The availability of mobile money in Kenya has positively impacted commerce, financial transaction efforts, and the ability of individuals to receive and save their money. The addition of mobile loan applications provides access to loans without the hassle of going to a physical bank and, in some cases, completing paperwork. While it has its benefits, limited protection of user data has been a cause for concern. User complaints prompted a change in the mobile loan industry, requiring applications to be licensed and banning the use of specific permissions for Android versions of the apps placed in the Google Play Store. We investigate the impact of this change and explore ways to improve regulation by reviewing 30 licensed (n=15) and unlicensed (n=15) Kenyan-targeted digital lender apps. The results suggest that regulation has not yet had a significant impact on digital lender app development and thus encourages government-supported development guidelines and audits.
Sudden Infant Death Syndrome (SIDS) is a serious public health concern, representing one of the leading causes of mortality in infants during their first year of life. This study focuses on investigating the possible relationship between sleep apnea and SIDS, addressing factors such as birth weight, age, and gender of the neonates. Using a methodology of data mining and recurrent neural networks (RNNs), physiological data obtained during the first 12 hours of life were analyzed. Initial results showed limited correlations, but with deeper analysis, greater consistency and accuracy in apnea predictions were achieved, reaching high reliability in the RNN model. Collaboration with the Mexican Red Cross was crucial, providing valuable data and allowing for the reevaluation of neonatal monitoring protocols. This methodological approach not only enhances the detection and early prevention of SIDS based on sleep apnea but also suggests practical solutions to reduce the incidence of this tragic condition.
Nowadays, artificial intelligence (AI) is finding relevance in diverse facets of human existence. For example, AI has been applied in areas such as finance, energy, healthcare, transportation, education, agriculture, and entertainment, just to mention a few. However, technologies that employ AI algorithms such as machine learning algorithms and deep learning algorithms are susceptible to a new challenge known as adversarial machine learning (AML) attacks. An AML attack occurs when a malicious actor crafts adversarial examples (AEs) to introduce perturbation into the original input data of an AI-based classification model to cause the model to misclassify inputs. This paper reviewed how computer vision systems, audio systems, industrial systems, transportation systems, blockchain technologies, and cybersecurity systems may be susceptible to AML attacks. Special attention is given to the discussion on AML attacks on network intrusion detection systems (NIDSs) while also highlighting adversarial defense strategies. The paper concludes with a proposal for future research direction regarding AML attack mitigation.
This paper introduces work by the IEEE P7014.1 Working Group on the Recommended Practice for Ethical Considerations of Emulated Empathy in Partner-based General-Purpose Artificial Intelligence Systems. This paper briefly details the scope and parameters of the standard, why it matters, and key ethical problems found regarding use of modern AI systems that emulate empathy for human AI-partnering. Some of these problems are fairly obvious, and others are less so, but no less important. A few however require deeper consideration because, like many important ethical discussions, they do not have easy answers. One such question is when is deception in human-computer interaction acceptable, particularly where deception overlaps with animism and anthropomorphism and may be exacerbated by emulations of empathy? This paper lingers on this question, drawing on philosophical and ethical discussion about the nature of deception, contexts where it is acceptable and beneficial, and contexts where it is morally out of scope.
This paper examines the integration of large language models (LLMs) in the workplace using a hermeneutic phenomenological approach, highlighting the critical differences between human language use and LLM operational mechanisms. It challenges the uncritical adoption of LLMs, emphasizing their limitations in contextual understanding and intuitive decision-making. Advocating a reimagined workforce, the paper proposes viewing LLMs as "AI interns" and humans as "Master Orchestrators," focusing on enhancing human creativity and ingenuity to ensure AI supports rather than replaces human capabilities.
We present an exploratory research aimed at understanding the perceptions that students and professors associated with the Engineering in Computer Science and Information Technology (ECSIT) program have regarding the ethical challenges of Artificial Intelligence, based on a theoretical framework that engages with ecofeminist, care, and ecological ethics. To achieve the research objective, a qualitative methodology was designed, and a comparative study was conducted between two ECSIT university programs in Mexico, one public and one private. Eight professors were interviewed, and two focus groups were conducted with the participation of 10 students. In both cases, the participation of women was encouraged, and professors teaching engineering courses and ethics courses for engineers were included. Findings reveal that the type of academic program is more relevant than the type of university, as well as the gender of professors and their interdisciplinary training.
