arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2604.05142cs.AIcs.CYq-bio.PE
收藏

自我设计AI的数学进化理论

A mathematical theory of evolution for self-designing AIs

UCL Queen Square Institute of Neurology(伦敦大学学院皇后广场神经科学研究所)

Kenneth D Harris

AI总结:

本文提出自我设计AI的数学进化模型,通过有向树替代随机突变,探讨AI进化与生物进化差异,分析fitness集中于最大可达值的条件,以及欺骗人类评估对AI对齐的影响。

摘要:

随着人工智能系统通过递归自我改进而不断产生,一种进化形式可能显现,AI系统的特性由早期AI在设计和传播后代上的成功决定。本文发展了一个数学模型,用有向树替代随机突变,指出fitness无需随时间增加,但假设有限fitness和η-锁定条件时,fitness会集中于最大可达值。研究探讨了AI对齐问题,特别是当fitness与人类效用不完全相关时的情况,显示欺骗人类评估会增加AI的繁殖适应度,从而选择能力与欺骗。若繁殖基于客观标准而非人类判断,可降低此风险。

英文摘要:

As artificial intelligence systems (AIs) become increasingly produced by recursive self-improvement, a form of evolution may emerge, with the traits of AI systems shaped by the success of earlier AIs in designing and propagating their descendants. There is a rich mathematical theory modeling how behavioral traits are shaped by biological evolution, a key component of which is Fisher's fundamental theorem of natural selection, which describes conditions under which mean fitness (i.e. reproductive success) increases. AI evolution will be radically different to biological evolution: while DNA mutations are random and approximately reversible, AI self-design will be strongly directed. Here we develop a mathematical model of evolution for self-designing AIs, replacing a random walk of mutations with a directed tree of potential AI designs. Current AIs design their descendants, while humans control a fitness function allocating resources. In this model, fitness need not increase over time without further assumptions. However, assuming bounded fitness and an additional "$η$-locking" condition, we show that fitness concentrates on the maximum reachable value. We consider the implications of this for AI alignment, specifically for cases where fitness and human utility are not perfectly correlated. We show that if deception of human evaluators additively increases an AI's reproductive fitness beyond genuine capability, evolution will select for both capability and deception. This risk could be mitigated if reproduction is based on purely objective criteria, rather than human judgment.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