Issue 01 • Fall 2025
A peer-reviewed, student-run journal that publishes original research, commentary, and analysis at the intersection of science, technology, and ethics. Our aim is to critically examine the design, governance, and real-world consequences of emerging technologies, while providing students with a credible platform to share their work with a global audience. Explore the abstracts below or download the full issue as a printable PDF.
We are now accepting submissions for the Winter issue of our journal. If you have research or insights to share, we encourage you to submit your work.
Submit Your ArticleThe Youth Tech and Ethics Institute’s Youth Journal of STEM and Society is where youths study the ethical responsibility of science and technology. Students ask challenging questions, where they consider responsibility, progress, and the kind of future we want to create. Our goal in these pages is to present concepts and create discussion regarding the contribution of STEM to the development of human society...
Agentic AI systems, autonomous or semi-autonomous agents that can execute key components of the scientific workflow, are reshaping how research is conducted. This paper examines the integration of such systems across five major stages of the scientific process: literature review, experimental design, data analysis, writing, and publishing and dissemination. Drawing from current tools, startups, and multi-agent frameworks, we assess both capabilities and constraints of AI in current research settings. Special attention is paid to the rise of closed-loop scientific systems, where AI not only assists but autonomously completes entire research cycles. While notable advances have been made, ethical and technical challenges must be resolved to realize AI’s full potential in responsible science. This study maps the landscape, identifies critical white spaces, and outlines the ethical implications of using AI in the research and publication process.
In neighborhoods bordering factories, highways, landfills, or in areas where police sirens are more frequent than birdsong, a saddening reality exists: environmental racism has not only damaged the body but the mind as well. Across the United States and beyond, communities of color are disproportionately exposed to neurotoxic pollutants in the environment around them, such as lead, mercury, and fine particulate matter, often a result of discriminatory zoning. The agents infiltrate not only the lungs and bloodstreams but also the delicate brain, impairing focus, cognitive development, and long-term neurological health. This form of harm, what some scholars call environmental violence, is a slow, systematic assault on the human mind. Yet in the age of technological advancements, artificial intelligence (AI) emerges as both a potential remedy and risk. This paper explores the current situation of environmental racism through various techniques to assist in solving the effects of environmental racism for the General Public.
This essay surveys the transformative role of AI across operations: process automation, predictive analytics, workforce planning, and marketing personalization. It outlines efficiency gains, governance needs, and the ethical guardrails required for responsible enterprise adoption.
Major technology firms increasingly tout sustainability, yet disclosures and performance vary widely. This paper reviews public commitments, supply-chain policies, and emissions data to construct a transparent scoring framework that distinguishes substantive action from greenwashing and recommends pathways for sector-wide improvement.
The rapid integration of artificial intelligence into computer science has transformed programming practices, introducing both opportunities and risks. While AI excels at repetitive coding tasks and enables new approaches such as “vibe coding,” it remains limited in creativity, accountability, and security. Studies show that AI-assisted programmers often generate less secure code and face greater risks of perpetuating bias due to flaws in training data. This paper explores the ethical implications of these developments, focusing on issues of accountability, security, bias, and the erosion of human expertise. Drawing on industry examples and recent research, the analysis argues that AI in computer science should be treated as a tool to augment, not replace, human judgment. Responsible adoption requires addressing bias, enforcing accountability, and ensuring that human creativity and critical thinking remain central to the field.
A new type of Artificial Intelligence is emerging - one that is trying to combat the shortcomings of current narrow AI systems. Referred to as "Artificial General Intelligence”, this concept is a vision where AI could be as adaptable and able to perform general tasks like humans. This new technology is only a theory currently, but could be a reality in as little as a few years, or may not even be achievable at all. This new field of AI, if achieved, could change the way that we live, either by creating a prosperous future or one that could lead to the extinction of humans entirely. This article’s purpose is to identify exactly what AGI is, what its impacts will be, and if it even is possible to achieve.
Public confusion caused by inconsistent recycling regulations continues to undermine the effectiveness of waste management programs. Traditional educational methods have shown only modest improvements, leaving contamination and improper disposal as widespread issues. This study examines the potential of artificial intelligence (AI) to reshape recycling behaviors by providing real-time, interactive feedback. A small-scale experiment in which an AI model trained on nearly 30,000 images guided participants in classifying household waste showed a marked improvement: the AI-assisted participant increased from 4 to 9 correct classifications out of 10, compared to the control participant, who improved from 4 to 7 using printed guides. These findings underscore AI’s ability to accelerate learning and enhance recycling accuracy, while commercial applications such as gamified platforms and accessibility-focused tools demonstrate scalability. Despite challenges including the digital divide, cost, and privacy concerns, AI offers a promising pathway to reduce confusion, improve motivation, and foster more sustainable recycling habits.
