The potential of artificial intelligence (AI) to improve decision making

: investigating the reliance on AI advice in a business context

Abstract

With the intention of overcoming human decision making biases, organizations are increasingly using AI as decision support. However, to unlock the full potential of AI-based advice to improve decision making, users must be willing to rely on it in the first place. To better understand people’s readiness to accept advice from AI, two experimental studies were conducted in the scope of this research. Study 1 examined whether people rely more on advice coming from AI or a human. People showed algorithm appreciation in both tasks – the performance evaluation of an employee and the closing price prediction of a stock. The effect was fully mediated by people’s trust in the source and varied across different levels of confidence in one’s own decision. Study 2 examined whether people also choose AI advice over human advice when presented with both options and whether they choose equally for themselves and for others. In this setting, algorithm appreciation persisted only for the stock price prediction task, irrespectively of who the decision was made for. Furthermore, several influencing factors were identified that point to domains where AI is most likely to be accepted and ways in which its benefits can be maximized. The results from these studies have clear implications for organizations that turn to Big Data and AI-generated advice to improve decision making, suggesting that AI might be a good addition to their daily operations.

Date of Award	27 Jan 2023
Original language	English
Awarding Institution	Universidade Católica Portuguesa
Supervisor	Filipa de Almeida (Supervisor)