‍

NEAR Protocol has embarked on a strategic mission to establish itself as a central hub for user-owned artificial intelligence (AI), adding to its impressive technology stack that is also an industry leader in sharding, data availability, chain abstraction, and more. These user-friendly features have led Near to become one of the most widely used and adopted blockchains in the industry, as detailed in the first portion of this report.

AI, a transformative technology of the modern era, has primarily seen advancements within centralized, profit-driven entities. NEAR aims to disrupt this trend by positioning its ecosystem at the forefront of decentralized AI research and applications.

Importantly, AI applications developed on NEAR will not be confined within its ecosystem; they will be accessible and composable across the broader blockchain landscape. This interoperability is poised to drive widespread adoption and innovation within the AI sector.

NEAR's long-term strategy involves the establishment of the NEAR.AI R&D Lab, spearheaded by NEAR co-founders Alex Skidanov and Illia Polosukhin. The lab’s initial objective is to develop an "AI Developer," capable of creating end-to-end Web3 applications from user intents. This endeavor will be supported by a dedicated team of AI researchers and significant investments in cutting-edge research. Given the founders' extensive AI backgrounds, NEAR is uniquely positioned to advance the vision of user-owned AI.

Growing Adoption in 2024

Throughout 2024, NEAR has seen a notable increase in adoption/user metrics driven by non-speculative use cases. Integral to the bump in activity are KaiKai and Hot Protocol, two of the most widely used decentralized applications (dApps) in the cryptocurrency space. NEAR currently boasts 1.4 million daily active users, positioning it at the forefront of all smart contract platforms. Despite the elevated usage, Near’s transaction throughput and fees remain competitive with leading platforms like Solana.

‍

‍

‍

Despite this significant user engagement, NEAR lags behind its competitors in terms of monetization, having generated only $4.1 million in fees year-to-date. While the platform demonstrates high transaction processing capabilities, it has yet to achieve the level of value accrual for token holders or stakers that is seen with its larger counterparts.

‍

The growing adoption of NEAR, even with its current monetization challenges, marks a crucial phase in its development. Should NEAR continue to expand its user base or successfully increase average transaction fees without diminishing network activity—much like the recent advancements seen with Solana—it has the potential to achieve meaningful value accrual.

Much of Near’s growth and adoption can attributed to its focus on customer UX and removing the many frictions that come with interacting with a traditional blockchain and crypto wallet. UX initiatives like Chain abstraction (covered in this former report), chain signatures, and the Blockchain Operating System (covered here) have all proven fruitful endeavors for the project. In addition to those rollouts, Near has also recently introduced even more user-friendly tech and wallet upgrades.

Upgrade from Four to Six Shards

The rapid growth and increased user engagement with the NEAR Protocol necessitated significant upgrades to its infrastructure. Each shard operates independently, processing transactions and maintaining its state. This segmentation allows for parallel processing, significantly enhancing the network's capacity and speed. Users access the required shard directly, bypassing the need to engage the entire blockchain, which leads to efficient data handling and improved user experience.

‍

In Q2 2024, Near successfully added two more shards, increasing the total to six and thereby enhancing the network's capacity by 50%. Anticipating further advancements, NEAR plans to launch stateless validation on Mainnet by the end of 2024. This upgrade is projected to enhance throughput for each shard by a factor of five and expand the number of shards available, facilitating a more robust and scalable network.

NEAR Protocol's implementation of sharding stands out due to its seamless interaction and synchronization among shards. Developers do not need to concern themselves with the specific shard on which a contract might reside. This approach ensures high security and interconnectivity, facilitating effective network scaling. As user demand grows and applications become more complex, additional shards can be integrated, much like expanding a library with more rooms to accommodate an increasing collection of books.

Introducing Telegram-based NEAR Wallet for mass adoption

In Q2 2024, the NEAR Wallet was integrated with Telegram. The wallet offers an easy way to transfer tokens like NEAR stablecoins and to mine HOT tokens directly via Telegram accounts. By utilizing NEAR Protocol's chain abstraction technology, the wallet connects users to dApps across various blockchains without the user’s knowledge. Features such as account abstraction and meta-transactions remove the need for repetitive signatures, while "User.tg" addresses enhance the user experience.

