IEEE Computational Intelligence Magazine - May 2023 - 60

Review
Article
Zhenghua Chen and Min Wu
Agency for Science, Technology and Research (A*STAR), SINGAPORE
Alvin Chan
, Xiaoli Li , and Yew-Soon Ong
Agency for Science, Technology and Research (A*STAR) and Nanyang
Technological University, SINGAPORE
Survey on AI Sustainability: Emerging Trends
on Learning Algorithms and Research Challenges
Abstract
A
rtificial Intelligence (AI) is
a fast-growing research
and development
(R&D) discipline which is
attracting increasing attention
because it promises to bring vast
benefits for consumers and businesses,
with considerable benefits
promised in productivity
growth and innovation. To
date, significant accomplishments
have been reported in
many areasthathavebeen
deemed challenging for machines, ranging
from computer vision, natural language
processing, audio analysis to smart
sensing and many others. The technology
trend in realizing success has developed
towards increasingly complex and largesize
AI models to solve more complex
problems at superior performance and
robustness. This rapid progress, however,
has taken place at the expense ofsubstantial
environmental costs and resources. In
addition, debates on the societal impacts
ofAI, such as fairness, safety, and privacy,
have continued to grow in intensity.
Theseissueshave reflected major concerns
pertaining to the sustainable development
ofAI. In this work, major trends
in machine learning approaches that can
address the sustainability problem of AI
have been reviewed. Specifically, the
emerging AI methodologies and algorithms
are examined for addressing the
sustainability issue of AI in two major
aspects, i.e., environmental sustainability
and social sustainability ofAI. Then, the
Digital Object Identifier 10.1109/MCI.2023.3245733
Date ofcurrent version: 6April 2023
ßSHUTTERSTOCK.COM/WETZKAZ GRAPHICS
major limitations of the existing studies
are highlighted, and potential research
challenges and directions are proposed
for the development ofthe next generation
of sustainable AI techniques. It is
believed that this technical review can
help promote a sustainable development
of AI R&D activities for the research
community.
I. Introduction
Artificial Intelligence (AI) is intelligence
demonstrated by computers or machines
as opposed to the natural intelligence presented
by humans1.It iscapable of performing
tasks that conventionally require
human intelligence, and even surpasses
human capabilities in many challenging
tasks that involve a large amount ofdata to
analyze and learn from. In recent years, it
has achieved remarkable success in many
areas, such as computer vision [1],natural
1https://en.wikipedia.org/wiki/
Artificial_intelligence
Corresponding author: Zhenghua Chen (e-mail:
chen0832@e.ntu.edu.sg)
60 IEEE COMPUTATIONAL INTELLIGENCE MAGAZINE | MAY 2023
language processing [2],recommendation
systems [3],
intelligent sensing [4],condition
monitoring systems [5],
advanced Web search engines
[6], speech recognition [7],
automated decision-making
[8],and competingat the
highest level in strategic game
systems [9],etc.AIplays an
important role in enhancing
human work productivity and
efficiency, reducing operational
costs in many timeconsuming
and labour-intensive applications
across different industry verticals and
government agencies. Notable examples
ofAI applications include assisting doctors
in diagnosis, developing self-driving cars,
monitoring the operating conditions of
expensive equipment, recommending
suitable products at the right time and
location, and facilitating human communication
and interaction via machine translation
tools. From Fig. 1,it can observean
increasing trend ofAI in not only the traditional
discipline of computer science, but
also other novel application domains,
including material science, telecommunications,
physics, environmental science,
and others. In this light, it is no doubt that
AI has significant impacts on and changes
even traditional fields ofstudies and industries.
All these will bring about a new era
of interactions and augmentations
between AI and human activities.
Apart from the technical achievements
of AI, there have been increasing
concerns about the sustainable development
ofAI R&D activities [10], [11].For
instance, researchers have reported that
1556-603X ß 2023 IEEE

https://orcid.org/0000-0002-1719-0328 https://orcid.org/0000-0003-0977-3600 https://orcid.org/0000-0002-1041-3891 https://orcid.org/0000-0002-0762-6562 https://orcid.org/0000-0002-4480-169X http://www.SHUTTERSTOCK.COM/WETZKAZ https://en.wikipedia.org/wiki/Artificial_intelligence https://en.wikipedia.org/wiki/Artificial_intelligence

IEEE Computational Intelligence Magazine - May 2023

Table of Contents for the Digital Edition of IEEE Computational Intelligence Magazine - May 2023