Blockchain and Artificial Intelligence: Two of the most epoch-defining technologies of our time. In each realm, each has been a powerful and groundbreaking force in its own right, like Godzilla and King Kong. However, Kong and Godzilla occasionally team up when faced with monsters neither of them can defeat alone. (They will do the same next year Godzilla x Kong: New Empire.🍿) Now, imagine the possibilities unlocked when you combine the best of both worlds—AI and blockchain—to tackle massive problems.
These are the possibilities I recently had the opportunity to explore when I moderated a panel at Coinbase’s inaugural Machine Learning (ML) and Blockchain Summit. Convening his four leaders from academia and industry, the panel uncovered opportunities at the intersection of these two rapidly evolving technologies. Our conversations ran along many themes, from how blockchain could accelerate his AI development, to the complexities of working with blockchain data, to the potential of large-scale language models (LLMs). was broken.
One of the big advantages of AI x Blockchain is the use of cryptographic digital signatures and timestamps to address the problem of fake data and content, which is becoming an exponentially bigger problem with the spread of AI. It’s about being able to clarify what’s real because it can counteract misinformation. and manipulated. At the same time, AI will improve blockchain network efficiency, enhance security, and unlock new capabilities such as enabling protocols to make decisions based on real-time on-chain data.
Rather than recap all possible synergies, I think it’s best to let your colleagues tell. Read their takes, edited for clarity and length.
Bhaskar Krishnamachari, Professor of Electrical and Computer Engineering and Computer Science, University of Southern California
In my view, there are two main areas where blockchain and AI intersect. The first applies ML models to tackle blockchain challenges, and the second uses blockchain to address pressing problems in AI.
In the first scenario, ML models can uncover complex patterns within blockchain data and help improve the performance of on-chain decentralized applications. Analyzing transaction data can expose potential fraud such as wash transactions and illegal fund transfers, and detect new security threats. ML models not only help ensure the security of blockchain networks, but they can also improve performance. For example, you can dynamically adjust transaction fees based on transaction volume or optimize system resources during peak usage periods.
There is not much discussion about how blockchain can help AI development. As the foundation for borderless, internet-native payment systems, blockchain can create financial incentives for those who contribute data and computing resources to train ML models. To enable this, we’ve been doing research at USC on decentralized data marketplaces.
In recent years, we’ve seen a handful of tech companies capture an ever-larger share of the world’s data and AI power. This raises concerns about privacy, bias, and security, all of which blockchain can address as a decentralized, transparent, and openly auditable system. For example, blockchain can track the origin of data used to train AI models and cryptographically verify its authenticity. By ensuring that these inputs are unaltered and unbiased, blockchain helps increase confidence in the recommendations provided by AI systems.
Leo Liang, Head of Data Platforms and Services, Coinbase
At Coinbase, most of the challenges my team faces are data related. Specifically, we need to extract the data from the blockchain and transform it into a format that can be used by ML models. I like to think of blockchain like an onion. The reason is that blockchain has countless layers of complexity. Its distributed nature means that data is spread over many nodes, each independently validating and adding new blocks. The complexity increases even further when multiple blockchains are involved. Now you are dealing with a network of interconnected onions. Synchronizing data and ensuring consistency across this vast and diffuse ecosystem is no small feat.
Furthermore, blockchain is a self-contained system and knowledge of the world cannot be accessed beyond its boundaries. For ML models to make real-world predictions, they need to combine on-chain data (data stored on the blockchain) and off-chain data (data outside the blockchain such as stock prices, exchange rates, weather patterns, etc.). there is. upon). Think of it like connecting a blockchain to the internet. It’s a fascinating yet challenging engineering puzzle.
Sam Green, Co-Founder and Head of Research at Semiotic Labs
At Semiotic Labs, he leads AI R&D efforts for The Graph, a decentralized protocol for manipulating and utilizing blockchain data. Simply put, the graph reads data from the blockchain, processes it, and creates an index similar to the alphabetical list at the end of an encyclopedia. This organizational structure simplifies the retrieval of data from the blockchain. By “indexing” blockchain data in this way, The Graph transforms it into a format that is easy to query, analyze, and apply in downstream applications.
Transactions in The Graph involve two main participants: data sellers (indexers) and data buyers (consumers). These entities interact through what we call “gateways”. When a consumer submits a query to the gateway, the gateway distributes the query among the indexers considering factors such as bid price, quality of service, and latency. Indexers generate revenue by processing these queries and distributing blockchain data to consumers. With the help of AI, we built an algorithmic pricing agent that helps indexers maximize their returns while ensuring consumers receive reliable, high-quality service.
In many ways, blockchain is an ideal environment for training AI agents. All rules defined by smart contracts and player actions recorded in transactions are publicly viewable on-chain. Since these rules and actions are known, we can create simulations of this blockchain “game”, use them to train AI agents, and then deploy these agents in a live setting. The secret is in a fast feedback loop. The faster the trial-and-error learning rate, the faster the agent can improve its performance.
As we look to the future, we see immense possibilities for integrating LLM into The Graph. Currently, users have to query The Graph with her specialized language called GraphQL. In contrast, LLM allows users to express their requests in natural language. By allowing anyone to interact with The Graph in plain English, LLM can further democratize access to blockchain data.
Paul Bohm, Founder of Teleport
Teleport is developing an open marketplace for ridesharing. Ride sharing is currently a closed system, making it difficult for users to switch between different services. If e-mail is shut down like rideshare, users of Microsoft’s Outlook Mail and Apple’s iCloud Mail will not be able to e-mail each other. Similarly, Apple’s Safari browser will not be able to communicate with Microsoft.com if the web is closed.
Opening up ridesharing means bringing ridesharing back to internet standards. An open system allows participants to choose from different apps from different vendors that communicate with each other. Closed markets often fail to set fair prices. Instead, they set the price themselves and maximize the value they can extract. Opening up ridesharing and removing this middleman means more money is paid to drivers, passengers pay less per ride, and more money remains in the local economy.
To be successful, an open marketplace must be trustworthy. Engineers often focus first on aspects of technology, such as speed and new features. But when building for the real world, you must start with your users’ needs for safety, security, and privacy. Only then can you determine the best technology to meet these needs. not the other way around.
These are just a few of the possibilities, and just the start of a conversation about what blockchain and machine learning together can unblock, enhance, enhance, and take to new heights. I’m sorry. Digital consensus technologies like blockchain enable the design of systems that are not only fair, reliable and secure, but are proven to be so in practice. As AI threatens to further erode trust, blockchain strengthens it and provides robust mechanisms to protect the integrity of sensitive data. AI, on the other hand, makes it possible to understand the depth of distributed data that makes blockchain unwieldy and unsuitable for mass adoption. Bringing artificial intelligence to a problem of this inhuman scale could bring blockchain to his billion users.
For blockchain and AI entrepreneurs, these are the heart-opening and exhilarating prospects of not just one technology but both working together and becoming more powerful. AI and blockchain. Godzilla and Kong. Atomic Fire and Gorilla Punch. This is how we take it to the next level. Go ahead and be a legend.
follow me twitter Or LinkedIn. check out my website.
