近日，一场由智源研究院发起的国际性会议在北京举行，图灵奖得主约书亚·本吉奥、杰弗里·辛顿、姚期智等数十位中外人工智能领域的权威专家齐聚一堂，共同签署了一份具有里程碑意义的《北京AI安全国际共识》。这份文件旨在规范AI技术的发展，保障全球社会的安全与福祉。
《共识》明确了人工智能的“风险红线”和“发展路线”，将AI的自我复制和应用于生化武器明确列为禁止行为。在“风险红线”部分，专家们一致认为，AI系统不应具有自主复制和自我改进的能力，以防止技术失控带来的潜在威胁。此外，AI技术也不得被用于寻求权力、协助违法或不道德行为，以及进行误导性的欺骗行为。
这一举措显示出国际社会对AI安全问题的高度重视，旨在通过制定全球性的规范，防止技术滥用，确保AI在为人类带来便利的同时，不会对社会秩序和人类安全构成威胁。《北京AI安全国际共识》的签署，标志着全球AI治理迈出了坚实一步，为未来AI技术的健康发展提供了重要的道德和法律指引。
英语如下：
**News Title:** “World’s Top Experts Agree: Ban on AI Self-Replication to Mitigate Biochemical Weapon Risks”
**Keywords:** AI safety consensus, expert signatures, prohibition on replication
**News Content:**
Title: Top Global Experts in Beijing Sign the “Beijing AI Safety International Consensus”: Establishing Red Lines for AI Development, Banning Self-Replication and Biochemical Applications
Recently, an international conference, initiated by the BAAI Institute, was held in Beijing, gathering renowned AI experts such as Turing Award winners Yoshua Bengio, Geoffrey Hinton, and Yao Qizhi, among several others, from China and abroad. Together, they signed a landmark “Beijing AI Safety International Consensus,” aimed at regulating AI technology development and ensuring global safety and well-being.
The Consensus outlines the “risk red lines” and “development pathways” for AI. It specifically prohibits self-replication and the application of AI in biochemical weapons. In the “risk red lines” section, experts uniformly agreed that AI systems should not possess the ability to autonomously replicate or self-improve, to prevent potential threats from technology失控. Additionally, AI technology must not be employed to seek power, facilitate illegal or unethical activities, or engage in deceptive misrepresentation.
This initiative demonstrates the international community’s high level of concern for AI safety, striving to establish global guidelines to prevent misuse and ensure that AI, while bringing convenience, does not pose threats to social order and human security. The signing of the “Beijing AI Safety International Consensus” marks a significant step forward in global AI governance, providing crucial ethical and legal guidance for the healthy development of AI technology in the future.
【来源】https://new.qq.com/rain/a/20240318A04IFG00
Views: 1