Let's step back for a moment and consider AI's journey to where it is today. From its dawn in the 1950s, as vast machines that processed limited tasks, to the present, where AI mimics human touches in creative outputs like text and images, we can say it's developing faster than ever—and it is a precursor to a greater convergence with Web3.

Web3 and AI: A Shared Destiny

From Turing to GPT-4

Artificial Intelligence (AI) has undergone a remarkable transformation since Alan Turing first pondered the question, "Can machines think?" in the early 1950s. Turing's groundbreaking work laid the theoretical foundation for AI, but it was the advent of stored-program computers that truly set the stage for AI's evolution. The 1956 Dartmouth Conference, organized by John McCarthy, marked the formal birth of AI as a field of study. Despite the conference's lack of concrete outcomes, it ignited a wave of research and development that would see AI oscillate between periods of intense progress and stagnation.

The 1960s and 1970s witnessed the creation of early AI programs like ELIZA and Shakey the Robot, which showed rudimentary natural language processing and robotic navigation capabilities. However, the field faced setbacks during the "AI Winter" of the 1970s and 1980s, a period marked by dwindling funding and skepticism about AI's potential. It wasn't until the late 1990s and early 2000s that AI began to regain momentum, thanks to advances in computational power and machine learning algorithms, and even then, it was an esoteric field reserved for the truly dedicated.

At the turn of the millennium, landmark achievements included IBM's Deep Blue defeating world chess champion Gary Kasparov in 1997, and the development of Kismet, a robot capable of recognizing and simulating human emotions. These milestones were followed by the rise of neural networks and deep learning, popularized by researchers like Geoffrey Hinton, which paved the way for modern AI applications.

The release of OpenAI's GPT-3 in 2020 and GPT-4 in 2023 marked significant leaps in generative AI, capable of producing human-like text and images, setting the stage for the next era of AI innovation and catapulting the technology into the hands of the masses.

‘The Future’ is now

The current state of AI is already remarkable, with the most awe inspiring examples coming from the area of Generative AI. Humans crave closure. Answers = closure = safety. So Gen AI’s ability to give answers, regardless of their truthfulness (closer to truthiness) is bound to astound. The way the technology really works, is that it learns patterns from vast datasets, using neural networks to identify relationships between data points. It then applies these patterns to generate new, contextually relevant content. This output is often directionally useful, with potential to simulate complex scenarios, generate synthetic data for training other AI models, and even assist in scientific research by proposing novel hypotheses, based on data patterns from past data.

So what next?

We have this call/response interaction in play, between humans and AI’s, with increasingly truthful/useful/reliable answers being generated. So the question becomes, how do we move from questions and answers, to directives and actions? Enter Autonomous AI agents - "AI to AI loops", where the outputs of one query become the inputs for the next, and the loop continues until the ‘success criteria’ is met. For example, take the task of "planning a vacation." An autonomous AI agent could:

1. Analyze user preferences and budget

2. Research destinations, flights, and accommodations

3. Compare options and optimize for cost and experience

4. Book flights and hotels

5. Generate an itinerary

This process involves multiple AI systems working together, each handling specific subtasks and passing information to the next, ultimately delivering a comprehensive solution with minimal human intervention. Right now, only some of this is possible… up until we hit step 4.

How does web3 play a role?

Why, you might ask, do Web3 and AI need to lock arms? Well, currency itself is at the heart of the matter—a tool to transact value. As AI begins to reach its rich maturity, it struggles to live within economic models constructed for humans, by humans. Web3 emerges as the solution, much like a missing jigsaw puzzle piece.

The Synergy of AI and Digital Currency

As AI systems become increasingly intelligent, their autonomy depends on their ability to transact and interact with the world at large. Digital currency offers a common medium for AI entities to exchange value, procure resources, or even incentivize certain actions within the ecosystem, or to keep it simple, ‘book flights and accommodation’.

