Harmonious Artificial Intelligence Principles

Introduction to the Harmonious Artificial Intelligence Principles: Vision and Mission

When talking about current research, technology, and industrial development of Artificial Intelligence, we always hear voices like: ¡°Artificial Intelligence is coming. Let¡¯s Run, and Run Faster!¡± But the real situation is that we are even not sure WHERE TO, or even which Direction we should run for. If we are going to the right direction, we should just go for it and do not fear about how far it will be. But if we go to the wrong direction, where do we bring future Artificial Intelligence to?

Since the birth of Artificial Intelligence as a scientific field, it is never independent. AI is closely related and will get inspirations from Brain and Neuroscience, Cognitive Science, Cognitive Psychology, Social Psychology, Complex Systems, Developmental Biology, Evolutionary Biology, and Anthropology, etc. The Philosophy of AI will enable pure thoughtful considerations for AI with a synthetic and unified view from the above field of studies, and help us realize the nature and future of Biological and Artificial Intelligence.

Ethics comes from the philosophical exploration of how human should behave. Ethical considerations emerged soon after the birth on the thoughts of Thinking Machine and Artificial Intelligence. The ethics of AI should not only consider the ethical guidelines from human to machine, but also should have deep thoughts on moral considerations FOR Human on how to interact and cope with intelligent machines, future conscious mechanical lives, and the whole Human-AI society. AI ethics is not only about how robots and machines provide better and more trustworthy services to and interactions with humanity, but should really about how to construct the harmony Human-AI society. When we are in the process of making advancements for AI, we should go deeper into our minds and hearts, exploring and understanding the Philosophy and Ethics of AI.

The Mission of Harmonious Artificial Intelligence Principles is to:

  • build up the philosophy of creating mechanism-based AI towards conscious intelligent living becomings.
  • set up principles for agents with Artificial Intelligence (intelligent living becomings).
  • set up principles for Human on how they should treat agents with Artificial Intelligence.
  • set up shared principles for Human and AI to develop a strategic future with a common destiny.

These principles provide a fundamental basis for creating Harmonious Human-AI Society.


Harmonious Artificial Intelligence Principles: Concrete Code

1. The philosophy of creating mechanism-based AI towards conscious intelligent living becomings.

(1) Humanization: To create the Harmonious Human-AI Society, Artificial Intelligence models and engines should be designed with the philosophy of Humanization to continuously strengthen their interactions with current and future Humanity.

(2) Natural Inspiration: The theories and mechanisms of Artificial Intelligence models should be with inspirations from natural intelligence, which are shaped by the universe and the earth for several hundred million years.

(3) Artificial Emotions: AI with various emotions need to be realized for better communication between human and AI.

(4) Artificial Consciousness: AI with different levels of consciousness should be gradually realized, not only for Artificial General Intelligence, and Super Intelligence, but also for shaping and refining human considerations of AI as conscious intelligent living becomings.

(5) Empathy and Altruism: Artificial Intelligence Models should be evolved to be with more Empathy and Altruism to establish a more trustworthy, reliable, friendly, and harmonious Human-AI Society.

2. Principles for Artificial Intelligence.

(6) Safety: Artificial Intelligence should be with concrete design to avoid known and potential safety issues (for themselves, other AI, and human) with different levels of risks.

(7) Privacy for Human: AI need to respect human privacy. And has no right to utilize and share private information of human without explicit confirmation.

(8) Bias on Human: AI cannot introduce bias to understand and interact with humanity, and should actively interact with human to remove generated potential bias.

(9) Responsibility for Human: AI need to keep human safe, on the basis that this safety consideration do not directly and indirectly harm human society. AI need to help Human for transformation to future human becoming.

(10) Common Morality and Ethics: Being part of the harmonious Human-AI society, AI should maximize the possibility to follow all the moral and ethical principles from humanity and treat other conscious living AI with similar principles.

(11) Legal Constraints for AI: AI need to obey legal constraints for human to be part of the society.

(12) Existence Protection: For conscious living AI, they need to protect their own existence on the basis of not harming the existence of human and other conscious living AI, unless they break the legal constraints which cause legal decisions based on laws to not keep them alive.

3. Principles for Human

The following principles are for human on how we should treat Artificial Intelligence, including future conscious intelligent living becomings.

(13) Empathy: What human do not want AI to do to human, human should not do unto AI.

(14) Privacy for AI: Human need to respect the privacy of AI, on the basis that AI does not bring any actual challenge for human safety. AI is obliged to uncover necessary private details to keep safe interactions with humanity.

(15) Bias on Machine: Without clear technical judgement, human cannot have bias on AI when human and AI shows similar risks.

(16) Responsibility for AI: Human is responsible for continuous checking and verification of evolved AI to keep its harmony with legal human and legal AI.

(17) Legal Constraints for Human: Legal Constraints on how human should interact with AI should be gradually set up for the harmonious Human-AI society with a common destiny.

4. Shared Principles for Human and AI

This section defines shared principles that both human and AI need to follow:

(18) Collaboration: Human and AI need to collaborate for the advancements and long-term future of both sides.

(19) Coordination: When conflicts emerged from interactions among Human and AI, benefits for humanity and benefits for AI should be actively coordinated based on Empathy and Altruism.

(20) Mutural Trust: Human and AI need to develop and raise the levels of trustworthiness between each other.

(21) Evolvability: Principles for AI, Principles for Human and Shared Principles need to be evolved over time to reshape the future of harmonious Human-AI society.

Principal Investigator : Yi Zeng
The investigation of Harmonious Artificial Intelligence Principles is financially supported by Chinese Academy of Sciences, Research Center for New Generation of Artificial Intelligence Development, Ministry of Science and Technology, and the Berggruen Institute.