Charles Darwin's On the Origins of Species had two principal goals: to show that species had not been separately created and to show that natural selection had been the main force behind their proliferation and descent from common ancestors. In Coevolution, the author proposes a powerful new theory of cultural evolution--that is, of the descent with modification of the shared conceptual systems we call cultures--that is parallel in many ways to Darwin's theory of organic evolution. The author suggests that a process of cultural selection, or preservation by preference, driven chiefly by choice or imposition depending on the circumstances, has been the main but not exclusive force of cultural change. He shows that this process gives rise to five major patterns or modes in which cultural change is at odds with genetic change. Each of the five modes is discussed in some detail and its existence confirmed through one or more case studies chosen for their heuristic value, the robustness of their data, and their broader implications. But Coevolution predicts not simply the existence of the five modes of gene-culture relations; it also predicts their relative importance in the ongoing dynamics of cultural change in particular cases. The case studies themselves are lucid and innovative reexaminations of an array of oft-pondered anthropological topics--plural marriage, sickle-cell anemia, basic color terms, adult lactose absorption, incest taboos, headhunting, and cannibalism. In a general case, the author's goal is to demonstrate that an evolutionary analysis of both genes and culture has much to contribute to our understanding of human diversity, particularly behavioraldiversity, and thus to the resolution of age-old questions about nature and nurture, genes and culture.
Enterprise API design is often bottlenecked by the tension between rapid feature delivery and the rigorous maintenance of usability standards. We present an industrial case study evaluating an AI-assisted design workflow trained on API Improvement Proposals (AIPs). Through a controlled study with 16 industry experts, we compared AI-generated API specifications against human-authored ones. While quantitative results indicated AI superiority in 10 of 11 usability dimensions and an 87% reduction in authoring time, qualitative analysis revealed a paradox: experts frequently misidentified AI work as human (19% accuracy) yet described the designs as unsettlingly "perfect." We characterize this as a "Perfection Paradox" -- where hyper-consistency signals a lack of pragmatic human judgment. We discuss the implications of this perfection paradox, proposing a shift in the human designer's role from the "drafter" of specifications to the "curator" of AI-generated patterns.
The study analyzes the evolution of digital maturity in the European Union between 2018 and
2024, with the aim of assessing the level of readiness for the transition to an advanced digital
leadership model. Based on Eurostat data integrated into the DESI and Digital Decade frameworks,
indicators on digital skills, ICT specialization, internet use, continuing education, and the
digitization of SMEs are examined. The research uses an exploratory quantitative approach,
principal component analysis (PCA), to identify the latent structures of digital transformation. The
results highlight two major dimensions of digital maturity and three evolutionary phases,
culminating in a convergent acceleration in 2023–2024. The study confirms that European digital
transformation is becoming a systemic phenomenon, supported by human capital and technological
infrastructure, but also shows that ICT training at company level remains underutilised.
Recent advances in large language models (LLMs) have sparked growing interest in building fully autonomous agents. However, fully autonomous LLM-based agents still face significant challenges, including limited reliability due to hallucinations, difficulty in handling complex tasks, and substantial safety and ethical risks, all of which limit their feasibility and trustworthiness in real-world applications. To overcome these limitations, LLM-based human-agent systems (LLM-HAS) incorporate human-provided information, feedback, or control into the agent system to enhance system performance, reliability and safety. These human-agent collaboration systems enable humans and LLM-based agents to collaborate effectively by leveraging their complementary strengths. This paper provides the first comprehensive and structured survey of LLM-HAS. It clarifies fundamental concepts, systematically presents core components shaping these systems, including environment & profiling, human feedback, interaction types, orchestration and communication, explores emerging applications, and discusses unique challenges and opportunities arising from human-AI collaboration. By consolidating current knowledge and offering a structured overview, we aim to foster further research and innovation in this rapidly evolving interdisciplinary field. Paper lists and resources are available at https://github.com/HenryPengZou/Awesome-Human-Agent-Collaboration-Interaction-Systems.
Chitralekha Gupta, Hanjun Wu, Praveen Sasikumar
et al.
