Scaling Decentralized Science: Layer‑2 Blockchain and AI
By: Rachael Massey, Ryan Manville, and Morgan Lin
Original Purpose of of DeSci
DeSci originated to address problems within the medical innovation community. Due to the heavy regulation, developing a medical device is a lengthy process. The average development time takes anywhere between three and seven years. Class III devices often exceed seven years. In addition to being lengthy, the research process is expensive. As with the duration, cost varies on risk level. Class I devices average between $200,000 and $2M. Class II devices cost about $2M- $30M+; Class III devices average between $5M to upwards of $100M. Oftentimes devices are introduced to other countries and commercialized there-first to generate revenue-before they are brought to the US. Generating revenue in other countries helps offset the expense of introducing a device to the US. The length to implementation and the high development cost are significant barriers to medical innovation.
These problems are a direct result of government regulation. The industry is paralyzed by ineffective and aging public policy. The involvement of the government in medical innovation even deters investors. Venture capitalists cite inconsistent and changing policies as a major reason for not investing in medical innovation. Moreover, innovators cannot rely on consistent public funding for research and development. Funds are competitive and not infinite. This problem has persisted for decades. Similarly to limited funds, data is not readily accessible to researchers. It is difficult for researchers across institutions to collaborate and share data. The resulting data silo prevents collaboration that can potentially accelerate development.
The current medical innovation industry is plagued by lengthy approval times, high costs, limited funding, and limited access to data. The effects of these problems on public health is immeasurable. The slower it takes to innovate, the more people suffer from life-threatening medical problems.
The answer to the ineffective medical industry are DAOs. As a decentralized organization, DAOs are the antithesis of BIO of the slow bureaucracy. They allow individuals to collaborate on research and share resources by pooling capital, labor, and data to generate new assets such as intellectual property (IP), applications, and agents. All members of the DAO share ownership of the assets. One popular example is cerebrumdao. This DAO shares resources with the goal of preventing neurodegeneration. Cerebrumdao successfully raised over $1.5M dollars and secured a partnership with the biotech company Fission Pharma.
Narrative
As introduced, DeSci has been focused on and constrained to mainly biotech initiatives. However, over the past year, much of DeSci’s focus has shifted to AI and machine learning (ML). OriginTrail, currently the biggest DeSci project by market cap, is focused purely on AI technological advancement through the use of blockchain technology. Additionally, Bio has increasingly been utilizing AI technologies to ease developmental hurdles, particularly in the development of novel drugs. This shift is important because biotech is constantly burdened by the “Valley of Death” problem, generally making it a risky investment. If this problem can be solved through the use of AI, biotech as a whole would benefit.
Although DeSci as a whole has seen an overall decrease in market capitalization between December 2024 and December 2025, this is to be expected with a grander restructuring of what scientific research token holders deem important. Bio’s nearly 93% decrease in market capitalization is surprising, but can be explained by a lack of liquidity and important developments within Bio, and a general downturn in the crypto market over the last year. However, DeSci projects focusing on AI, such as Alchemist AI and Pythia, have been able to outperform the market. Sasha Shilina, a DeSci expert and post-doctoral researcher at Moscow State University believes that DeSci and AI will eventually be unified into a single blockchain framework called DeScAI, and the increasing dominance of AI based DeSci projects shows this.
In the near-term, these DeScAI projects should be expected to grow, as traditional AI adjacent investments like Nvidia have, due to capital spillovers from traditional AI and the broader AI boom. DeScAI could offer an alternative pathway for AI startups that do not have the necessary capital to utilize expensive centralized AI data systems.
Introduction to Major Protocols
OriginTrail, the leading DeScAI project by market cap, seeks to change how AI processes and receives data through the development of what they call the Decentralized Knowledge Graph (DKG), which aims to decentralize the storage and retrieval of structured knowledge by anchoring verifiable data assets across a distributed network of nodes — enabling AI systems to query trusted, provenance-tracked information rather than relying on opaque, centralized data sources.
