Introduction
Artificial Intelligence (AI) is one of the most exciting technological advancements of our time. It has the power to significantly boost human efficiency and facilitate groundbreaking discoveries, especially in the medical field. Though AI is already a key player in many industries today, its impact is expected to grow exponentially. According to PwC, AI could become a $15 trillion industry by 2030. AI is coming
Yet, as promising as AI is, it faces substantial challenges. As it becomes more powerful, the AI industry is increasingly centralized, concentrating influence within a few corporations, which may have negative consequences for society. Additionally, this centralization raises issues related to deepfakes, biases embedded in algorithms, and data privacy risks. Fortunately, crypto—especially its foundational traits of decentralization and transparency—can provide solutions to many of these issues.
In this analysis, we explore the problems that come with centralized AI and how decentralized AI could resolve some of these challenges. We also discuss the present-day relationship between AI and crypto, showcasing the early-stage crypto applications that have demonstrated potential for adoption.
The Dangers of Centralized AI
Today’s AI landscape is marked by a few key challenges, many of which are tied to the highly concentrated nature of its development. The significant capital requirements and network effects needed for AI innovation mean that only large corporations can afford to stay at the forefront, leaving small businesses and academic researchers struggling to access the resources they need. This stifles competition and slows overall innovation in the field.
Because of this, a small group of companies—like OpenAI and Google—now hold a disproportionate amount of control over AI development. This consolidation of power raises critical governance questions. For instance, earlier this year, Google’s AI image generator, Gemini, was found to contain racial biases and historical inaccuracies, shedding light on how companies can manipulate their models to serve their own interests. In another example, the abrupt firing of OpenAI’s CEO, Sam Altman, by a board of six people in November 2023, highlighted the fact that very few people wield immense control over companies at the cutting edge of AI development.
As AI becomes an even more integral part of our world, many worry that a single corporation could dominate the decision-making process around AI, quietly shaping the future to serve its interests while potentially harming society. Companies may impose strict guardrails, make decisions behind closed doors, or alter models for their benefit at the cost of public trust.
How Decentralized AI Offers Solutions
Decentralized AI is an approach that leverages blockchain technology to distribute ownership and governance of AI, aiming to improve transparency and accessibility. According to Grayscale Research, decentralized AI has the potential to break down the barriers created by centralized entities and place these critical decisions into the hands of the public.
By using blockchain, developers who may have otherwise struggled to access the resources they need could now gain entry. Blockchain-powered systems reduce the cost of building and monetizing AI solutions, creating a level playing field for independent developers. This could fuel greater innovation and competition and act as a counterbalance to the models created by large tech corporations.
Moreover, decentralized AI democratizes the financial opportunities connected to AI development. Currently, only a small group of accredited investors and venture capitalists can profit from AI’s financial upside, as most AI startups rely on private capital. In 2022 alone, $47 billion was invested in AI startups, followed by another $42 billion in 2023. However, decentralized AI assets allow anyone to own a share of the future of AI, enabling broader access to its financial rewards.
The Current State of Crypto and AI
Although still in the early stages, the intersection of crypto and AI is gaining traction, and the market’s response has been optimistic. In 2024, AI-related crypto assets returned 20% by May, outperforming most sectors of the crypto market except for currencies. Data from Kaito shows that AI is now the most talked-about theme on social platforms, outpacing other emerging topics such as decentralized finance (DeFi), Layer 2 solutions, memecoins, and real-world assets.
Several key figures have also begun advocating for decentralized AI as a solution to the problems posed by centralized AI. In March 2024, Emad Mostaque, founder of Stability AI, left his company to focus on decentralized AI, stating, "It is now time to ensure AI remains open and decentralized." Similarly, crypto entrepreneur Erik Vorhees recently launched Venice.ai, a privacy-focused AI platform that features end-to-end encryption.
Breaking Down the AI-Crypto Intersection
The intersection between AI and crypto can be categorized into three main areas:
1. Infrastructure for AI Development
This category includes networks that offer an open, permissionless architecture specifically designed for developing AI services. These platforms focus on providing the infrastructure and incentives to support various AI applications, not just a single product or service.
Notable Examples:
- NEAR Protocol: Founded by one of the co-creators of the Transformer architecture that powers models like ChatGPT, NEAR is focused on user-owned AI. In mid-2024, NEAR launched an incubator program to support the development of foundational models, data platforms for AI, and marketplaces for AI computing power.
- Bittensor: This platform uses the TAO token to incentivize AI development, hosting 38 different subnets that cater to use cases ranging from chatbots to financial predictions and image generation. Developers can query these subnets through a permissionless API, allowing them to build a wide range of AI applications.
2. Resources for AI Development
AI requires vast amounts of computational power and data storage, and this category focuses on the decentralized assets that provide these resources.
Notable Examples:
- Render (RNDR) and Akash (AKT): These decentralized GPU marketplaces offer AI developers access to computing power for tasks like model training and 3D rendering. Render alone provides around 10,000 GPUs, catering primarily to artists and generative AI projects.
- Filecoin (FIL) and Arweave (AR): These platforms provide decentralized, secure data storage, offering alternatives to centralized solutions like AWS. They also enhance data integrity and security by eliminating single points of failure.
Additionally, services like Grass and Masa allow individuals to monetize their data for AI training while maintaining control over their personal information, helping to level the playing field for smaller developers who lack access to real-time data like Google or Bing.
3. Addressing AI-Related Issues
The rise of AI has intensified problems related to deepfakes, misinformation, and bots. Several crypto projects are working on solutions to these challenges by focusing on content verification and ensuring trust in AI models.
Notable Examples:
- Worldcoin (WLD): This project aims to prove a person’s humanity using biometric identifiers, tackling the issue of bots.
- Origin Trail (TRAC) and Numbers Protocol (NUM): These platforms focus on establishing content provenance, ensuring that digital content—whether AI-generated or not—can be verified and trusted.
Additionally, platforms like Modulus Labs and Zama are developing solutions that use cryptography and zero-knowledge proofs to ensure the integrity of AI models, addressing the concern of manipulated AI outputs.
Conclusion
While decentralized AI assets have made significant strides, we are still in the early stages of this evolving space. Prominent voices in the tech world, such as venture capitalist Fred Wilson, have likened AI and crypto to "two sides of the same coin," predicting that Web3 technologies will help instill trust in AI. As AI continues to mature, these decentralized crypto solutions will become increasingly relevant, creating a mutually beneficial relationship between the two technologies.
AI is already reshaping our world, and its potential impact is only just beginning to be felt. By leveraging the inherent strengths of blockchain technology—decentralization, transparency, and accessibility—crypto can play a crucial role in mitigating the risks posed by AI, ensuring that this transformative technology benefits everyone, not just a select few.