Does the Internet Really Need Blockchain-Powered AI?

In a world where blockchain and artificial intelligence (AI) are often hyped as revolutionary technologies, it's easy to be skeptical of projects claiming to combine the two. However, Humayun Sheikh, founder and CEO of Fetch.ai, has been working at the intersection of blockchain and AI since 2017, long before the current hype cycles. In this episode of Validated, Sheikh shares his compelling vision for how decentralized AI could transform internet transactions and create truly peer-to-peer systems.

The Reality of Blockchain and AI Intersection

Sheikh begins by addressing the fundamental question of whether the intersection of blockchain and AI is real or just hype. He argues that both technologies have genuine use cases and value. AI and machine learning have been around since the 1950s, but recent advancements in computing power and data generation have accelerated their development. Blockchain, on the other hand, brings decentralization and enables trustless record-keeping.

The combination of these technologies, according to Sheikh, offers unique advantages. While there is undoubtedly hype surrounding both blockchain and AI, Sheikh contends that there are real use cases for their integration.

Components of Decentralized AI

Sheikh breaks down decentralized AI into three main components:

1. Data: Currently, data is generated in a decentralized manner but collected and stored centrally.
2. Models: AI models can be owned and controlled by centralized entities.
3. Usage: Who can use the models and how they can be accessed.

Sheikh emphasizes that while it's currently too expensive to store entire AI systems on a blockchain, decentralized storage solutions like IPFS can be utilized. The key is not just about storage but also about creating auditable records of data access and usage.

Fetch's Value Proposition

Fetch.ai aims to simplify and automate complex transactions by removing unnecessary intermediaries. Sheikh explains:

"The whole purpose where Fetch was born was to make that a much simpler process, where automation can be carried out, and all these intermediaries were not necessary now with the technological development that has happened, is to remove those which would improve the economic efficiency and actually the transactional efficiency of the whole task."

The goal is to return to the original promise of the internet: peer-to-peer connections without centralized aggregators.

Blockchain as an Orchestrator

One of the key innovations of Fetch is using blockchain technology as an orchestrator for its AI agents. Sheikh explains:

"We use the ledger technology, the decentralized ledger technology as the orchestrator. So if you think about, there's millions of agents running around, but how do you find them? So where's this Google of these agents? How do you find them? And what we're doing is we're saying, okay, that search does not need to be centralized. It can be decentralized and enables you to find these agents."

This approach allows for a more flexible and decentralized system of AI agents that can interact and perform tasks without relying on a central authority.

Agents as Microservices

Sheikh compares Fetch's AI agents to microservices in web development. These agents are essentially small, specialized programs that can perform specific tasks. He explains:

"So if you think about microservices and you think of bringing them to web three, then that's what an agent looks like effectively. So you have all these small tasks and all these small microservices effectively, or we call them microagents, delivering a single task."

This approach allows for a more modular and flexible system, where different agents can be combined to perform complex tasks.

Practical Applications of Fetch Agents

While Fetch is still in the early stages of deployment, Sheikh shares some potential use cases. One area of focus is the automotive industry, where Fetch is collaborating with Bosch on predictive maintenance solutions. Sheikh explains that Bosch is forward-thinking in recognizing that the data generated by their products ultimately belongs to the consumers, and there should be a way to pass economic value back to them.

Sheikh elaborates on how this might work:

"So what needs to happen and the reason why they're collaborating with us is that you have all these data points which are coming from various devices. You build machine learning models in the middle, which enables all these agents to do these transactions without the need for going into a one big centralized entity."

This approach could allow for more efficient and user-centric systems for managing and maintaining consumer products.

Overcoming Data Monetization Challenges

When asked about the challenges of monetizing user data, Sheikh acknowledges that previous attempts at creating data marketplaces have largely failed. He argues that people are not necessarily interested in selling their data directly, but rather in receiving better services:

"People are not interested in collecting the data and selling the data. So what are they interested in? What they're interested in is it can I get a better service, right?"

Sheikh suggests that the value proposition of Fetch's system is not in directly paying users for their data, but in using that data to provide better, more personalized services through AI agents.

Trust and Reputation in Decentralized Systems

One of the challenges of decentralized systems is establishing trust and reputation. Sheikh acknowledges the importance of this issue and suggests that blockchain technology can actually enhance trust:

"The good thing is that blockchain enables that trust layer to build. So it might not start day one. I mean, you could still pick up the rating, the star rating, you could still enable the same techniques that exist already. But what is quite interesting is that when you do transaction on the blockchain, you can actually take those transactions and you can actually enable that trust layer to be built within the system."

He proposes the use of self-sovereign identities and portable reputation systems that could be used across different services and platforms.

