Sign in Subscribe
CULT COINS

CULT COINS

AI-Native Art

Introduction

Ikigai Labs sits at the forefront of the intersection between art, technology, and the evolving Web3 landscape. With a curated collection of Web3-native art spanning from 2017 to 2025, Ikigai Labs celebrates the pioneering creativity of the early NFT era while looking ahead to the future of Real-World Art (RWA) integrated with TRACE technology. But beyond our digital museum walls, a new frontier is emerging—Cult Coins.

Cult Coins: An Evolution in Art and Technology

Cult Coins represent a new era of AI-native digital assets. Unlike traditional NFTs, which are static representations, Cult Coins are dynamic, AI-driven characters and personas with evolving skill sets, hosted on the Solana blockchain. These coins are more than just collectibles; they are living entities that grow and adapt over time, reflecting the cutting edge of AI-native art and digital asset technology.

Curating the Future

To fully realize this vision, a strategic approach to curation is essential. While the temptation exists to mint all 100+ AI characters we've developed, the most impactful path forward is to curate this vast collection down to a manageable number—24. Each character will be carefully selected not just for their visual appeal, but for the story they tell, embodying the innovation and spirit of AI-native art.

The Launch Strategy

The launch of Cult Coins will be executed in phases, beginning with the deployment of 8 tickers on Moonshot, our chosen platform. Each ticker will represent a unique AI persona, complete with an image, ticker name, description, and integration with our LiveTheLifeTV social channels, including Discord, X (formerly Twitter), and Facebook.

Interactive Experiences

What sets Cult Coins apart is their interactivity. Each AI persona will have its own dedicated channel on LiveTheLifeTV’s Discord, where collectors and fans can engage directly. These characters are not static; they can respond to questions, participate in discussions, and even broadcast messages on X, making them living, evolving entities within the Web3 space.

Real-World Integration

While Cult Coins exist in the digital realm, their influence will extend into the physical world. Selected characters from the curated collection will be minted on the OG Ethereum network and printed on high-quality materials through our partnership with WhiteWall. These prints will not only serve as valuable art pieces for collectors but also blur the lines between the digital and physical worlds.

Conclusion

Ikigai Labs is more than just a museum; it’s a launchpad for the next generation of AI-native art. As we prepare to mint and launch Cult Coins, we're not simply creating another set of collectibles; we're crafting a narrative, curating an experience, and pushing the boundaries of what art can be in the Web3 era. Our focus now is on fine-tuning this narrative, selecting the personas that will lead this new frontier, and preparing to launch them into the world. Stay tuned to see which characters make the cut and how you can become part of this AI-native revolution.

Fine-Tuning AI Models for Cult Coin Personas

To bring each Cult Coin persona to life, a robust dataset capturing both the skills and personality traits of each role is essential. This dataset will be transformed into JSONL files for training models on platforms like OpenAI's GPT, Cohere, or similar. The following outlines the process:

1. Data Collection

1.1 Identify Core Competencies and Personality Traits

  • Core Competencies: Define key skills for each role. For instance:
    • Design Persona: Creative thinking, UX/UI design, design tools.
    • Engineering Persona: Problem-solving, coding, algorithm understanding.
    • Operations Persona: Process optimization, project management, workflow automation.
  • Personality Traits: Assign distinct voice traits to each role. For example:
    • Design Persona: Innovative, empathetic, artistic.
    • Engineering Persona: Logical, detail-oriented, methodical.
    • Operations Persona: Organized, pragmatic, efficient.

1.2 Gather Relevant Data

  • Books and Articles: Extract relevant information and language style.
  • Blogs and Social Media: Use content from industry leaders.
  • Job Descriptions and Interviews: Analyze industry-specific roles and phrases.
  • Internal Documentation: Leverage any available internal documents.

1.3 Synthetic Data Generation

  • Prompt-Based Generation: Use large language models to generate additional training data.
  • Persona-Specific Conversations: Create dialogues where personas answer common questions or interact with other roles.

2. Data Structuring

2.1 Create a Structured Dataset

  • Text Segmentation: Organize data into segments representing each competency or trait.
  • Labeling: Label each segment according to its role, skill, and personality trait.
  • Data Format: Structure data in a table with columns for Input Text, Role, Skill/Competency, and Personality Trait.

2.2 Data Augmentation

  • Paraphrasing: Increase dataset size by creating variations of the same data points.
  • Scenario Simulation: Generate hypothetical scenarios for persona interactions.

3. Converting Data to JSONL Format

3.1 Structure of JSONL Files

Each line should represent one training example in the format:

{"prompt": "<INPUT TEXT>", "completion": "<EXPECTED OUTPUT>"}

3.2 Data Transformation

  • Scripted Conversion: Convert structured datasets into JSONL format using Python.
  • Validation: Ensure each JSONL entry is well-formed.

4. Model Training and Fine-Tuning

4.1 Platform Selection

Choose a fine-tuning platform based on specific needs and budget.

4.2 Model Configuration

  • Hyperparameter Tuning: Set appropriate learning rates, batch sizes, and epochs.
  • Training Iterations: Determine the number of epochs based on dataset quality.

4.3 Evaluation and Iteration

  • Testing: Use unseen prompts to evaluate the model’s performance.
  • Feedback Loop: Adjust the dataset or fine-tuning approach to improve the model.

5. Deployment and Monitoring

5.1 Integration with Moonshot and Other Platforms

  • API Deployment: Deploy fine-tuned models as APIs for integration with Moonshot.
  • Persona Channels: Ensure models interact correctly within Discord channels.

5.2 Continuous Learning

  • User Feedback: Collect feedback to refine and update the models.
  • Dataset Expansion: Continuously expand and refine datasets, retraining models as needed.

Next Steps:

  1. Start collecting and organizing data for each role and personality.
  2. Write scripts to transform this data into JSONL files.
  3. Begin fine-tuning on your chosen platform, starting with one persona as a test case.

Ikigai Labs is committed to pushing the boundaries of AI-native art. Through careful curation, strategic launch planning, and robust AI development, we are set to redefine the intersection of art, technology, and Web3. Join us as we embark on this revolutionary journey with Cult Coins, where art meets intelligence in the digital age.