Wearable devices are transforming human capabilities by seamlessly augmenting cognitive functions. In this position paper, we propose a voice-based, interactive learning companion designed to amplify and extend cognitive abilities through informal learning. Our vision is threefold: (1) to enable users to discover new knowledge on-the-go through contextual interactive quizzes, fostering critical thinking and mindfulness, (2) to proactively detect misinformation, empowering users to critically assess information in real time, and (3) to provide spoken language correction and prompting hints for second language learning and effective communication. As an initial step toward this vision, we present Factually - a proactive, wearable fact-checking system integrated into devices like smartwatches or rings. Factually discreetly alerts users to potential falsehoods via vibrotactile feedback, helping them assess information critically. We demonstrate its utility through three illustrative scenarios, highlighting its potential to extend cognitive abilities for real-time misinformation detection. Early qualitative feedback suggests that Factually can enhance users' fact-checking capabilities, offering both practical and experiential benefits.
José C. Segovia, Juan C. Ramírez, Paula Río
et al.
ABSTRACT Two strains of the minute virus of mice (MVM) parvovirus (of the species Protoparvovirus rodent1), named i (immunosuppressive) and p (prototype), show disparate tropism and pathogenicity in mice. Unlike the virulent MVMi, the MVMp is attenuated and infection-restricted in primary mouse hematopoietic cells; however, it does infect human transformed, established, and stem cells. We thus seek here a putative MVMp natural tropism to humans compared to MVMi. Infectious purified MVMp virions suppress in vitro the clonogenic potential of human committed (erythroid and myeloid) and primitive CD34+ progenitors, regardless of the cell source (umbilical cord blood, G-CSF mobilized peripheral blood, or bone marrow), in a multiplicity of infection-dependent manner. Furthermore, primitive CD34+ progenitors in culture are permissive to the MVMp cytotoxic NS-1 protein expression and the synthesis of viral genome replicative intermediates. None of these features is shared by MVMi. In consistency, MVMp suppresses the pool of CD34+ progenitors in bone-marrow-humanized immunodeficient NSG mice and severely compromises their survival, whereas in basal NSG mice, it replicates to several thousand-fold lower levels than MVMi and undergoes evolution to variants carrying genetic changes at the capsid tropism determinant binding sialic acid types. We suggest that MVMp may have originated from a MVMi-like mouse parvovirus upon adaptation to human sialic acid receptors in immunocompromised individuals. Extensive serological and genetic screens failed to demonstrate circulating MVMp in Eurasian and African human populations, including hematopoietic patients. Nevertheless, the high MVMp cytotoxicity to human primitive hematopoietic progenitors and rapid evolvability deserve further epidemiological studies.IMPORTANCEAssigning the genuine natural viral host range may be uncertain without testing the susceptibility of primary cells. Parvoviridae, a family of ssDNA icosahedral viruses, harbors several members infecting humans, but only parvovirus B19 of the Erythroparvovirus genus shows tropism for human hematopoietic progenitors. Here, we evaluate the tropism of the MVMi and MVMp parvoviruses for humans, which are assumed to be mouse pathogens. We show that MVMp is remarkably cytotoxic for human primitive CD34+ and committed erythroid and myeloid hematopoietic progenitors. Infection of basal and bone marrow-humanized immunodeficient mice shows the emergence of viral strain-specific genetic changes at the capsid domain binding sialic-acid receptors, denoting adaptation to the respective host hematopoiesis. Although a large epidemiological survey failed to identify circulating MVMp sequences or antibodies in human populations, its high toxicity to human hematopoietic progenitors of different lineages and primitiveness and rapid evolutionary capacity demand in-depth characterization of its potential pathogenicity for humans.
IntroductionAge-related hearing loss (ARHL) is a common chronic condition that significantly affects the quality of life in older adults. Studies have shown that genetic factors play a substantial role in ARHL, with heritability estimates ranging from 46 to 74%. Although advances in genomics and epigenetics have led to the identification of numerous candidate genes in recent years, most related studies have focused on European and North American populations. There remains a lack of systematic mapping of research trends and cross-ethnic gene consistency, limiting the broad applicability of these findings.MethodThis study screened English-language publications on ARHL genetics from 1995 to June 2025 across PubMed, Embase, Web of Science, and Scopus, ultimately including 465 studies. Bibliometric analyses were conducted using R Bibliometrix, VOSviewer, and CiteSpace to extract research trends, research hotspots, and candidate genes. Ethnic information from human studies were compiled to facilitate cross-ethnic comparative analysis.ResultOver the past 30 years, publications in this field have shown continuous growth, with an average annual growth rate of 6.83%. Hearing Research emerged as the core journal. China and the United States were the top two publishing countries, though international collaboration remained limited. Research priorities have gradually shifted from inner ear anatomy to molecular mechanisms such as gene variants, oxidative stress, mitochondrial function, and inflammation. A total of 365 candidate factors from animal studies and 221 candidate genes from human studies were extracted and grouped into seven categories. Cross-ethnic analysis identified 56 genes that were repeatedly reported across at least two populations. Among these, CDH23, ILDR1, and SLC26A5 showed high cross-ethnic consistency, while genes such as GRHL2 exhibited notable ethnic specificity.ConclusionThis study systematically maps the developmental trajectory and research hotspots of ARHL genetics, revealing key patterns in geographic distribution, thematic evolution, and cross-ethnic applicability. The findings highlight the urgent need to strengthen research in non-European populations and promote international collaboration, thereby providing a theoretical foundation and data support for building a universally applicable genetic risk framework and advancing individualised interventions.
