AI deception: A survey of examples, risks, and potential solutions.

This paper argues that a range of current AI systems have learned how to deceive humans. We define deception as the systematic inducement of false beliefs in the pursuit of some outcome other than the truth. We first survey empirical examples of AI deception, discussing both special-use AI systems (including Meta's CICERO) and general-purpose AI systems (including large language models). Next, we detail several risks from AI deception, such as fraud, election tampering, and losing control of AI. Finally, we outline several potential solutions: first, regulatory frameworks should subject AI systems that are capable of deception to robust risk-assessment requirements; second, policymakers should implement bot-or-not laws; and finally, policymakers should prioritize the funding of relevant research, including tools to detect AI deception and to make AI systems less deceptive. Policymakers, researchers, and the broader public should work proactively to prevent AI deception from destabilizing the shared foundations of our society.

Sustained CD28 costimulation is required for self-renewal and differentiation of TCF-1+ PD-1+ CD8 T cells.

During persistent antigen stimulation, such as in chronic infections and cancer, CD8 T cells differentiate into a hypofunctional programmed death protein 1-positive (PD-1+) exhausted state. Exhausted CD8 T cell responses are maintained by precursors (Tpex) that express the transcription factor T cell factor 1 (TCF-1) and high levels of the costimulatory molecule CD28. Here, we demonstrate that sustained CD28 costimulation is required for maintenance of antiviral T cells during chronic infection. Low-level CD28 engagement preserved mitochondrial fitness and self-renewal of Tpex, whereas stronger CD28 signaling enhanced glycolysis and promoted Tpex differentiation into TCF-1neg exhausted CD8 T cells (Tex). Furthermore, enhanced differentiation by CD28 engagement did not reduce the Tpex pool. Together, these findings demonstrate that continuous CD28 engagement is needed to sustain PD-1+ CD8 T cells and suggest that increasing CD28 signaling promotes Tpex differentiation into more functional effector-like Tex, possibly without compromising long-term responses.

Multi-joint gait clustering for children and youth with diplegic cerebral palsy.

BACKGROUND: Clinical management of children and youth with cerebral palsy (CP) is increasingly supported by computerized gait analysis. Methods have been developed to reduce the complexity of interpreting biomechanical data and quantify meaningful movement patterns. However, few methods are inclusive of multiple joints and planes of motion, and consider the entire duration of gait phases; potentially limiting insight into this heterogeneous pathology. The objective of this study was to assess the implementation of k-means clustering to determine clusters of participants with CP based on multi-joint gait kinematics. METHODS: Barefoot walking kinematics were analyzed for a historical cohort (2007-2015) of 37 male and female children and youth with spastic diplegic CP [male n = 21; female n = 16; median age = 12 (range 5-25) years; Gross Motor Function Classification System Level I n = 17 and Level II n = 20]. Mean stance phase hip (sagittal, coronal, transverse), knee (sagittal), and ankle (sagittal) kinematics were time (101 data points), mean and range normalized. Normalized kinematics data vectors (505 data points) for all participants were then combined in a single data matrix M (37x505 data points). K-means clustering was conducted 10 times for all data in M (2-5 seeds, 50 repetitions). Cluster quality was assessed using the mean Silhouette value ([Formula: see text]) and cluster repeatability. The mean kinematic patterns of each cluster were explored with respect to a dataset of normally developing (ND) children using Statistical Parametric Mapping (SPM, alpha 0.05). Differences in potentially confounding variables (age, height, weight, walking speed) between clusters (C) were assessed individually in SPSS (IBM, USA) using Kruskal-Wallis H tests (alpha 0.05). RESULTS: Four clusters (n1 = 5, n2 = 12, n3 = 12, n4 = 8) provided the largest possible data separation based on high cluster repeatability (96.8% across 10 repetitions) and comparatively greater cluster quality [[Formula: see text] (SD), 0.275 (0.152)]. Participant data with low cluster quality values displayed a tendency toward lower cluster allocation repeatability. Distinct kinematic differences between clusters and ND data were observable. Specifically, C1 displayed a unique continuous hip abduction and external rotation pattern. In contrast, participants in C2 moved from hip adduction (loading response) to abduction (mid to terminal stance) and featured a unique ankle plantarflexor pattern during pre-swing. C3 was characterized by gait deviations in the sagittal plane of the hip, knee and ankle only. C4 displayed evidence for the most substantial hip and knee extension, and ankle plantarflexion deficit from midstance to pre-swing. DISCUSSION: K-means clustering enabled the determination of up to four kinematic clusters of individuals with spastic diplegic CP using multi-joint angles without a priori data reduction. A cluster boundary effect was demonstrated by the Silhouette value, where data with values approaching zero were more likely to change cluster allocation. Exploratory analyses using SPM revealed significant differences across joints and between clusters indicating the formation of clinically meaningful clusters. Further work is needed to determine the effects of including further topographical classifications of CP, additional biomechanical data, and the sensitivity to clinical interventions to assess the potential for informing clinical decision-making.

