The five Is: Key principles for interpretable and safe conversational AI
Paper i proceeding, 2021

In this position paper, we present five key principles, namely interpretability, inherent capability to explain, independent data, interactive learning, and inquisitiveness, for the development of conversational AI that, unlike the currently popular black box approaches, is transparent and accountable. At present, there is a growing concern with the use of black box statistical language models: While displaying impressive average performance, such systems are also prone to occasional spectacular failures, for which there is no clear remedy. In an effort to initiate a discussion on possible alternatives, we outline and exemplify how our five principles enable the development of conversational AI systems that are transparent and thus safer for use. We also present some of the challenges inherent in the implementation of those principles.

Interpretable AI

conversational agents

multimodal human-computer interaction

dialogue managers

Författare

Mattias Wahde

Chalmers, Mekanik och maritima vetenskaper, Fordonsteknik och autonoma system

Marco Virgolin

Chalmers, Mekanik och maritima vetenskaper, Fordonsteknik och autonoma system

ACM International Conference Proceeding Series

50-54
9781450385930 (ISBN)

4th International Conference on Computational Intelligence and Intelligent Systems, CIIS 2021
Virtual, Online, Japan,

Ämneskategorier

Annan data- och informationsvetenskap

Programvaruteknik

Systemvetenskap

DOI

10.1145/3507623.3507632

Mer information

Senast uppdaterat

2022-05-20