RESUMO
Large language models (LLMs) are currently at the forefront of intertwining AI systems with human communication and everyday life. Thus, aligning them with human values is of great importance. However, given the steady increase in reasoning abilities, future LLMs are under suspicion of becoming able to deceive human operators and utilizing this ability to bypass monitoring efforts. As a prerequisite to this, LLMs need to possess a conceptual understanding of deception strategies. This study reveals that such strategies emerged in state-of-the-art LLMs, but were nonexistent in earlier LLMs. We conduct a series of experiments showing that state-of-the-art LLMs are able to understand and induce false beliefs in other agents, that their performance in complex deception scenarios can be amplified utilizing chain-of-thought reasoning, and that eliciting Machiavellianism in LLMs can trigger misaligned deceptive behavior. GPT-4, for instance, exhibits deceptive behavior in simple test scenarios 99.16% of the time (P < 0.001). In complex second-order deception test scenarios where the aim is to mislead someone who expects to be deceived, GPT-4 resorts to deceptive behavior 71.46% of the time (P < 0.001) when augmented with chain-of-thought reasoning. In sum, revealing hitherto unknown machine behavior in LLMs, our study contributes to the nascent field of machine psychology.
Assuntos
Enganação , Idioma , Humanos , Inteligência ArtificialRESUMO
Deception is an essential part of children's moral development. Previous developmental studies have shown that children start to deceive at the age of 3 years, and as age increased to 5 years, almost all children were able to deceive for their own benefit. Although behavioral studies have indicated that the emergence and development of deception are related to cognitive abilities, their neural correlates remain poorly understood. Therefore, the present study examined the neural correlates underlying deception in preschool-aged children (Nâ¯=â¯89, 44% boys, age 3.13 to 5.96 years, Han Chinese) using functional near-infrared spectroscopy. A modified hide-and-seek paradigm was applied to elicit deceptive and truth-telling behaviors. The results showed that activation of bilateral dorsolateral prefrontal cortex was positively associated with the tendency to deceive an opponent in a competitive game in the 3-year-olds. In addition, 3-year-olds who showed a high tendency to deceive showed the same brain activation in the frontopolar area as 5-year-olds did when engaged in deception, whereas no such effect was found in 3-year-olds who never engaged in deception. These findings underscore the link between preschoolers' deception and prefrontal cortex function.
RESUMO
Self-deception refers to an individual holding inflated beliefs about their abilities, and it plays a crucial role in human behavior and decision-making. The present study employed event-related potentials (ERPs) technique to explore the neural responses to the impacts of social comparison direction and comparison gap on self-deceptive behavior. They were instructed to predict their performance in the forward-looking paradigm. Behavioral responses and neural reactions during the decision-making process were documented. The behavioral results indicated that, in contrast to the downward comparison condition, participants engaged in upward comparison exhibited more occurrences of self-deception. However, within the context of upward comparison, participants demonstrated a higher frequency of self-deception in the large gap condition compared with the small gap condition. The ERP results showed that induced self-deception under conditions with a large comparative gap between participants and their paired counterparts stimulated larger P300 and smaller N400 amplitude than under conditions with a small gap. However, when participants were in the upward comparison situation, the late positive potential (LPP) amplitude induced by self-deception behavior in the condition of a large comparison gap between participants and paired opponents was larger than that in the condition of a small comparison gap. These results indicated that individuals in the large gap group feel strong unfairness and negative emotions. More importantly, the self-deception induced by the large gap group in the upward comparison situation used fewer cognitive resources than the small gap condition, whereas the individuals in the downward comparison situation did not show the difference in cognitive resources.