Alchemist AI leverages a modular architecture built around on-chain knowledge primitives, enabling AI agents to access, compose, and act upon structured real-world data in a trustless environment. Its core function is to serve as an inference and reasoning layer for decentralized applications, allowing AI models to draw from verified data pipelines for use cases ranging from autonomous DeFi strategy execution to enterprise-grade data verification workflows.
Pythia seeks to bridge the gap between artificial and organic intelligence by translating complex on-chain and off-chain signals into human-legible data. Through its architecture, Pythia enables both AI systems and end users to interpret the reasoning behind data-driven outputs, supporting applications in predictive analytics, governance tooling, and real-time decision support. Additionally, Neiry, the company behind Pythia, has created the first AI-brain-linked rat using Pythia’s technology.
Analysis of Major Protocols
OriginTrail has partnered with many leading corporations in the AI sphere such as Google, Microsoft, Oracle, and Nvidia. Additionally, it has been integrated into the U.S. supply chain system through incorporation of its technologies in the Supplier Compliance Audit Network (SCAN), which includes members such as Costco and Walmart. In this sense, OriginTrail has already overcome the barrier of widespread enterprise adoption. Alchemist AI is largely a consumer based product and there are some doubts about its scalability for enterprise use. Alchemist AI is largely one-dimensional, without anything being developed outside of its no-code development environment, which may be unappealing for more professional use cases. Pythia’s technology remains unproven at a large scale and the leading edge biotech nature of Pythia makes it highly susceptible to the Valley of Death problem, even with the backing of DeSci funding. Additionally ethical dilemmas present in their research may lead to regulation dampening its effectiveness, particularly in the EU.
Generating Alpha
The convergence of AI and DeSci creates a unique opportunity for accelerated innovation. Therefore, Villanova Blockchain believes that AI based DeSci platforms will be able to generate alpha. The ability to generate alpha is largely dictated by an increased access to novel information to inform financial decisions. The purpose of DeSciAI platforms is to democratize information. AI systems are able to rapidly recall and analyze large amounts of data. Moreover, the decentralized governmental structure of DeSci platforms unlocks and previously recondite research. Traditional medical/scientific institutions create barriers to data through centralization. The combination of AI and DeSci results in enhanced access to information. DeScAI platforms will be able to generate alpha because their financial decisions will be supported by a deeper network of information. The more access to data an organization has, the more accurate and lucrative its decisions will be.
Another factor that indicates DeScAI platforms will generate alpha is the speed at which decisions are made. These platforms are able to shorten the decision making cycle, resulting in properly timed financial decisions and capturing value sooner. DeSci platforms already compress the decision making cycle through their DAO block-chain based voting systems. AI’s analytical capabilities increase the speed at which research is processed. Consequently, governance token holders are adequately prepared to vote sooner. There is a direct relationship between faster decision making and an increase in profitability.
Blockchain Infrastructure Overview: Trilemma
The success of DeScAI platforms relies on robust blockchain infrastructure to scale efficiently while maintaining security and decentralization. The blockchain trilemma suggests that a blockchain can typically achieve at most two of its three core properties: security, decentralization, and scalability. For example, Bitcoin and Ethereum prioritize security and decentralization, ensuring the networks are resistant to attacks and not controlled by any single party. However, this design comes at the cost of throughput, meaning these blockchains can only process roughly 7–15 transactions per second (TPS). During high network activity, this congestion results in high transaction fees, making frequent interactions like microgrants, governance votes, or tokenized incentives impractical.
Layer-1 Blockchains
Layer-1 (L1) blockchains, such as Bitcoin and Ethereum, provide the foundation for security and decentralization. Ethereum introduced smart contracts in 2015, which enabled decentralized applications and specialized blockchains like Solana, Avalanche, and Cardano. While L1s maintain consensus, resist censorship, and store full state, their low throughput limits high-frequency, low-cost operations — creating a bottleneck for DeScAI platforms that require constant, coordinated collaboration.
