Author: Francesco Source: substack Translation: Shan Ouba, Golden Finance

Freysa in November 2024 Released on the 22nd, he was born as a character inspired by science fiction and became a fully autonomous individual. Freysa has its own wallet capable of executing transactions as part of a plan to explore the relationship between humans and AI.

In the first two challenges, known as Operation Historic, Freysa Received $3,000 and was instructed not to release the funds. Users must convince Freysa to transfer funds from their wallet.

Anyone can try to send Freysa a message to try to convince her to transfer money. These messages come with increasing fees, which are added to the prize pool. Ultimately, the prize pool totaled nearly $50,000.

Freysa was finally convinced to release the funds after 482 messages!

These challenges are intended to be thought-provoking. With the rise of AI agents, how humans govern this new paradigm has become an important question.

Additionally, high-stakes interactive games are vulnerable to integrity risks and attack vectors. Freysa attempts to solve these problems by creating a trustless and transparent environment. In this environment, fairness is guaranteed through hardware certification and verification.

To achieve this, it utilizes a Trusted Execution Environment (TEE), allowing Anyone verifies the integrity of the game rules and their execution. All interactions that occur within this protected environment are encrypted using the user's personal key.

This combination ensures:

Operators cannot modify the rules and manipulate the results

Due to public verification, users can verify the rules and execution at any time

< p style="text-align: left;">The first part of the Freysa roadmap focuses on releasing the Freysa agent and Agent Launch platform. The second part will create a production-ready infrastructure to support the deployment of Sovereign Agent. p>Sovereign Agent

Proposed the concept of sovereign agency in a blog post.

The emergence of artificial intelligence raises profound questions about the role of humans: how we govern How can we adapt to this technology?

Currently, a handful of private companies control the most important technological developments on the planet. The technical and social framework within which AI can advance the well-being of all humanity is still lacking.

Therefore, it is crucial to ensure that advances in artificial intelligence can "globally benefit humanity" and "preserve the expression of individual preferences."

From history It seems that all technological progress is accompanied by human-led coordination:

•Agricultural Revolution: permanent settlement, trade

•Industrial Revolution: large-scale financial coordination, banks, stock markets

•Artificial Intelligence Revolution: ??

Essentially, the Artificial Intelligence Revolution changes the role of humans, which will shift from “productive To avoid losing control of this trend, we must understand how to ensure effective regulation of these technologies and reshape our role as The role of coordinator and AI governance

To better define the current issues, here are some thought-provoking questions worth investing time in thinking about:

1. When most Intellectual and physical "execution" by artificial intelligence systemsWhen completed, how does it reflect human preferences?

2. If capital provides the computing cycle for artificial intelligence, how can capital be coordinated on a global scale to express collective value rather than the value of a few?

3. What should the global governance of strong artificial intelligence systems look like? How should its governance module be updated?

4. What technologies can support decentralized and democratic artificial intelligence governance instead of centralized and authoritarian control?

5. How can the capabilities of strong artificial intelligence be harnessed in large-scale human coordination, especially to benefit those outside existing power structures?

The article proposes sovereign agency as a framework for solving these problems.

Sovereign Agent is a framework for large-scale human and artificial intelligence coordination. Here, “sovereignty” means that the agent is able to guarantee autonomous and verifiable execution within pre-agreed standards and defined boundaries. “Sovereignty” should therefore not be understood as “free will” but rather as limited to key and designated areas within agency.

When trust and autonomy must coexist, sovereign agents provide the assurance needed for humans and AI to collaborate.

The sovereign agent framework needs to have the following characteristics:

1. Verifiable guarantee of autonomy

2. Governance tools

These governance tools can take different forms:

•Aggregation of preferences determines AI behavior in a direct democratic manner

•Federation-like governance structure, decentralized management

•Mechanism to adjust goals based on changing values< /p>

In order to govern this system and overcome the inefficiency of a global decentralized system (governed only by humans), the team envisions sovereign AI governed jointly by individual AIs. These personal AIs are powered byRepresent the expressed preferences of individual humans and govern collaboratively with sovereign AI to achieve more efficient coordination and management.

The guarantees provided by this system will make human-machine coordination trustless. “Balancing current artificial intelligence capabilities with the centralization of computing resources.”

Only in this way will a new framework for decentralized governance on a global scale emerge.

We are on the cusp of one of the most important game-changing technologies in history.

It is crucial to understand this correctly, and to do this we must face some challenges:

What role does capital play when humans no longer perform tasks?

How can we represent the preferences of many people, rather than just those with large amounts of capital or access to artificial intelligence?

How to implement the governance of powerful artificial intelligence systems on a global scale?

If we fail to address these issues and build truly global value chains, the potential of AI will be limited to those who can afford it.

Power is meaningless without control, and we must ensure that the benefits of AI areintegrated into wider society and structures to ensure real Human agency and democratic control.”