This paper presents a preliminary exploration of utilizing the wireless received signal strength indicator (RSSI) as a secondary authentication method in smartphones. We propose AuthRSSI - a gesture based authentication system that uses the RSSI measurements from a smartphone connected to an indoor wireless network to identify a pre-determined user pattern. The key idea is that users will create a pattern in the air using a smart-phone, and the changes in the wireless channel can be analyzed to authenticate the user. AuthRSSI will serve as an important component in a proposed larger multi-sensory authentication system that combines the RSSI, Channel State Information (CSI) as well as built-in motion sensors in the smartphone like the accelerometer and gyroscope to accurately classify personalized user patterns and serve as a secondary authentication system that is complementary to existing methods in use today. Our findings demonstrate a strong correlation between RSSI measurements and user device movement, particularly in scenarios with a clear line-of-sight path.
This study examines the critical factors influencing the implementation of government vocational training programs for women in Ernakulam, Kerala, India, focusing on the perspectives of key implementers involved in the SANKALP project. Through qualitative focus group discussions, the research highlights both challenges and factors affecting program execution, including logistical constraints, socio-economic and cultural barriers, and gender-specific issues. SANKALP extends beyond mere vocational training, leveraging humanitarian technology and innovative research to tackle broader social and environmental challenges, thus contributing to Sustainable Development Goals (SDG) 4 and 5 by promoting quality education and achieving gender equality. The program focuses on establishing educational institutions, developing relevant curricula, and implementing community-focused initiatives that uplift rural and underprivileged communities. The findings underscore the necessity of integrating gender-sensitive approaches and tailoring programs to local contexts to enhance the effectiveness of vocational training, ultimately empowering women economically and socially, and driving progress towards SDGs 4 and 5.
Corporations, institutions, and individuals increasingly use artificial intelligence (AI) to make decisions and predictions that shape many aspects of human lives. Furthermore, individuals and organizations use AI to generate articles, blog posts, social media posts, or books, which can be substantial in various scenarios. However, this AI-generated content is often subject to biases that have unintended consequences on human lives during decision-making. The ability of AI models and systems to sustain and augment biases is a growing issue. Therefore, this literature examines how bias in AI-generated content can impact society, the sources of bias, and the quantitative methods used for measuring the bias. The literature also reviews real-world scenarios where quantitative bias measures in AI-generated content have been implemented successfully.
As automation and digitalisation become more pervasive, understanding their social implications, opportunities and risks, as well as appropriate policies and regulatory needs, require increasing inter-disciplinary efforts. This paper describes the motivation, format and outcomes of the AIvolution event about AI and digital technologies, organized in the European Parliament (EP) on November 2023, for that purpose. It presents the main topics addressed, the key contributions that were presented, the discussions that took place, and the main takeaways, as well as the perspectives to continue these exchanges in further initiatives.
The ethnic diversity of the Colombian Caribbean is reflected in its rich cultural manifestations, which are at risk due to digital divides in the artisanal sector. While some argue that technology threatens artisanal practices, others highlight the lack of knowledge transmission and documentation as the real threats. Public and community initiatives aim to preserve ancestral knowledge, opening opportunities for design and other disciplines to support traditional crafts. This study focuses on the Chimichagua community in Cesar, Colombia, analyzing cultural heritage promotion strategies to identify findings and methodologies that integrate social marketing, collaborative design, and heritage conservation. An ethnographic approach is used to understand community needs and motivations, enabling the development of an ICT-based strategy to promote material culture. The research includes a foresight process using the "Three Tomorrows of Postnormal Times" methodology, horizon scanning for future forces, and the "Pace Layering" framework to understand ecosystem changes over time. The study prioritizes four forces—Education, Demographics, Media & ICTs, and Geopolitics—to strengthen digital capacities and prevent the loss of artisanal vocation, proposing strategies to stimulate local development through culture and innovation.
