Estimating the Brittleness of AI: Safety Integrity Levels and the Need for Testing Out-Of-Distribution Performance

Abstract

Test, Evaluation, Verification, and Validation (TEVV) for Artificial Intelligence (AI) is a challenge that threatens to limit the economic and societal rewards that AI researchers have devoted themselves to producing. A central task of TEVV for AI is estimating brittleness, where brittleness implies that the system functions well within some bounds and poorly outside of those bounds. This paper argues that neither of those criteria are certain of Deep Neural Networks. First, highly touted AI successes (eg. image classification and speech recognition) are orders of magnitude more failure-prone than are typically certified in critical systems even within design bounds (perfectly in-distribution sampling). Second, performance falls off only gradually as inputs become further Out-Of-Distribution (OOD). Enhanced emphasis is needed on designing systems that are resilient despite failure-prone AI components as well as on evaluating and improving OOD performance in order to get AI to where it can clear the challenging hurdles of TEVV and certification.

References in corpus (7)

• Very Deep Convolutional Networks for Large-Scale Image Recognition
• Self-training with Noisy Student improves ImageNet classification
• Fixing the train-test resolution discrepancy: FixEfficientNet
• Pretrained Transformers Improve Out-of-Distribution Robustness
• Benchmarking Adversarial Robustness
• Anomalous Example Detection in Deep Learning: A Survey
• Adversarial Examples for Cost-Sensitive Classifiers

Cited by in corpus (3)

• Actionable Principles for Artificial Intelligence Policy: Three Pathways
• Evaluating the Robustness of Collaborative Agents
• Improved OOD Generalization via Adversarial Training and Pre-training