On June 26th, the Oracle Protocol ORA announced that it had completed a $20 million financing round, and this article provides a detailed introduction to ORA’s core technology and products. This article is based on an ORA-authored article titled “ORA: Ethereum’s TrustlessAI,” compiled, translated, and written by Chain Catcher.
Table of Contents
Introduction to ORA
1. Oracle ORA’s On-chain Artificial Intelligence
a) zkML: Zero-Knowledge Machine Learning
b) opML: Optimistic Machine Learning
c) opp/ai: Optimistic Privacy-Preserving Artificial Intelligence
2. ORA’s On-chain Artificial Intelligence Oracle (OAO)
a) Introduction
b) OAO and Ethereum and Optimism
c) Building with OAO
3. Initial Model Offering (IMO) by ORA
a) Introduction
b) Core Components of IMO
c) IMO Summary
a) Introduction to OpenLM
b) OpenLM IMO
c) OpenLM IMO Events
d) OpenLM IMO Roadmap
On June 26th, the Oracle Protocol ORA announced the completion of a $20 million financing round, with investors including Polychain, SevenX Ventures, HF0, and Hashkey Capital.
At the intersection of blockchain and artificial intelligence, ORA is pioneering new frontiers. By combining AI, rich data sources, and scalable computing power, ORA not only breaks the limitations of smart contracts but also empowers developers with innovation space.
This article provides a detailed introduction to ORA’s core technology and products, including Zero-Knowledge Machine Learning (zkML), Optimistic Machine Learning (opML), and Optimistic Privacy-Preserving Artificial Intelligence (opp/ai). Reading this article will give you an understanding of how ORA is driving the development of on-chain AI.
ORA is Ethereum’s trustless artificial intelligence that enables AI and scalable computation on the blockchain.
By providing AI, richer data sources, and scalable computation, ORA breaks the limitations of smart contracts, allowing developers to innovate freely.
ORA’s work has gained the trust of institutions such as Compound, Ethereum Foundation, Uniswap, and Optimism.
ORA makes on-chain artificial intelligence a reality. ORA has contributed a series of front-end research and products to the field of on-chain AI. The products include zkML, opML, and opp/ai.
zkML is an encryption method for on-chain machine learning.
ORA is a pioneer in the zkML field and has invented the world’s first zkML library: Keras2Circom. According to research benchmarks supported by the Ethereum Foundation, Keras2Circom outperforms other zkML frameworks.
Furthermore, Keras2Circom is production-ready and has undergone extensive testing in the ecosystem.
opML is an economically efficient machine learning framework that can now be implemented for on-chain machine learning.
ORA is the inventor and creator of opML, both from an academic research and open-source implementation perspective.
Unlike zkML, which can generate encrypted proofs for ML computations that are concise enough to be verified on-chain, the current computing power is not efficient and economical enough to generate proofs, especially for large AI models.
In contrast, opML can efficiently bring large models, such as Grok’s 31.4 billion parameter model, onto the chain, reducing management costs by over 1,000,000 times.
Learn more in our paper: https://arxiv.org/abs/2401.17555.
opp/ai, Optimistic Privacy-Preserving Artificial Intelligence, combines zkML for privacy protection and opML for efficiency improvement.
opp/ai is a hybrid model tailored for on-chain AI.
ORA is the inventor of opp/ai.
Compared to pure zkML, opp/ai performs better while maintaining the same privacy characteristics.
Read more here and learn more in our paper: https://arxiv.org/abs/2402.15006.
ORA’s Onchain AI Oracle (OAO) is ORA’s verifiable and decentralized AI oracle.
OAO enables anyone to use on-chain AI reasoning on any blockchain.
OAO uses opML to establish verifiable proofs for ML computations that can be verified on the blockchain. This opens up new possibilities for AI in the blockchain.
OAO has been running on Ethereum and Optimism based on opML and generating verifiable AI reasoning.
Last month, ORA OAO was deployed on the Ethereum mainnet.
Recently, in collaboration with Optimism, ORA OAO has been integrated into the Optimism mainnet.
Read more about the OAO deployment announcement on Ethereum and Optimism.
As a developer, you can refer to the following resources to build AI-driven on-chain applications:
OAO Repository: OAO.
OAO Documentation: AI Oracle.
Video tutorials on building with OAO: How to interact and build with OAO.
Ideas built with OAO for inspiration: So you want to build with Onchain AI.
ORA Ecosystem project reference: awesome-ora.
In the era of AI, ORA introduces a new mechanism called Initial Model Offering (IMO).
TL;DR: IMO tokenizes AI models on the chain.
For AI models, IMO provides sustainable funding for open-source AI models.
For the ecosystem, IMO helps coordinate value alignment, continuous contribution, and incentive measures.
For token holders, IMO allows anyone to capture the value of on-chain AI models through revenue and inference assets such as ERC-7007.
Many open-source AI models face challenges in monetizing their contributions, leading to a lack of motivation for contributors and organizations. As a result, the AI industry is currently dominated by closed, for-profit companies. The key to winning with open-source AI models is the need to raise more funds and build openly.
With IMO, we can win the battle of open-source AI. IMO enables effective fundraising for open-source AI by fostering long-term interests, encouraging participation, and funding the open-source AI community. When we have better open-source models than proprietary ones, we win.
