Manus achieved breakthrough results in the GAIA Benchmark, sparking controversy over the development path of AI.

Manus demonstrates outstanding performance in the GAIA benchmark tests, surpassing other large language models of its class. This achievement indicates that Manus has the capability to independently complete complex tasks, such as handling multinational business negotiations, including contract clause analysis, strategy formulation, and solution generation, among various stages. Manus's advantages mainly lie in three aspects: dynamic goal decomposition, cross-modal reasoning, and memory-augmented learning. It can break down complex tasks into hundreds of executable subtasks while processing multiple types of data, and continuously improve decision-making efficiency and reduce error rates through reinforcement learning.

The emergence of Manus has once again sparked discussions within the industry regarding the development path of AI: will the future move towards a unified model of Artificial General Intelligence (AGI), or will it be dominated by the collaboration of Multi-Agent Systems (MAS)?

This issue involves Manus's design philosophy, suggesting two possible development directions:

1. AGI Path: Continuously enhancing the capabilities of a single intelligent system to gradually approach the comprehensive decision-making level of humans.

2. MAS Path: Position Manus as a super coordinator to command thousands of agents in specialized fields to work together.

On the surface, this is a disagreement about technological pathways, but it actually reflects a fundamental contradiction in AI development: how to balance efficiency and safety? As a single intelligent system approaches AGI, the risk of opacity in its decision-making process increases; meanwhile, while multi-agent collaboration can disperse risk, it may miss critical decision-making opportunities due to communication delays.

The progress of Manus has inadvertently amplified the inherent risks of AI development. For example:

1. Data privacy issues: In the medical field, Manus may need to access patients' genomic data in real time; in financial negotiations, it may involve undisclosed financial information of enterprises.

2. Algorithmic Bias: During the recruitment process, Manus may provide unfair salary recommendations to specific groups; during the legal contract review, there may be up to a 50% misjudgment rate on terms for emerging industries.

3. Adversarial attack vulnerability: Hackers may implant specific audio signals, causing Manus to incorrectly assess the opponent's bidding range during negotiations.

These issues highlight a concerning trend: the more powerful the intelligent systems, the broader their potential attack surface.

In the Web3 space, security has always been a topic of great concern. The "impossible triangle" theory proposed by Ethereum founder Vitalik Buterin (that blockchain networks cannot simultaneously achieve security, decentralization, and scalability) has inspired the development of various cryptographic technologies:

1. Zero Trust Security Model: Following the principle of "never trust, always verify", strict identity verification and authorization are performed for each access request.

2. Decentralized Identity (DID): A standard for identity recognition that does not require a centralized registration authority, enabling a new model of decentralized digital identity management.

3. Fully Homomorphic Encryption (FHE): Allows computations on encrypted data, suitable for cloud computing and data outsourcing scenarios that require protection of the original data.

Among these technologies, fully homomorphic encryption is considered a key technology for solving security problems in the AI era. It can provide protection on the following levels:

1. Data layer: All information input by users (including biometrics, voice, etc.) is processed in an encrypted state, and even Manus itself cannot decrypt the original data.

2. Algorithm level: Implement "encrypted model training" through FHE, so that even developers cannot directly understand the AI's decision-making process.

3. Collaborative level: Communication between multiple agents uses threshold encryption, which ensures that even if a single node is compromised, it will not lead to global data leakage.

Although Web3 security technologies may seem distant for the average user, they are closely related to everyone's interests. In this challenging environment, continuously strengthening security measures is key to avoiding becoming "chives".

Several projects worth noting include:

• uPort: Possibly the earliest decentralized identity project launched on the Ethereum mainnet.
• NKN: Has made achievements in the zero-trust security model.
• Mind Network: As the first FHE project launched on the mainnet, it has established collaborations with several well-known institutions.

Although security projects are often not favored by speculators, they are crucial for the long-term development of AI and blockchain technology. As AI technology continues to approach human intelligence levels, unconventional defense systems are becoming increasingly important. FHE not only addresses current security issues but also lays the foundation for the future era of strong AI. On the road to AGI, FHE has transformed from an option to a necessity for survival.