In this era, significant transformations in industries and tool utilization are driven by AI/Large Language Models (LLMs) and advancements in Machine Learning. There’s a growing emphasis on MLOps for managing and deploying these AI models, along with a focus on distributed inferences. Concurrently, the imperative for secure on-chain computation is escalating. Our paper introduces an innovative framework that integrates blockchain technology, particularly the Cosmos SDK, to facilitate distributed AI inferences on edge devices. This system, built on WebAssembly (WASM), enables interchain communication and deployment of WASM modules executing AI inferences across multiple blockchain nodes. We critically assess this system’s safety, scalability, and model security, with a special focus on its portability and engine-model agnostic deployment on edge devices.

Large language models (LLMs) have showcased remarkable skills, rivaling or even exceeding human intelligence in certain areas. Their proficiency in translation is notable, as they may replicate the nuanced, preparatory steps of human translators for high-quality outcomes. Although there have been some notable work exploring using LLMs for code to code translation, there has not been one for smart contracts, especially when a target language is unseen to the LLM. In this work, we aim to introduce our novel framework SolMover consisting of two different LLMs working in tandem in a framework to understand coding concepts and then use that to translate code to an unseen language. We explore the human-like learning capability of LLMs in this paper with a detailed evaluation of the methodology to translate existing smart contracts from Solidity to a low-resource one called Move. Specifically, we enable one LLM to understand coding rules for the new language to generate a planning task, for the second LLM to follow, which does not have planning capability but does have coding. Experiments show that SolMOver brings significant improvement over gpt-3.5-turbo-1106 and outperforms both Palm2 and Mixtral-8x7B-Instruct. Our further analysis shows us that employing our bug mitigation technique even without the framework still improves code quality for all models.

With the recent advances in concepts like decentralized “cloud” and blockchain-enabled decentralized computing environments, the legacy modeling and orchestration tools developed to support centrally managed cloud-based ICT infrastructures are challenged by such a new paradigm built on top of decentralization. On the other hand, decentralized “cloud” and computing infrastructures need to support many Dapp use cases. As the complexity of these targeted application scenarios increases, there is an urgent need for developing automation and modeling tools for deploying and managing decentralized infrastructures. Instead of creating such tools from scratch, a natural approach is extending mature infrastructure modeling tools for Dapps and decentralized computing environments. To this end, in this work, we have developed extensions to the TOSCA domain-specific language to support smart contract specification of decentralized computing infrastructures for supporting Dapps, where smart contracts or chain codes manage a decentralized computing environment. The result is blockchain-based orchestration and automation for decentralized “cloud” and computing environments that use existing infrastructure as code tools to deploy and manage decentralized applications.

Researchers have started to recognize the necessity for a well-defined ML governance framework based on the principle of decentralization and comprehensively defining its scope of research and practice due to the growth of machine learning (ML) research and applications in the real world and the success of blockchain-based technology. In this paper, we study decentralized ML governance, which includes ML value chain management, decentralized identity for the ML community, decentralized ownership and rights management of ML assets, community-based decision-making for the ML process, decentralized ML finance, and risk management.

Trusted execution environment (TEE) technology has many uses, such as protecting data in the cloud and improving security for industrial IoT. However, there are technical challenges that limit its widespread adoption. These challenges include the fact that different TEE vendors have incompatible solutions, and devices equipped with the same TEE technology may belong to different owners, making it difficult to establish trust between them. To address these challenges and fully utilize TEE technology, a decentralized coordination mechanism called DHTee is proposed. DHTee uses blockchain technology to support key TEE functions in a heterogeneous TEE environment, especially attestation service. Devices equipped with TEE can interact securely with the blockchain to determine whether potential collaborating devices meet the requirements. DHTee is also flexible and can support new TEE schemes without affecting existing TEEs.

Function-as-a-service (FaaS) is an emerging computation architecture, which provides high scalability and flexibility. All the existing F aaS systems are owned and managed by a single cloud service provider. While this is not an issue for most existing enterprise applications, such character is not compatible with the decentralization principle of dApp/Web3 applications, more of which are being deployed in the cloud environment. Therefore, there is an urgent need to build a decentralized FaaS, which is managed by multiple cloud service providers and allows a decentralized application to take advantages of FaaS. In this research paper, we propose DeFaaS, a novel system for managing decentralized FaaS using blockchain technology and decentralized API management, where functions are executed on a distributed network of nodes by multi-cloud data centers, rather than on a centralized server. This allows for greater scalability and flexibility, as well as improved security and reliability.

With the recent advance in concepts like decentralized "cloud" and blockchain-enabled decentralized computing environments, the legacy modeling and orchestration tools developed to support centrally managed cloud-based ICT infrastructures are challenged by such a new paradigm built on top of decentralization. On the other hand, decentralized "cloud" and computing infrastructures need to support many Dapp use cases. As the complexity of these targeted application scenarios increases, there is an urgent need for developing automation and modeling tools for deploying and managing decentralized infrastructures. Instead of creating such tools from scratch, a natural approach is extending mature infrastructure modeling tools for Dapps and decentralized computing environments. To this end, in this work, we have developed extensions to the TOSCA domain-specific language to support smart contract specification of decentralized computing infrastructures for supporting Dapps, where smart contracts or chain codes manage a decentralized computing environment. The result is blockchain-based orchestration and automation for decentralized "cloud" and computing environments, which is a step forward for achieving full decentralization in general-purpose computing.

In this paper, we present the design and implementation of a privacy preserving event based UTXO (Unspent Transaction Output) transaction system. Unlike the existing approaches that often depend on smart contracts where digital assets are first locked in a vault, and then released according to event triggers, the event based transaction system encodes event outcome as part of the UTXO note and safeguards event privacy by shielding it with zero-knowledge proof based protocols such that associations between UTXO notes and events are hidden from the validators. Without relying on any triggering mechanism, the proposed transaction system separates event processing from the transaction processing where confidential event based UTXO notes (event based UTXOs or conditional UTXOs) can be transferred freely with full privacy in an asynchronous manner, only with their asset values conditional to the linked event outcomes. The main advantage of such design is that it enables free trade of event based digital assets and prevents the assets from being locked. We implemented the proposed transaction system by extending the Zerocoin data model and protocols. The system is implemented and evaluated using xJsnark.