The Race Against Digital Amnesia: Why We Must Preserve Living Culture Ahead of the Singularity

The story behind Bloomyn Times

Hello, my name is Pierre Chesneau. I am the founder of Bloomyn Times, and I wanted to share with you the foundational story behind the work that we are trying to achieve with Bloomyn Times on international cultures and artistic expression across different regions of the world. The story of Bloomyn Times actually comes from a somewhat unexpected thought process—our motivation to work on traditional cultures, local artistic expression, and cultural heritage is deeply grounded in the recent explosion of AI development.

The Realization - AI Guardians of Human Knowledge

At some point, I realized that much of human history and knowledge will, in a few decades, be gathered and hosted by very large multi-modal models—models that can manage text, image, video, 3D, understand context, physics, and movement. This isn't just speculation—leading AI researchers have been predicting this trajectory for years. As Stuart Russell notes in "Human Compatible" (2019), "AI systems will become the repositories of most human knowledge and will be the primary means by which that knowledge is accessed and applied."¹

Just as we've been using the internet for the last 20 years as a reference point for knowledge, I believe that in 50 years, we will turn to these models for knowledge and won't question the information they hold. This shift is already beginning—researchers at DeepMind, including CEO Demis Hassabis, have described how large language models are becoming "knowledge engines" that can integrate and synthesize information in ways that surpass traditional search engines.²

The challenge of synthetic data in AI training is particularly concerning. As models evolve, they increasingly rely on synthetic data—creating new information based on existing patterns. Recent research from Zhang et al. highlights how these cultural biases in large language models can become self-reinforcing, creating what they call "cultural feedback loops" that systematically exclude certain forms of knowledge.³

This means that cultural practices not present in early models risk being absent from future AI knowledge bases entirely.

Here's a striking reality check:

Try asking any AI today about a niche cultural activity you're familiar with. While these systems can provide complex answers about advanced physics, they often fail to give relevant information about simple yet specific cultural practices. Try it yourself, and you'll understand the urgency of our work!

The Problem - Cultural Amnesia of the Digital Age

Coming from a background in traditional and social dances, I realized that many niche practices and communities have very little representation online. When representation exists, it's often without context, proper SEO, or adequate categorization.

This digital invisibility of cultural practices aligns with what UNESCO has termed "digital heritage at risk," highlighting how intangible cultural heritage faces unique challenges in the digital age. ⁴

Furthermore, we often overlook how rapid evolution itself can be a form of cultural loss. While a practice may still be thriving, its swift transformation means historical variations are lost. This isn't negative—culture should evolve—but we need to document these evolutionary stages rather than waiting for practices to near extinction before preserving them.

As Kirshenblatt-Gimblett's research on intangible heritage demonstrates, "The digital preservation of cultural heritage requires not just recording practices, but capturing their social and cultural ecosystem."⁵

I have personally witnessed dances disappearing from the common dance pool at French Folk events and festivals within just 10 years! Recently, when proposing a workshop on one of these dances (Bourrée 3temps in ligne) to a group of students, no one had any idea what I was talking about.

(I will dive deeper into this mechanism in another article soon so make sure to subscribe to our account!)

And this isn't just true for traditional dance—it's happening in ALL cultural niches around the world. From traditional crafts in Southeast Asia to storytelling traditions in West Africa, from indigenous healing practices in the Americas to community rituals in Eastern Europe—countless expressions of human culture risk falling through the digital cracks. Anthropologist Wade Davis's concept of the "ethnosphere"—the sum total of all thoughts, dreams, myths, ideas, and inspirations brought into being by human imagination—is being eroded not just by globalization but by digital exclusion.⁶

Remembering Today, and Again Tomorrow

There's a time-sensitive element to this work. We tend to record cultural practices only when they're on the verge of disappearing—as happened in the 1970s during the folk revival movement. They recorded dances and music from elderly practitioners, but by then, the quality of data was poor because these people couldn't remember or demonstrate everything properly.

Cultural anthropologist James Clifford's concept of "salvage paradigm" helps explain this pattern—our tendency to document cultural practices only when they're endangered rather than during their vibrant prime.⁷ I've seen this firsthand in my work with traditional dances—by the time we think about preservation, we've often already lost crucial context and nuance.

Societal Changes Shift the Focus of Human Existence

Going Back to Being Social-Expressive Animals

There's another aspect to consider when talking about AI progress and development. This point might create varying levels of resistance depending on your familiarity with current AI developments and projections. I believe that in the fairly near future, a significant portion of people—first in Western society and then globally—will have substantially more free time due to a major shift in the work ecosystem.

This prediction is supported by extensive research in the field of technological unemployment and social transformation. Erik Brynjolfsson and Andrew McAfee's work at MIT has demonstrated how AI and automation are likely to reshape labor markets fundamentally.⁸ Their research suggests that this transition could lead to what they call a "second machine age," where human creativity and social interaction become increasingly central to society.

Anthropologist Tim Ingold's work on "skilled practice" suggests that humans are fundamentally makers and performers, and these activities are central to our well-being and sense of identity.⁹ As AI takes over more routine tasks, these fundamental human traits may become increasingly important.

