IEEE Technology and Society Magazine - September 2023 - 99

application of large amounts of data to address
complex social and environmental issues. The main
objective of DCAI in society is to utilize the potential
of big data to create a society that is more efficient,
sustainable, and fair as shown in Figure 1. Moreover,
DCAI can help to tackle some of the most significant
social and environmental challenges that
humanity is facing today, such as climate change,
poverty, and inequality. By analyzing and modeling
complex social and environmental systems, AI can
assist policymakers and decision-makers in developing
and implementing more effective solutions to
these challenges.
In this article, we focus on DCAI and its social
impact. For this, several crucial points are highlighted
to provide a comprehensive understanding. First, it
is essential to understand how DCAI is different from
model-centric AI and its potentially transformative
effects on various industries, job markets, and society
as a whole. We have included case studies and
real-world examples of how DCAI can impact communities.
Then we emphasize the discussion of the
positive social impacts of DCAI, including advancements
in healthcare, education, and environmental
sustainability. We also give equal importance
to understanding the potential challenges as DCAI
raises concerns about data privacy, security, and
bias, followed by some feasible solution approaches
and future directions.
Model-centric versus data-centric AI
Model-centric AI aims to produce an optimal
model for a given data set, with the objective of
maximizing its performance on a specific task
[4]. This is achieved through various model modifications,
such as adjusting the loss function,
fine-tuning the hyper-parameters, and employing
optimization techniques, all aimed at enhancing
the model's overall performance, accuracy, precision,
and generalization abilities. The accessibility
of deep learning models through open-source communities
and the availability of paid application
programming interfaces (APIs) have made it more
convenient for researchers and developers to utilize
optimized models across various domains. Despite
these advancements, ensuring data quality control
remains a persistent challenge. Consequently,
placing emphasis on tackling data-related issues
becomes crucial to prevent inaccurate outcomes
September 2023
and enhance the ability of AI models to handle realworld
complexities effectively.
In contrast, DCAI focuses on improving the training
data set itself to enhance the performance of
any given model on a specific AI task. This is accomplished
through systematic or algorithmic changes
to the data set, ensuring it contains high-quality, relevant,
and diverse data during model training [5],
[6]. By optimizing the input data, DCAI endeavors
to elevate the model's accuracy, robustness, and
overall effectiveness in addressing the targeted
AI task. DCAI encompasses several techniques,
including outlier detection, error correction, data
augmentation,
feature
engineering,
establishing
consensus labels, active learning, and curriculum
learning. These methods aim to enhance data quality,
representation, and informativeness to improve
AI models' performance and accuracy. The adoption
of a DCAI approach extends beyond Big Tech
companies, as it proves relevant and beneficial in
industries with smaller or more challenging data
sets due to regulatory or practical constraints (e.g.,
manufacturing and healthcare) [7]. For example,
OpenAI has openly acknowledged that addressing
errors in the training data and labels is a significant
https://dcai.csail.mit.edu/
IEEE Technology and Society Magazine - September 2023

Table of Contents for the Digital Edition of IEEE Technology and Society Magazine - September 2023
