[论文解读] The Future of Fundamental Science Led by Generative Closed-Loop Artificial Intelligence

Héctor Zenil, Jesper Tegnér+18 more|arXiv (Cornell University)|Jul 9, 2023

Scientific Computing and Data Management被引用 11

一句话总结

本观点概述了一个以AI驱动、自动化、生成的闭环科学发现的愿景，该系统能够提出假设、进行实验，并寻求揭示超越当前AI应用的基本解释模型。

ABSTRACT

Recent advances in machine learning and AI, including Generative AI and LLMs, are disrupting technological innovation, product development, and society as a whole. AI's contribution to technology can come from multiple approaches that require access to large training data sets and clear performance evaluation criteria, ranging from pattern recognition and classification to generative models. Yet, AI has contributed less to fundamental science in part because large data sets of high-quality data for scientific practice and model discovery are more difficult to access. Generative AI, in general, and Large Language Models in particular, may represent an opportunity to augment and accelerate the scientific discovery of fundamental deep science with quantitative models. Here we explore and investigate aspects of an AI-driven, automated, closed-loop approach to scientific discovery, including self-driven hypothesis generation and open-ended autonomous exploration of the hypothesis space. Integrating AI-driven automation into the practice of science would mitigate current problems, including the replication of findings, systematic production of data, and ultimately democratisation of the scientific process. Realising these possibilities requires a vision for augmented AI coupled with a diversity of AI approaches able to deal with fundamental aspects of causality analysis and model discovery while enabling unbiased search across the space of putative explanations. These advances hold the promise to unleash AI's potential for searching and discovering the fundamental structure of our world beyond what human scientists have been able to achieve. Such a vision would push the boundaries of new fundamental science rather than automatize current workflows and instead open doors for technological innovation to tackle some of the greatest challenges facing humanity today.

研究动机与目标

• 激发AI从辅助当前工作流程转向发现新的科学规律的必要性。
• 提出一个闭环、自治框架，将AI与实验室自动化整合，用于端到端的科学发现。
• 讨论生成式AI和大型语言模型如何将高层猜想转化为自动化循环中的可计算模块。
• 在AI主导的科学中，解决因果性、可解释性、可重复性和治理等挑战。

提出的方法

• 提出一个形式化的面向AI驱动的科学发现的闭环迭代周期，包含假设生成、模型构建、实验和知识整合。
• 主张将因果分析与生成建模结合，以揭示透明、可解释的表示和潜在规律。
• 建议利用LLMs和GenAI将高层猜想分解为循环内的可计算组件。
• 强调需要人机协作和元AI策略来管理知识表示和代理交互。
• 讨论历史背景和先前的AI在科学中的努力，将所提出的框架置于更广泛的轨迹之中。

实验结果

研究问题

• RQ1 一个AI驱动的闭环系统要在科学中发现新的基本解释模型，需要哪些条件？
• RQ2 如何将生成式AI、LLMs以及符号/因果方法整合，以生成可解释、可测试的科学假设？
• RQ3 为了实现自主AI驱动的科学，需要解决哪些治理、可重复性与信任方面的考量？
• RQ4 在科学探究的不同循环中，人类和AI应扮演哪些角色，以最大化发现与可靠性？

主要发现

• 如果将AI嵌入闭环框架中，可以自动进行观测、假设生成、实验和文献整合的循环。
• 当前的AI方法缺乏用于基本发现所需的高级抽象和因果推理的机制；整合生成式与符号方法可以填补这一空缺。
• LLMs可以充当界面和经纪人，将人类研究目标转化为循环内的模块化、可计算任务。
• 主动学习和因果分析对于避免数据偏差、过拟合和模型崩溃在AI驱动的科学中至关重要。
• 协同的人机方法可能带来更快的发现、更高的可重复性和科学实践的民主化。