Abstract:Over the past few years, ubiquitous, or pervasive computing has gained popularity as the primary approach for a wide range of applications, including enterprise-grade systems, consumer applications, and gaming systems. Ubiquitous computing refers to the integration of computing technologies into everyday objects and environments, creating a network of interconnected devices that can communicate with each other and with humans. By using ubiquitous computing technologies, communities can become more connected and efficient, with members able to communicate and collaborate more easily. This enabled interconnectedness and collaboration can lead to a more successful and sustainable community. The spread of ubiquitous computing, however, has emphasized the importance of automated learning and smart applications in general. Even though there have been significant strides in Artificial Intelligence and Deep Learning, large scale adoption has been hesitant due to mounting pressure on expensive and highly complex cloud numerical-compute infrastructures. Adopting, and even developing, practical machine learning systems can come with prohibitive costs, not only in terms of complex infrastructures but also of solid expertise in Data Science and Machine Learning. In this paper we present an innovative approach for low-code development and deployment of end-to-end AI cooperative application pipelines. We address infrastructure allocation, costs, and secure job distribution in a fully decentralized global cooperative community based on tokenized economics.

What problem does this paper attempt to address?

The paper aims to address several key issues in the current development and deployment of artificial intelligence (AI) applications: 1. **High Costs**: The cost of developing and running AI applications is very high, especially in terms of cloud infrastructure, which requires not only expensive hardware resources but also a significant amount of expertise. 2. **High Skill Requirements**: Developing production-level AI systems requires a variety of complex skills, including expertise in data science and machine learning, which raises the development threshold. 3. **Environmental Impact**: GPU computing infrastructure consumes a large amount of energy, leading to a significant carbon footprint. 4. **Decentralization Needs**: Traditional AI systems still rely on centralized cloud architectures, lacking true decentralization features. To address these issues, the authors propose an innovative approach to democratize AI through low-code development and deployment of end-to-end AI collaborative application pipelines. Specifically, this approach includes the following aspects: - **Decentralized Infrastructure Allocation**: Utilizing distributed computing resources, such as personal computers and gaming consoles, to distribute tasks in a decentralized manner. - **Economic Incentive Mechanisms**: Using blockchain technology and token economics to incentivize participants to contribute computing resources. - **Secure Task Distribution**: Ensuring the security of tasks to prevent code injection and other malicious activities. - **Multi-Framework Support**: Supporting multiple AI frameworks, such as TensorFlow and PyTorch, and providing plugin interfaces to support non-AI applications. - **Low-Code Solutions**: Simplifying the development process so that non-professionals can easily build and deploy complex application systems. In summary, the goal of this paper is to reduce the cost and threshold of AI application development through decentralized technologies, thereby achieving the popularization and sustainable development of AI.

Naeural AI OS -- Decentralized ubiquitous computing MLOps execution engine

An Overview and Solution for Democratizing AI Workflows at the Network Edge

OML: Open, Monetizable, and Loyal AI

An Optimization Toolchain Design Of Deep Learning Deployment Based On Heterogeneous Computing Platform

Enabling more efficient and cost-effective AI/ML systems with Collective Mind, virtualized MLOps, MLPerf, Collective Knowledge Playground and reproducible optimization tournaments

CLAN: Continuous Learning using Asynchronous Neuroevolution on Commodity Edge Devices

Edge-to-Edge Cooperative Artificial Intelligence in Smart Cities with On-Demand Learning Offloading

Decentralized & Collaborative AI on Blockchain

Virtualizing AI at the Distributed Edge towards Intelligent IoT Applications

AIOS: LLM Agent Operating System

LLM as OS, Agents as Apps: Envisioning AIOS, Agents and the AIOS-Agent Ecosystem

Network AI Management & Orchestration - A Federated Multi-task Learning Case.

RedCoast: A Lightweight Tool to Automate Distributed Training of LLMs on Any GPU/TPUs

Toward Among-Device AI from On-Device AI with Stream Pipelines

Deep Learning on Operational Facility Data Related to Large-Scale Distributed Area Scientific Workflows

FusionAI: Decentralized Training and Deploying LLMs with Massive Consumer-Level GPUs

NOMA-Based Multi-User Mobile Edge Computation Offloading via Cooperative Multi-Agent Deep Reinforcement Learning

opML: Optimistic Machine Learning on Blockchain

Accelerated Cloud for Artificial Intelligence (ACAI)

Rethinking Mobile AI Ecosystem in the LLM Era

Online Learning for Orchestration of Inference in Multi-User End-Edge-Cloud Networks