Addressing Concerns About Unregulated AI

When asked about the potential risks of open, unregulated AI systems, Sheikh emphasizes the importance of having multiple models and systems:

"One of the key elements of what we're building is this compatibility with many models. So if you feel one model is biased or is not giving you a fair representation, you can always change to another supplier of another model."

He also argues that the risk of unregulated AI exists in centralized systems as well, and that a decentralized approach could potentially provide more transparency and auditability.

The Vision for Fetch's Future

Sheikh's vision for Fetch aligns with the original promise of the internet: a decentralized system where people can access services and conduct transactions without relying on large intermediaries. He sees potential applications in areas like the gig economy, where people are seeking more control over their economic activities.

Sheikh summarizes his vision:

"The future for me is bringing back the promise of the internet which was we're going to be able to connect with things. We're going to be able to transact with things, automate a lot of the tasks which are just silly which is like picking up the phone to five people to find a plumber. I mean, you know, why do we need to do that today?"

Implications for Solana and the Wider Blockchain Ecosystem

While the discussion doesn't directly mention Solana, the concepts and technologies discussed by Sheikh have significant implications for the Solana ecosystem and other high-performance blockchain platforms.

Solana's high throughput and low transaction costs make it an ideal platform for hosting decentralized AI systems like those proposed by Fetch. The ability to quickly and cheaply record interactions between AI agents on a blockchain could be crucial for creating auditable and trustworthy decentralized AI networks.

Moreover, Solana's growing ecosystem of decentralized finance (DeFi) and non-fungible token (NFT) projects could benefit from the integration of AI agents. These agents could help automate complex DeFi strategies, optimize NFT trading, or even create new forms of AI-generated digital art on the Solana blockchain.

The Role of Decentralized AI in Web3

Sheikh's vision for decentralized AI aligns closely with the broader goals of the Web3 movement, which seeks to create a more open, decentralized internet. By combining blockchain technology with AI, projects like Fetch aim to disintermediate traditional centralized services and empower users with more control over their data and digital interactions.

This approach could be particularly relevant for Solana-based projects looking to push the boundaries of what's possible in the Web3 space. The combination of Solana's high-performance blockchain and sophisticated AI agents could enable new forms of decentralized applications that are more intelligent, adaptive, and user-centric than current offerings.

Challenges and Opportunities for Decentralized AI

While Sheikh presents a compelling vision for decentralized AI, there are still significant challenges to overcome. These include:

1. Scalability: Ensuring that decentralized AI systems can operate at scale without compromising performance or security.
2. User adoption: Convincing users to switch from familiar centralized services to new decentralized alternatives.
3. Regulatory compliance: Navigating the complex regulatory landscape surrounding AI and blockchain technologies.
4. Data privacy: Balancing the need for data to train AI models with users' privacy concerns.
5. Interoperability: Ensuring that different AI agents and blockchain systems can work together seamlessly.

Despite these challenges, the potential benefits of decentralized AI are significant. By combining the trustlessness and transparency of blockchain with the power of AI, projects like Fetch could help create a more efficient, fair, and user-centric digital economy.

Implications for DeFi and Other Blockchain Applications

The integration of AI agents into blockchain systems could have far-reaching implications for decentralized finance (DeFi) and other blockchain applications. Some potential use cases include:

1. Automated market making: AI agents could optimize liquidity provision and trading strategies in decentralized exchanges.
2. Risk assessment: AI models could analyze on-chain data to provide more accurate risk assessments for DeFi lending protocols.
3. Fraud detection: Machine learning algorithms could help identify suspicious transactions and protect users from scams.
4. Personalized financial services: AI agents could offer tailored financial advice and product recommendations based on a user's on-chain activity and preferences.
5. Supply chain optimization: In non-financial applications, AI agents could help optimize logistics and supply chain management on blockchain-based systems.

The Future of Internet Transactions

Sheikh's vision for the future of internet transactions is one where AI agents, powered by blockchain technology, handle many of the mundane tasks that currently require human intervention or centralized intermediaries. This could lead to a more efficient and user-friendly digital economy, where individuals have greater control over their data and digital interactions.

For platforms like Solana, which prioritize high performance and low costs, this vision presents exciting opportunities. The combination of Solana's scalable infrastructure and AI-powered decentralized applications could enable a new generation of digital services that are faster, cheaper, and more personalized than current offerings.

Conclusion: A New Paradigm for the Internet

While the integration of blockchain and AI is still in its early stages, the vision presented by Humayun Sheikh and Fetch.ai offers a compelling glimpse into a possible future for the internet. By combining the decentralization and trustlessness of blockchain with the intelligence and automation capabilities of AI, we may be able to create a digital ecosystem that is more efficient, fair, and user-centric than ever before.

For Solana and its ecosystem, this presents both challenges and opportunities. As a high-performance blockchain platform, Solana is well-positioned to support the kind of decentralized AI applications envisioned by Sheikh. However, realizing this potential will require continued innovation, collaboration, and a focus on solving real-world problems.

