选择性注意在人工语法学习中的作用——回应Eitam，Schul和Hassin(2009)中的两点疑问

从语法规则复杂性（复杂语法vs.简单语法）的角度考察选择性注意在人工语法学习中的必要性,并且比较两种非法序列下的成绩差异来检验被选择忽视的语法规则是否能被习得却未能在测验阶段体现出来。结果表明纵使降低被选择注意的语法的复杂性,被忽视的语法也未被成功习得,只有被选择注意的语法才能被习得;两种非法序列下的正确率无显著差异,即非法序列b所遵循的被忽略的语法未能在分类判断中起作用。选择性注意是语法规则被习得的关键。     

The influence of complexes on implicit learning

A century ago, Jung looked into the unconscious through complexes by using word association tests. Jung wrote, 'modern psychology with its investigation of complexes has opened up a psychic taboo area riddled with hopes and fears', and complexes remain an unexplored taboo area of research. In the present study, we have investigated the influence of complexes on unconscious cognitive processing, in particular on implicit learning. We have found that complexes shown to disturb conscious cognitive processing in fact enhanced the attention of the subjects and their performance on an implicit learning task. These results suggest that complexes are not just abstractions, but have various actual influences on both conscious and unconscious processing.     

Trust in hybrid human-automated decision-support

Research has examined trust in humans and trust in automated decision support. Although reflecting a likely realization of decision support in high-risk tasks such as personnel selection, trust in hybrid human-automation teams has thus far received limited attention. In two experiments (N₁ = 170, N₂ = 154) we compare trust, trustworthiness, and trusting behavior for different types of decision-support (automated, human, hybrid) across two assessment contexts (personnel selection, bonus payments). We additionally examined a possible trust violation by presenting one group of participants a preselection that included predominantly male candidates, thus reflecting possible unfair bias. Whereas fully-automated decisions were trusted less, results suggest that trust in hybrid decision support was similar to trust in human-only support. Trust violations were not perceived differently based on the type of support. We discuss theoretical (e.g., trust in hybrid support) and practical implications (e.g., keeping humans in the loop to prevent negative reactions).     

Enlarging the model of the human at the heart of human-centered AI: A social self-determination model of AI system impact

Increasing awareness of the harms that artificial intelligence (AI) systems can cause has inspired a movement towards creating more human-centered AI (HCAI). One way in which AI systems can be made more human-centered is by focusing on the effects they have on people's needs. However, existing theories of technology impact in HCAI drawn from human-computer interaction (HCI) and related fields such as psychology are not able to account for the ways in which both needs and the impact of technology on these are dynamically shaped by the social context. To address this limitation, in this paper we outline a Social Self-Determination Model (SSDM) of AI system impact. SSDM proposes that people's needs for self-determination can be individual or collective in a particular context depending on features of the person and their environment. Accordingly, because AI systems treat people differently depending on the groups they belong to, they can make needs for individual or collective self-determination psychologically relevant and either fulfill or hamper these. SSDM can be used to help designers and developers create more human-centered AI systems by quantifying the effects of these systems and the social environment in which they are embedded on people's fundamental psychological needs and wellbeing.     

Robustness of the rule-learning effect in 7-month-old infants: A close,multicenter replication of Marcus et al. (1999)

We conducted a close replication of the seminal work by Marcus and colleagues from 1999, which showed that after a brief auditory exposure phase, 7-month-old infants were able to learn and generalize a rule to novel syllables not previously present in the exposure phase. This work became the foundation for the theoretical framework by which we assume that infants are able to learn abstract representations and generalize linguistic rules. While some extensions on the original work have shown evidence of rule learning, the outcomes are mixed, and an exact replication of Marcus et al.'s study has thus far not been reported. A recent meta-analysis by Rabagliati and colleagues brings to light that the rule-learning effect depends on stimulus type (e.g., meaningfulness, speech vs. nonspeech) and is not as robust as often assumed. In light of the theoretical importance of the issue at stake, it is appropriate and necessary to assess the replicability and robustness of Marcus et al.'s findings. Here we have undertaken a replication across four labs with a large sample of 7-month-old infants (N = 96), using the same exposure patterns (ABA and ABB), methodology (Headturn Preference Paradigm), and original stimuli. As in the original study, we tested the hypothesis that infants are able to learn abstract “algebraic” rules and apply them to novel input. Our results did not replicate the original findings: infants showed no difference in looking time between test patterns consistent or inconsistent with the familiarization pattern they were exposed to.     

Philosophers ought to develop,theorize about,and use philosophically relevant AI

The transformative power of artificial intelligence (AI) is coming to philosophy—the only question is the degree to which philosophers will harness it. This paper argues that the application of AI tools to philosophy could have an impact on the field comparable to the advent of writing, and that it is likely that philosophical progress will significantly increase as a consequence of AI. The role of philosophers in this story is not merely to use AI but also to help develop it and theorize about it. In fact, the paper argues that philosophers have a prima facie obligation to spend significant effort in doing so, at least insofar as they should spend effort philosophizing.     

Künstliche Intelligenz und Philosophie

Artificial Intelligence can be considered as the so far last attempt to decode the anthropological comparison between human beings and machines. Thereby it also represents in a prominent way what can be called systemic thought. Searle's conclusive argument against strong AI (that is the idea of computers having intention in a literal way) refers to his precise distinction between syntax and semantics. This difference obviously opposing some of Searle's other essential ideas will only convince if it also explains the genetic-pragmatic aspect. A theory explaining the life of mind and the possibility of understanding needs to combine representation and intention with the subjective causation of signs. At the same time they have to be contextualized within a model of teleologically interpreted life recognized with the help of self-experience and self-reflection. This suggests that AI is a simulation which wrongly believes to be a real duplication. Actually it is a semiotic reduction (syntax and semantic surface of signs only) and a psychological compensation (Turing test) connected with a genetic or abductive fallacy. The biological decontextualization, in fact the elimination of teleology and intention, the deconstruction of subjectivity, the loss of the genetic-pragmatic dimension and the abductive fallacy induce the strong AI to confuse its surface-illusion of simulated understanding with the real process itself.     

ARTIFICIAL INTELLIGENCE AND DE LAS CASAS: A 1492 RESONANCE

Abstract. A comparison is made between two unlikely debates over intelligence. One debate took place in 1550 at Valladolid, Spain, between Bartolomé de las Casas and Juan Gines de Sepúlveda over the intelligence of the Amerindian. The other debate is contemporary, between John Searle and various representatives of the "strong" artificial intelligence (AI) community over the adequacy of the Turing test for intelligence. Although the contemporary debate has yet to die down, the Valladolid debate has been over for four hundred years. The question asked here is whether the contemporary debate can profit from the previous one. The common bond providing the basis for contrast is the issue of the "other" which is present in both debates. From this contrast, the observation is made that the question of meaning is intimately tied to the question of intelligence.