April 29, 2025 • By Ethan Gach

Two veteran game directors, Sven Vincke of Larian Studios (Baldur's Gate 3) and Yoko Taro (Nier Automata), have shared their thoughts on the impact of AI on the gaming industry. Vincke believes that AI will not replace human game developers, as it can only perform certain tasks and not replicate the craftsmanship that goes into creating a game. He sees AI as a tool to aid in game development, such as speeding up grunt work and flagging potential plot holes. On the other hand, Yoko Taro has a more pessimistic view, stating that AI will make many game creators unemployed in the future. He thinks that AI will be able to generate content that caters to users' preferences, potentially bypassing the need for human game makers. Taro's comments were made during a roundtable discussion about the potential of AI in game development, where he suggested that AI could lead to a future where game creators are no longer needed. The differing opinions of these two directors highlight the uncertainty and debate surrounding the role of AI in the gaming industry. While some see AI as a tool to enhance game development, others worry about its potential to replace human creators.

April 29, 2025

An automated machine learning program developed by researchers from Edith Cowan University (ECU) in conjunction with the University of Manitoba has been able to identify potential cardiovascular inci…

April 29, 2025 • By Bill Desowitz

The Academy Scientific and Technical Awards will honor 37 recipients for 14 achievements in film technology. The awards, hosted by Diego Luna, recognize innovations such as fire stunt safety, microphone synchronization, camera stabilization, and digital character simulation. Winners include Jayson Dumenigo for safer fire stunt materials, Thijs Vogels and team for Disney's machine learning denoiser, and Nir Averbuch and team for seamless microphone blending. Other winners include Curt Schaller for a camera stabilization system, Dave Freeth for a hand-held stabilization system, and Essex Edwards and team for a digital character simulation system. The awards also recognize advancements in machine learning denoisers, fire stunt gels, and modular motion bases. The winners will receive plaques or certificates for their contributions to the film industry.

April 29, 2025 • By Muiris O'Cearbhaill

Intel is facing financial difficulties and plans to cut jobs to reduce bureaucracy and keep up with the fast-paced AI industry. The company reported a loss of approximately €720m in the first quarter of this year and expects further losses. Intel has struggled to compete with competitors such as ARM and Nvidia, and has lost market share in the PC industry to AMD. The company is attempting to refocus on engineering and streamline its decision-making processes, with a renewed focus on AI chip development. Intel's new facility in Kildare, Ireland, will focus on manufacturing AI chips, and the company's CEO believes this could yield positive results for the workforce in Ireland. However, the company's financial struggles and job cuts have raised concerns about the impact on employees.

April 29, 2025 • By Ian Barker

A recent study by Optiv and the Ponemon Institute found that 79% of organizations are changing their cybersecurity budgets, with 71% increasing their budgets to an average of $24 million. Despite this, 66% of respondents reported an increase in cybersecurity incidents over the past year. The study highlights a shift towards data-driven decision-making, with 67% of organizations using risk and threat assessments to inform budget decisions. Additionally, 58% of organizations are now outsourcing to managed security service providers (MSSPs), particularly for cloud security guidance. The use of AI/ML to prevent cyberattacks is also on the rise, with 46% of respondents adopting these technologies to improve operational efficiency and maintain competitive advantage. However, 74% of respondents identified a lack of understanding of potential vulnerabilities as their biggest challenge to effective vulnerability management.

April 29, 2025 • By Tolu Olarewaju, Economist and Lecturer in Management, Keele University

The trade war between China and the US may have a significant impact on Nigeria. In 2024, 27.8% of Nigeria's imports came from China, while US exports to Nigeria reached $4.2 billion. If Chinese products are diverted to Nigeria due to US tariffs, it could benefit consumers but harm local entrepreneurs. Nigerian goods that could be replaced by cheaper Chinese products include textiles, furniture, footwear, and beauty products. Nigeria's weak infrastructure, high inflation, and poor business environment make it challenging for local entrepreneurs to compete. To mitigate this, the government could implement targeted tariffs and invest in sectors with growth potential. Entrepreneurs can also identify niche market needs, focus on customer service, embrace innovation, diversify supply chains, and explore new export markets. By adapting to global shifts and focusing on value creation, Nigerian businesses can thrive. This includes investing in light manufacturing or local processing to create value-added products, which offer better margins and market protection.

