The risks, challenges, and future development directions of decentralized AI

Introduction

Decentralized AI, although having many advantages, also faces many risks and challenges. As the third article in this series, this article will analyze these challenges for you and look ahead to the future development direction of decentralized AI.

We also welcome entrepreneurs and project parties in this direction to contact us.

Opportunities for the development of AI Agent

AI Agent is the natural evolution of large models. By introducing memory mechanisms, task decomposition, and planning capabilities, AI Agent is able to perceive the environment, make autonomous decisions, and execute complex tasks.

Existing large-scale models can generate text and solve problems, but they do not yet have complete task planning and execution capabilities. AI Agent will fill this gap and enhance AI’s performance in complex tasks.

If AI is nuclear energy, then it should not be controlled by a few. Decentralized AI Agents will ensure fairness and transparency of AI technology through blockchain and encryption technologies.

In the future agent society, decentralized AI will become an inevitable trend to solve the problems faced by existing centralized AI systems.

Development Opportunities for Data Annotation:

Data preparation includes data collection, cleaning, annotation, and enhancement. The diverse needs of AI for data have increased the reliance on high-precision and highly customized data annotation. The lengthy work cycles and high labor costs of data annotation have limited the development of the AI industry.

Web3 can connect to a large number of AI data collection and annotation personnel from various regions around the world through economic incentives, allowing them to benefit from data contribution.

Case Study: Ocean Protocol, Data Trading Marketplace

Operation Mechanism

• Providers: Data providers can issue and sell their own data tokens to generate revenue.

• Consumers: Purchase or earn the required data tokens to gain access.

• Marketplaces : Refers to an open, transparent, and fair data trading market provided by Ocean Protocol or third parties, connecting providers and consumers worldwide, and offering various types and domains of data tokens.

• Network: Refers to the decentralized network layer provided by Ocean Protocol.

• Curator: Refers to a role in an ecosystem that is responsible for selecting, managing, and reviewing datasets. They are responsible for reviewing information such as the source, content, format, and license of the dataset to ensure that it meets standards and can be trusted and used by other users.

• Verifier: Refers to a role in an ecosystem that is responsible for verifying and auditing data transactions and data services.

Summary: AI Agent and decentralized data labeling are currently two popular directions in DeAI, and many entrepreneurial teams are developing in this area.

Risks and Challenges Faced by Decentralized AI

• Limitations of Web3 Empowerment for AI : Due to the limited number of encrypted users in Web3, the radiance of the economic incentive mechanism is relatively small. This limits the rapid development of decentralized AI, requiring more user participation and acceptance.
• Challenges of Zero Knowledge Proof Technology: issues such as quantization accuracy, hardware requirements, and adversarial attacks. Zero Knowledge Proof Technology (ZKP) has long-term significance in achieving model verifiability, but currently faces technical challenges and implementation challenges.
• Cost advantage attractiveness: If the supply of computing resources in the market is relieved, the value and cost advantage of decentralized computing power networks will be weakened. This requires decentralized AI to continuously improve efficiency and reduce costs to maintain its competitiveness.
• Efficiency and Cost Issues of AI Combined with Cryptography：The efficiency of using zero-knowledge proof technology or fully homomorphic encryption (FHE) technology to perform privacy computing tasks is much lower than plaintext execution. Due to the high computational demand of AI, the addition of cryptographic technology will further increase the cost, which may be difficult to implement in practice.
• The deep fake problem of AI: The significant communication bottleneck problem in AI model training. Frequent exchange of model parameters and gradient information consumes a large amount of network bandwidth, resulting in high communication overhead. At the same time, the synchronization problem of each node will also affect the training results, requiring frequent data verification and synchronization operations.
• The popularization of AI has led to an increased risk of deep fakes. In the scenario of Web3 empowered by AI, it is necessary to guard against AI forgery risks.

The Direction of Decentralized AI Future Development

Model Layer: With the increasing prevalence of AI agents, future users will rely on AI agents to help them complete tasks, which is the key to connecting the model layer and the application layer. The diversification of models is gradually forming on various platforms, and the cost of large models is continuously decreasing. It still takes time to develop “dark horse” applications.

Training Layer: Decentralized Training AI models exist the possibility of implementation, but as the inference demand is much greater than the training demand, the training layer will rely more on centralized computing power.

Compute Layer: Decentralized computing power effectively reduces GPU usage costs, and enterprise-level GPUs meet current computing power requirements. In the future, with the implementation of on-device models, consumer-grade GPUs will have their place.

Data Layer: The difficulty of obtaining public data is increasing, and decentralized data collection and data annotation will become important ways for future AI model data sources and processing.

Conclusion

Decentralized AI as an emerging technological trend, although the road is full of challenges, it has enormous development potential. With the continuous advancement of technology and the gradual maturation of the market, decentralized AI is expected to play a greater role in the future. We need to continue to pay attention to these challenges and seek innovative solutions to promote the development of decentralized AI. In this regard, we believe that decentralized AI has its place in the four aspects of models, training, data, and computing power, especially DeAI is one of the most visible and value-generating directions.