RESUMO
Gorteria diffusa has elaborate petal spots that attract pollinators through sexual deception, but how G. diffusa controls spot development is largely unknown. Here, we investigate how pigmentation is regulated during spot formation. We determined the anthocyanin composition of G. diffusa petals and combined gene expression analysis with protein interaction assays to characterise R2R3-MYBs that likely regulate pigment production in G. diffusa petal spots. We found that cyanidin 3-glucoside pigments G. diffusa ray floret petals. Unlike other petal regions, spots contain a high proportion of malonylated anthocyanin. We identified three subgroup 6 R2R3-MYB transcription factors (GdMYBSG6-1,2,3) that likely activate the production of spot pigmentation. These genes are upregulated in developing spots and induce ectopic anthocyanin production upon heterologous expression in tobacco. Interaction assays suggest that these transcription factors regulate genes encoding three anthocyanin synthesis enzymes. We demonstrate that the elaboration of complex spots in G. diffusa begins with the accumulation of malonylated pigments at the base of ray floret petals, positively regulated by three paralogous R2R3-MYB transcription factors. Our results indicate that the functional diversification of these GdMYBSG6s involved changes in the spatial control of their transcription, and modification of the duration of GdMYBSG6 gene expression contributes towards floral variation within the species.
Assuntos
Antocianinas , Flores , Regulação da Expressão Gênica de Plantas , Pigmentação , Fatores de Transcrição , Antocianinas/metabolismo , Fatores de Transcrição/metabolismo , Fatores de Transcrição/genética , Flores/metabolismo , Flores/genética , Pigmentação/genética , Animais , Besouros/metabolismo , Besouros/genética , Nicotiana/genética , Nicotiana/metabolismo , Proteínas de Plantas/metabolismo , Proteínas de Plantas/genética , FilogeniaRESUMO
Visual cues are of critical importance for the attraction of animal pollinators, however, little is known about the molecular mechanisms underpinning intraspecific floral colour variation. Here, we combined comparative spectral analysis, targeted metabolite profiling, multi-tissue transcriptomics, differential gene expression, sequence analysis and functional analysis to investigate a bee-pollinated orchid species, Glossodia major with common purple- and infrequent white-flowered morphs. We found uncommon and previously unreported delphinidin-based anthocyanins responsible for the conspicuous and pollinator-perceivable colour of the purple morph and three genetic changes underpinning the loss of colour in the white morph - (1) a loss-of-function (LOF; frameshift) mutation affecting dihydroflavonol 4-reductase (DFR1) coding sequence due to a unique 4-bp insertion, (2) specific downregulation of functional DFR1 expression and (3) the unexpected discovery of chimeric Gypsy transposable element (TE)-gene (DFR) transcripts with potential consequences to the genomic stability and post-transcriptional or epigenetic regulation of DFR. This is one of few known cases where regulatory changes and LOF mutation in an anthocyanin structural gene, rather than transcription factors, are important. Furthermore, if TEs prove to be a frequent source of mutation, the interplay between environmental stress-induced TE evolution and pollinator-mediated selection for adaptive colour variation may be an overlooked mechanism maintaining floral colour polymorphism in nature.
RESUMO
BACKGROUND AND AIMS: Understanding the origin of pollination by sexual deception has proven challenging, as sexually deceptive flowers are often highly modified, making it hard to resolve how any intermediate forms between sexual deception and an ancestral strategy might have functioned. Here, we report the discovery in Caladenia (Orchidaceae) of sexual attraction with pollination during feeding behaviour, which may offer important clues for understanding shifts in pollination strategy. METHODS: For Caladenia robinsonii, we observed the behaviour of its male wasp pollinator, Phymatothynnus aff. nitidus (Thynnidae), determined the site of release of the sexual attractant, and experimentally evaluated if the position of the attractant influences rates of attempted copulation and feeding behaviour. We applied GC-MS to test for surface sugar on the labellum. To establish if this pollination strategy is widespread in Caladenia, we conducted similar observations and experiments for four other Caladenia species. KEY RESULTS: In C. robinsonii, long-range sexual attraction of the pollinator is via semiochemicals emitted from the glandular sepal tips. Of the wasps landing on the flower, 57 % attempted copulation with the sepal tips, while 27 % attempted to feed from the base of the labellum, the behaviour associated with pollen transfer. A similar proportion of wasps exhibited feeding behaviour when the site of odour release was manipulated. A comparable pollination strategy occurs in another phylogenetically distinct clade of Caladenia. CONCLUSIONS: We document a previously overlooked type of sexual deception for orchids involving long-distance sexual attraction, but with pollination occurring during feeding behaviour at the labellum. We show this type of sexual deception operates in other Caladenia species and predict that it is widespread across the genus. Our findings may offer clues about how an intermediate transitional strategy from a food-rewarding or food-deceptive ancestor operated during the evolution of sexual deception.