Additionally, the NEAR Wallet supports $HOT-based games. $HOT, a digital token on NEAR, can be easily mined through this Telegram-based wallet system. Users can create wallet addresses linked to their Telegram handles, simplifying interactions and transactions. The mining process is straightforward, allowing users to collect tokens regularly. Initially, transactions are free, with the option to pay fees in $HOT tokens as users advance. This integration with NEAR Protocol offers a secure and efficient environment for managing and mining $HOT tokens, making it appealing for users seeking an easy entry into the crypto world without significant upfront investments.

‍

AI

Finally, Near has recently begun to transition to incorporating AI into its already proven L1. The crypto x AI intersection has been a leading narrative in 2024, promising legions of benefits and improvements across both industries. Before digging into Near’s specific approach, let’s review some of the talking points in the crypto x AI narrative.

Early Crypto x AI 

The combination of cryptocurrencies and AI has led to decentralized networks designed specifically for AI use cases. Blockchain's tamper-evident nature ensures the authenticity and integrity of AI datasets, reducing the risk of using falsified or biased data. This transparency is crucial for sectors like finance, healthcare, and legal systems, where trust in AI operations is paramount.

Blockchain also enables privacy-preserving AI systems through encryption techniques and decentralized storage, ensuring sensitive data remains under owner control. Decentralized AI networks can distribute computational resources more equitably among participants, preventing monopolization by powerful entities. This model promotes a more inclusive and democratic AI landscape, balancing power dynamics and fostering innovation.

AI and cryptocurrency’s convergence has the potential to address pressing digital challenges, such as:

• Deepfakes: AI's pattern recognition capabilities combined with blockchain's tamper-proof characteristics can effectively identify and nullify deepfakes.
• Data Privacy: Cryptographic techniques in blockchain safeguard user information while enabling AI systems to learn from data without compromising privacy.
• Power Concentration: Decentralization distributes control across multiple nodes, balancing power dynamics and mitigating monopolistic practices in tech industries.

Projects such as Bittensor, Fetch.ai, Akash Network, Render Network, and Gensyn showcase the promising early applications of integrating AI and blockchain technology. For instance, Bittensor is a decentralized network designed to democratize AI by establishing a platform for various decentralized commodity markets unified under a single token system. This project utilizes unique incentive mechanisms and a subnetwork architecture to create a competitive environment, encouraging the development and enhancement of AI models.

The convergence of AI with blockchain addresses scalability and accessibility issues prevalent in traditional cloud-based AI services. Crypto-economic models, such as those employed by platforms like Gensyn, use tokens to incentivize computational resource contributions. This approach reduces AI development costs and lowers entry barriers, fostering a diverse and dynamic ecosystem of AI innovation. By distributing computational tasks globally, decentralized AI can scale effectively, meeting the increasing demand for AI services without centralized system constraints.

Potential Niches for the Crypto x AI Combination

Market Trend Forecasting

AI algorithms analyze vast data sets to identify patterns and predict market trends. This capability helps traders make informed decisions and optimize their investment strategies. By leveraging AI, investors gain access to predictive insights, reducing the uncertainty inherent in cryptocurrency markets. These algorithms consider considerable variables, including historical data, market sentiment, and real-time transactional information, providing a comprehensive view of market dynamics.

Security Enhancement and Verification

Blockchain technology's decentralized nature offers security benefits over centralized systems in certain scenarios. A significant impact area is data management and security. AI systems, which require vast amounts of data for learning and improvement, benefit from blockchain's secure and transparent data sharing across platforms and stakeholders. This ensures data integrity and promotes collaborative AI research and development, breaking down data silos that hinder innovation.

Integrating AI's computational power and cryptographic techniques further strengthens transaction security by identifying and mitigating vulnerabilities, ensuring the integrity and confidentiality of blockchain transactions. Additionally, Decentralized computing resources, facilitated by blockchain, allow AI developers to distribute model training tasks across multiple nodes, speeding up the process. Token-based economies incentivize participation in AI networks, rewarding individuals who contribute data, computational power, or algorithm improvements. This incentive structure encourages a collaborative and inclusive approach to AI development.