Cognitive Computing, a branch of artificial intelligence that simulates human thought processes, holds a great promise for transforming mental health care through diagnosis, personalized treatment, and consistent monitoring. However, as these systems tend to handle sensitive mental health data, concerns about privacy and data security are becoming paramount. This article explores the potential of cognitive computing to revolutionize mental health care while critically examining the challenges and future directions for safeguarding patient privacy.
Artificial intelligence (AI) has increasingly entered both everyday life and mental health care, becoming an unseen assistant to some and offering new opportunities for diagnosis, treatment, and accessibility. AI technologies are reshaping mental health delivery. These tools show promise in addressing shortages of clinicians, reducing costs, and personalizing interventions, but they also raise ethical challenges. Concerns around privacy, algorithmic bias, accountability, transparency, and the risk of dehumanizing care highlight the tension between innovation and trust. Applications such as Woebot demonstrate how AI-guided interactions can provide support, while the “black box” nature of many systems complicates informed consent and patient understanding. A survey of 500 U.S. adults revealed both optimism and caution: nearly half viewed AI as beneficial, yet most emphasized the importance of confidentiality, autonomy, transparency in risk assessment, and clinician accountability for misdiagnosis. These findings show that while AI can enhance access and supplement therapeutic alliances, it must be deployed with robust safeguards. Proposed solutions include developing explainable AI, clear accountability frameworks, strong data protection measures, inclusive implementation, and governance standards to validate safety and equity. Ultimately, the success of focused AI in mental health depends on balancing its transformative potential with ethical oversight to preserve trust, protect patients, and ensure that technology augments the human connection central to mental health care.
Superconductors, materials that exhibit zero electrical resistance below a critical temperature (Tc), are widely regarded as a cornerstone for future technologies including quantum computing, sustainable energy grids, and frictionless transportation systems. However, the discovery of new superconductors has historically been slow, limited by experimental cost and the challenges of theoretical prediction. This paper presents an interdisciplinary framework that integrates principles from physics and chemistry with modern machine learning methods to accelerate superconductor discovery. Using a dataset of over 21,000 materials, we engineered chemically informed features and trained a deep neural network to predict Tc with high accuracy (MAE ≈ 5K, RMSE ≈ 9K, R² ≈ 0.92). Beyond performance, our approach interprets model behavior through the lens of BCS theory, bridging data-driven insights with physical mechanisms. The societal potential of this work lies in its ability to reduce barriers to innovation, offering a scalable path toward materials that can enable cleaner energy transmission and more accessible advanced technologies. We also consider the challenges of relying on data-driven discovery in critical fields, underscoring the need for responsible and equitable development of AI-driven materials research.
Artificial intelligence is rapidly expanding its role in mental health, from conversational chatbots like Woebot and Wysa to diagnostic tools that analyze language, voice, and behavior. These technologies offer new opportunities for early detection, intervention, and emotional support, yet they also raise concerns about bias, privacy, and overreliance on machines. While AI can complement clinical care and increase access, it cannot replace the trust and nuance provided by human professionals. To ensure responsible use, AI in mental health must prioritize transparency, equity, and ethical safeguards while supporting—not substituting—the therapeutic relationship.
TransforMerger is a transformer-based system created to improve the flexibility of human-robot interaction (HRI) by taking in multimodal inputs (voice and gesture) and fusing these inputs into a single sentence. This employs probabilistic embeddings and takes a contextual scene to handle multimodal ambiguity. This sentence is sent to a Large Language Model (LLM) for processing. The experiment was conducted under real-world environments, demonstrating its robustness to misalignment, ambiguity, and missing information as it outperformed traditional baselines (especially in scenarios requiring more contextual knowledge). Thus, highlighting its potential to advance multimodal communications in HRI scenarios and its potential to advance various disciplines.
The tumour microenvironment (TME) is an increasingly recognised factor in cancer development, tumour progression and therapeutic resistance, comprising immune cells, stromal fibroblasts, vasculature and extracellular matrix components. TME significantly influences tumour behaviour through cell–cell interactions and molecular signalling. Recent advanced biotechnology has increased our understanding by incorporating TME profiling into techniques like microscopy techniques, cell culture models and molecular analysis methods. These innovations have not only deepened our knowledge, but also allowed us to establish new therapeutic targets and predict treatment responses especially in immunotherapy. This article aims to provide an overview of TME, the common techniques involved in TME profiling as well as to demonstrate the process from overcoming challenges like tumour heterogeneity and therapeutic resistance, to ultimately improving patient outcomes.