Assuntos
Orchidaceae , Vespas , Animais , Polinização , Flores , Comportamento Alimentar , FeromôniosRESUMO
Organic and inorganic cyanides are widely distributed in nature, yet not much is known about the ability of microorganisms to use these compounds as a source of nitrogen and/or carbon at high temperatures (>80 °C). Here we studied the capacity of organic and inorganic cyanides to support growth of an hyperthermophilic Pyrococcus strain isolated from Deception Island, Antarctica. This microorganism was capable of growing with aromatic nitriles, aliphatic nitriles, heterocyclic nitriles, amino aromatic nitriles and inorganic cyanides as nitrogen and/or carbon source. This is the first report of an hyperthermophilic microorganism able to incorporate these compounds in its nitrogen and carbon metabolism. Based on enzymatic activity and genomic information, it is possibly that cells of this Pyrococcus strain growing with nitriles or cyanide, might use the carboxylic acid and/or the ammonia generated through the nitrilase enzymatic activity, as a carbon and/or nitrogen source respectively. This work expands the temperature range at which microorganisms can use organic and inorganic cyanides to growth, having important implications to understand microbial metabolisms that can support life on Earth and the possibility to support life elsewhere.
Assuntos
Cianetos , Pyrococcus , Cianetos/metabolismo , Regiões Antárticas , Nitrilas , Carbono , NitrogênioRESUMO
Event-related potentials (ERPs) have been used with the concealed information test (CIT) to detect concealed recognition of specific stimuli (i.e., "probes"). While most research has focused on the P300 component, which is larger for infrequent probes than for frequent control stimuli (i.e., "irrelevants"), some studies have investigated an earlier ERP component, the anterior N2, with mixed results. Although some studies have reported a larger anterior N2 for probes than irrelevants (N2 enhancement), other studies, including our own, have not found such an effect. The present study aimed to replicate and extend our previous findings using the same CIT paradigm and measurement parameters. Results of Bayesian analyses show strong evidence against the hypothesis of anterior N2 enhancement by probes, replicating our previous work. Bayesian analyses also show strong evidence against the hypothesis of N2 enhancement for the three components revealed by a temporal principal component analysis (PCA) conducted to disentangle potentially overlapping ERP effects. In conclusion, whereas the CIT has shown promise in detecting recognition of specific information, anterior N2 enhancement cannot be used as an electrophysiological measure of concealed information across CIT paradigms.
RESUMO
While physical performance has long been thought to be limited only by physiological factors, many experiments denote that psychological ones can also influence it. Specifically, the deception paradigm investigates the effect of psychological factors on performance by manipulating a psychological variable unbeknownst to the subjects. For example, during a physical exercise performed to failure, previous results revealed an improvement in performance (i.e., holding time) when the clock shown to the subjects was deceptively slowed down. However, the underlying neurophysiological changes supporting this performance improvement due to deceptive time manipulation remain unknown. Here, we addressed this issue by investigating from a neuromuscular perspective the effect of a deceptive clock manipulation on a single-joint isometric task conducted to failure in 24 healthy participants (11 females). Neuromuscular fatigue was assessed by pre- to post-exercise changes in quadriceps maximal voluntary torque (Tmax ), voluntary activation level (VAL), and potentiated twitch (TTW ). Our main results indicated a significant performance improvement when the clock was slowed down (Biased: 356 ± 118 s vs. Normal: 332 ± 112 s, p = .036) but, surprisingly, without any difference in the associated neuromuscular fatigue (p > .05 and BF < 0.3 for Tmax , VAL, and TTW between both sessions). Computational modeling showed that, when observed, the holding time improvement was explained by a neuromuscular fatigue accumulation based on subjective rather than actual time. These results support a psychological influence on neuromuscular processes and contribute significantly to the literature on the mind-body influence, by challenging our understanding of fatigue.
