In the past two years, as artificial intelligence has risen, AI has penetrated various industries including manufacturing, e-commerce, advertising, and pharmaceuticals, with the cryptocurrency sector being no exception. The integration of artificial intelligence and blockchain has led us to witness unique digital assets — AI crypto tokens.
Its popularity began at the end of 2022, with the explosive success of OpenAI's chatbot ChatGPT, leading many to realize that artificial intelligence is no longer just a feature in movies; more applications have entered reality, and AI has become an efficient productivity tool in various industries.
The AI frenzy has also influenced institutional giants. For example, Google announced it would develop its own AI chatbot, Bard. Additionally, noteworthy news includes Microsoft's acquisition of OpenAI for $10 billion and its proposal to integrate it into its Bing search engine. This sustained mainstream interest in AI technology has led to explosive growth in the market value of various AI tokens, with some increasing by as much as 1600%!
So what is an AI token? How does it combine with Web3, and what is the future direction? Let's discuss these issues below.
What is an AI token?
AI tokens are crypto assets that integrate AI principles into blockchain technology. The AI elements of such tokens enable them to develop better automated strategies that can solve specific problems. Due to their intelligence more effectively adapting to market conditions, they have advantages over other crypto assets.
AI tokens are cryptocurrencies that support AI-based projects, applications, and services in the blockchain ecosystem. They primarily serve three key roles:
Facilitating transactions, they serve as exchange mediums within the AI-driven platform, allowing users to pay service fees, access data, and participate in platform activities.
They can also serve as governance tokens, where these tokens grant their holders governance rights, allowing them to participate in shaping the development of AI projects or platforms.
They can also be used as rewards to incentivize users to contribute to AI protocols or projects, generally earned by providing data, computing resources, etc.
AI + Web3 infrastructure
The main projects in the infrastructure layer of the AI + Web3 industry are primarily framed around decentralized computing networks, with low cost as the main advantage and token incentives as the primary method for network expansion, serving AI + Web3 clients as the main goal.
Infrastructure is a certain growth direction for AI development
Explosive growth in AI computing power demand
In recent years, the demand for computing power has rapidly increased, especially after the release of large language models (LLMs). The demand for AI computing power has ignited the high-performance computing market. OpenAI's data shows that since 2012, the computing usage for training the largest AI models has grown exponentially, doubling every 3-4 months, significantly outpacing Moore's Law.
At the same time, the need for vast amounts of data places demands on storage and hardware memory, especially during the model training phase, which requires a large amount of parameter input and data storage. The storage chips used in AI servers mainly include: High Bandwidth Memory (HBM), DRAM, and SSDs, which need to provide greater capacity, higher performance, lower latency, and faster response speeds for AI server work scenarios.
Supply-demand imbalance drives high computing power costs
With the development of large models, computational complexity has also surged, necessitating more high-end hardware to meet model training demands. For example, in the case of GPT-3, with 13 million unique users accessing it, the corresponding chip demand is over 30,000 A100 GPUs. Thus, the initial investment cost will reach an astonishing $800 million, with daily model inference costs estimated at $700,000. As a result, the rising demand for high-end GPUs and supply chain disruptions are driving up the high prices of current hardware like GPUs.
AI infrastructure occupies the core value growth of the industrial chain
According to a report by Grand View Research, the global cloud AI market is estimated to be worth $62.63 billion in 2023, and is projected to grow to $647.6 billion by 2030, with a compound annual growth rate of 39.6%. This data reflects the growth potential of cloud AI services and their significant share in the entire AI industrial chain.
Narrative logic of AI + Web3 infrastructure projects
The demand for distributed AI infrastructure is strong and has long-term growth potential, making it a field that is easy to narrate and favored by capital. Currently, the main projects in the infrastructure layer of the AI + Web3 industry are primarily framed around decentralized computing networks, with low cost as the main advantage and token incentives as the primary method for network expansion, serving AI + Web3 clients as the main goal. It mainly includes two levels:
1. A pure decentralized cloud computing resource sharing and leasing platform: There are many early AI projects, such as Render Network, Akash Network, etc.;
Computing power resources are the main competitive advantage: core competitive advantages and resources often involve access to a large number of computing power providers, quickly establishing their foundational networks while offering user-friendly products to clients.
Low product thresholds and fast launch speeds: For mature products like Render Network and Akash Network, tangible growth data is already visible, providing a certain competitive advantage.
Homogenization of new entrants' products: Due to the current hot spots in the sector and the low thresholds for such products, a large number of projects focused on shared computing power and computing power leasing have recently emerged; however, the products are relatively homogeneous, necessitating more differentiated competitive advantages.
Tending to serve clients with simple computing needs: For instance, Render Network primarily serves rendering needs, while Akash Network's resource offerings include more CPU.
2. Providing decentralized computing + ML workflow services: There are many recently funded emerging projects, such as Gensyn, io.net, Ritual, etc.;
Decentralized computing elevates valuation foundations. Since computing power is a certain narrative in AI development, projects with a foundation in computing power often have more stable and high-potential business models, resulting in higher valuations compared to purely intermediary layer projects.
Intermediary layer services provide differentiated advantages. The services of the intermediary layer are parts of these computing power infrastructures that have competitive advantages, such as oracles and validators that synchronize on-chain and off-chain computations for AI, as well as deployment and management tools for the overall AI workflow.