RNA-based therapeutics are revolutionizing medicine by targeting RNA, the blueprint of protein production, rather than proteins themselves. This allows scientists to silence harmful genes, correct faulty transcripts, and regulate protein synthesis at its source. The success of COVID-19 mRNA vaccines and therapies like patisiran and nusinersen demonstrates their potential across infectious disease, cancer, and rare genetic disorders. Advances in delivery systems have improved stability and broadened applications, while innovations such as self-amplifying and circular RNA promise even greater impact. Despite challenges in cost, accessibility, and global equity, RNA therapies represent a shift toward precision medicine and the possibility of personalized treatments once thought impossible.
Early detection of cancer significantly increases patient survival rates. However, delays in diagnostics, especially in underserved areas, remain a major obstacle. This paper explores the growing role of AI in medicine, mainly in convolutional neural networks supporting radiologists in identifying tumors from medical images. By improving accuracy, reducing diagnostic time, and extending care into low-resource settings, AI-assisted imaging offers a powerful tool for the future of cancer diagnosis.
Alzheimer’s disease, the leading cause of dementia, affects over 139 million people worldwide and imposes a profound social and economic burden. Current treatments can only slow progression or manage symptoms but cannot restore lost neuronal connections. Brain-Computer Interfaces (BCIs), particularly stimulation-based systems, are emerging as a potential therapy to bypass damaged neural circuits, enhance memory consolidation, and even retrieve lost memories. This paper reviews the history and mechanisms of BCIs, explores their experimental applications to memory restoration, and highlights both the promise and limitations of stimulation of BCIs in Alzheimer’s treatment. While preliminary studies show improved memory performance in controlled trials, translation to Alzheimer’s patients remains hindered by biological, technical, and ethical challenges. Ultimately, BCIs offer an exciting but cautious path forward, requiring interdisciplinary collaboration to ensure safe, equitable, and meaningful clinical application.
Decisions about climate change infrastructure demand long-term foresight, as projects in transportation, water, energy, and urban development must endure for decades while facing an uncertain climate future. This study highlights the importance of adaptive, integrated strategies that link land-use, energy, and transportation planning while considering socio-economic equity and resilience across vulnerable populations. Using lessons drawn from experimental interventions in Mumbai and comparative policy frameworks, the analysis demonstrates that hybrid green-gray solutions, multipurpose infrastructure, and smart-grid innovations enhance both technical performance and community well-being. At the same time, challenges such as unequal access, fragmented governance, and slow technological updating limit resilience. The evidence suggests that successful pathways require multidisciplinary coordination, integration of adaptation with mitigation, and lifecycle planning that leverages replacement opportunities after extreme events. Ultimately, the research shows that resilient climate infrastructure emerges not only from technical advances but from inclusive strategies that align ecological knowledge, social equity, and adaptive governance.
Solar roadways represent an innovative approach to renewable energy generation by embedding photovoltaic panels into transportation infrastructure. Beyond producing electricity, these systems aim to provide multifunctional benefits such as lighting, heating, and smart traffic management. Case studies from China, Germany, and pilot projects in the United States highlight both the promise and challenges of implementation, including high costs, durability concerns, and efficiency losses compared to traditional solar farms. While the concept remains largely experimental, ongoing research and technological advances suggest potential pathways for integrating solar roadways into future sustainable infrastructure.
A geoengineering technique called Stratospheric aerosol injection (SAI) is a mechanism of injecting reflective particles like sulfate aerosols into the stratosphere to cool Earth by 1–2°C. Conversely, it raises ethical concerns about moral hazard, inequity, and consent (Robock, 2018). This article examines whether SAI can be ethical if mitigation fails and who should govern its deployment. From analysis, case studies including Harvard’s SCoPEx, paused due to Sámi Indigenous protests, and the UK’s SATAN, banned for unauthorized experiments, reveals governance gaps. SAI is at risk of delaying decarbonization and adversely harming vulnerable regions such as African river basins facing droughts, South Asian monsoons risking disruption, and South Central American ecosystems experiencing precipitation shifts. The proposed Ethical Geoengineering Governance Framework (EGGF) states six principles to promote inclusive decision-making and prioritization of Indigenous, African Union, and regional representation. This structure advances climate justice and provides a path for ethical SAI deployment in a warming world.