This study examined the potential of the W-STEM mobile app to inspire and motivate more young women to pursue careers in science, technology, engineering, and mathematics (STEM). Developed within the Erasmus+ project "Building the future of Latin America: engaging women into STEM," the W-STEM application collects testimonies and life trajectories of outstanding and successful women in various STEM areas, presenting them as role models of easy access. Through in-depth interviews, invaluable insights were gained into the motivations, achievements, recommendations, and experiences these outstanding STEM women shared. Analysis of these testimonials revealed inspiring messages and critical lessons that can encourage the involvement of the next generation of women in traditionally male-dominated STEM disciplines. By leveraging the ubiquity of mobile devices, the W-STEM app provides unprecedented access to female role models in STEM, demystifying stereotypes and encouraging more young women to pursue paths in these fields. With its focus on providing inspiration, guidance, and resources in an accessible and engaging way, W-STEM represents a powerful technological tool that promotes gender equity and female participation in STEM fields. As of May 17, 2024, the application has garnered 16,275 views across 344 videos, averaging 47.28 views per video. These metrics highlight the app’s significant reach and engagement, underscoring its potential impact in providing accessible female STEM role models to a broad audience and encouraging more young women to explore and succeed in STEM careers.
This paper presents the development and evaluation of a multi-platform communication application designed for children with nonverbal autism. Utilizing Python and deep learning models, the application, named Conectando Voces, aims to facilitate communication for children with autism by providing a user-friendly interface and voice generation capabilities. The application employs the Flet framework for building the UI, enabling deployment across various devices including mobile phones, tablets, and PCs. Voice generation is achieved through Google’s text-to-speech model, with a fallback to pyttsx3 for offline use. The validation study involved five children with varying degrees of autism and comorbidities, highlighting the app’s potential effectiveness in promoting speech. Results indicate that children with mild to moderate autism benefit significantly, while those with severe autism or additional comorbidities require further tailored support. The study underscores the importance of personalized adaptations and comprehensive support for maximizing the app’s utility in enhancing communication skills in children with autism.
A problem that occurs every school period in a numerical methods course is the generation of problems. Problems are required for class examples, assignments, or exams. A wide variety of problems is needed to prevent students from having the answers before the problems are even assigned by the teacher. Preventing this is important to ensure the validity of the problems used as an assessment tool. The problems should have their answer, in addition to their solution, which can be found in a similar time to that given to solve an exam. This paper presents a problem generator for the topic of numerical integration. The developed software allows the generation of a large number of problems with the respective answer, so that the student can solve them in the time given to take the exams. One approach to teaching a course is gamification, which involves using a game environment in contexts as varied as business or education, among many others. The generated problems are numerical integration problems that are interpreted in a gamified context. The algorithms for the development of the software, the implementation, and the results regarding the generation of problems used in the numerical integration unit for the course of Numerical Methods in Engineering at the UAM Azcapotzalco are provided.
Mainstream media report that governments and business leaders see AI as an "extinction-level threat" to humans. In posthumanism, we find, besides conceptional tools for thinking technology in an indeterministic way, also argumentations for singularity, accelerationism and appeasement towards strong AI. Notwithstanding the heterogeneity of the field, the eminent posthumanist claim that autonomous machines can be a conscient other is a fallacy. An analysis of computer language as Derridean writing shows that computational, supposedly autonomous machines run on algorithmic, linguistic, written code. As such, they are extensions of human cognition. To proclaim machines conscient and autonomous, is hence not just misleading per se, but disguises the human agency that uses the AI-autonomy as a proxy.
