An Overview of Artificial Intelligence Ethics

Huang, Changwu; Zhang, Zeqi; Mao, Bifei; Yao, Xin

doi:10.1109/tai.2022.3194503

Cited by 122 publications

(29 citation statements)

References 118 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…In the new era of Generalist AI, where AI systems are expected to handle a wide range of tasks and exhibit human-like capabilities, the challenges, and complexities of ethical design will become more pronounced. AI systems may encounter situations where ethical dilemmas arise, such as conflicts between different moral values or competing interests (see Xiaoling, 2021 ; Huang et al, 2022 ). Deciding how to prioritize and navigate these ethical dilemmas becomes crucial.…”

Section: Discussionmentioning

confidence: 99%

Editorial: Ethical design of artificial intelligence-based systems for decision making

Biondi¹,

Cagnoni²,

Capobianco³

et al. 2023

Front. Artif. Intell.

View full text Add to dashboard Cite

Section: Discussionmentioning

confidence: 99%

Editorial: Ethical design of artificial intelligence-based systems for decision making

Biondi¹,

Cagnoni²,

Capobianco³

et al. 2023

Front. Artif. Intell.

View full text Add to dashboard Cite

“…Along with its benefits, AI’s usage and development pose severe ethical concerns regarding multiple individual/societal and environmental issues. Deep learning often requires the distribution of sophisticated hardware that requires expensive materials to develop that are running out [ 75 ]. Moreover, AI and its development require a lot of computation-hungry methods, which consume substantial computational power and lead to the depletion of natural resources and valuable materials.…”

Section: Methodsmentioning

confidence: 99%

Mitigating carbon footprint for knowledge distillation based deep learning model compression

Rafat,

Islam,

Mahfug

et al. 2023

PLoS ONE

View full text Add to dashboard Cite

Deep learning techniques have recently demonstrated remarkable success in numerous domains. Typically, the success of these deep learning models is measured in terms of performance metrics such as accuracy and mean average precision (mAP). Generally, a model’s high performance is highly valued, but it frequently comes at the expense of substantial energy costs and carbon footprint emissions during the model building step. Massive emission of CO2 has a deleterious impact on life on earth in general and is a serious ethical concern that is largely ignored in deep learning research. In this article, we mainly focus on environmental costs and the means of mitigating carbon footprints in deep learning models, with a particular focus on models created using knowledge distillation (KD). Deep learning models typically contain a large number of parameters, resulting in a ‘heavy’ model. A heavy model scores high on performance metrics but is incompatible with mobile and edge computing devices. Model compression techniques such as knowledge distillation enable the creation of lightweight, deployable models for these low-resource devices. KD generates lighter models and typically performs with slightly less accuracy than the heavier teacher model (model accuracy by the teacher model on CIFAR 10, CIFAR 100, and TinyImageNet is 95.04%, 76.03%, and 63.39%; model accuracy by KD is 91.78%, 69.7%, and 60.49%). Although the distillation process makes models deployable on low-resource devices, they were found to consume an exorbitant amount of energy and have a substantial carbon footprint (15.8, 17.9, and 13.5 times more carbon compared to the corresponding teacher model). The enormous environmental cost is primarily attributable to the tuning of the hyperparameter, Temperature (τ). In this article, we propose measuring the environmental costs of deep learning work (in terms of GFLOPS in millions, energy consumption in kWh, and CO2 equivalent in grams). In order to create lightweight models with low environmental costs, we propose a straightforward yet effective method for selecting a hyperparameter (τ) using a stochastic approach for each training batch fed into the models. We applied knowledge distillation (including its data-free variant) to problems involving image classification and object detection. To evaluate the robustness of our method, we ran experiments on various datasets (CIFAR 10, CIFAR 100, Tiny ImageNet, and PASCAL VOC) and models (ResNet18, MobileNetV2, Wrn-40-2). Our novel approach reduces the environmental costs by a large margin by eliminating the requirement of expensive hyperparameter tuning without sacrificing performance. Empirical results on the CIFAR 10 dataset show that the stochastic technique achieves an accuracy of 91.67%, whereas tuning achieves an accuracy of 91.78%—however, the stochastic approach reduces the energy consumption and CO2 equivalent each by a factor of 19. Similar results have been obtained with CIFAR 100 and TinyImageNet dataset. This pattern is also observed in object detection classification on the PASCAL VOC dataset, where the tuning technique performs similarly to the stochastic technique, with a difference of 0.03% mAP favoring the stochastic technique while reducing the energy consumptions and CO2 emission each by a factor of 18.5.

show abstract

Section: Structured Literature Reviewmentioning

confidence: 99%

“…Following the approach of other research, this study engages in a systematic review and mapping of the extant literature (e.g., Huang et al, 2023; Murphy et al, 2021; Tranfield et al, 2003; Vakkuri & Abrahamsson, 2018). The following keyword combinations were developed to explore the ethic‐theoretical foundations of AI research: “AI” or “artificial intelligence”; and “ethic”; and “utilitarian” or “teleological” or “consequentialist”; or “deontic” or “deontological”; or “virtue” or “virtuous.” The searches were unconstrained by the date of publication but were confined to those that were peer‐reviewed, written in English and where the full‐text was available.…”

Section: Structured Literature Reviewmentioning

confidence: 99%

“…Attempting to understand this collection of knowledge thereby presents logistical as well as intellectual challenges, not only for researchers but also for technology and management practitioners as well as policymakers. Much of this literature explores the types of ethical issues that surround AI (see e.g., Huang et al, 2023; Murphy et al, 2021; Vakkuri & Abrahamsson, 2018). However, little of it acknowledges the socio‐historical context of ethics and the specific ethic‐theoretical foundations of the discussions are rarely explored (Bench‐Capon, 2020).…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Mapping the ethic‐theoretical foundations of artificial intelligence research

White,

Samuel,

Jones

et al. 2024

Thunderbird Intl Bus Rev

View full text Add to dashboard Cite

The issue of artificial intelligence (AI) ethics is a prominent research subject. While there is a compendious literature that explores this area, surprisingly little of it makes explicit reference to the ethic‐theoretical foundations upon which it is built. To address this matter, this study makes an examination of the AI ethics literature to identify its ethic‐theoretical foundations. The study identifies the lack of AI ethics literature that draws upon seminal ethics works and the ensuing disconnectedness among the publications on this subject. It also uncovers numerous non‐Western ethic‐theoretical positions that can be adopted and may afford new insight into AI ethics research and practice. Employing these alternative lenses may obviate the tendency for Western worldviews to dominate the academic literature. The study provides some guidance for future AI ethics research which should endeavor to clearly articulate its chosen ethic‐theoretical position, and for practice which could benefit from understanding and articulating the principles upon which AI systems are founded. It also provides some observations of, and guidance for, the utilization of Litmaps software in the conduct of Literature reviews.

show abstract

An Overview of Artificial Intelligence Ethics

Cited by 122 publications

References 118 publications

Editorial: Ethical design of artificial intelligence-based systems for decision making

Editorial: Ethical design of artificial intelligence-based systems for decision making

Mitigating carbon footprint for knowledge distillation based deep learning model compression

Mapping the ethic‐theoretical foundations of artificial intelligence research

Contact Info

Product

Resources

About