- The digital revolution started a peak conversation and adoption of research thoughts around how advanced data analytics and artificial intelligence can make the society with better living standards. Especially in the case of healthcare, artificial intelligence is much discussed and applicative. The low middle-income economies have the challenge and strategic importance towards adopting new digital technologies that require both hardware and software up-gradation. Responsible AI is an emerging area of discussion, where the ethical capability of AI in relatedness with moral responsibility in emerging technology is discussed (Tigard, 2020).
- Recent studies and practical implementation of models have addressed that AI can lead to unintentional biases, discrimination, unexpected results, and lack of transparency as to understand how results are obtained. The hidden layers of the model are expected to be explained that how they transform function and provide the output. In responsible AI, we discuss the possible standards to reduce the biases and promote interpretability and transparency of the outcomes to ensure the security and authenticity of results. Trocin et al., 2021 in their study provide in-depth analysis and propose a critical need to understand responsible AI in the field of digital health, where we have to be aware of the potential harm of the AI models to the patients and other care actors, to behave accordingly.
- Another application of responsible AI is in the field of safety applications where AI is intended to implement for social good. In this area, AI provides relatively better services in Transportation, Energy Management, Survellience and Defense systems. However, responsible AI deals with the social risks related to all the above areas where biases, imputations, imperatives of human rights, and transparency of results are discussed.
- Therefore, the decision of adopting AI models in different fields requires interdisciplinary understanding and analysis.
Assistant Professor at Siddharth University
3 年Very useful