In contrast to conventional systems, this study examines the long-term financial advantages of green infrastructure in urban engineering. It concludes that green infrastructure offers significant yearly economic and social returns, such as increased property values, job creation, and enhanced climate resilience, and reduces costs by almost 49% using global case studies, lifecycle cost analysis, and economic impact assessment. Important initiatives in places like Philadelphia and New York show high returns on investment and wide-ranging multiplier effects on local economies. The findings unequivocally demonstrate that green infrastructure is a sensible and strategic option for contemporary urban development since it is not only better for the environment but also for the economy. (Keywords: Green Infrastructure, Economics Benefits, Cost Savings, Return on Investment(ROI)).
Failure is often viewed as defeat, but in science it is a catalyst for progress. From the null results of the Michelson–Morley experiment to NASA’s Mars Polar Lander disaster and the downfall of Theranos, scientific and technological failures have repeatedly shaped new discoveries, regulations, and innovations. These examples reveal that while failure can be costly and discouraging, it provides opportunities to learn, correct errors, and push knowledge forward. Just as famous scientists like Thomas Edison emphasized persistence, failure remains essential to the very process of discovery.
Burnout among STEM students is a growing concern, often attributed to academic pressure, mental exhaustion, and cognitive overload. This paper explores two contributing psychological factors - altruism and stress-induced memory failure (academic amnesia)- and examines how to create a pathway that is the opposite of student burnout in STEM fields. Gathering from existing literature on medical knowledge, psychological studies of empathy and stress. This work encourages institutions to rethink their ethical structure and academic expectations for STEM scholars, an environment’s atmosphere to better accommodate students' well-being.
Disability and mortality from neurodegenerative diseases such as Parkinson’s disease (PD) have risen significantly since 1990. PD involves dopamine dysregulation, oxidative stress, neuroinflammation, and Lewy body formation. Emerging concerns point to the effects of repetitive short-form video consumption on dopamine regulation, as platforms like TikTok and Reels promote frequent reward cycles that may drive receptor desensitization and chronic dopamine fluctuations. This study examines whether such dysregulation contributes to PD-related mechanisms, including dopaminergic neuron loss in the substantia nigra, oxidative stress, neuroinflammation, and protein aggregation. Using rodent models, Enzyme-Linked Immunosorbent Assay (ELISA) and immunohistochemistry (IHC) will quantify and localize biomarkers after exposure. Findings aim to clarify the potential link between digital media use and neurodegeneration, offering insights for prevention and treatment strategies.
As commercialized space travel grows, concerns over its environmental impact have expanded, particularly regarding the accumulation of space debris. Space trash, originating from satellites, rockets, and failed missions, has doubled across sectors since 2010 and poses risks to navigation, satellites, and essential services such as GPS and weather forecasting. Without binding international regulations, cleanup remains limited, heightening the threat of the “Kessler Syndrome,” in which collisions create cascading debris that could make future space travel nearly impossible. This article examines the ethical, environmental, and technological challenges of managing orbital debris and emphasizes the need for global responsibility and innovative solutions.
Neuroscience tools like brain-stimulating headsets, wave-tracking apps, and cognitive enhancers are no longer limited to medical use—they’re being adopted by healthy students and professionals chasing productivity. While they can help treat ADHD, anxiety, or stroke recovery, using them for enhancement raises tough questions. Who gets access, and does it widen inequality? Will people feel pressured to use them just to keep up? What happens to identity and achievement when success feels engineered? With safety still uncertain, especially for developing brains, these technologies demand careful ethical reflection before we trade authenticity and well-being for efficiency.
This paper concerns itself with exploring the synthesis of various topics discussed in length at the Active Inference Institute, including Friston’s Free Energy Principle, Rudrauf’s Projective Consciousness Model and Worden’s Projective Wave Model. Through a deeper examination of the correlations between these closely-linked theories we may develop a stronger understanding of the nature of human consciousness, resolving problems in modern logic regarding the topic (eg., dispelling claims of AI consciousness) and holding potential for practical applications in fields such as AI and fMRI.
Why Trust is a Design Problem. Trustworthy systems are made, not found—from dataset curation to model deployment, institutional incentives shape outcomes. This issue curates research and perspectives that move beyond slogans to concrete practice.
Citable, archived layout with page numbers, figures, and references.