Minima and the Artificial Superintelligence Alliance (ASI) announced a strategic partnership on platform X, aiming to enable AI models to run directly on user devices (such as smartphones) through blockchain technology, achieving true decentralization and autonomous control. This collaboration is hailed as a milestone in decentralized AI, potentially reshaping the future landscape of the AI industry.
1. Background: The strong alliance between Minima and ASI
1. Minima: Innovator of lightweight blockchain Minima is a company focused on blockchain technology, with its core product being a Layer 1 blockchain protocol. Unlike traditional blockchains, Minima's design allows each device (including smartphones and IoT devices) to run a full node. This 'embedded blockchain' technology makes it unique in the field of edge computing. According to Minima's official website, its protocol only requires 300MB RAM to run a full node, suitable for large-scale edge networks, and can support AI, smart cities, and autonomous systems in the future. This lays a solid foundation for the realization of decentralized AI. 2. ASI: Pioneer of decentralized AI The Artificial Superintelligence Alliance (ASI) is an alliance composed of Fetch.ai, SingularityNET, Ocean Protocol, and Cudos, aimed at advancing artificial general intelligence (AGI) and ultimately achieving superintelligence (ASI). ASI has created a unified AI token $ASI by merging the tokens of these projects (such as $FET), emphasizing decentralization, privacy, and ethics. According to industry data, the ASI ecosystem integrates technologies like decentralized economy, data sharing, and cloud computing, dedicated to building an open AI ecosystem. Notably, the $FET token was already listed on Binance in 2019, currently valued at 2.1 billion USD.
2. Partnership Details: AI moves from the cloud to the edge According to Minima's post on X (Minima_Global), this partnership includes the following key components: Running ASI:One LLMs
ASI's ASI:One large language models (LLMs) will run directly on Minima's nodes. This means users do not need to rely on centralized cloud services; AI models can run locally on smartphones. This 'Edge AI' approach not only reduces latency but also significantly enhances data privacy. Integritas data verification
Minima has launched a data verification service called Integritas, designed to ensure that the inputs and outputs of AI models are verifiable. Through cryptographic techniques, Integritas can prove the accuracy and integrity of the data in a trustless environment, which is crucial for decentralized AI. Based on a token infrastructure
The partnership will establish a 'tokenized infrastructure', which is a decentralized infrastructure based on blockchain tokens, ensuring legality and compliance. This may incentivize users to contribute computing resources or data through tokens, creating a self-sustaining AI network. Integration of the ASI ecosystem
This collaboration will integrate projects like Fetch.ai ($Fet), Ocean Protocol (Ocean), and Cudos within the ASI ecosystem, constructing a broader AI and data economy. This will further expand the application scenarios for decentralized AI. 3. Technical Feasibility: Challenges and Opportunities coexist 1. The feasibility of running AI on edge devices Running AI models on edge devices requires overcoming technical challenges, such as model compression and resource optimization. Research indicates that tools like TensorFlow Lite and Google’s Learn2Compress are advancing this field. Minima's lightweight design (300MB RAM) enables the possibility of running small AI models, but further breakthroughs are needed to run large language models (LLMs). 2. Advantages and limitations of decentralization Blockchain-assisted edge computing can enhance security, but consensus mechanisms (like PoW) may lead to latency and energy consumption issues. Minima’s lightweight protocol may alleviate these problems, but the actual effectiveness needs to be validated through testing. If successful, this technology will provide users with unprecedented privacy and control.
