[论文解读] AI-GAs: AI-generating algorithms, an alternate paradigm for producing general artificial intelligence

Jeff Clune|arXiv (Cornell University)|May 27, 2019

Machine Learning and Data Classification参考文献 189被引用 52

一句话总结

本文主张以 AI-GAs 作为通往通用人工智能的另一条路径，概述三大支柱——元学习架构、元学习学习算法，以及生成高效学习环境——并讨论为何 AI-GAs 可能是最快的路线，同时仍然重视手工路径。

ABSTRACT

Perhaps the most ambitious scientific quest in human history is the creation of general artificial intelligence, which roughly means AI that is as smart or smarter than humans. The dominant approach in the machine learning community is to attempt to discover each of the pieces required for intelligence, with the implicit assumption that some future group will complete the Herculean task of figuring out how to combine all of those pieces into a complex thinking machine. I call this the "manual AI approach". This paper describes another exciting path that ultimately may be more successful at producing general AI. It is based on the clear trend in machine learning that hand-designed solutions eventually are replaced by more effective, learned solutions. The idea is to create an AI-generating algorithm (AI-GA), which automatically learns how to produce general AI. Three Pillars are essential for the approach: (1) meta-learning architectures, (2) meta-learning the learning algorithms themselves, and (3) generating effective learning environments. I argue that either approach could produce general AI first, and both are scientifically worthwhile irrespective of which is the fastest path. Because both are promising, yet the ML community is currently committed to the manual approach, I argue that our community should increase its research investment in the AI-GA approach. To encourage such research, I describe promising work in each of the Three Pillars. I also discuss AI-GA-specific safety and ethical considerations. Because it it may be the fastest path to general AI and because it is inherently scientifically interesting to understand the conditions in which a simple algorithm can produce general AI (as happened on Earth where Darwinian evolution produced human intelligence), I argue that the pursuit of AI-GAs should be considered a new grand challenge of computer science research.

研究动机与目标

• 展示手工 AI 方法及其在组装实现通用 AI 所需的众多构件中的挑战。
• 介绍 AI-GAs 作为一种可扩展的替代方案，它通过学习来构建通用 AI，而无需逐一手工设计每个组件。
• 概述对 AI-GAs 至关重要的三大支柱，并讨论预期的研究方向与安全考量。
• 倡导将研究投资转向 AI-GAs，同时承认手工路径的价值。

提出的方法

• 描述三大支柱：（1）元学习架构，（2）元学习学习算法，（3）生成高效学习环境。
• 将 AI-GA 与手工方法进行比较，并讨论架构、学习者和环境的自动化学习如何使通用 AI 成为可能。
• 回顾现有工作并提出每一支柱的研究方向，包括架构搜索、学习优化器，以及课程/数据生成。
• 主张将 AI-GAs 作为一项伟大挑战进行框架化讨论，并讨论潜在的安全与伦理考量。

实验结果

研究问题

• RQ1 一个 AI-generating algorithm (AI-GA) 能否从零开始通过三大相互关联的支柱来构建通用 AI？
• RQ2 元学习架构、元学习学习算法以及自动环境/课程生成存在哪些研究方向和技术挑战？
• RQ3 在什么条件下，AI-GAs 可以在产生通用智能方面超越手工 AI 路径？

主要发现

• 架构搜索已开始在 CIFAR 和 ImageNet 等基准测试上超越手工设计的架构，表明第一支柱具有潜力。
• 元学习方法（如 MAML 和基于 RNN 的元学习器）可以带来更快的适应和内部引导的学习策略，诸如可微塑性和神经调制等方法提升持续学习。
• 生成高效学习环境和课程被确认为最少被探索但最难的支柱，有潜力通过任务分布和训练数据对学习过程产生有意义的影响。
• AI-GA 框架可能实现可扩展、开放式的进展，并为可能的智能的起源和空间提供洞见，同时即使长期目标未完全实现也具有短期价值。
• 本文讨论了 AI-GAs 独有的安全与伦理考量，并主张将 AI-GAs 视为一项伟大科学挑战以及通往通用 AI 的潜在最快路径。