deAI – Part 1: Exiting the Matrix
Decentralized artificial intelligence (deAI) has long been dismissed as impractical and infeasible, relegated to the realm of hobbyist experimentation. A year ago, these perspectives were not entirely unfounded. However, breakthroughs in distributed training have radically reshaped the narrative. What once seemed like speculative fiction is now emerging as a tangible reality. deAI has transcended its conceptual roots—it is no longer a distant vision but a transformative force that will redefine what AI could be. We’ll delve deeper into “how” later in this report, but first, let’s focus on understanding “why.”
AI originated through inevitable and disparate efforts to solve practical problems but to also discover the furthest extents of what could. The resulting chaotic frenzy of innovation within the chamber of 21st century capitalism comes with a lack of a comprehensive plan. The ensuing AI product of today is centralized within entities like OpenAI. Nation states are reluctant to depend on foreign organizations for critical AI development, while public trust in tech giants has waned as profit-driven agendas increasingly clash with societal interests. If “power tends to corrupt, and absolute power corrupts absolutely,” then what technological revolution in history has concentrated as much potential power into the hands of so few? And if that is the case, what challenge could be more urgent or more deserving of our collective focus than addressing this profound imbalance?
The Age of Enlightenment started an intellectual movement, philosophical insight empowered by new technology, birthing the American Republic and originated the thoughts that shaped the contemporary world order. We are entering a new era where AI will become synonymous with “intelligence”, one that could erode human competence and the essence of the human condition into data. AI may even induce the same form of religious fervor that the Age of Enlightenment itself superseded, of which science replaced faith as the dominant framework for human consciousness. The Age of Intelligence will have both many goals and unintended outcomes, but retaining knowledge, freedom, and happiness remain constant along with (a new variation) the separation of intelligence and state.
This is not to detail a sympathetic letter to doomists and alarmists, but to present a practical solution. The challenge of viable alternatives extends beyond merely advancing AI capabilities but establishing sustainable economic models that foster open innovation. Without a solution, the Age of Intelligence has the potential to steer us backwards and undo the work of Enlightenment.
The emergence of deAI represents the beginning of a definite exit from this paradigm — a framework that empowers researchers to work transparently and collaboratively, advancing collective human progress while maintaining economic sustainability. This new paradigm enables contributors to monetize their efforts in alignment with their ethical values, free from the constraints of profit-driven corporate and authoritarian government agendas.
In essence, deAI is a more practical and effective path to achieving the objectives of “AI Safety” whilst remaining inherently aligned with the principals of the e/acc (Effective Accelerationism) movement. The scientific world is impelled to explore the furthest extents of technology and its achievements whilst technologists are driven by commercial vistas of outrageous proportions. deAI is pioneering a transformative model for AI development—one that harmonizes the drive for innovation and profits whilst being committed to societal well-being.
From a crypto-native perspective, AI most likely accelerates the migration of the world’s financial and societal infrastructure to blockchain. Agentic volume will surpass and come to dominate user activity driven by a combination of intent-based abstractions and the complete overhaul of Web2 transactions. Blockchain becomes the perfect substrate for near frictionless agentic life as much as it is for human-to-human interaction. Blockchain’s distributed architecture ensures resilience, accountability and trustlessness, and breaks down a centuries worth of infrastructure built for humans that AI agents are discriminated against. AI needs crypto as much as crypto needs AI; a symbiotic relationship that is a distillation of what Web3 really is going forward.
deAI also leverages a globally distributed pool of talent, drawing on the expertise of researchers, developers, and domain specialists from diverse geographic and cultural backgrounds. This decentralized approach harnesses a broader spectrum of knowledge and innovation, enabling the conception of ideas and solutions that might never emerge within the confines of centralized teams. The collaborative dynamics inherent in this global framework foster creativity and accelerate the development of groundbreaking advancements. By redefining how AI is conceptualized, built, and refined, deAI drives a transformative shift in the pace and scope of innovation.
For those well-versed in machine learning (ML), several significant challenges present themselves when contemplating decentralization. Chief among these is communication overhead—most ML models are trained in massive data centers utilizing high-speed interconnections, effectively allowing a cluster of GPUs to be treated as a singular massive computer. Approaches such as Distro, DiLoCo, and Swarm Parallelism have made strides in addressing this issue. With new research emerging on a monthly basis, these methodologies have laid the foundational groundwork to demonstrate that minimizing communication overhead is a feasible route toward the decentralization of AI development.
Furthermore, ongoing research into model merging, mixture of experts, federated learning, and sparsity signals a promising trajectory for the future of deAI. Another major challenge appears to be in the verification of compute in a decentralized environment to ensure robustness in AI development. Advances in these fields continue to expand the boundaries of what was once thought possible.
In this 3-part report, we will delve into an in-depth exploration of several of the aforementioned approaches to provide a nuanced understanding of the complexities inherent in this emerging field.
Continue to Part 2: deAI – Decentralized Training
Disclosure: Unless otherwise indicated, the views expressed in this post are solely those of the author(s) in their individual capacity and are not the views of Big Brain Holdings or its affiliates. Certain information contained herein may have been obtained from third-party sources, including from portfolio companies of funds managed by Big Brain Holdings. Big Brain Holdings believes that the information provided is reliable but has not independently verified the non-material information and makes no representations about the enduring accuracy of the information or its appropriateness for a given situation. Charts and graphs provided within are for informational purposes solely and should not be relied upon when making any investment decision. Any projections, estimates, forecasts, targets, prospects, and/or opinions expressed in this blog are subject to change without notice and may differ or be contrary to opinions expressed by others.
The content is provided for informational purposes only, and should not be relied upon as the basis for an investment decision, and is not, and should not be assumed to be, complete. The contents herein are not to be construed as legal, business, or tax advice. You should consult your own advisors for those matters. References to any securities or digital assets are for illustrative purposes only, and do not constitute an investment recommendation or offer to provide investment advisory services. Any investments or portfolio companies mentioned, referred to, or described are not representative of all investments in vehicles managed by Big Brain Holdings, and there can be no assurance that the investments will be profitable or that other investments made in the future will have similar characteristics or results.
This blog does not constitute investment advice or an offer to sell or a solicitation of an offer to purchase any limited partner interests in any investment vehicle managed by Big Brain Holdings. An offer or solicitation of an investment in any Big Brain Holdings investment vehicle will only be made pursuant to an offering memorandum, limited partnership agreement and subscription documents, and only the information in such documents should be relied upon when making a decision to invest.
Past performance does not guarantee future results. There can be no guarantee that any Big Brain Holdings investment vehicle’s investment objectives will be achieved, and the investment results may vary substantially from year to year or even from month to month. As a result, an investor could lose all or a substantial amount of its investment. Investments or products referenced in this blog may not be suitable for you or any other party.