
As AI takes on literature review, hypothesis development, analysis and even peer review, the core skills of the scientist risk fading, much as mental arithmetic did after the calculator. Sir Peter Gluckman asks what a scientist will be, and how one will be trained, if critical thinking is increasingly offloaded to machines. Science needs to reflect urgently on which practices and norms must remain human, and on what that means for training, for society and for policy. For a full article, please click here.
European science advice is mourning the loss of an ideal, simpler world and is only part way to deciding what to let go of, what to hold on to and what to create. Claire Craig, Rokia Ballo and Ruth Morgan argue that this “post-naïve” moment calls for the confidence to be uncomfortable, not retreat. Drawing on the Frontiers Policy Labs and INGSA Europe blog collection, they point to shifting models of advice, the risks and benefits of AI, the social contract for science, and the role of early career researchers. For a full article, please click here.
The first UN Global Dialogue on AI Governance, held in Geneva in July 2026, gave the UN a standing political forum and an independent scientific panel on AI. Pavlina Ittelson and Seydina M. Ndiaye argue that a seat at the table is not enough if countries lack the capacity to shape the agenda. The next Secretary-General should treat capacity-building as part of the governance architecture and measure success by capability gained, not by attendance.For a full article, please click here.
Adding AI to science is not automatically progress. Across funders, grant applications are surging, polished text is cheap, and the proxies for quality (effort, experience, originality) are eroding. James Wilsdon argues that more output can mean dilution unless we change what we count. Science needs to measure contribution rather than rank, fund validation as seriously as ideation, and disclose where AI is used in assessment. For a full article, please click here.
Background: Attention-deficit/hyperactivity disorder (ADHD) is one of the most common neurodevelopmental disorders of childhood and is associated with substantial academic, behavioral, social, emotional, and health-related burdens. Contemporary care extends beyond symptom identification to include developmentally informed assessment, recognition of comorbidities, shared decision-making, and measurement-based multimodal treatment. Objective: To provide an updated, clinically oriented synthesis of current evidence on the epidemiology, neurobiology, diagnosis, differential diagnosis, comorbidities, treatment, monitoring, and long-term management of ADHD in children and adolescents. Methods: This narrative clinical review used a focused PubMed/MEDLINE search covering January 1, 2018, through August 22, 2026, supplemented by foundational publications and authoritative guidance from the American Academy of Pediatrics, National Institute for Health and Care Excellence, US Food and Drug Administration, Centers for Disease Control and Prevention, and National Institute of Mental Health. The search identified 183 records for initial consideration. Evidence was selected purposively according to pediatric relevance, methodological rigor, clinical applicability, and regulatory status. Results: ADHD is a heterogeneous disorder arising from complex genetic, neurobiological, and environmental influences. Diagnosis remains clinical and requires developmentally inappropriate symptoms, functional impairment, childhood onset, and manifestations in more than one setting, while excluding alternative explanations. Rating scales support but do not replace comprehensive clinical assessment. Comorbid learning, anxiety, depressive, oppositional, autism spectrum, tic, sleep, and substance-use disorders can substantially influence presentation and management. Behavioral parent training is recommended as first-line treatment for preschool-aged children. For most school-aged children and adolescents with clinically significant impairment, stimulants provide the strongest evidence for reducing core symptoms, whereas atomoxetine, extended-release guanfacine, extended-release clonidine, and extended-release viloxazine are appropriate alternatives for selected patients. Effective management integrates medication when indicated with psychoeducation, behavioral strategies, school-based interventions, and treatment of comorbidities. Monitoring should address symptoms, functioning, adherence, appetite, growth, sleep, blood pressure, heart rate, mood, and medication misuse or diversion. Digital interventions may offer adjunctive benefits, but current evidence does not support replacing established treatments. Conclusions: Pediatric ADHD should be managed as a chronic, developmentally evolving condition. Accurate diagnosis, individualized multimodal treatment, systematic monitoring, and planned transition to adult services can improve symptoms and functioning. Objective biomarkers, pharmacogenomics, neurostimulation, microbiome-directed therapies, and most digital interventions remain adjunctive or investigational.