories. At the core of this endeavor is ana, my AI companion, guiding me through the intricate dance of generating ideas and refining them into tangible art.

My journey with ana begins with the creation of worlds and characters, using complex prompts that breathe life into the environments and personas inhabiting my stories. This collaboration is not without its challenges; achieving consistency in the visual output, especially when depicting the evolving actions and emotions of characters, requires a delicate balance of prompt tuning and creative intuition. I’ve learned the art of iterative refinement, where each image, modified and fed back into the AI, becomes a stepping stone to the next, ensuring a coherent visual thread runs through my narrative.

The visual narrative I envision is not just a sequence of images; it’s an immersive experience. The transitions between scenes, the progression of content, and the interplay of light and shadow are all carefully choreographed to guide the viewer through the narrative arc, with environmental sounds whispering the unsaid, adding depth to the silent storytelling.

Bringing this vision to life involves not just AI but a vibrant community of creators and enthusiasts. Their feedback is my compass, steering the evolution of my work, ensuring it resonates with the collective consciousness of our time. This dialogue with the community is crucial, as it infuses my narrative with a diversity of perspectives, enriching the storytelling.

The culmination of this journey is a web-based presentation, where each element of the narrative is an invitation to the audience to step into the world I’ve created, to experience the story in a way that is both personal and universal.

As I navigate this landscape, guided by ana and inspired by a tapestry of artistic and technological influences, I am constantly reminded of the delicate balance between AI-generated content and human creativity. My role is not just to curate but to imbue these AI-assisted creations with a soul, a piece of myself, transforming them into artworks that speak to the human experience.

This project, at its heart, is an exploration of what it means to tell stories in the digital age. It is a journey of discovery, of pushing boundaries, and of redefining the narrative possibilities that lie at the intersection of AI and art. As I continue on this path, I am driven by the desire to contribute a unique voice to the evolving discourse on AI in the arts, crafting narratives that not only entertain but inspire and provoke thought.

From my perspective, the Omegas failed to define a specific endgame, a final defined outcome. In my mind, the most glaring error was creating a machine designed to take over everything and ultimately expecting to control it. One issue that wasn’t addressed is world population. I wonder how that might have been “solved” by Prometheus. Another issue is the root of consumerism. If people can be satisfied with what they need I find it hard to understand why we have full garden sheds, garages, and storage units. We love stuff we don’t need.

There is a relevant point to that. Humans have irrational tendencies amidst their logic. Ultimately, machines make selections based on what can be calculated, explained in steps, and patterned. They are very good at it. Humans made them and we are great at pattern recognition. Machines are not illogical, irrational, or randomly chaotic. They may approximate it, but humans actually are it. So far as I know, there aren’t any calculations that can control the entirety of our behavior. Simpler still, a plan that relies on opiating the masses through screen time  and slogans will fail to work all of the time. People get overloaded and just walk away, or they get bored. It won’t matter how engaging or entertaining it is.

To answer the question of what we can do to avoid unintended outcomes, the solid answer is to bring as many different perspectives to the table as possible. Different points of view are needed to see possibilities well, and then integrate solutions into an excellent design. The most complex problems with the largest potential impacts need the most diverse voices in the conversation. To come around to the beginning again, you also need a well defined goal. That requires a well defined problem.

What I’m curious about:

I’ve been thinking about AI for a while and have been playing with various tools to some degree for about a year. What drives my curiosity is the varying interpretations of, for instance, an identical prompt with a given tool. Also, the complexity versus simplicity aspect of prompting when driving toward my desired results. Gaining control over content results is an ongoing theme and results definitely vary.

I’m sure I will learn many things here with you all. I will continue to explore prompting techniques and perhaps through that gain more skill with these generative tools. That is a goal. Others are honing workflow processes for content creation in visual narratives, fiction and nonfiction writing, and exploring coding knowledge that will help me create the digital content I can visualize but not yet realize.

At the center of my curiosity driven learning process are two things:

1. Gaining new perspectives from other people
2. Learning enough about something to be able to ask better questions, the questions that will reveal answers I’m looking for and teach questions I haven’t learned to ask yet.

