Electrophysiological indexes of option characteristic processing

Psychological and Brain Responses to Artificial Intelligence's Violation of Community Ethics

Human moral reactions to artificial intelligence (AI) agents' behavior constitute an important aspect of modern-day human-AI relationships. Although previous studies have mainly focused on autonomy ethics, this study investigates how individuals judge AI agents' violations of community ethics (including betrayals and subversions) compared with human violations. Participants' behavioral responses, event-related potentials (ERPs), and individual differences were assessed. Behavioral findings reveal that participants rated AI agents' community-violating actions less morally negative than human transgressions, possibly because AI agents are commonly perceived as having less agency than human adults. The ERP N1 component showed the same pattern with moral rating scores, indicating the modulation effect of human-AI differences on initial moral intuitions. Moreover, the level of social withdrawal correlated with a smaller N1 in the human condition but not in the AI condition. The N2 and P2 components were sensitive to the difference between the loyalty/betrayal and authority/subversion domains but not human/AI differences. Individual levels of moral sense and autistic traits also influenced behavioral data, especially on the loyalty/betrayal domain. In our opinion, these findings offer insights for predicting moral responses to AI agents and guiding ethical AI development aligned with human moral values.

Uncertainty deconstructed: conceptual analysis and state-of-the-art review of the ERP correlates of risk and ambiguity in decision-making

Risk and uncertainty are central concepts of decision neuroscience. However, a comprehensive review of the literature shows that most studies define risk and uncertainty in an unclear fashion or use both terms interchangeably, which hinders the integration of the existing findings. We suggest uncertainty as an umbrella term that comprises scenarios characterized by outcome variance where relevant information about the type and likelihood of outcomes may be somewhat unavailable (ambiguity) and scenarios where the likelihood of outcomes is known (risk).These conceptual issues are problematic for studies on the temporal neurodynamics of decision-making under risk and ambiguity, because they lead to heterogeneity in task design and the interpretation of the results. To assess this problem, we conducted a state-of-the-art review of ERP studies on risk and ambiguity in decision-making. By employing the above definitions to 16 reviewed studies, our results suggest that: (a) research has focused more on risk than ambiguity processing; (b) studies assessing decision-making under risk often implemented descriptive-based paradigms, whereas studies assessing ambiguity processing equally implemented descriptive- and experience-based tasks; (c) descriptive-based studies link risk processing to increased frontal negativities (e.g., N2, N400) and both risk and ambiguity to reduced parietal positivities (e.g., P2, P3); (d) experience-based studies link risk to increased P3 amplitudes and ambiguity to increased frontal negativities and the LPC component; (e) both risk and ambiguity processing seem to be related with cognitive control, conflict monitoring, and increased cognitive demand; (f) further research and improved tasks are needed to dissociate risk and ambiguity processing.

The hierarchical sensitivity to social misalignment during decision-making under uncertainty

Social misalignment occurs when a person’s attitudes and opinions deviate from those of others. We investigated how individuals react to social misalignment in risky (outcome probabilities are known) or ambiguous (outcome probabilities are unknown) decision contexts. During each trial, participants played a forced-choice gamble, and they observed the decisions of four other players after they made a tentative decision, followed by an opportunity to keep or change their initial decision. Behavioral and event-related potential data were collected. Behaviorally, the stronger the participants’ initial preference, the less likely they were to switch their decisions, whereas the more their decisions were misaligned with the majority, the more likely they were to switch. Electrophysiological results showed a hierarchical processing pattern of social misalignment. Misalignment was first detected binarily (i.e. match/mismatch) at an early stage, as indexed by the N1 component. During the second stage, participants became sensitive to low levels of misalignment, which were indexed by the feedback-related negativity. The degree of social misalignment was processed in greater detail, as indexed by the P3 component. Moreover, such hierarchical neural sensitivity is generalizable across different decision contexts (i.e. risky and ambiguous). These findings demonstrate a fine-grained neural sensitivity to social misalignment during decision-making under uncertainty.