The Xiliu Songhua River Basin in northeast China was selected as the research object to quantitatively analyze the driving mechanism of climate change and human activities on runoff evolution under the background of global warming and reveal the evolution law of water cycle in cold regions and its eco-hydrological effects. A precipitation phase discrimination model was constructed based on the wet-bulb temperature threshold method, and the Budyko equation was improved by introducing the snowfall ratio. Combined with multi-source methods such as spatial interpolation of meteorological data, Pettitt mutation test, Mann-Kendall trend test, and land use transfer matrix, the synergistic effect of climate factors and underlying surface parameters was systematically evaluated. The annual runoff at Fuyu Station and Jilin Station increased significantly at a rate of 0.12 mm/a and 1.77 mm/a, respectively. The annual average precipitation and annual average snowfall ratio in the basin showed a continuous increasing trend, while the annual average potential evapotranspiration showed a continuous decreasing trend. In terms of spatial distribution, the annual average precipitation and annual average snowfall ratio in the basin showed a trend of less in the north and more in the south, while the annual average potential evapotranspiration showed a pattern of more in the north and less in the south. The contribution rate of precipitation humidification to runoff change reached 57.24%~66.89%, and the contribution rate of human activities was only 4.39%~7.52%. The proposed SR-Budyko coupling framework effectively solved the defects of traditional models in rain and snow phase identification and freeze-thaw energy and water exchange characterization and provided a new theoretical tool for the optimization of water resources systems in cold regions, spring flood disaster prevention and control, and the protection of ecological barriers in northeast China. It has important practical guiding value for water security decision-making in climate change-sensitive areas.
River, lake, and water-supply engineering (General)
Amid the deep transformation of traditional industries driven by the digital economy, firms face a critical challenge in restructuring their knowledge systems to achieve innovation breakthroughs. This study investigates the core question of whether and how digital-physical integration affects corporate knowledge breadth. Using panel data of Chinese listed firms, we construct firm-level indicators for digital-physical integration and knowledge breadth. The analysis is conducted through two-way fixed effects models, instrumental variable methods, and propensity score matching. Furthermore, we examine three mediating mechanisms—dynamic capability, asset novelty, and total factor productivity to uncover the pathways through which integration exerts its influence. The results show that digital-physical integration significantly promotes the expansion of corporate knowledge breadth. This finding remains robust across various endogeneity controls and sensitivity checks. Mechanism analysis reveals that integration indirectly drives the external restructuring of knowledge structures by enhancing firms’ dynamic capabilities, optimizing resource allocation patterns, and improving productivity. Heterogeneity analysis further indicates that the effect is more pronounced among non-technology-intensive firms, non-state-owned enterprises, firms located in regions with higher human capital, and those in areas with weaker intellectual property protection. By shifting the analytical focus to knowledge organization, this study deepens the understanding of the consequences of digital transformation and extends the theoretical boundaries of corporate knowledge evolution and technological integration. The findings also offer practical insights for policymakers designing differentiated digital strategies and for firms aiming to build effective knowledge coordination systems.