Layer-2 Blockchains
Thus, L2s handle overcome these limitations by processing transactions off-chain while still settling on L1. Throughput can scale to hundreds or thousands of TPS, enabling faster and cheaper interactions essential for DeScAI. L2s allow researchers and contributors to participate in frequent governance votes, distribute microgrants, and collaborate on tokenized research without prohibitive costs. L2 solutions include:
State/Payment Channels: Parties lock funds in a smart contract and transact off-chain directly. While the effect is extremely high throughput, participants must remain online or active, and routing transactions between multiple users can be complex. This is useful for continuous Desci engagement.
Sidechains: Polygon PoS, Avalanche C-Chain, and BNB Chain, are independent chains that run Ethereum-like virtual machines. They offer fast transactions and low fees, with throughput claims up to ~65,000 TPS. However, they rely on their own validators for security, so collusion can put funds at risk.
Optimistic rollups: Arbitrum, Optimism, and Base, batch transactions and essentially assume they are valid when posting to the L1. Furthermore, malicious batches can be challenged using fraud proofs in a designated window, and they support full EVM compatibility and very low fees. Full EVM compatibility and low fees support DeSci governance and funding workflows.
ZK-Rollups: zkSync Era, Polygon zkEVM, and StarkNet, generate cryptographic proofs for batches of transactions that the L1 can verify instantly. High security and instant finality benefit sensitive DeSci data.
Arbitrum One
Arbitrum One is an established optimistic rollup on Ethereum, meaning it processes transactions off-chain but still relies on Ethereum for security. Arbitrum One is a well-established optimistic rollup with the highest total value locked (TVL) among L2s, which is around $16–19 billion, roughly 40% of the L2 market. It uses multi-step and interactive fraud proofs, which reduce gas usage and supports full EVM equivalence, meaning developers can run their existing Ethereum contracts without modification. While Arbitrum’s throughput is moderate (~20 TPS), its main strength is deep liquidity and a mature DeFi ecosystem. This allows large transactions with minimal slippage and seamless interaction with hundreds of dApps, AMMs, and lending platforms. Arbitrum One allows DeSci projects to run multiple research and governance apps without worrying about high fees, while all apps share the same Ethereum security.
Optimism
Optimism is another optimistic rollup with a TVL around $6–8 billion. It uses a single-round fraud proof, which simply means that Optimism only allows one challenge step, so the system is simpler but slightly slower to detect cheating compared to Arbitrum’s multi-step approach. Because of the simpler fraud proof, historically fees were a bit higher than Arbitrum’s, but recent updates have made fees closer to Arbitrum. Furthermore, its key differentiator is Superchain vision, which connects multiple L2s under a shared software layer, in which contracts and transactions move between L2s with minimal friction. L2s in the Superchain can exchange data, tokens, or smart contract calls easily, which helps developers scale across chains without rebuilding everything.
Base
Base is a newer Layer‑2 blockchain built by Coinbase, specifically designed for retail and consumer-focused applications. Unlike many L2s, Base has no native token and leverages Coinbase’s 110 million-user base for fast and easy onboarding, making it highly accessible for everyday users. At its peak, Base processed up to $55 billion in stablecoin transfers weekly, accounting for roughly 18% of global stablecoin volume, while maintaining extremely low fees (~$0.01 per transaction) by processing transactions in batches and having Coinbase cover some costs. While Coinbase controls the sequencer (the system that orders transactions), the security of funds still depends on Ethereum, so users’ money remains safe. The trade-off is that Base’s sequencer is controlled by Coinbase, and the ecosystem is anchored by Coinbase products.
DeSci Context
Decentralized science (DeSci) projects are increasingly leveraging Layer-2 (L2) solutions to overcome Ethereum mainnet limitations, supporting high-frequency interactions, microgrants, tokenized incentives, and data-intensive workflows without prohibitive gas fees.