April 29, 2025 • By harold-fritts

Hewlett Packard Enterprise (HPE) has announced significant advancements in its HPE Aruba Networking and HPE GreenLake cloud platforms. The updates include AI-driven security, zero-trust networking, and compliance tools to help enterprises modernize secure connectivity and hybrid cloud operations. Key features include cloud-based access control, network access control, and enhanced policy management to support zero-trust network access. The HPE Aruba Networking EdgeConnect SD-WAN now includes advanced Secure Access Service Edge (SASE) features and Adaptive DDoS defense powered by machine learning. HPE GreenLake has also been enhanced with new capabilities, including a "digital circuit breaker" to temporarily disconnect from the public internet when network threats are detected, and air-gapped cloud management for sovereign environments and private clouds. Additionally, HPE has introduced new cybersecurity services to help enterprises evaluate, adopt, and integrate sovereign security solutions into their risk management frameworks. These services guarantee regulatory alignment and operational control for sensitive environments. The updates reflect HPE's commitment to edge-to-cloud security and align with global and industry regulations.

April 29, 2025 • By Arleo Dordar, Forbes Councils Member, Arleo Dordar, Forbes Councils Member https://www.forbes.com/councils/forbesbusinesscouncil/people/arleodordar/

The article discusses the future of FinTech and business borrowing, with a focus on predictive lending, autonomous banking, and one-click loans. The market for AI in the financial technology sector is expected to grow to $61.6 billion by 2032, with a growth rate of nearly 20% per year. Predictive lending uses AI to analyze and anticipate a business's needs based on cash flow, trends, and market conditions, while autonomous banking uses AI to manage every aspect of the lending process. One-click loans integrate AI financial technology into platforms that businesses already use, such as e-commerce platforms. However, there are challenges to be addressed, including data security, regulatory hurdles, and the potential for over-borrowing. To successfully integrate automated lending tools, companies should clearly define their end users' capital needs, evaluate the tool's data inputs and risk modeling logic, and pilot the tools in controlled stages. The future of business borrowing is expected to be transformed by predictive lending, autonomous banks, and one-click loans, allowing businesses to anticipate financial pressures and focus on growth and innovation.

April 29, 2025 • By Pepe Escobar

Enjoy in-depth, acute analysis of the most pressing local, regional and global trends at Sputnik! https://sputnikglobe.com/20250429/pepe-escobar-china-steps-up-its-game-in-the-global-ai-race-1121952…

April 29, 2025 • By Sehoon Park, Soomin Chung, Yisak Kim, Sun-Ah Yang, Soie Kwon, Jeong Min Cho, Min Jae Lee, Eunbyeol Cho, Jiwon Ryu, Sejoong Kim, Jeonghwan Lee, Hyung Jin Yoon, Edward Choi, Kwangsoo Kim, Hajeong Lee

A recent study developed a deep-learning-based model to predict postoperative acute kidney injury (PO-AKI) in non-cardiac major surgeries. The model integrates preoperative clinical characteristics with minute-scale intraoperative vital signs, outperforming conventional models. The study analyzed data from 110,696 patients across three hospitals, with 51,345 patients in the development cohort and 59,351 in the external validation cohorts. The model's performance was evaluated using the area under the receiver operating characteristic curve (AUROC), with results showing an AUROC of 0.795 in the discovery cohort and 0.762 and 0.786 in the validation cohorts. The study found that adding 11 key clinical variables, such as age, sex, and estimated glomerular filtration rate, improved the model's performance. The deep-learning-based model also demonstrated comparable predictive power to preoperative-only models. The study's results suggest that the integration of preoperative and intraoperative data can enhance predictive performance for PO-AKI risk prediction, providing a comprehensive approach to evaluating PO-AKI risk and offering better clinical decision-making. However, the study's limitations include its retrospective design and the exclusion of some potential AKI-associated variables. Overall, the study's findings have implications for the development of more accurate PO-AKI prediction models, which can help improve patient outcomes and reduce healthcare costs. The study's results also highlight the importance of considering both preoperative and intraoperative factors in PO-AKI risk prediction.