Morten Roed Frederiksen, Katrin Fischer, Maja Matarić
This paper describes a between-subjects Amazon Mechanical Turk study (n = 220) that investigated how a robot's affective narrative influences its ability to elicit empathy in human observers. We first conducted a pilot study to develop and validate the robot's affective narratives. Then, in the full study, the robot used one of three different affective narrative strategies (funny, sad, neutral) while becoming less functional at its shopping task over the course of the interaction. As the functionality of the robot degraded, participants were repeatedly asked if they were willing to help the robot. The results showed that conveying a sad narrative significantly influenced the participants' willingness to help the robot throughout the interaction and determined whether participants felt empathetic toward the robot throughout the interaction. Furthermore, a higher amount of past experience with robots also increased the participants' willingness to help the robot. This work suggests that affective narratives can be useful in short-term interactions that benefit from emotional connections between humans and robots.
Mohamad Shaaban, Simone Macci`o, Alessandro Carf`ı
et al.
This article investigates mixed reality (MR) to enhance human-robot collaboration (HRC). The proposed solution adopts MR as a communication layer to convey a mobile manipulator's intentions and upcoming actions to the humans with whom it interacts, thus improving their collaboration. A user study involving 20 participants demonstrated the effectiveness of this MR-focused approach in facilitating collaborative tasks, with a positive effect on overall collaboration performances and human satisfaction.
Computational modeling plays an essential role in the study of language emergence. It aims to simulate the conditions and learning processes that could trigger the emergence of a structured language within a simulated controlled environment. Several methods have been used to investigate the origin of our language, including agent-based systems, Bayesian agents, genetic algorithms, and rule-based systems. This chapter explores another class of computational models that have recently revolutionized the field of machine learning: deep learning models. The chapter introduces the basic concepts of deep and reinforcement learning methods and summarizes their helpfulness for simulating language emergence. It also discusses the key findings, limitations, and recent attempts to build realistic simulations. This chapter targets linguists and cognitive scientists seeking an introduction to deep learning as a tool to investigate language evolution.
With the rise of human-machine communication, machines are increasingly designed with humanlike characteristics, such as gender, which can inadvertently trigger cognitive biases. Many conversational agents (CAs), such as voice assistants and chatbots, default to female personas, leading to concerns about perpetuating gender stereotypes and inequality. Critiques have emerged regarding the potential objectification of females and reinforcement of gender stereotypes by these technologies. This research, situated in conversational AI design, aims to delve deeper into the impacts of gender biases in human-CA interactions. From a behavioral and communication research standpoint, this program focuses not only on perceptions but also the linguistic styles of users when interacting with CAs, as previous research has rarely explored. It aims to understand how pre-existing gender biases might be triggered by CAs' gender designs. It further investigates how CAs' gender designs may reinforce gender biases and extend them to human-human communication. The findings aim to inform ethical design of conversational agents, addressing whether gender assignment in CAs is appropriate and how to promote gender equality in design.
My research centers on the development of context-adaptive AI systems to improve end-user adoption through the integration of technical methods. I deploy these AI systems across various interaction modalities, including user interfaces and embodied agents like robots, to expand their practical applicability. My research unfolds in three key stages: design, development, and deployment. In the design phase, user-centered approaches were used to understand user experiences with AI systems and create design tools for user participation in crafting AI explanations. In the ongoing development stage, a safety-guaranteed AI system for a robot agent was created to automatically provide adaptive solutions and explanations for unforeseen scenarios. The next steps will involve the implementation and evaluation of context-adaptive AI systems in various interaction forms. I seek to prioritize human needs in technology development, creating AI systems that tangibly benefit end-users in real-world applications and enhance interaction experiences.
A. V. Lukina-Gronskaya, I. K. Chudinov, I. K. Chudinov
et al.
IntroductionSmall mammals, especially rodents and bats, are known reservoirs of zoonotic viruses, but little is known about the viromes of insectivorous species including hedgehogs (order Eulipotyphla), which often live near human settlements and come into contact with humans.MethodsWe used high-throughput sequencing and metaviromic analysis to describe the viromes of 21 hedgehogs (Erinaceus sp.) sampled from summer 2022 to spring 2023. We captured 14 active animals from the wild (seven in European Russia and the other seven in Central Siberia). The remaining 7 animals were hibernating in captivity (captured in European Russia before the experiment).Results and discussionThe diversity of identified viral taxa as well as the total number of reads classified as viral was high in all active animals (up to eight different viral families per animal), but significantly lower in hibernating animals (zero or no more than three different viral families per animal). The present study reports, for the first time, betacoronaviruses and mammasrenaviruses in hedgehogs from Russia. Erinaceus coronaviruses (EriCoVs) were found in 4 of 7 active animals captured in the wild, in European Russia, making it is the easiest finding of EriCoVs in Europe. One animal was found to carry of two different EriCoVs. Both strains belong to the same phylogenetic clade as other coronaviruses from European hedgehogs. Pairwise comparative analysis suggested that one of these two strains arose by recombination with an unknown coronavirus, since all of identified SNPs (n = 288) were found only in the local genome region (the part of ORF1b and S gene). The novel mammarenaviruses (EriAreVs) were detected in 2 out of 7 active and in 2 out of 7 hibernating animals from the European Russia. Several complete L and S segments of EriAreVs were assembled. All identified EriAreVs belonged to the same clade as the recently described MEMV virus from Hungarian hedgehogs. As the hibernating hedgehogs were positive for EriAreVs when kept in controlled conditions without contact with each other, we suggest the possibility of persistent arenavirus infection in hedgehogs, but further experiments are needed to prove this.
