AREA I: Augmentation of the human mental system
AREA I Overview
General Principle: Human cognition is a mental system involving a number of interrelated activities supporting thinking, affect (emotion, trust, beliefs, and values), and action (doing): deciding (planning and choosing), perceiving (observing), reasoning and analyzing (figuring out and solving problems), conceptualization, contemplation, and questioning, investigating (examining), imagining (and dreaming), and responding (includes adapting). In every human activity this mental system is dynamically in play across these activities at various levels of biological neural activation. A specific mental activity does not operate in complete isolation in the human mental system but rather interdependently. Thus, affective dispositions, involving for example trust, beliefs, and values, can influence thinking activities associated with observing, reasoning, analyzing, questioning, deciding, and responding.
Relevance for operational applications: Any strategy dependent upon improving human decision-making needs to account for what is likely to be involved when augmenting mental activities with artificial intelligence (AI) and machine learning (ML). This level of accounting for the mental system in operational strategy is the application of ‘meta’ cognitive knowledge (metacognition) about how a mental system (augmented by AI) is likely to operate beneficially, or not, in support of decision-making and decisionmakers (to include the context of decision-making among and across teams). The quality of AI metacognition, addressing how well the mental system is accounted for and properly augmented by AI in operational applications, is expected to evolve over time. Greater precision will be necessary in how advances in cognitive sciences are applied to improve and align advantages offered by AI for properly supporting human-machine teaming and augmenting human decision-making. The area of AI metacognition is not only reserved for humans, but can also be expected to grow in relevance to smart machines as AI-based systems improve with means to address how well their supporting artificial mental systems are performing in relation to higher goals and then make appropriate adjustments on the basis of accurate self-assessment. Additionally, strategists will need to have adequate levels of AI metacognition to be able to anticipate scenarios involved not only with effective use of human-centered AI systems, but also be able to factor into their strategic thinking the potential, forms, and countermeasures for adversarial use of AI systems that may be employed with human-out-the-loop processes (e.g. AI-enabled autonomous systems).
Near term considerations: AI metacognition among strategists, regarding how best to employ and use AI in support of decision-making and strategy in operational applications is just coming into existence and beginning to display signs of future potential. Thus, rapid and comprehensive human-machine teaming with AI in operational applications, in support of decision-making and strategy, ought to proceed carefully on the basis of thorough simulation at this nascent stage of employment.