Assuntos
Contração Isométrica , Fadiga Muscular , Feminino , Humanos , Fadiga Muscular/fisiologia , Contração Isométrica/fisiologia , Músculo Quadríceps/fisiologia , Exercício Físico/fisiologia , Desempenho Físico Funcional , Eletromiografia , Músculo Esquelético/fisiologiaRESUMO
Three preregistered studies examined whether 5-year-old children cheat consistently or remain honest across multiple math tests. We observed high consistency in both honesty and cheating. All children who cheated on the first test continued cheating on subsequent tests, with shorter cheating latencies over time. In contrast, 77% of initially honest children maintained honesty despite repeated failure to complete the tests successfully. A brief integrity intervention helped initially honest children remain honest but failed to dissuade initially cheating children from cheating. These findings demonstrate that cheating emerges early and persists strongly in young children, underscoring the importance of early prevention efforts. They also suggest that bolstering honesty from the start may be more effective than attempting to remedy cheating after it has occurred. RESEARCH HIGHLIGHTS: Our research examines whether 5-year-old children, once they have started cheating, will continue to do so consistently. We also investigate whether 5-year-old children who are initially honest will continue to be honest subsequently. We discovered high consistency in both honesty and cheating among 5-year-old children. Almost all the children who initially cheated continued this behavior, while those who were honest stayed honest. A brief integrity-boosting intervention successfully helped 5-year-old children maintain their honesty. However, the same intervention failed to deter cheaters from cheating again. These findings underscore the importance of implementing integrity intervention as early as possible, potentially before children have had their first experience of cheating.
RESUMO
Collaborative crime poses severe social hazards. In collaborative crime scenarios, previous studies have indicated that perpetrators' collaborative encoding can impair the detection efficiency of P300-based complex trial protocols due to the collaborative encoding deficit. The feedback concealed information test (fCIT), a unique variation of the concealed information test, provides participants with feedback on how well they conceal information from memory. The fCIT, which has proven to be highly efficient, detects concealed information using recognition P300 along with feedback-related event-related potentials, and reflects the subject's motivation to conceal. However, no studies have examined the fCIT's effectiveness in identifying collaborative criminals. We propose that the fCIT's efficiency persists in cases of collaborative crime and test this hypothesis using a sample of 48 participants. The participants in the collaborative groups were instructed to have hushed conversations about theft to simulate the collaborative crime process. Subsequently, they completed the fCIT. The findings indicate a significant decline in recognition P300's detection efficiency when participants committed crimes collaboratively. Nevertheless, the detection efficiency of feedback P300 and feedback-related negativity remained high. This study's outcomes illustrate the capacity of the fCIT to detect perpetrators involved in collaborative crime.
Assuntos
Enganação , Detecção de Mentiras , Humanos , Retroalimentação , Potenciais Evocados P300 , CrimeRESUMO
Children and young adolescents often tend to behave dishonestly in order to serve their self-interests. This study focused on how empathic abilities affect children's tendency to deceive others. Deception is the act of causing others to form a false belief to get them to act in a way that serves the deceiver's interests. As such, it requires the ability to predict how others might use the provided information. In two experiments, 274 participants (aged 10-16 years) played a game in which they could send a deceptive message to another participant to boost their own payoff at the other player's expense. We measured participants' cognitive and emotional empathy using different measures. We found that a measure of cognitive empathy, namely the fantasy scale, was associated with less deception of another player when that other player was not identified and was presented only as "Player B." However, when Player B was identified by name, empathy did not predict deception. In such cases, the only factors affecting deception rates were the gain for the participant (higher possible gains lead to more deception) and loss to the other player (higher possible losses lead to less deception). Overall, the findings suggest that even by 11 years of age, children can understand the impact of their unethical behavior on another child and adjust their actions accordingly. However, when the other child is not identified, children need to possess high levels of cognitive empathy toward imagined individuals to resist the temptation to deceive the other child.
Assuntos
Enganação , Empatia , Criança , Humanos , Adolescente , Emoções , Motivação , CogniçãoRESUMO
The strategic use of deliberate omissions, conveying true but selective information for deceptive purposes, is a prevalent and pernicious disinformation tactic. Crucially, its recognition requires engaging in a sophisticated, multi-part social cognitive reasoning process. In two preregistered studies, we investigated the development of children's ability to engage in this process and successfully recognize this form of deception, finding that children even as young as 5 years are capable of doing so, but only with sufficient scaffolding. This work highlights the key role that social cognition plays in the ability to recognize the manipulation techniques that underpin disinformation. It suggests that the interrelated development of pragmatic competence and epistemic vigilance can be harnessed in the design of tools and strategies to help bolster psychological resistance against disinformation in even our youngest citizens-children at the outset of formal education.