IMO tokenizes ownership of open-source AI models and shares their profits with token holders.
IMO is permissionless, so anyone and any community can conduct IMO for their AI models.
In IMO, there are two core components:
Verifiable on-chain artificial intelligence models
On-chain revenue sharing
For Onchain AI, we need to link tokens to the right AI models and reasoning. ORA is achieving this through the aforementioned ORA OAO, the world’s first AI oracle applicable to any AI model.
For on-chain revenue sharing, ORA introduces ERC-7641: Intrinsic RevShare tokens as the foundation of IMO, enabling fair revenue sharing. IMO tokens are not AI meme coins but represent shares of AI models. Holders can benefit from the revenue of AI models through OAO and inference assets such as ERC-7007.
IMO is permissionless, so anyone and any community can conduct IMO for their AI models.
IMO tokenizes specific AI models, providing:
Communities with efficient fundraising for open-source.
Incentives for contributors to continue improving globally accessible models.
Opportunities for token holders to gain income from on-chain AI model usage.
Meaningful IMO requires revenue sharing mechanisms for Onchain AI and IMO tokens:
Onchain AI can be achieved through ORA OAO.
Revenue comes from model usage in ORA OAO and inference assets jointly developed by ORA.
Revenue sharing mechanisms can be implemented through ERC-7641 written by ORA.
$OLM on the mainnet: 0xe5018913F2fdf33971864804dDB5fcA25C539032LP
Address on the mainnet: 0xB57CFe32dbF5227d9c9A03eF3D2E9E53cb4524B0
OpenLM is an executable language modeling (LM) repository designed to facilitate medium-sized LM research.
OpenLM is open, high-performance, decentralized AI, as opposed to closed, centralized, licensed LLM funded and supervised by large tech oligarchs.
OpenLM is an open-source model built by a decentralized team under the MIT license. Their contributions are recognized through this donation, which will help motivate and foster the development of open-source LM accessible to anyone.
On April 10th, we released the world’s first IMO for OpenLM: OpenLM IMO.
OpenLM IMO tokenizes OpenLM’s open-source models.
Just as Prometheus stole fire for civilization, ORA demonstrates the benefits IMO brings to the open-source AI community through the first IMO for OpenLM.
$OLM is not ORA’s token. It represents the tokenization of OpenLM models on Ethereum and is shared among OpenLM contributors and the community. However, ORA’s Onchain AI Oracle can tokenize AI models.
IMO will be the cornerstone of the OpenLM journey, providing funding and incentives for the sustainable development of the open-source AI model ecosystem, democratizing access, and promoting continuous growth.
IMO for OpenLM has the potential to revolutionize the landscape of LLM through the power of open-source software and communities.
Read more about OpenLM IMO here.
The inaugural OpenLM IMO was a huge success.
All 500 million tokens were sold at a price of 150 ETH (approximately $500,000) within 5 minutes. This demonstrates the high demand in a low supply situation. $OLM continues to trade on decentralized exchange Uniswap.
Unfortunately, due to overwhelming demand exceeding the limited supply, the majority of transactions attempted to purchase $OLM at the start of the sale were not successful.
This is the first IMO, and there will be more in the future. Future IMOs will have different parameters to allow more valuable individuals and wallets to participate.
The goal of IMO is to nurture open-source communities and strengthen ecosystems. OpenLM IMO provides more resources for the OpenLM open-source community.
This is a flagship demonstration of IMO technology. It is the beginning of a more sustainable, competitive, and open-source future for AI.
The roadmap to achieving a successful OpenLM IMO is as follows:
1. OAO Integration
To realize revenue sharing for $OLM, ORA integrates OpenLM AI models into the Onchain AI Oracle. This way, every interaction between users and OpenLM models can generate on-chain revenue tracked by OAO.
2. OpenLM Ecosystem Expansion Package
For AI models, IMO provides sustainable funding for open-source AI models. With the efforts of OpenLM IMO, various projects can contribute to the revenue pool for $OLM.
3. Donations to the OpenLM Community
As previously announced and planned, 10% of $OLM IMO tokens will be allocated to existing and future OpenLM contributors.
We are finalizing all the details of the support for the open and decentralized AI community.
4. Revenue Sharing Activities
OpenLM RevShare tokens ($OLM) have the following features:
Revenue Sharing: A portion of the revenue from OAO or other sources will be allocated to the revenue-sharing pool of $OLM tokens. Token holders can claim their share of revenue every 90 days after a snapshot.
Burn: A portion of the revenue from OAO or other sources will be allocated to the burn pool of $OLM tokens. Token holders can burn their tokens as part of a buyback mechanism. The burn mechanism will support a “floor” for the OpenLM token.
Related Reports
Is Ethereum’s Vitalik Rich? Arkham Complete Analysis: ETH Holdings Rank 2nd in the World, Only Losing to This Person
Reuters is also optimistic: Ethereum’s Spot ETF might be listed as early as July 4th, how much new capital will it bring?
ETH Surges above $3400: SEC’s Gensler: Ethereum Spot ETF Approval “Progressing Well,” Stable Listing on July 2nd