Current advances in robotic decision-making: Is there such a thing as an intuitive robot?

The ‘rise of the machines’ in the workplace is difficult to ignore, with industrial robots now being used by almost a third of the economy. In relative comparison, other areas of robot integration (e.g., service robots), are still in their infancy. However, recent advances in the field of cognitive robotics are rapidly redefining what robots can do, allowing them to seemingly learn and reason in complex environments. The recent advent of AlphaGo©, a computer program able to master the complex ‘Go’ board game at a world class level, has been touted as a watershed moment in the design of intuitive robots. Indeed in making decisions, such machines possess a super-human cognitive stamina to visualize the likely outcomes for an array of different options, calculate probabilities regarding their success, and use knowledge previously learned from training, play, and even their own ‘imagined’ play. Elements of these functions are somewhat aligned to popular conceptions of human expertise and intuition (e.g., the recognition of familiar cases); however others contrast substantially (e.g., the deliberation between options). Ultimately, these machines may be challenging our understandings of expertise and intuition, and refocussing the decision-making paradigm in the process. Aims/objectives: The current paper aims to: 1) discuss the validity of robotic ‘intuition’; 2) forecast the impact of intuitive robotics on decision-making theory, and; 3) explore the implications of intuitive robots to the workplace, and in particular their potential impact on the traditionally ‘human’ professions. Method: The paper reviews past and current thinking surrounding information processing and decision-making, offering a decomposition of robotic intuition compared to human intuition. Conclusion: The paper illuminates persistent differences between humans and robots and concludes that although robots do not yet operate like intuitive humans, their superior processing capabilities may result in a paradigm shift away from current naturalistic decision-making approaches. This has implications for current approaches to decision support and training. Finally, the paper concludes that traditionally ‘human’ professions may be impacted more (and sooner) by advances in cognitive robotics than currently projected.