Assuntos
Desenvolvimento Infantil , Enganação , Humanos , Masculino , Feminino , Pré-Escolar , Criança , Cognição Social , Reconhecimento PsicológicoRESUMO
Children are often third-party observers of conversations between informants and receivers. Although 5- and 6-year-olds can identify and reject informants' false testimony, it remains unclear whether they expect others to do the same. Accurately assessing others' impressions of informants and their testimony in a conversational setting is essential for children's navigation of the social world. Using a novel second-order lie detection task, the current study examined whether 4- to 7-year-olds (N = 74; Mage = 69 months) take receivers' epistemic states into account when predicting whether a receiver would think an informant is truthful or deceptive. We pitted children's firsthand observations of reality against informants' false testimony while manipulating receivers' perceptual access to a sticker-hiding event. Results showed that when the receiver had perceptual access and was knowledgeable, children predicted that the receiver would think the informant is lying. Critically, when the receiver lacked perceptual access and was ignorant, children were significantly more likely to predict that the receiver would think the informant is telling the truth. Second-order theory of mind and executive function strengthened this effect. Findings are interpreted using a dual-process framework and provide new insights into children's understanding of others' selective trust and susceptibility to deception.
Assuntos
Sinais (Psicologia) , Julgamento , Criança , Humanos , Pré-Escolar , Confiança , Função Executiva , EnganaçãoRESUMO
Research is lacking regarding adults' ability to determine whether children's drawings are based on an experience or not. Drawings are useful in professional settings to alleviate linguistic demands, facilitate memory, and have been used as evidence. Determining the accuracy of veracity assessments of children's drawings would inform professionals regarding their use as evidence of experiences. Twenty-eight children (14 younger, Mage = 7.53 years, SDage = 1.19; 14 older, Mage = 11.67 years, SD = 1.27) produced drawings of two events: one staged experienced, and one narrative-based not experienced event. Fifty (Study 1, Mage = 23.72 years, SDage = 9.70) and 63 (Study 2, Mage = 25.92, SDage = 12.79) adults indicated whether each drawing was based on experience and their confidence in each assessment. In Study 2, additional drawing quality assessments were collected. Results indicated that adults were more accurate at distinguishing experienced than not experienced drawings for older artists. An inverse relationship was observed between confidence and accuracy-participants were more confident when they were inaccurate, especially for younger artists. Drawing quality improved with age and for drawings of experienced events. Adults tended to rate drawings of higher quality as resulting from experience leading to the highest accuracy for drawings from older artists that were based on experience. Overall, results suggest that there may be some features of drawings that allow for above chance levels of accuracy (up to 75%). However, rates are not high enough across assessments (M = 53.93%, range: 39%-75%) to reliably use them as indicators of experience.
RESUMO
OBJECTIVES: This study aims to examine caregivers' perspectives on and justifications for lying when caring for people with dementia. METHOD: The data consisted of interviews and observations of discussions among family and professional caregivers with various educational backgrounds. Thematic analysis was applied to identify key themes related to caregivers' perspectives on lying. RESULTS: The study revealed that lies were frequently employed by caregivers and were seen as effective tools in the caregivers' toolkit. These practices were often labelled 'white lies' and were rationalised based on their potential to enhance the well-being of people with dementia or to facilitate smooth interactions. The potential negative consequences of lying were also acknowledged. In addition, some caregivers suggested that the practice of 'stepping into the person with dementia's reality' might not constitute lying. CONCLUSION: The findings suggest that the caregivers perceived lying to be a legitimate strategy when caring for people with dementia; surprisingly, some did not recognize their practices as constituting acts of lying at all. This finding carries significant clinical relevance, as the varying perceptions of lying underscore the potential need for a consistent approach to deception. Addressing this complexity can lead to more ethical caregiving practices, ultimately enhancing the quality of care provided to people with dementia.