Robust identity verification through cryptography is also crucial, particularly for distinguishing between human and AI interactions. Blockchain supports effective verification mechanisms, such as:

• Cryptographic Proof via Digital Signatures: Digital signatures are generated using private keys known only to the creator and verified with publicly available public keys to ensure transaction and interaction authenticity.
• IPFS and Merkle Trees: The integration of IPFS (InterPlanetary File System) and Merkle Trees verifies the data set and AI model integrity. Any data alteration triggers an update in the Merkle Trees, preserving content integrity and serving as a verification mechanism.
• zkML (Zero-Knowledge Machine Learning): zkML provides cryptographic proof of AI models without revealing sensitive details, enhancing privacy and security in applications such as financial services, smart contracts, and legal processes.
• Decentralized AI: Distributes data across a network, enhancing data security and user privacy. This model mitigates the risks associated with data breaches and gives users greater control over their data. Federated learning exemplifies this approach, training AI models on decentralized devices or servers without centralizing personal data. 

Blockchain's immutability ensures that AI models and datasets stored on the network remain unchanged, fostering trust in data integrity. Researchers can securely share datasets using blockchain, maintaining privacy while facilitating collaboration. This secure sharing environment accelerates AI research and development, driving innovation in the field.

AI Agents: Enhancing Blockchain Functionality

The convergence of artificial intelligence (AI) and blockchain technology is paving the way for innovative advancements, particularly in the development of AI agents. These agents are sophisticated autonomous systems that enhance blockchain performance and broaden its application spectrum. Initially designed as basic task bots, AI agents have evolved into highly autonomous entities capable of performing complex tasks. They facilitate transaction processing, manage and exchange assets, and potentially, could become primary users of blockchain networks.

Unlike traditional conversational AI models and chatbots, crypto AI agents are engineered for intricate, multi-step processes. They excel in real-time analysis and decision-making, especially within the crypto investment arena. These agents continuously monitor market conditions, track price movements, and analyze extensive data from social media and news outlets. Their rapid data processing capabilities allow them to identify trading opportunities and execute trades with precision and speed beyond human capabilities. Their autonomous operation around the clock makes them particularly well-suited for the volatile nature of the crypto market.

Moreover, crypto AI agents adapt and learn from user behavior and feedback. When assigned a task, they determine the necessary steps, simulate various scenarios, and adjust strategies to minimize risks and maximize profits. This includes managing trading portfolios, balancing risk and reward, and operating continuously to ensure optimal outcomes.

Real-World Applications of Crypto AI Agents

1. Managing DeFi Portfolios: AI agents revolutionize decentralized finance (DeFi) portfolio management by monitoring positions across multiple chains and protocols. They prevent liquidations, manage funding costs, and track price movements to avoid potential losses. These agents balance and diversify portfolios based on market conditions and user objectives, ensuring optimal performance and effective risk management. Research AI agents further support this by analyzing market opportunities, understanding user preferences, and providing insights from diverse data sources, thus aiding in trading and portfolio management.

2. Executing Trades and Providing Investment Advice: In trading, AI agents make split-second decisions using real-time data and trading bots. They analyze vast amounts of information to identify trading opportunities, optimize strategies, and execute trades with unmatched precision and speed. Additionally, they offer personalized investment advice by assessing market trends, conditions, and social media sentiment, helping investors make informed decisions free from emotional biases.

3. Automating Tasks: AI agents automate various tasks within the crypto space, such as claiming airdrops and participating in governance votes on behalf of users. This automation frees users to focus on more strategic activities, ensuring that opportunities are not missed and that governance participation remains consistent and efficient.

Examples of teams building Crypto AI agents on Near include:

• Nevermined
• Bitte
• Valory
• Theoriq
• Almanak

Near AI

NEAR Protocol’s scalability and user-friendliness make it an ideal platform for integrating AI. AI can automate tasks, enhance decision-making, and personalize user experiences within DAOs, DeFi, and onchain games. It can also analyze blockchain data for anomalies and suspicious activities, thereby improving network security and providing real-time insights and market analysis for crypto traders and investors. 