April 29, 2025 • By Jianqiang Sun, Sunao Ochi, Takehiko Yamanaka

A recent study analyzed Japan's nationwide crop disease and pest (CDP) survey data to evaluate the potential of leveraging historical data for forecasting CDP occurrences. The study found that a simple algorithm, averaging data from the past five years, outperformed more complex models. However, the prediction error remained substantial, with a root mean squared error (RMSE) of 6.2 ± 19.6. The study highlights the challenges of precise forecasting and the need for fundamental reforms in the survey methodology, including the integration of modern technologies such as IoT/ICT and artificial intelligence. The results suggest that the current survey system, which has been in place for over half a century, may not be effective in achieving its goals, and a more modern and efficient approach is needed to ensure food security.

April 29, 2025 • By Rui-Si Hu, Kui Gu, Muhammad Ehsan, Sayed Haidar Abbas Raza, Chun-Ren Wang

Researchers have developed a deep learning model called deepBCE-Parasite to predict linear B-cell epitopes (BCEs) in parasites. BCEs are crucial for developing vaccines, therapeutic antibodies, and diagnostic tools. The model uses a Transformer-based architecture and achieved an accuracy of 81% and an AUC of 0.90 in both 10-fold cross-validation and independent testing. The model was trained on a dataset of 5,752 positive and 5,752 negative BCEs and was compared to traditional machine learning models using 12 handcrafted features and four conventional machine learning algorithms. The results showed that deepBCE-Parasite outperformed the traditional models. As a case study, the model was applied to predict BCEs in the liver fluke Fasciola hepatica, a parasite that causes fascioliasis, a neglected tropical disease. The model predicted eight peptide sequences derived from the leucine aminopeptidase (LAP) protein, and dot-blot immunoassays confirmed that seven of these peptides exhibited specific binding to antibodies in F. hepatica-positive ovine sera. The study demonstrates the potential of AI in advancing epitope prediction in parasitology, providing a rapid, scalable, and cost-effective strategy for discovering immune targets. The deepBCE-Parasite model can be used to predict BCEs in other parasites, offering a valuable tool for developing vaccines, therapeutic antibodies, and diagnostic tools.

April 29, 2025 • By Magnus Ölander, Daniel Rea Vázquez, Karsten Meier, Aakriti Singh, Amanda Silva de Sousa, Fabiola Puértolas-Balint, Milica Milivojevic, Lieke Mooij, Johanna Fredlund, Eduard Calpe Bosch, María Rayón Díaz, Moa Lundgren, Karin van der Wal, Shaochun Zhu, André Mateus, Bjoern O. Schroeder, Jeremy R. Lohman, Barbara S. Sixt

Researchers have identified over 60 compounds that can prevent the growth of Chlamydia trachomatis, a bacterium that causes millions of infections worldwide. These compounds are chemically diverse, non-toxic to human cells, and highly potent in preventing the growth of Chlamydia in cell cultures. Some compounds blocked Chlamydia development reversibly, while others eradicated both established and persistent infections. The top molecules displayed broad activity against diverse Chlamydia strains and species. The most potent antichlamydial compound was found to inhibit fatty acid biosynthesis via covalent binding to the active site of Chlamydia FabH, highlighting a possible way to selectively treat Chlamydia infections. The study used a combination of experimental and virtual screening to identify these compounds, and the results provide a promising starting point for the development of new, selective therapies against Chlamydia.

April 29, 2025 • By Mariane Branco Alves, Rafael Santos Erbisti, Aline Araújo Nobre, Taynãna César Simões, Alessandre de Medeiros Tavares, Márcia Cristina Melo, Rodrigo Moreira Pedreira, Jan Pierre Martins de Araújo, Marilia Sá Carvalho, Nildimar Alves Honório

Researchers used Bayesian space-time models to analyze dengue case reports in Natal, Brazil, from 2015 to 2018. The study aimed to identify factors influencing dengue occurrence, predict case counts, and forecast dengue cases by neighborhood. The results showed that dengue risk increased with previous week's cases, Aedes egg positivity index, and mean daytime temperature. The study also found that sociosanitary factors, such as density of impoverished population, contributed to dengue risk. The researchers used a combination of entomological, climatic, and sociosanitary indicators to predict dengue cases and identified high-risk neighborhoods. The study's findings can inform targeted interventions and support operational control efforts to mitigate the incidence and impact of urban arboviruses like dengue, chikungunya, and Zika.