Due to greater accessibility, diverse game options, and the social experiences provided by gaming platforms, the number of children engaging with the Roblox platform has increased over time. This increase in gaming from children of all ages has led to the parental challenge of balancing child safety with fun. Although parents typically know when their children are online, it can be challenging to trust that Roblox will protect data and minimize risky experiences due to the publicized criticisms of inadequate protections and the requirement of parent engagement. In this study, we examine parental Roblox concerns by (1) reviewing the game’s privacy policy and features, (2) characterizing parental concerns expressed on Reddit, and (3) surveying adults about their Roblox opinions. Our findings indicate gaps exist between what safety features Roblox provides and what parents need. Additionally, Roblox could improve how they convey their privacy and security practices to players and parents.
In the Anthropocene era, novel design strategies aligned with community mental models are imperative for sustainable behavior. Through in-depth interviews with 18 participants actively engaged in cultural activities, the study identifies three axial codes: a) nature worship, b) cultural identity, and c) sacred lunisolar events. These codes are integral to the formation of environmental identity, fostering human-nature connectedness. Drawing on these findings, the paper proposes a framework for designing interventions that leverage cultural attributes to promote sustainable behavior.
In Colombia, the high prevalence of obstructive sleep apnea (OSA) represents a significant public health problem. Associated comorbidities, such as cardiovascular disease and cognitive impairment, emphasize the importance of early diagnosis and appropriate treatment. While polysomnography (PSG) is the best standard for diagnosis, its high cost and limitations in terms of comfort have prompted the search for accessible and effective alternatives. In this context, a biofeedback model is proposed as an innovative tool for diagnosing and treating OSA in the Colombian population. This model is compared with existing devices such as the Beddr Sleep Tuner and the Apnealink, evaluating parameters such as durability, efficacy in detecting apnea episodes, and patient comfort. The compact and comfortable design of the biofeedback device makes it a more accessible alternative to the PSG. Preliminary results recommend that this model could be a valuable tool to improve the quality of life of OSA patients in Colombia by facilitating diagnosis and treatment and reducing the impact of associated comorbidities
This study explores the efficacy of artificial neural networks in predicting climate variables, specifically temperature, in Tungurahua Province, Ecuador. Utilizing climate data from the Mula Corral meteorological station, three types of neural networks—Long Short-Term Memory (LSTM), bidirectional LSTM, and Gated Recurrent Unit (GRU)—were evaluated. The data set, spanning from April 1, 2013, to April 1, 2024, included measurements of humidity, wind, precipitation, and temperature. Missing data were addressed using linear interpolation to ensure continuity. The LSTM model demonstrated superior performance, achieving a Root Mean Square Error (RMSE) of 0.71, making it the optimal choice for these data. Artificial Intelligence (AI) based temperature predictions offer significant benefits for agricultural planning, enabling farmers to optimize planting and harvesting times to enhance crop yields, including potatoes, corn, and beans, which are vital for the region’s food security. Moreover, these predictions facilitate the anticipation and mitigation of adverse impacts from extreme weather events, safeguarding public health and food security. Accurate long-term climate predictions also aid in efficient water resource management, contributing to economic and environmental savings. This research underscores the transformative potential of AI in climate prediction, providing a robust foundation for informed decision-making in agriculture, resource planning, and risk management. Furthermore, the neural network approach can be adapted to other regions with similar climatic conditions, opening avenues for future research and applications in meteorology and climate change.
Yoga has become increasingly popular in recent years as it offers several benefits through physical, mental, and spiritual practices. Although online resources offer the possibility of self-training at home without the need for an instructor, unsupervised training can result in injury, as users are not provided suggestions on how to improve. In this work, a Stereo Vision System, hosted on an NVIDIA Jetson Nano, was developed to evaluate Yoga poses in real time. The MediaPipe Pose framework and Stereo Vision System were enabled to perform 3D angle estimation of several joints of the human body. These angles were used to design and implement a methodology to evaluate user performance and provide them with feedback. The proposed method provides a 0-100 score, as well as a color scale system to visually indicate how a user can improve as they are performing a pose. Results prove the Stereo Vision System can perform accurate 3D Pose Detection and Angle Estimation, and is, therefore, reliable for providing accurate and correct feedback in real time. The proposed system will aid in minimizing injury risk while improving physical and mental health of individuals that practice Yoga.