In 2023, NEAR introduced NEAR Tasks, an AI marketplace that facilitates task listings, work reviews, and crypto rewards. This platform allows for the creation of self-service AI models, which are evaluated based on human feedback, enabling verified experts to build reputations and earn rewards for their contributions. Several use cases for AI within the Near ecosystem are already evident, such as:

• AI-powered Oracles: Providing reliable and secure data feeds for DeFi applications built on NEAR.
• Personalized Onboarding: Enhancing the onboarding process for new users, making it smoother and more user-friendly.
• AI-driven NFT Marketplaces: Improving the functionality and user experience of NFT platforms.

To realize its AI ambitions, the NEAR Foundation plans to make substantial investments in core infrastructure. Key focus areas include data collection and crowdsourcing, curation and rewarding of creators, access to computational resources, novel monetization methods, and ensuring the verifiability of training and inference processes. By addressing these elements, NEAR aims to create a robust platform for AI applications that leverage user-owned infrastructure.

User-Owned AI

In 2024, Near has focused on “user-owned AI” and has emerged as a leading consumer blockchain for mainstream applications, processing millions of transactions daily. NEAR's ecosystem has consistently prioritized usability and the facilitation of mainstream applications. However, AI represents the most significant technological disruption of the coming decade. The bulk of AI development currently occurs within centralized, profit-driven enterprises, a trend NEAR aims to counteract. Near Protocol advocates for a user-owned AI model, aiming to shift the power balance from corporate-owned AI systems to those optimized for individual users. The ambition is to democratize AI, ensuring that its benefits are accessible to all users rather than being monopolized by a handful of corporations. 

The next phase of NEAR’s development is to establish it as the hub for User-Owned AI, a vision that underscores the empowerment of individuals over the control of intelligence. User-owned AI refers to intelligent tools designed to enhance the well-being and success of individual users and their communities, prioritizing their needs over corporate profits. By empowering users with control and ownership, it leads to increased productivity, more accurate search results, highly personalized experiences, and innovative economic opportunities, all while maintaining a strong focus on privacy.

To realize this vision, the NEAR Foundation is committed to creating the premier ecosystem for the next wave of AI research and applications. This involves substantial investment in foundational infrastructure, including data collection and crowdsourcing, creator incentives, access to computational resources, and innovative monetization strategies. Ensuring the verifiability of AI training and inference processes will also be crucial to maintaining transparency and trust. The objective is to make NEAR the go-to platform for AI applications and consumer use cases, seamlessly integrated into the Web3 ecosystem.

‍

‍

Near’s AI Efforts and Recent R&D

Near Protocol has been pioneering AI applications with its crowdsourcing platform, Near Crowd, which has been operational for over three years. Near Crowd leverages the decentralized nature of blockchain to facilitate data collection from a broad base of contributors. With around 2,000 active users daily, the platform demonstrates the power of blockchain in organizing large-scale, collaborative efforts efficiently. This initiative exemplifies how blockchain can reduce operational overhead and streamline processes that would traditionally require significant administrative resources.

Finally, in 2024, NEAR launched the NEAR.AI R&D Lab, co-led by Alex Skidanov and another industry expert. The lab's inaugural project, an “AI Developer,” aims to address the challenge of building comprehensive Web3 applications from user intents. To support this initiative, NEAR plans to recruit a team of AI researchers and invest in pioneering research. One area of potential focus is leveraging AI to enhance code and blockchain security. As AI models improve in detecting vulnerabilities in code, Near Protocol is exploring ways to enhance the security of smart contracts through formal verification methods. Ensuring that AI-written smart contracts are secure and free from vulnerabilities is crucial for maintaining trust and reliability in blockchain systems. This focus on security highlights Near Protocol's commitment to building a safe and resilient blockchain infrastructure

Horizon Incubator

In June 2024, the NEAR Foundation initiated the NEAR Horizon AI Incubation Program, selecting six pioneering teams to advance its AI ecosystem. These teams will focus on building core infrastructure, including foundational models, model training, and developer tools, to support user-owned AI that is scalable, efficient, and secure.