But digital currency is broad, so let’s look at the likely scenarios:

Scenario #1: integrating with existing fiat systems

AI's integration with traditional fiat currencies faces significant regulatory hurdles, slowing its implementation within existing financial frameworks, however, when/if it does happen, it would allow AI systems to manage budgets, optimize investments, and handle transactions using familiar monetary systems. As with all policy, progress will be hampered by strict oversight. AI must navigate a complex landscape of international exchange rules, banking regulations, and legislative processes that move at a glacial pace. While this sounds like a negative outlook, I believe in policy that is well thought out, and solid, and in an ideal world, this one would be the real scenario. However, we are moving at the pace of technology, not legislation.

Scenario #2: Cryptocurrencies and Independent AI Economies

Cryptocurrencies offer AI a decentralized, borderless alternative. By utilizing crypto, AI can engage in rapid, peer-to-peer transactions without intermediaries. This opens up possibilities for AI-to-AI commerce, automated smart contracts, and operation in areas where traditional banking is limited. However, the volatility of many cryptocurrencies and regulatory uncertainties pose challenges for widespread adoption.

There’s potential in this emerging ecosystem, for AI agents to engage in a dynamic marketplace where value is generated through solely AI-driven activities such as data processing, predictive modeling, and automated decision-making. The key innovation lies in how this value is redistributed. Smart contracts can be programmed to automatically allocate rewards to various stakeholders, including:

1. Data contributors who provide training data

2. Developers who create AI models and applications

3. Node operators who maintain the network infrastructure

4. End-users who benefit from AI services

This system ensures that value is not concentrated in the hands of a few large corporations but is instead distributed across the entire network of participants. For example, a decentralized AI marketplace could allow individuals to monetize their personal data by selling it directly to AI models for training purposes, receiving cryptocurrency rewards in return. Furthermore, these economies can implement automated Universal Basic Income (UBI) systems, where a portion of the value generated by AI activities is redistributed to all network participants. This creates a more equitable economic model that benefits the broader community. By leveraging the transparency and immutability of blockchain, these AI economies also foster trust and accountability. Every transaction and value transfer is recorded on the distributed ledger, allowing for fair and transparent redistribution of rewards. In essence, this fusion of AI and blockchain technologies is creating a new economic paradigm where value creation and distribution are more democratic, transparent, and equitable, moving away from the centralized models that currently dominate the AI landscape.

Scenario #3: Governments work to retain control in an inevitable transition

Central Bank Digital Currencies (CBDCs) represent a nation's response to the digital currency movement, offering a digital form of fiat money. They aim to provide the benefits of cryptocurrency—fast transactions, security, and reduced costs—while remaining under the periphery of a country's monetary policy. For AI, CBDCs could become the preferred medium for transactions—being stable, regulated, and universally accepted within a country's ecosystem. This scenario is the most likely for widespread adoption, and opens the door to AI dominating trading floors and financial asset portfolios. For example, a system could automatically manage an energy producers portfolio based on real-time economic data to continually hedge.

In reality, we're likely to see a messy coexistence of all these options. AI systems will need to be versatile, capable of operating across fiat, crypto, and CBDC ecosystems as different use cases and regulatory environments demand.

Looking Ahead

As AI expands, we are likely to see them become economic agents in their own right, complete with the ability to manage resources, participate in markets, and drive value. Whether through CBDCs, cryptocurrencies, or fiat (if regulations allow), the ability for AI to engage autonomously in the financial system marks a profound step forward in our ability to get things done. It will not just change how businesses operate but also challenge our very notions of value, currency, and economic participation.

Fusing Web3 with AI will create a future where AI has economic agency, disrupting and redefining the world as we know it.

This future is uncertain, with potential for massive prosperity, or fallout - maybe both.

My questions are; will we have systems that are auditable? Are we standing in our own way? Is all value destined to accrue to the owners of only a handful of companies? How do we create prosperous markets that free people to lead with creativity, without needing to fulfill the mundane aspects of delivery? And most importantly, what is the nature of control in a world where humans are not always the decision makers?