Will existing forms of artificial intelligence (AI) lead to genuine intelligence? How is AI changing our society and politics? This essay examines the answers to these questions in Brian Cantwell Smith's The Promise of Artificial Intelligence and Mark Coeckelbergh's The Political Philosophy of AI with a focus on their central concern with judgment—whether AI can possess judgment and how developments in AI are affecting human judgment. First, I argue that the existentialist conception of judgment that Smith defends is highly idealized. While it may be an appropriate standard for intelligence, its implications for when and how AI should be deployed are not as clear as Smith suggests. Second, I point out that the concern with the displacement of judgment in favor of “reckoning” (or calculation) predates the rise of AI. The meaning and implications of this trend will become clearer if we move beyond ontology and metaphysics and into political philosophy, situating technological changes in their social context. Finally, I suggest that Coeckelbergh's distinctly political conception of judgment might offer a solution to the important boundary-drawing problem between tasks requiring judgment and those requiring reckoning, thus filling a gap in Smith's argument and clarifying its political stakes.

In deception research, little consideration is given to how the framing of the question might impact the decision-making process used to reach a veracity judgment. People use terms such as “sure” to describe their uncertainty about an event (i.e., aleatory) and terms such as “chance” to describe their uncertainty about the world (i.e., epistemic). Presently, the effect of such uncertainty framing on veracity judgments was considered. By manipulating the veracity question wording the effect of uncertainty framing on deception detection was measured. The data show no difference in veracity judgments between the two uncertainty framing conditions, suggesting that these may operate on a robust and invariant cognitive process.

Six experiments investigated how the distance between one’s initial opinion and advice relates to advice utilization. Going beyond previous research, we relate advice distance to both relative adjustments and absolute adjustments towards the advice, and we also investigate a second mode of advice utilization, namely confidence shifts due to social validation. Whereas previous research suggests that advice is weighted less the more it differs from one’s initial opinion, we consistently find evidence of a curvilinear pattern. Advice is weighted less when advice distance is low and when it is high. This is in particular because individuals are much more likely to retain their initial opinions in the light of near advice. Also, absolute opinion adjustments towards the advice increases in a monotone fashion as advice distance increases. This finding is in contrast to the predictions of the theoretical framework previous studies on advice distance are based on, social judgment theory. Instead, they data are more in line with a simple stimulus-response model suggesting that absolute adjustments towards the advice increase with advice distance but—potentially—with diminished sensitivity. Finally, our data show that advice can be utilized even when it receives zero weight during belief revision. The closer advice was to the initial opinions, the more it served as a means for social validation, increasing decision-makers’ confidence in the accuracy of their final opinions. Thus, our findings suggest that advice utilization is a more complex function of advice distance than previously assumed.

One of the core challenges of decision research is to identify individuals’ decision strategies without influencing decision behavior by the method used. Bröder and Schiffer (2003) suggested a method to classify decision strategies based on a maximum likelihood estimation, comparing the probability of individuals’ choices given the application of a certain strategy and a constant error rate. Although this method was shown to be unbiased and practically useful, it obviously does not allow differentiating between models that make the same predictions concerning choices but different predictions for the underlying process, which is often the case when comparing complex to simple models or when comparing intuitive and deliberate strategies. An extended method is suggested that additionally includes decision times and confidence judgments in a simultaneous Multiple-Measure Maximum Likelihood estimation. In simulations, it is shown that the method is unbiased and sensitive to differentiate between strategies if the effects on times and confidence are sufficiently large.

Two studies examined the relationship between individual differences in cognitive reflection (CRT) and the tendency to accord genuinely moral (non-conventional) status to a range of counter-normative acts — that is, to treat such acts as wrong regardless of existing social opinion or norms. We contrasted social violations that are intrinsically harmful to others (e.g., fraud, thievery) with those that are not (e.g., wearing pajamas to work and engaging in consensual acts of sexual intimacy with an adult sibling). Our key hypothesis was that more reflective (higher CRT) individuals would tend to moralize selectively — treating only intrinsically harmful acts as genuinely morally wrong — whereas less reflective (lower CRT) individuals would moralize more indiscriminately. We found clear support for this hypothesis in a large and ideologically diverse sample of American adults. The predicted associations were not fully accounted for by the subjects’ political orientation, sensitivity to gut feelings, gender, age, educational attainment, or their placement on a sexual morals-specific measure of social conservatism. Our studies are the first to demonstrate that, in addition to modulating the intensity of moral condemnation, reflection may also play a key role in setting the boundaries of the moral domain as such.

Using a cognitive task (mental calculation) and a perceptual-motor task (stylized golf putting), we examined differential proficiency using the CWS index and several other quantitative measures of performance. The CWS index (Weiss & Shanteau, 2003) is a coherence criterion that looks only at internal properties of the data without incorporating an external standard. In Experiment 1, college students (n = 20) carried out 2- and 3-digit addition and multiplication problems under time pressure. In Experiment 2, experienced golfers (n = 12), also college students, putted toward a target from nine different locations. Within each experiment, we analyzed the same responses using different methods. For the arithmetic tasks, accuracy information (mean absolute deviation from the correct answer, MAD) using a coherence criterion was available; for golf, accuracy information using a correspondence criterion (mean deviation from the target, also MAD) was available. We ranked the performances of the participants according to each measure, then compared the orders using Spearman’s rs. For mental calculation, the CWS order correlated moderately (rs =.46) with that of MAD. However, a different coherence criterion, degree of model fit, did not correlate with either CWS or accuracy. For putting, the ranking generated by CWS correlated .68 with that generated by MAD. Consensual answers were also available for both experiments, and the rankings they generated correlated highly with those of MAD. The coherence vs. correspondence distinction did not map well onto criteria for performance evaluation.

Risk and reward are negatively correlated in a wide variety of environments, and in many cases this trade off approximates a fair bet. Pleskac and Hertwig (2014) recently proposed that people have internalized this relationship and use it as the basis for probability estimation and subsequent choice under conditions of uncertainty. Specifically, they showed that risky options with high-value outcomes are inferred to have lower probability than options offering a less valuable reward. We report two experiments that test a simple corollary of this idea. In both studies, participants estimated the magnitude of prizes offered by lotteries with known win-probabilities. The relationship between estimates and probabilities followed the power relationship predicted by the risk-reward heuristic, albeit with a tendency to overestimate outcome magnitude. In addition, people’s estimates predicted their willingness to take the gamble. Our results provide further evidence that people have internalized the ecological relationship between risk and reward in financial lotteries, and we suggest that this relationship exerts a wide-ranging influence on decision-making.

The results of song contests offer a unique opportunity to analyze possible distortions arising from various biases in performance evaluations using observational data. In this study we investigate the influence of contestants’ order of appearance on their ranking. We found that, in the New Wave Song Contest, expert judgments were significantly influenced by the contestant’s running number, an exogenous factor that, being assigned randomly, clearly did not influence the output quality. We also found weaker statistical evidence of such an ordering effect in Eurovision Song Contest finals of 2009–2012.