Among the selected participants are:

• Mizu: aims to create a synthetic data layer to foster a collaborative environment for AI developers.
• Pond specializes in foundational crypto models that utilize graph neural networks to interpret on-chain data and predict behaviors.
• Nevermined: focusing on developing payment solutions for AI-commerce to streamline transactions for AI developers.
• Hyperbolic: will provide an open-access AI cloud and GPU marketplace to support AI inference services.
• Ringfence: enables creators to monetize their content while retaining ownership, similar to a Web3 version of DALLE-3.
• Exabits: offers GPU computing resources in a data center setting for AI training and inference.

NEAR will work closely with each of these teams to develop key use cases essential to its AI ecosystem. Over a two-month period, each project will be explored in detail as part of the Near AI x HZN program.

Mizu Spotlight

Mizu aims to establish the first blockchain-based synthetic AI data layer. This innovation is poised to create large-scale, verifiable datasets that address the limitations inherent in traditional AI training data. Mizu's vision mirrors the collaborative framework of code repositories like GitHub but for data. This community-driven approach to data curation could become foundational for AI development.

The depletion of real-world data for AI training is fast approaching, with projections indicating a potential shortfall by 2026. Synthetic data, which is artificially generated to replicate real-world data characteristics, presents a viable solution. It offers several advantages, including rapid generation, customizable properties, and enhanced privacy. Notably, advanced AI models such as GPT-5 and Tesla's Full Self-Driving (FSD) V12 already rely significantly on synthetic data. 

The Mizu Data Network is a decentralized framework where developers can deploy and manage data workflows transparently and securely. This network operates through DataVM (Data Virtual Machine), analogous to the Ethereum Virtual Machine (EVM) but tailored for data workflows. Within this ecosystem, developers can execute customized data tasks, which are processed by Mizu data nodes.

Mizu transforms static datasets, such as those from CommonCrawl and HuggingFace, into dynamic, permissionless data repositories. These repositories consist of several critical components:

• Smart Account: Governed by a smart contract, each repository can hold tokens and manage user permissions, ensuring secure governance and access control.
• Datasets: These are the core data resources, available in various formats and domains, serving as the foundation for AI applications.
• Data Index: This component provides specific views or aggregated results of the data. Validation rules applied to the data index ensure the quality and consistency of accessible information.
• Validation Rules: These rules determine what data can be added to the repository, validated either by Large Language Models (LLMs) or pre-compiled validators, thus maintaining data integrity.

Conclusion

NEAR Protocol has positioned itself as a transformative force within the blockchain and AI landscape, demonstrating a clear vision for the future of decentralized, user-owned AI. Throughout 2024, NEAR has seen significant advancements in user engagement and infrastructure, highlighted by its strategic upgrades and the integration of AI into its ecosystem. Despite the current monetization challenges, NEAR's robust user growth and technological innovations, such as the expansion to six shards and the launch of the NEAR Wallet on Telegram, underscore its commitment to enhancing user experience and accessibility. These developments have solidified NEAR’s reputation as a leader in blockchain usability and scalability.

Looking ahead to a future where crypto and AI are intertwined, the introduction of the NEAR.AI R&D Lab and the Horizon AI Incubation Program further cements NEAR's dedication to advancing AI research and applications. By focusing on user-owned AI, NEAR aims to democratize AI development, ensuring that its benefits are accessible to all rather than being monopolized by a few centralized entities.

As NEAR continues to invest in foundational infrastructure and cutting-edge research, it is well-positioned to become the go-to platform for AI applications and consumer use cases within the Web3 ecosystem. The protocol's focus on scalability, security, and user empowerment will be crucial in achieving its vision of a decentralized, user-owned AI future, marking NEAR as a pivotal player in the evolving landscape of blockchain and AI technology.

‍

‍

Disclaimer: This report was commissioned by NEAR Foundation. This research report is exactly that — a research report. It is not intended to serve as financial advice, nor should you blindly assume that any of the information is accurate without confirming through your own research. Bitcoin, cryptocurrencies, and other digital assets are incredibly risky and nothing in this report should be considered an endorsement to buy or sell any asset. Never invest more than you are willing to lose and understand the risk that you are taking. Do your own research. All information in this report is for educational purposes only and should not be the basis for any investment decisions that you make.