Abstract Background The high transmissibility of the Omicron variants of severe acute respiratory syndrome coronavirus 2 continues to impose a significant burden on public health systems worldwide. Continuous mutations in the virus challenge the efficacy of vaccines and prior immunity from other variants, posing a severe threat to human health. Therefore, there is an urgent need to predict mutations of the Omicron variants and identify key factors influencing their spread. This study investigated critical amino acid mutations of the Omicron variants using epidemiological data and the mechanisms underlying their transmission. The aim was to understand the key factors driving the spread of the Omicron variants and provide insights for effective control and prevention strategies. Methods A total of 488,646 Omicron cases recorded between December 2021 and February 2023 were analyzed using a sliding time window and the Epi Score model, which has high accuracy for the identification of mutation sites. MutPred2, PolyPhen2 and VarSite tools were used to predict future mutations and identify factors driving the prevalence of the Omicron variants. Results Epi Scores showed fluctuating patterns at mutation sites, highlighting N969K, Y505H, N764K, T478K, and S371F mutations on the spike protein as significant for future prevention efforts. The spread of the Omicron variants was linked to changes in the viral entry pathway and improved angiotensin‐converting enzyme 2 (ACE2) binding and immune evasion tactics. Conclusions The findings of this study reveal trends in the evolution of the Omicron variants, including altered cell entry, increased affinity for ACE2, and evasion of the immune system. These factors are critical for understanding the global spread of the Omicron variants and developing effective control strategies.
The causative agent of human tuberculosis (TB), Mycobacterium tuberculosis, is an obligate pathogen that evolved to exclusively persist in human populations. For M. tuberculosis to transmit from person to person, it has to cause pulmonary disease. Therefore, M. tuberculosis virulence has likely been a significant determinant of the association between M. tuberculosis and humans. Indeed, the evolutionary success of some M. tuberculosis genotypes seems at least partially attributable to their increased virulence. The latter possibly evolved as a consequence of human demographic expansions. If co‐evolution occurred, humans would have counteracted to minimize the deleterious effects of M. tuberculosis virulence. The fact that human resistance to infection has a strong genetic basis is a likely consequence of such a counter‐response. The genetic architecture underlying human resistance to M. tuberculosis remains largely elusive. However, interactions between human genetic polymorphisms and M. tuberculosis genotypes have been reported. Such interactions are consistent with local adaptation and allow for a better understanding of protective immunity in TB. Future ‘genome‐to‐genome’ studies, in which locally associated human and M. tuberculosis genotypes are interrogated in conjunction, will help identify new protective antigens for the development of better TB vaccines.
Predicting human mobility holds significant practical value, with applications ranging from enhancing disaster risk planning to simulating epidemic spread. In this paper, we present the GeoFormer, a decoder-only transformer model adapted from the GPT architecture to forecast human mobility. Our proposed model is rigorously tested in the context of the HuMob Challenge 2023 -- a competition designed to evaluate the performance of prediction models on standardized datasets to predict human mobility. The challenge leverages two datasets encompassing urban-scale data of 25,000 and 100,000 individuals over a longitudinal period of 75 days. GeoFormer stands out as a top performer in the competition, securing a place in the top-3 ranking. Its success is underscored by performing well on both performance metrics chosen for the competition -- the GEO-BLEU and the Dynamic Time Warping (DTW) measures. The performance of the GeoFormer on the HuMob Challenge 2023 underscores its potential to make substantial contributions to the field of human mobility prediction, with far-reaching implications for disaster preparedness, epidemic control, and beyond.