AI Value Alignment Problem: The Clear and Present Danger

Artificial intelligence (AI) refers to the development of machines that are able to think and act like humans. Current advances in AI are targeting the advancements of systems that perform at super human level in particular tasks or domains. The critical challenge in developing AI is ensuring that the technology is safe and can be trusted. Because AI systems are capable of making decisions and taking actions without human intervention, it is crucial to ensure that they do not make mistakes or act in ways that are harmful to people or the environment. In this paper, we aim to present a clear conceptual understanding of the AI value alignment problem, the need to solve this problem, the major issues involved, approaches to tackle them and the way forward.