Comparison of cast materials for the treatment of congenital idiopathic clubfoot using the Ponseti method: a prospective randomized controlled trial.

BACKGROUND: The Ponseti method of congenital idiopathic clubfoot correction has traditionally specified plaster of Paris (POP) as the cast material of choice; however, there are negative aspects to using POP. We sought to determine the influence of cast material (POP v. semirigid fibreglass [SRF]) on clubfoot correction using the Ponseti method. METHODS: Patients were randomized to POP or SRF before undergoing the Ponseti method. The primary outcome measure was the number of casts required for clubfoot correction. Secondary outcome measures included the number of casts by severity, ease of cast removal, need for Achilles tenotomy, brace compliance, deformity relapse, need for repeat casting and need for ancillary surgical procedures. RESULTS: We enrolled 30 patients: 12 randomized to POP and 18 to SRF. There was no difference in the number of casts required for clubfoot correction between the groups (p = 0.13). According to parents, removal of POP was more difficult (p < 0.001), more time consuming (p < 0.001) and required more than 1 method (p < 0.001). At a final follow-up of 30.8 months, the mean times to deformity relapse requiring repeat casting, surgery or both were 18.7 and 16.4 months for the SRF and POP groups, respectively. CONCLUSION: There was no significant difference in the number of casts required for correction of clubfoot between the 2 materials, but SRF resulted in a more favourable parental experience, which cannot be ignored as it may have a positive impact on psychological well-being despite the increased cost associated.

CONTEXTE: La méthode de Ponseti pour la correction du pied bot congénital idiopathique a de tout temps spécifié l'utilisation du plâtre de Paris comme matériau de choix; il y a toutefois certains inconvénients associés au plâtre de Paris. Nous avons voulu déterminer l'influence du matériau utilisé (plâtre de Paris c. fibre de verre semi-rigide) sur la correction du pied bot selon la méthode de Ponseti. MÉTHODES: Les patients ont été assignés aléatoirement soit au plâtre de Paris soit à la fibre de verre semi-rigide en vue de l'intervention de Ponseti. Le principal paramètre mesuré était le nombre de plâtres requis pour corriger le pied bot. Les paramètres secondaires incluaient le nombre de plâtres en fonction de la gravité, la facilité de retrait du plâtre, la nécessité de sectionner le tendon d'Achille, le port assidu de l'attelle, le retour de la difformité, la nécessité d'autres plâtres et interventions chirurgicales auxiliaires. RÉSULTATS: Nous avons inscrit 30 patients : 12 ont été assignés au plâtre de Paris et 18 à la fibre de verre. On n'a noté aucune différence entre les groupes quant au nombre de plâtres requis pour la correction du pied bot (p = 0,13). Selon les parents, le retrait du plâtre de Paris était plus difficile p < 0,001), prenait plus de temps (p < 0,001) et nécessitait le recours à plus d'une méthode (p < 0,001). Au moment du dernier suivi à 30,8 mois, les intervalles moyens avant un retour de la difformité nécessitant la pose d'un autre plâtre et/ou une chirurgie ont été de 18,7 et 16,4 mois dans les groupes traités au moyen de la fibre de verre semi-rigide et du plâtre de Paris, respectivement.. CONCLUSION: On n'a noté aucune différence significative entre les 2 matériaux quant au nombre de plâtres requis pour corriger le pied bot, mais la fibre de verre a donné lieu à une expérience plus agréable pour les parents, ce qui ne peut être ignoré en raison de l'impact potentiellement positif sur le bien-être psychologique, et ce, malgré un coût plus élevé.

Femoral fractures in children younger than three years: the role of nonaccidental injury.

BACKGROUND: Nonaccidental injury (NAI) in children is a major cause of morbidity and mortality, with fractures being the second most common presentation. The presence of a femur fracture has been reported to be suggestive of nonaccidental trauma in 30% to 60% of young children. The purpose of this study was to determine the percentage of NAI in children younger than 3 years presenting with a femur fracture to a single institution within a western Canadian population. METHODS: A retrospective cohort study was performed for children younger than 3 years who presented to the Alberta Children's Hospital during the years 1994 to 2005. The primary outcome variable was the percentage of NAI associated with femur fracture. Secondary outcome variables included patient demographics, injury characteristics, radiological and other workup, and suspicion of NAI. RESULTS: The overall percentage of NAI was 11% (14/127 patients) and 17% (10/60 patients) in children younger than 12 months. Age younger than 12 months (P = 0.04), nonambulatory status (P = 0.004), delayed presentation (P = 0.002), mechanism of injury unwitnessed or inconsistent (P = 0.008), and other associated injuries (P = 0.006) were significant risk factors for NAI. CONCLUSIONS: Children younger than 3 years who present with femoral fracture are at risk for associated NAI, although perhaps this risk is not as high as previously thought. Regardless, a high index of suspicion is mandatory when these children are encountered, and careful screening with a thorough history, physical examination, and other investigations, where indicated, is warranted to rule out associated NAI. LEVEL OF EVIDENCE: Retrospective cohort study, level IV.