VitaDAO’s VITA token is deployed on Ethereum mainnet as well as on Layer‑2 networks like Base and Optimism, allowing users to save on transaction costs while still relying on Ethereum for settlement security. In 2023, VitaDAO ran an initiative called “$VITA Goes Optimism”, guiding users to bridge VITA to Optimism so they could participate in governance and provide liquidity with gas fees often under $0.15 per transaction, compared to typical Ethereum mainnet fees of $10–$50 per transaction at network peaks.
Similarly, HairDAO’s token trading activity on Base costs around $0.01 per transaction, far cheaper than on Ethereum mainnet, which encourages more frequent participation from holders and traders.
Why Ethereum & L2s Dominate Builder Activity in DeSci
Ultimately, platforms such as Molecule, ResearchHub, and VitaDAO are fundamentally built on Ethereum smart contracts, which provide a flexible and secure foundation for decentralized scientific applications. By combining Ethereum’s capabilities of programmable governance and tokenization of intellectual property with Layer-2 scaling solutions, DeSci protocols can significantly reduce transaction costs, increase throughput, and improve the user experience for scientists and contributors who may not be crypto-native.
Conclusion
Overall, Decentralized Science (DeSci), empowered by AI and blockchain technology, represents a transformative approach to overcoming longstanding barriers in medical and scientific innovation. By enabling collaboration through DAOs, integrating AI to accelerate research, and leveraging Layer-2 blockchain solutions to scale efficiently, DeSci platforms like OriginTrail, Alchemist AI, and Pythia demonstrate how decentralized systems can make scientific progress faster, more accessible, and more equitable.
Work Cited Page
Chainlink. “Layer 1.” Chainlink, 2026, https://chain.link/article/layer-1.
Gemini. “Layer 2 Scaling: ZK Rollup vs. Optimistic Rollup on Ethereum.” Gemini Cryptopedia, 2026, https://www.gemini.com/cryptopedia/layer-2-scaling.
AMBCrypto. “How Ethereum Quietly Crushed Its $50 Gas Problem in 2026.” AMBCrypto, 2026, https://ambcrypto.com/how-ethereum-crushed-gas.
Sharma, Sankalp. “Scaling Ethereum: When to Choose Layer 2 vs Alt L1 vs Sidechains.” Dev.to, 2026, https://dev.to/sankalp_sharma_634764b990/scaling-ethereum-layer-2.
Messari. “Defining Layer 1 Blockchains.” Messari, 2026, https://messari.io/copilot/share/defining-layer-1-blockchains.
Paxos. “Understanding the Different Blockchain Types.” Paxos Blog, 2026, https://www.paxos.com/blog/understanding-blockchain-types.
Techskill Brew. “Layer 2 Blockchain Scaling Solutions: Channels, Sidechains, Rollups, and Plasma.” Medium, 2026, https://medium.com/techskill-brew/layer-2-scaling-solutions.
XT.com. “Top Six Layer-2 Networks for Trading, DeFi, and Payments in 2026.” Medium, 2026, https://medium.com/@XT_com/top-six-layer-2-networks.
Payram. “Arbitrum vs Optimism vs Base.” Payram Blog, 2025, https://www.payram.com/blog/arbitrum-vs-optimism-vs-base.
CoinGecko. “HairDAO (HAIR).” CoinGecko, 2026, https://www.coingecko.com/en/coins/hairdao.
VitaDAO. “VDP-159: Bridge VITA to Solana.” VitaDAO Governance Forum, 2023, https://gov.vitadao.com/t/vdp-159-bridge-vita-to-solana/1754.
AInvest. “Ethereum 2026 Flow Overhaul: ZK Proofs & Proto-Danksharding.” AInvest, 2026, https://www.ainvest.com/news/ethereum-2026-flow-overhaul