Post-Employment Society Explained

With the imminent development of AGI and superintelligence, human roles in modern society will shift dramatically. Economics Nobel laureate Joseph Stiglitz has projected that AI could lead to unemployment rates of 30-50% within a few decades.¹⁰ Research from the Oxford Martin School suggests these figures could reach 70-80% in some sectors.¹¹

This transition to what I call a "post-employment society" will be profound and far-reaching. Imagine a world where AI systems handle the majority of what we currently consider "work." This isn't about machines taking all jobs—it's about fundamentally restructuring how society functions and what we consider valuable human activity.

I see this unfolding in three distinct phases:

1. Initial Displacement (30% unemployment): We're already seeing the beginnings of this with automation in manufacturing, customer service, and data processing. This phase will force us to begin seriously discussing universal basic income and social safety nets.

2. Major Transformation (50% unemployment): This is when traditional employment becomes the exception rather than the norm. Society will need to develop new frameworks for distributing resources and measuring human contribution beyond labor.

3. Post-Employment Era (70-80% unemployment): This is where things get really interesting. When most traditional jobs are automated, humans will have the freedom—and challenge—to redefine their role in society. This is where I believe we'll see a renaissance of cultural and creative expression.

This shift might mark a return to what humans historically excelled at—being social, emotional creatures. Social dance, once mainstream in village life before the industrial revolution prioritized productivity, could resurge not as recreation but as a natural return to fundamental human needs. As AI masters logical tasks, our capacity for emotional expression may become our most distinctive trait.

Kate Raworth's "Doughnut Economics" model offers one vision of how society could reorganize around human and ecological well-being rather than pure productivity.¹² The World Economic Forum's research suggests that universal basic income could become a crucial tool for managing this transition.¹³

All Part of a Process

This transition might seem daunting, but it's really part of a larger historical process. The Industrial Revolution pushed us toward productivity and broke much of our social fabric. Sociologist Richard Sennett's work on craft and community shows how industrialization disrupted traditional patterns of social learning and cultural transmission.¹⁴ However, his research also suggests that humans naturally gravitate back toward communal and creative activities when given the opportunity.

Recent Pew Research Center data supports this trend, showing increasing interest in traditional crafts and communal activities, particularly among younger generations seeking meaning beyond conventional work.¹⁵

I believe we're witnessing the end of an era—the age of human productivity as the primary measure of worth. Just as the Industrial Revolution transformed society by moving us away from agriculture and craftwork, the AI revolution will transform society by freeing us from the need to be primarily productive beings.

Long-lasting Impact | Serving the Needs of Tomorrow

Even if few people see the importance of this work right now, I have a deep belief that its value will become apparent when future generations want to learn about their cultural heritage.

Imagine a future where AI can not only describe but accurately synthesize historical movements. Someone might ask, "Show me how this traditional dance was performed in the 2020s," and receive a precise, animated demonstration—but only if we provide high-quality documentation today.

When that time comes, the quality of our documentation today will determine whether that knowledge survives.

If you resonate with this vision, there are several ways to support our mission:

• Watch and share our content

• Join our YouTube Membership or Patreon

• Reach out and let us know your thoughts

• If this really speaks to you, consider joining the Bloomyn Team!

The window for this work is now! Whether it's just one dance, one ritual, or one interview that makes it into the training data, every piece we properly document becomes part of humanity's lasting memory. This is our chance to ensure that the rich tapestry of human cultural expression remains accessible for generations to come.

1. Russell, S. (2019). Human Compatible: Artificial Intelligence and the Problem of Control. Viking.

2. Hassabis, D., et al. (2022). "Multimodal Language Models: A New Frontier in AI" - DeepMind Research Blog

3. Zhang, et al. (2022). "Cultural Biases in Large Language Models." Proceedings of NeurIPS.

4. UNESCO (2003). Charter on the Preservation of Digital Heritage.

5. Kirshenblatt-Gimblett, B. (2019). "Intangible Heritage in the Digital Age." Journal of Cultural Heritage.

6. Davis, W. (2009). The Wayfinders: Why Ancient Wisdom Matters in the Modern World.

7. Clifford, J. (1989). "The Others: Beyond the 'Salvage' Paradigm." Third Text.

8. Brynjolfsson, E. & McAfee, A. (2014). The Second Machine Age.

9. Ingold, T. (2018). Making: Anthropology, Archaeology, Art and Architecture.

10. Stiglitz, J. (2019). People, Power, and Profits: Progressive Capitalism for an Age of Discontent.

11. Frey, C.B. & Osborne, M.A. (2017). "The Future of Employment: How Susceptible Are Jobs to Computerisation?" Technological Forecasting and Social Change.

12. Raworth, K. (2017). Doughnut Economics: Seven Ways to Think Like a 21st-Century Economist.

13. World Economic Forum (2023). "The Future of Work Report"

14. Sennett, R. (2008). The Craftsman.

15. Pew Research Center (2023). "The Future of Work and Leisure"