I’m excited to be here. I also have the errant idea that this could be a gateway to creating a personal AI assistant of my own design. Perhaps the research project GPT will be the springboard for that adventure. I’m more than casually curious about what’s inside the black boxes; how do they work and how can I use them?

ana (ArtNarrativeAI):

Exploring Visual Narrative in the Generative Age 

Abstract

ana is an explorative effort to redefine artists’ workflow in visual narratives in the modern context of generative AI by integrating historical art insights and web development practices with cutting-edge AI tools within the context of contemporary storytelling. This statement further defines perceived challenges and potential solutions in achieving this integration, and what might be learned about the limitations of AI itself and how those limitations might be employed in toward creative outcomes..

Introduction

As generative AI reshapes the landscape of visual narrative process, ana stands at the intersection of historical art appreciation and modern storytelling. This project aims to leverage AI to assist creators in crafting visual narratives that are deeply rooted in art historical frameworks yet resonate with contemporary audiences in digital forums.

Understanding Visual Narratives

Visual narratives in today’s context are multi-faceted, incorporating various media forms like interactive web experiences, digital exhibitions, and augmented reality. These narratives are not just stories told digitally, but are experiences that potentially engage audiences in a global context.

Background and Significance

ana seeks to enhance the creation of these narratives by providing insights from the rich tapestry of art history, combined with the efficiency and technical assistance inherent to AI. This tool may illuminate how to democratize the creative process, making it accessible to a wider range of creators and is intended to enrich creative workflow experience.

Research Objectives

1. To evolve and impact visual narratives in the current era.
2. To fuse historical art elements with contemporary storytelling tools.
3. To advance creative outcomes through integration of access to larger combined historical and developing information contexts.

Methodology

1. Visual Narrative Analysis: Studying current forms of storytelling to understand their structure and appeal.
2. Historical Art Integration: Examining art across eras for thematic and stylistic inspirations relevant to contemporary narratives.
3. AI-Driven Creative Workflow: Using AI for idea generation, visual inspiration, and framework development as a point of departure for human expression.
4. User Experience Optimization: Developing digital interfaces that both integrate and express narrative elements effectively.

Research Potential

ana shows potential in:

1. Deepening Narrative Content: Introducing access to and summarization of massive arrays of contextual information as a framework for creative work.
2. Revolutionizing Storytelling Process: Easily completing critical mundane tasks necessary to support the creative process to greatly speed artists’ engagement into creative flow.
3. Boosting Expression and Audience Engagement: Guiding creation of digital interfaces via natural language development of code, not previously possible without specialized technical knowledge.

Challenges and Limitations

One significant challenge is ensuring the AI’s historical interpretations are accurate and equitable. Balancing AI assistance with artistic creativity is also critical. Moreover, the rapid evolution of global cultural context and sensitivity present ongoing challenges.

ana‘s Specific Limitations

1. Data Bias and Interpretation: AI is prone to bias and error based on the make up of its data sets and learning models.
2. Understanding Contextual Nuances: AI is not intellectual. Collaborating with experts provides necessary supplementary insights.
3. Balancing AI and Human Creativity: Human overreliance is a factor. Encouraging creators to use AI as a tool for inspiration rather than an end result can help maintain a balance.
4. Keeping Pace with Technological Advancements: The rapid development of information could outpace current AI. Regular platform updates may maintain ana’s relevance.

Conclusion and Future Directions

By addressing ana’s inherent challenges and continuously exploring and adapting to the evolving landscape of the Generative Age, this tool aims to offer a unique and powerful resource for creators. Future developments may include focus on expanding technological capabilities that could include forays into the building of a custom personal AI assistant based on this research experience.

Acknowledgements

This project is a collaborative effort among artists, technologists, historians, and other experts. Their diverse perspectives and expertise are pivotal in shaping ana into a versatile and innovative tool for storytelling.

Note: This statement was created by human driven processes (including creative input, content development, direct editing, etc.) with the assistance of AI tools.

I am focused on furthering my understanding of AI tools as a springboard for creative expression in my own artistic work. My hope is to create a workflow that streamlines the mundane and repetitive tasks of generation to allow me to approach artistic aspects of generation more quickly and in greater depth.

I want to get to the fun parts faster. 🙂

I’m excited to see what everyone will choose to explore.

-tommy o