July 7 News.United NationsSecretary-General Antonio Guterres, First United Nations Conference this Monday in Geneva, SwitzerlandAIThe Global Dialogue on Governance made a statement calling for broad and comprehensive global control of artificial intelligence。
Gutierrez said to the participants, “We may be the last generation to set the conditions for humans to coexist with machines“The construction of an artificially intelligent system of global governance is an urgent task。
Gutierrez mentioned that 1 billion people need only 2 hours of artificial intelligence technology, and the Internet has spent 15 years; that the existing regulatory system is not ready for machines that can make decisions on their own; that there is a high concentration of computing, data, and talent resources in artificial intelligence industries, and that most countries lack a voice; that artificial intelligence creates seemingly false information that erodes the foundations of trust; and that humans usually have high expectations and trust in artificial intelligence technology。
Of course, if well used and widely shared, artificial intelligence “can compress decades of development” and become “the great equality of the twenty-first century”. However, until that was achieved, the security of the technology should be thoroughly tested and the attribution of legal responsibility clearly defined。
The Secretary-General said:The development of artificial intelligent governance systems has four top priorities:
Safety
An artificial smart regulatory system that is not “symmetrically” in each country would entail additional costs and would be difficult to implement. All countries should reach consensus on front-line technical management, and systems with global coverage must meet standards of global trust。
No child should become a “experiment” of unregulated artificial intelligence, and States should adopt the Commitment to the Safety of Artificial Intelligence Children: Artificial intelligence developers need to prove that technology is safe for children and zero tolerance for sexual abuse. When children show signs of suffering, the system must stop operating and provide real support. When children are harmed, the excuse of “calculations” cannot be used。
Red Line
Human rights cannot be compromised. Artificial intelligence must not deprive human dignity or exacerbate discrimination. In every high-risk decision-making, such as justice, health care and law enforcement, machines can provide information, but humans must make decisions - and take responsibility。
quantitative (science)
Public support for artificial intelligence capacity-building in developing countries is only one “rounded error” compared to the $500 billion that the private sector invested in AI infrastructure last year. The international community must not allow the digital divide to become one of artificial intelligence. Gutierrez will soon submit to the United Nations General Assembly a proposal for a global artificial intelligence fund。
Transparent
Gutres reaffirmed its artificial-intellectual-environment transparency initiative last month, requiring each major artificial-intellectual company to measure and publicly disclose the full ecological footprint of its system and to commit to using renewable energy to power all data centres by 2030。
MOREOVER, ALTHOUGH THE FOCUS OF THE DIALOGUE WAS ON CIVILIAN ARTIFICIAL INTELLIGENCE, THE USE OF AI TECHNOLOGY IN THE MILITARY SPHERE HAD NOT BEEN OVERLOOKED BY THE SECRETARY-GENERAL OF THE UNITED NATIONS. GUTIERREZ SAID LATER IN HIS SPEECHAutonomous killing robots are not just morally repugnant, politically unacceptable,It also needs to be formally banned at the international legal level.