Cognitive science faces ongoing challenges in research integration, formalization, conceptual clarity, and other areas, partly due to its multifaceted and interdisciplinary nature. Recent advances in artificial intelligence, particularly the development of language models, offer tools that may help address these longstanding issues. Specifically, they can help map fragmented literatures, formalize verbal theories, identify overlap among constructs and measures, generate predictions across tasks, and extract cultural or ecological structure from naturalistic data. However, these opportunities come with risks, including oversimplification, opacity, deskilling, and bias. Taken together, we conclude that language models could serve as tools for a more integrative and cumulative cognitive science when used judiciously to complement, rather than replace, human agency.

In recent years, several claims have been put forth suggesting that animals of different species address each other using receiver-specific calls that function as individual names. If accurate, this would demonstrate a capacity for symbolic social reference that has rarely been documented in nonhuman communication systems. To evaluate such claims, we propose three minimal criteria for identifying names: individual reference, symbolic representation, and shared meaning. We then review the evidence relevant to these criteria in the species for which name-like calls have been suggested: spectacled parrotlets, common marmosets, African elephants, and bottlenose dolphins. Together, these criteria provide a clear and theoretically grounded framework for exploring the use of name-like calls in nonhuman animals.

People change across daily situations and life transformations, yet maintain a stable sense of self. Explaining this stability requires understanding how identities, traits, values, and memories interact-what changes together, what resists change, and how. We propose that the self is represented as a dynamic causal network of self-aspects linked by beliefs about how they influence one another. Coherence arises when activated self-aspects align with internal beliefs and contextual demands. When disrupted, the network restores coherence through activation shifts that preserve structure or through structural reorganization that reshapes it. This perspective distinguishes the self from associative, categorical, or hierarchical accounts by specifying directional dependencies that constrain change, and highlights how modifying connections among self-aspects promotes coherence and well-being.

Music is pervasive across human history, cultures, and development, yet its evolutionary origins remain unresolved due to the interplay of cultural and biological influences. We argue that identifying uniquely human musicality requires cross-primate comparative research that integrates rhythm and melody processing with exposure, computational principles, and intrinsic reward.

Brain-computer interfaces (BCIs) have achieved transformative success in restoring movement and communication. However, extending these approaches to decoding or recovery of cognitive function, such as attention or memory, poses fundamentally new challenges. Cognitive BCIs will need to contend with distributed and dynamic neural processes that differ sharply from the more localized, stable representations underlying motor and language control, imposing new technical and conceptual demands. Conversely, neuromodulation, long used in neurological and psychiatric therapies, offers a complementary methodological path and initial translational applications through causal modulation of cognitive circuits. Integrating these approaches into adaptive, closed-loop systems could allow cognitive BCIs to restore mental function and bridge systems neuroscience and next-generation neurotherapeutics capable of monitoring and shaping human cognition in real time.

Over the past decade, comparisons between deep neural networks (DNNs) and the human brain have become central to cognitive neuroscience. Early work focused on vision, driven by the success of convolutional neural networks in object recognition, before such comparisons later gained traction in language with the rise of large-scale language models. These comparisons have validated existing hypotheses and generated new ones, challenging views of information processing, connectivity, and computational goals. Despite progress, debates persist over the interpretability and validity of mapping DNNs to brains, underscoring the need for more refined models and methods. Looking ahead, integrating cross-modal insights from vision and language, together with improved modeling and experimental frameworks, promises to advance the mechanistic understanding of cognition.

Many worry that AI use will erode human cognition. We discuss evidence suggesting that offloading cognition to AI can impede skill acquisition and lead to skill decay, but risks depend on how AI is used. We also consider whether our basic cognitive abilities might prove more resilient to this erosion.

Why do some people remain mentally healthy under stress, while others develop symptoms? Recent observational and interventional evidence indicates that positive appraisal style-the tendency to interpret threats in a mildly positive fashion-is a key resilience factor and a promising target for preventive interventions.

Adolescence is a period of vulnerability for internalizing mental health disorders. One underappreciated mechanism of internalizing psychopathology is higher-order social cognition (i.e., cognitive processing of information about the self and others), which is highly salient during adolescence. We propose that the protracted development of social cognition during adolescence results in highly individualized, multidimensional profiles that can be characterized across four core dimensions (Propensity, Degree, Positivity, and Negativity). A social cognitive profile with greater Propensity, Degree, and Negativity, and decreased Positivity, along with associated decreases in Accuracy, may confer greater risk for internalizing psychopathology, particularly if applied inflexibly across contexts. Future research may test the utility of these profiles for stratification, prediction, or intervention targeting.

Noninvasive methods are making it possible to study neural replay in humans. Although focused on alignment with rodent findings, human research often tacitly assumes that replay is linked to conscious, deliberate thought. This assumption represents a missed opportunity to investigate the links between replay and consciousness, as well as cross-species differences more broadly.

Reciprocity theory holds a hallowed status in evolutionary social sciences because it provides a way to explain cooperation among unrelated individuals. Theorists have developed a large body of mathematical work based on the iterated prisoner's dilemma game that explores which strategies are evolutionarily stable in a range of conditions, including scenarios involving behavioral errors. However, there is a mismatch between the vigorous pace of theoretical research and the trickle of empirical studies that have tested predictions generated by the theoretical work. Furthermore, findings from the empirical work on dyadic cooperation in humans diverge from theoretical predictions in significant ways. This discrepancy indicates that the standard theoretical framework is poorly equipped to account for key aspects of real-world cooperation. To better align with real-world observations, theoretical explorations of reciprocity need to shift their focus toward understanding how actors choose partners and maintain relationships.