Machine Thinking for Telehealth – Toward the Next Generation of Healthcare

The revival of artificial intelligence (AI) a few years ago has often been attributed to the success of machine learning (ML), especially the branch of ML known as deep learning (DL). Since the introduction of workable DL, numerous success stories have been widely reported: board games, autonomous driving vehicles, natural language translation, voice recognition, etc. However, for machines to exhibit a high degree of intelligence, they need more than the ability to learn. Indeed, heavy reliance on ML has resulted in widely publicized mishaps that negatively affect people in unexpected ways. For healthcare and other mission-critical applications, intelligent machines need to perform dependably to earn and maintain the trust of human users. To this end, it is necessary to take a holistic view toward AI, one that encompasses multiple facets of what constitutes intelligence. Specifically, this paper previews what thinking machines would be capable of by taking into consideration a machine's ability to abstract and reason, in addition to its ability to learn.