I still remember the moment I realized Artificial Intelligence (AI) was going to be THE thing. If you’re a tech/AI enthusiast and haven’t watched AlphaGO yet, do yourself a favor. This documentary follows the battle between a Google DeepMind AI model and the Korean rockstar Lee Sedol in a five-stage match of the ancient Chinese board game called Go. This could be an ESPN 30 for 30 documentary—the quality, tension, and repercussions of the man-vs-machine tussle are eye-opening.

I dove down the rabbit hole after watching this, immersing myself in neuroscience books, research papers, and videos on what it takes to build an artificial brain. The deeper I went, the more obvious it became that society was well on its way to fundamentally changing how we interact with technology.

AI has emerged as the most impactful and transformative technology we have in shaping the future of humanity. It will be integral for world leaders, international commissions, tech pioneers, and entrepreneurs to collectively ensure that AI becomes a force for societal betterment. This collective effort should be a primary focus as we move forward at a breakneck pace.

We are approaching a tipping point, where the widespread usage of commercialized AI products, like ChatGPT and Midjourney, are driving mass adoption, but the players who develop and control this technology are concentrated in the few.

This has happened before, it is an unfortunate part of human nature. Modern tech behemoths have hoarded our social media data targeting our vulnerabilities, and financial corporations are leveraged without impunity only gratifying those at the top. With another tipping point in sight, do we learn the lessons from the past and harness emerging technologies to circumvent the potential pitfalls of centralized AI? Or in hindsight, will we look back and tell our children we should have done better?

Let’s have a look at the case for decentralizing AI, and how the convergence of web3 could propel this technology to its full potential.

AI has been around for a while, it has been a field of study and development since the 1950s. Over the past 15 years, fueled by social media interaction, the tidal wave of big data, and streamlined analytics, we have witnessed a rapid rise and implementation of commercialized AI products. From Siri telling us whether it's going to rain today, to facial recognition logging us into our bank accounts, and ChatGPT sharing the perfect margarita recipe, AI applications have become more sophisticated and deeply integrated into our everyday lives.

The potential is huge. PwC has shared that AI could contribute up to $15.7tn to the global economy in 2030. The AI consumer market, driven by Generative AI, is set to boom in tandem, with Bloomberg predicting it to grow from $40bn in 2022 to over $1.3tn in 2032 – a 3000% increase. For a technology that is set to reshape global commerce, the desire to capture that value is all too appealing for corporations.

The generative AI foundational models and platforms market is highly concentrated between OpenAI, Microsoft, Google and Amazon, the amount they control – 84%. This dominance in such a transformative technology is troubling. Some argue that it is still early stages, or that allowing innovation leaders to control the market is beneficial. Is this really the case?

Generally, centralized control of industries and the subsequent formation of monopolies suppress innovation and raise barriers to entry for entrepreneurs. In the context of AI, this centralization leads to issues like siloed data access, inflexible models that struggle to adapt to diverse real-world scenarios, and a lack of transparency and accountability, which erodes trust in AI.

AI also presents a unique philosophical challenge with the threat of "value lock-in" (first introduced to me by the titan of a book "What We Owe the Future" by Scott Macaskill). If Artificial General Intelligence (AGI) becomes a reality, which is a type of artificial intelligence that matches or surpasses human capabilities, do we want its formative stages shaped by corporations, a capitalist system, or a socialist system?

This debate extends far beyond applications and market share, and as current AI systems start to evolve into their next evolution, the solution to these challenges may lie in the growth of decentralized technologies fueled by web3.

The Generative AI market is evolving quickly. At its core it harnesses various models that serve distinct purposes, each contributing unique capabilities to the industry. The widespread use of Large Language Models (LLMs) like ChatGPT and Retrieval-Augmented Generative (RAG) models, have sparked products that induce human-like responses by incorporating external knowledge through retrieval mechanisms. Although exciting, their proficiency in linguistic tasks does not extend to real-world action and interaction. An exciting evolution of these models is the incoming boom of AI Agents.

AI Agents are independent software entities or autonomous systems that use AI to perform tasks, make better decisions and interact in complex environments. They collect and analyze data, recognize patterns, and carry out actions independently in order to achieve defined goals. These agents are set to revolutionize tasks requiring real-world action, decision-making, and dynamic interaction.

Soon, you’ll be able to drag and drop all your invoices and receipts into a personalized AI agent, who will assess and submit your taxes instantly, a dream in my snail mail bureaucratic German tax system. As AI systems storm into this next evolution, we need to look at combining its capabilities with other emerging technologies to ensure it reaches its full potential.

Over the past decade, entrepreneurs with grand visions have been driving a paradigm shift towards decentralization. Blockchain serves as the foundational layer of this shift, with its applications eclipsing its original purpose as digital cash. Decentralization has driven heightened collaboration within communities, and provided equitable access to ideas, products, and technology. This transparent and open-source approach ensures that innovations are inclusive and rapidly developed. As the AI segment continues to grow, it is crucial to integrate decentralized technologies into its development to address the above challenges.

By converging these technologies, we can unlock AI systems and collectives that blend the principles of decentralization, transparency, and user empowerment. These systems governance and development can be distributed across entities with diverse incentives and priorities, better aligning with individual needs rather than imposing homogeneous outcomes. This creates dynamic applications rather than a handful of prevailing models that dominate their markets. By decentralizing AI, we restrict any one entity from imposing a single set of incentives, constraints or goals – key for such a transformative technology.

Currently AI functions as black-box systems whose inner workings and development are hidden from the public. To effectively decentralize AI, open source developers need to coordinate to build machine learning models that can learn from each other over time. This collaborative approach is key to creating AI platforms that can rival centralized alternatives. These platforms and their collaborative environments will enable composable (meaning interchangeable lego blocks of code) and extensible (easily adding more lego blocks to the code) features to be added to AI systems. This will accelerate AI technological advancements and ensure that these innovations are widely accessible, rather than gated behind corporate walls.

The centralization of AI presents mad risks, limiting creativity and raising barriers for new entrants. By learning from past mistakes and leveraging the principles of web3, we can avoid these traps and ensure that AI reaches its full potential. The convergence of web3 and AI not only addresses current challenges but also unlocks new possibilities for extensible, and composable AI systems. By putting our heads together, we can build a future where AI serves as a collective asset, empowering all of us rather than the cringe jogger and jeans billionaires.

As we explore the case for decentralized AI, we must remember that the choices we make today will shape the technological landscape for generations to come. Let's embrace this opportunity to create more innovative, inclusive, and equitable AI systems now.

Bless – Tanu