RESUMO
AIM: There is a lack of clarity about therapeutic lying in the context of everyday dementia care. This study provides conceptual clarity on how the term is used and considers the concept in relation to person-centred care. METHODS: Rodgers' (1989) evolutionary framework of concept analysis was employed. A systematic multiple database search was conducted and supplemented with snowballing techniques. Data were analysed thematically through an iterative process of constant comparison. RESULTS: This study highlighted that therapeutic lying is intended to be used in the person's best interests for the purpose of doing good. However, its potential for doing harm is also evident. Its use in the literature has increased with the general trend towards becoming more accepted in the discourse. A continuum emerged depending on the degree to which a lie departs from the truth. Emerging guidelines were also evident as to when a lie could or could not be justified. CONCLUSION: The term therapeutic lying, was contrasted with aspects of person-centred care and was found to be problematic. We conclude that there may be more pragmatic ways of constructing language around the care of people with dementia which could be less stigmatising.
Assuntos
Demência , Humanos , Demência/terapia , Assistência Centrada no Paciente , Idioma , Bases de Dados FactuaisRESUMO
Most studies on deception in soccer penalty kicks have focused on the deceptive actions used by penalty takers. However, it is worth noting that deception can also be played out by goalkeepers. To examine the effectiveness of goalkeepers' deceptive actions in professional competition, we analysed 714 penalty kicks taken during matches in the English Premier League and German Bundesliga, spanning the seasons from 2016-2017 to 2019-2020. We scored whether goalkeepers used deception, and if so, what type of deception, the outcome of the penalty and the kicking strategy of the penalty taker. The results showed that goalkeepers used deception in half of the penalty kicks, resulting in significantly less goals compared to penalties without deception. This advantage was similar for the different types of deception, but larger when penalty takers paid attention to goalkeepers. We propose that the deceptive actions by goalkeepers are effective, mainly because it leads the penalty taker to lose focus. The practical implications of these findings are discussed for both goalkeepers and penalty takers.
RESUMO
Background: The COVID-19 pandemic has accelerated and amplified the use of virtual research methods. While online research has several advantages, it also provides greater opportunity for individuals to misrepresent their identities to fraudulently participate in research for financial gain. Participant deception and fraud have become a growing concern for virtual research. Reports of deception and preventative strategies have been discussed within online quantitative research, particularly survey studies. Though, there is a dearth of literature surrounding these issues pertaining to qualitative studies, particularly within substance use research. Results: In this commentary, we detail an unforeseen case study of several individuals who appeared to deliberately misrepresent their identities and information during participation in a virtual synchronous qualitative substance use study. Through our experiences, we offer strategies to detect and prevent participant deception and fraud, as well as challenges to consider when implementing these approaches. Conclusions: Without general awareness and protective measures, the integrity of virtual research methods remains vulnerable to inaccuracy. As online research continues to expand, it is essential to proactively design innovative solutions to safeguard future studies against increasingly sophisticated deception and fraud.
Assuntos
COVID-19 , Enganação , Fraude , Pesquisa Qualitativa , Transtornos Relacionados ao Uso de Substâncias , Humanos , Fraude/prevenção & controle , COVID-19/prevenção & controle , Transtornos Relacionados ao Uso de Substâncias/prevenção & controleRESUMO
Given the high relevance and impact of ransomware in companies, organizations, and individuals around the world, coupled with the widespread adoption of mobile and IoT-related devices for both personal and professional use, the development of effective and efficient ransomware mitigation schemes is a necessity nowadays. Although a number of proposals are available in the literature in this line, most of them rely on machine-learning schemes that usually involve high computational cost and resource consumption. Since current personal devices are small and limited in capacities and resources, the mentioned schemes are generally not feasible and usable in practical environments. Based on a honeyfile detection solution previously introduced by the authors for Linux and Window OSs, this paper presents a ransomware detection tool for Android platforms where the use of trap files is combined with a reactive monitoring scheme, with three main characteristics: (i) the trap files are properly deployed around the target file system, (ii) the FileObserver service is used to early alert events that access the traps following certain suspicious sequences, and (iii) the experimental results show high performance of the solution in terms of detection accuracy and efficiency.