Modulus Labs: All You Need to Know About the AI Accountability Platform

In the past decade, the AI sector and the blockchain ecosystem have been two of the fastest-growing technological ecosystems. Despite both industries being in their early phases of discovery and development, they have undoubtedly positively impacted the tech world. But what if we could have an integration between blockchain technology and AI?

Interestingly, that’s precisely what Modulus Labs is all about. This article focuses on Modulus Labs, what the startup has accomplished in the blockchain and AI space, the problems Modulus solves, and its expansion plans.

Image Source: Modulus.xyz

What is Modulus Labs?

Modulus Labs is a company focused on taking artificial intelligence on-chain. Modulus Labs’ vision has made it famous as “the accountable AI company” because it imbibes accountability into machine learning. It was established in 2021 but came to the limelight in 2022 and has become one of the fastest-growing blockchain-AI startups.

To accomplish this, the company has extensively built blockchain solutions to combat challenges in harmonizing AI data with blockchain technology. So far, the major challenge in merging AI with on-chain technology has been maintaining decentralization and data accuracy. Note that Modulus is the machine learning (ML) platform, while Modulus Labs is the company responsible for developing the technology.

Modulus: How does it work?

Modulus is a machine learning platform that uses cryptography to verify that AI results weren’t forged. It is the most specialized ZK prover for AI. With the advancement of AI, more individuals and organizations are relying on artificial intelligence. However, AI feeds could be manipulated and thus inaccurate in some cases.

Just like a Twitter verification badge can help users differentiate a real celebrity account from fake accounts, Modulus enables blockchain users and web3 developers to filter inaccurate feeds from accurate AI-generated feeds.

Zero-Knowledge Proofs on Modulus

Modulus uses a cryptographic verification system to ensure that web3 protocols can depend on data from AI while maintaining decentralization and privacy. To achieve this, Modulus utilizes zero-knowledge machine learning (ZK-ML) techniques. So far, zero-knowledge techniques are the best method for verifying accuracy without tampering with privacy. ZK proofs of machine learning models ensure that computations are hidden from the verifier. In turn, verifiers can also prove a data set’s correctness without access to the computations. Interestingly, the process of ZK proving is highly efficient. Zero-knowledge proving costs less while maintaining decentralization and data security.

ZK-ML is an upgrade to traditional opaque gateways. It gives users the assurance that a generated model is tamper-proof. With ZK-ML, developers can build better products, especially for sectors where trust is needed, like the medical and financial sectors.

Remainder – Modulus’ Specialized Prover

Interestingly, Modulus Labs has products that are either in development or in their early phases. One such product is Remainder. Remainder is Modulus’ specialized AI prover that executes zero-knowledge verification. Remainder enables developers to build smart contracts that can access AI data securely and without the need to give up decentralization.

When Modulus is integrated into a decentralized application, the dApp feeds Modulus with data related to its function. In turn, Remainder (Modulus’ AI prover) executes multiple cryptographic checks in the background to craft a proof of correctness for the data received. Finally, Remainder’s proofs are aggregated and sent on-chain for approval or final verification. Modulus utilizes the Ethereum network for AI proving. Once the process has been completed, we have mathematical certainty that data received from the sender (which in this case is the dApp) is accurate and has not been manipulated.

Remainder’s version one (V1) has been released and utilized by some Modulus partners. These partners include Upshot, Worldcoin, Ion Protocol, Polychain Monsters, Spectral Finance, and AI Arena. Below is a list of Modulus’ current partners and the nature of the product they will integrate Remainder in.

Modulus’ Partners

Modulus partners will be able to utilize Remainder to achieve:

• Trustless NFT Appraisals (Upshot)
• Authentic Machine Artists (Polychain Monsters)
• Tamper-proof Intelligent Finance (Spectral Finance and Ion Protocol)
• Private Identity Authentication (Worldcoin)
• Trustworthy AI Game Economies (AI Arena)

History of Modulus

Modulus Labs was founded in 2021. Daniel Shorr, Nicholas Cosby, and Ryan Cao co-founded the startup. According to Shorr, they were obsessed with artificial intelligence and web3. This triggered various experiments that eventually birthed Modulus.

Notably, in 2020, during the COVID-19 pandemic, Shorr and his colleagues were drawn to the exciting world of cryptocurrencies. But while crypto tokens and their valuations attracted many individuals, the technology particularly intrigued this group. Shorr and his team consumed as many whitepapers as possible, eventually specializing in zero-knowledge cryptography.

In 2021, they aimed to build a complicated project that would also be very handy. Their first experiment led to the development of Rocky.bot, an AI software built to predict ETH’s price. Shorr, Cosby, and Cao funded Rocky.bot. However, they ran out of experimental funds along the line and called for willing donors to sustain their experiment.

The bot caught the eyes of many investors who donated willingly to sustain their experiment. However, the donations received in thousands of dollars only sustained their AI-trading experiment for so long. Although the Rocky.bot experiment failed, it birthed new ideas in the founders’ minds. The experiment of bringing AI on-chain became exciting, triggering the new concept – Modulus.

Hence, the team came together to build software that would verify feeds from AI and enable smart contract developers to integrate artificial intelligence into their web3 products. This kicked off in 2021.

The Team

Image Source: Modulus.xyz

Modulus Labs’ leading team members are Daniel Shorr, Nicholas Cosby, and Ryan Cao. Daniel Shorr is the company’s co-founder and chief executive officer. Shorr attended Stanford University and was famous for his leadership and public speaking skills.

Similarly, Ryan Cao, who has a master’s degree in computer science, also graduated from Stanford University. Cao holds the position of Co-Founder, Chief Financial Officer, and Chief Technology Officer. At the same time, Nicholas Cosby is the Co-Founder and Senior Data Engineer at Modulus, also from the United States.

Other team members include Giorgios Zirdelis, Ben Wilson, Makis Arsenis, Vishruti Ganesh, Ende Shen, and Cami Ramos. These co-builders have been tagged with a nickname coined from the project’s name – Moodlers. Finally, the Modulus team has three members on its board of advisors. They are Zhang Zhenfei, Riad Wahby, and Rand Hindi.

Funding Rounds and Key Investors

Modulus Labs’ earliest investors are Floodgate and Taproot Wizards. Floodgate is a renowned investment company that spreads its investment across various sectors, while Taproot Wizards is an NFT company.

The startup has attracted investment from Inflection.xyz, Alliance, Stanford, and GCR. More recently, though, Modulus Labs raised $6.3 million in another seed round of funding. The recent investment was announced on November 1st, 2023, with Variant and 1kx leading the funding round. Modulus Labs also has several other individual angel investors backing the project.

Modulus Labs: The Road Ahead

The Modulus team strongly believes in an AI-filled future. The tech ecosystem will continue to advance, and Modulus aims to play its part in revolutionizing the AI sector. In 2024, Modulus Labs will release the Modulus API, and its codebase will be made open-source. Modulus Labs also intends to integrate the product into various dApps in the first and second quarter of 2024.

Conclusion

Modulus is the first on-chain AI project and has quickly gained traction. Currently, Modulus has no direct competitors. However, several other platforms utilize zero-knowledge machine learning in the blockchain space. Thanks to ZK technology, Modulus’ vision to attain AI accountability via a secure, scalable, and decentralized process is achievable.

Interestingly, the platform also has an excellent user interface to enable seamless integration for product builders. As the world of artificial intelligence expands exponentially, the Modulus team looks forward to seeing more developers embrace its unique on-chain and AI solution to data proving.