Emotion Artificial Intelligence (Emotion AI) is one of the most developing and fastest-growing technologies globally (Gartner, 2019). As Gartner (2019) predicts, “By 2022, 10% of personal devices will have emotion AI capabilities.” Emotion AI will be the next phase of technology development in which companies can integrate emotional understanding and consciousness into AI and humanize this technology for the good of humanity. The industry is projected to grow at USD 25 billion by 2023, with a CAGR of 17% (2017 – 2023) (Market Research Future, n.d). However, despite this technology’s potential, there are still many discussions raising psychological, technical, and ethical concerns about the limitations and feasibility of building such a platform to detect emotions. 

Crawfold (2019) argues that this field was built on “markedly shaky foundations,” meaning no scientific consensus on emotion detections. In this paper, I say that this scientific dissensus is the core constraint of this industry, leading to technical and commercial problems. I used the word “constraint” instead of “obstacle” because I believe this problem is not something businesses can overcome easily, nor can scientists figure it out in a short period. The scientific dissensus negatively influences all steps in the process of developing the Emotion AI model: including input choices, data acquisition, data preparation, feature engineering, model training, model deployment, and output display. It also leads to challenges to regulate this field and raises issues about usefulness, data protection, and ethics. These issues reduce the public adoption and acceptance of Emotion AI applications, affecting revenue generation of businesses working in this field.

Therefore, this project aims to explain these core challenges and develop strategies to overcome or work around them. In the first part of the paper, I will discuss how the core constraint – the lack of scientific consensus – happens and influences both the product development and the commercialization process of an Emotion AI application. The second part of the paper aims to provide directions that a business can follow in dealing with this industry’s uncertainty. Although these directions may only help companies to deal with these problems at a macro level, they could be a starting point for developing sustainable strategies to support businesses in the long term.

Download the full report at: https://cutt.ly/AxDcO59