April 29, 2025 • By Mohammed Okmi, Tan Fong Ang, Muhammad Faiz Mohd Zaki, Chin Soon Ku, Koo Yuen Phan, Irfan Wahyudi, Lip Yee Por

A recent systematic literature review examined the use of mobile phone network data (MPND) in analyzing and managing the COVID-19 pandemic. The review identified 55 studies that utilized MPND, with 46 being quantitative and 9 qualitative. The studies were categorized into five main groups: monitoring and tracking human mobility patterns, investigating the correlation between mobility patterns and COVID-19 spread, analyzing economic recovery, assessing factors associated with non-pharmaceutical intervention (NPI) compliance, and investigating the impact of COVID-19 lockdowns on human behavior and economic activity. The review found that MPND played a crucial role in tracking human mobility patterns and monitoring the spread of COVID-19. However, the use of MPND also raised ethical and privacy concerns, highlighting the need for balanced approaches that address these concerns while maintaining the effectiveness of MPND in public health interventions. The study's findings suggest that NPI measures had a significant impact on reducing human movement and dynamics, but demographics, political party affiliation, socioeconomic inequality, and racial inequality affected population adherence to NPI measures. The review also emphasized the importance of developing new privacy-preserving methodologies for MPND analysis to ensure public trust and compliance with data protection regulations. Overall, the review highlights the potential of MPND in informing public health responses and policy decisions, while also emphasizing the need for careful consideration of ethical and privacy concerns. Future research should focus on developing innovative solutions that balance the benefits of MPND with the need to protect individual privacy and maintain public trust.

April 29, 2025 • By Verena Fischer, Anita Ignatius, Katharina Schmidt-Bleek, Georg Duda, Melanie Haffner-Luntzer

Researchers used artificial intelligence (AI) to analyze immune cell populations in the fracture hematoma and bone marrow of osteoporotic and non-osteoporotic mice. The AI-based clustering software, Cytolution, identified distinct subclusters of immune cells, including granulocytes, macrophages, B cells, and T cells. The study found that osteoporotic mice had an increased abundance of a specific B cell subpopulation and a reduced abundance of a particular granulocyte subpopulation in the early fracture hematoma. The AI-based approach provided a more nuanced understanding of immune cell phenotypes during bone regeneration, highlighting the potential of AI in identifying subtle differences in immune cell populations that may impact fracture healing. The study's findings suggest that AI-based clustering may be a powerful tool for analyzing flow cytometry data and understanding the complex interactions between immune cells and bone cells during fracture healing.

April 29, 2025 • By Ayaz Nanji

A recent survey of 150 US CEOs from companies with $500 million or more in annual revenue found that 71% would give their CMO's performance a grade of "A" or "B". 24% would give an "A" grade, while 47% would give a "B" grade. However, CEOs are less likely to give high grades for integrating AI/machine-learning and driving company growth. Half of CEOs believe their CMO plays it safe, while 39% think they play big and inspire others. 51% of CEOs consider their CMO central to the company's growth strategy, while 49% see them as peripheral.

April 29, 2025 • By Musab Wedyan, Yu-Chen Yeh, Fatemeh Saeidi-Rizi, Tai-Quan Peng, Chun-Yen Chang

A recent study compared the evaluations of visual urban scenes by human participants and a large language model (LLM) called GPT-4o, focusing on the context of urban walkability. The research involved 174 human participants and GPT-4o evaluating street-level images based on key dimensions of walkability, including overall walkability, feasibility, accessibility, safety, comfort, and liveliness. The findings revealed that GPT-4o and human participants aligned in their evaluations of overall walkability, feasibility, accessibility, and safety. However, notable differences emerged in the assessment of comfort and liveliness, with human participants demonstrating broader thematic diversity and addressing a wider range of topics. The study concludes that human input remains essential for fully capturing human-centered evaluations of walkability and underscores the importance of refining LLMs to better align with human perceptions in future walkability studies. The research highlights the limitations of GPT-4o in accurately perceiving urban environments and points to opportunities for refining LLM models to better align with human perspectives. The study's results have implications for urban planning and design, emphasizing the need to consider human-centered perceptions and subjective experiences in creating more pedestrian-friendly environments. By combining the strengths of LLMs with human input, researchers and urban planners can develop more effective and inclusive approaches to assessing and improving walkability in urban areas. Overall, the study contributes to the growing body of research on the application of LLMs in urban studies, highlighting both the potential and limitations of these models in capturing human perceptions and experiences of urban environments.