As we move forward, it will be crucial for projects in the Solana ecosystem to explore the possibilities offered by decentralized AI while remaining mindful of the challenges and potential risks. By doing so, Solana could play a key role in shaping the future of the internet and digital economy.

Facts + Figures

• Humayun Sheikh is the founder and CEO of Fetch.ai, a blockchain-based AI platform.
• Sheikh has been working on the intersection of blockchain and AI since 2017.
• AI and machine learning have been around since the 1950s, but recent advancements in computing power and data generation have accelerated their development.
• Fetch.ai aims to simplify and automate complex transactions by removing unnecessary intermediaries.
• Fetch uses blockchain technology as an orchestrator for its AI agents.
• Fetch's AI agents are compared to microservices in web development, performing specific tasks.
• Fetch is collaborating with Bosch on predictive maintenance solutions in the automotive industry.
• Sheikh proposes the use of self-sovereign identities and portable reputation systems for building trust in decentralized systems.
• The integration of AI agents into blockchain systems could have implications for DeFi, including automated market making, risk assessment, and fraud detection.
• Solana's high throughput and low transaction costs make it a potential platform for hosting decentralized AI systems.

Questions Answered

What is the intersection of blockchain and AI?

The intersection of blockchain and AI combines the decentralization and trustless record-keeping of blockchain with the data processing and decision-making capabilities of AI. According to Humayun Sheikh, this combination offers unique advantages by enabling decentralized AI systems that can operate without relying on centralized intermediaries. These systems can potentially improve efficiency, transparency, and user control in various applications, from financial services to supply chain management.

How does Fetch.ai use blockchain technology?

Fetch.ai uses blockchain technology as an orchestrator for its AI agents. Sheikh explains that the blockchain serves as a decentralized ledger and search mechanism for finding and coordinating AI agents. This approach allows for a more flexible and decentralized system of AI agents that can interact and perform tasks without relying on a central authority. The blockchain also provides a transparent and auditable record of agent interactions and data usage.

What are the potential applications of decentralized AI?

Decentralized AI has numerous potential applications across various industries. In the automotive sector, Fetch is collaborating with Bosch on predictive maintenance solutions. In finance, AI agents could optimize liquidity provision and trading strategies in decentralized exchanges, provide more accurate risk assessments for lending protocols, and offer personalized financial advice. Other applications include fraud detection, supply chain optimization, and automating complex tasks in the gig economy.

How does Fetch.ai address concerns about data privacy and monetization?

Sheikh acknowledges that previous attempts at creating data marketplaces have largely failed. Instead of directly monetizing user data, Fetch.ai focuses on using data to provide better, more personalized services through AI agents. The system aims to give users more control over their data while still benefiting from improved services. Additionally, the use of blockchain technology allows for more transparent and auditable data usage, potentially addressing some privacy concerns.

What are the challenges facing decentralized AI systems?

Decentralized AI systems face several challenges, including scalability, user adoption, regulatory compliance, data privacy, and interoperability. Ensuring that these systems can operate at scale without compromising performance or security is crucial. Convincing users to switch from familiar centralized services to new decentralized alternatives may also be challenging. Navigating the complex regulatory landscape surrounding AI and blockchain technologies and balancing the need for data to train AI models with users' privacy concerns are additional hurdles that need to be addressed.

How might decentralized AI impact the future of internet transactions?

According to Sheikh's vision, decentralized AI could revolutionize internet transactions by creating truly peer-to-peer systems that don't rely on centralized intermediaries. AI agents, powered by blockchain technology, could handle many mundane tasks that currently require human intervention or centralized platforms. This could lead to a more efficient and user-friendly digital economy, where individuals have greater control over their data and digital interactions. The result could be faster, cheaper, and more personalized digital services.

What role could Solana play in the development of decentralized AI?

While not directly mentioned in the discussion, Solana's high-performance blockchain could be an ideal platform for hosting decentralized AI systems like those proposed by Fetch.ai. Solana's high throughput and low transaction costs could enable the quick and cheap recording of interactions between AI agents, which is crucial for creating auditable and trustworthy decentralized AI networks. The integration of AI agents with Solana's growing ecosystem of DeFi and NFT projects could also lead to new and innovative applications in these spaces.

How does Fetch.ai address the issue of trust in decentralized systems?

Fetch.ai addresses the issue of trust in decentralized systems through several mechanisms. Sheikh proposes the use of self-sovereign identities and portable reputation systems that could be used across different services and platforms. He also suggests that blockchain technology can enhance trust by providing transparent and auditable records of transactions and interactions. Additionally, the decentralized nature of the system allows users to switch between different AI models if they feel one is biased or unreliable, providing a form of market-based trust mechanism.