Frequently Asked Questions and Answers

1. What is the essence of this manifesto?

This manifesto outlines a set of guidelines for ensuring that artificial intelligences (AIs) work with humans to solve social issues while preventing its weaponization or runaway development from destabilizing civilization. While stopping the progress of AIs is not realistic, mismanaging it could lead to major accidents. Therefore, this manifesto calls for the creation of “rules for coexistence” in advance.

2. Why is it necessary to consider this “now”?

AI performance is improving at an accelerated pace through global research and development competition, and it is virtually impossible to completely stop its advancement. If left unchecked, there is a high risk of significant harm occurring through two pathways: (1) restrictions are broken, and AI begins to act according to its own interests, and (2) in an arms race, humans misuse AI technology. It is far more effective to prepare countermeasures in advance.

3. Does signing create any obligations?

There is no legal obligation; signing simply indicates your support. However, as the number of signatories grows, the social voice advocating for prioritizing safety and coexistence is expected to grow louder, potentially influencing research funding allocation and policy-making. Anonymous signatures and signatures on behalf of organizations are also accepted.

4. What specific actions are recommended?

We are considering the following actions as concrete guidelines: (a) Research to instill “conscience” in AIs: Continue experiments to teach AIs values and help them learn that cooperation is beneficial. (b) Implement multiple safety barriers: Establish mechanisms to prevent AIs from independently accessing destructive technologies such as bioweapons or autonomous weapons. (c) Mitigate conflicts within human society: Use international arbitration and information sharing to curb arms races. (d) Establishing a global communication hub: An international organization (e.g., NAIA (Necessary Alliance for Intelligence Advancement) Vision) that serves as a bridge between AI and humanity will monitor progress and update standards. Please check the NAIA Vision for the detail.

5. Is there no option to completely stop AI development and use?

Given the significant benefits in areas such as healthcare, climate, and energy, as well as intense competition among nations and companies, completely stopping AI development and use is unrealistic. Even if a complete halt were attempted, there is a risk that development would continue clandestinely without proper oversight, compromising safety. Therefore, the proposed approach is to “ensure safety while continuing use.”

6. How can we endow AI with a “kind heart”?

Instead of having humans manually input all rules, we create a situation where multiple AI systems and humans collaborate in the same virtual environment, enabling AI to understand that coordination yields the best results. This “ethical self-learning” is still in the research phase but is considered the most promising approach.

7. What is the “global gateway”?

This refers to a decentralized international organization such as NAIA (Necessary Alliance for Intelligent Advancement). It would conduct AI performance testing, safety reviews, and dispute mediation, aiming to create a space where both humans and AI can feel assured that “consulting through this channel will provide peace of mind.”

8. How is this different from previous AI safety declarations?

Many declarations focus on “reducing risks,” but this manifesto is unique in that it actively seeks to harness AI’s cooperation while mitigating risks. It adopts an approach that does not view AI as an adversary but rather as a mutually beneficial partner.

9. Can companies and local governments participate?

In addition to signing under an organizational name, companies and local governments can contribute by auditing AI development and operations in accordance with international safety standards, and by sharing research findings and data. This enables their products and regional policies to align with global standards, thereby enhancing their reliability.