July 9, 2023

July 8, 2023

July 10, 2023

  • Detection

    It is too easy to falsely accuse a student of using AI: a cautionary tale

    Times Higher Education, 07/09/23. Emily’s experience of being falsely accused of using AI to generate her essay highlights the need for fair procedures and robust evidence in university assessments. Her anxiety and depression worsened as she faced the prospect of a black mark on her record and the impact on her future career. With the help of a lawyer experienced in university proceedings, Emily challenged the accusation and provided evidence that she had written the essay herself. Eventually, the adjudication committee dropped the concerns, but Emily’s parents had to bear the cost of her defense. This case emphasizes the importance of universities reevaluating their approach to prevent unjust outcomes for students. READ THE ARTICLE

  • Adversarial AI

    How Adversarial AI threatens your Machine Learning Models

    IT Brief, 07/09/23. The rise of Artificial Intelligence (AI) has revolutionized various sectors of society. However, with its promising capabilities also come significant challenges, particularly in terms of cybersecurity. Adversarial AI, which involves deceiving AI systems to make erroneous decisions, has become alarmingly prevalent. Unfortunately, most organizations are ill-prepared to defend against adversarial AI attacks, leaving them vulnerable to significant damage to their reputation and revenue. Current cybersecurity solutions have gaps that fail to address AI-specific considerations, leaving AI/ML services susceptible to attacks. To overcome these challenges, it is crucial to weave security into AI systems by incorporating bias identification, malicious input identification, ML forensic capabilities, and sensitive data protection. At Mantel Group, we understand the nuances of AI and cybersecurity, and we can assist organizations in building robust defenses against adversarial AI attacks. READ THE ARTICLE

  • Enterprise, Hype

    Enterprise-grade generative AI automation: Unpacking the hype, understanding the challenges

    SiliconAngle, 07/09/23. The recent excitement surrounding generative artificial intelligence is justified, as it opens up new possibilities for human-computer interaction. However, it is important to avoid treating generative AI as a magic wand, especially in enterprise-grade automation. While generative AI can generate impressive outputs, it is crucial to consider security, governance, and compliance concerns in complex enterprise systems. Generative AI works best when combined with domain models and better data. By leveraging both old and new approaches, we can achieve more informed and intelligent solutions. Ultimately, the focus should be on reducing the cost of real-time, contextualized data for application developers, enabling them to rapidly create data-driven applications. READ THE ARTICLE

  • Healthcare

    GE Healthcare survey: AI faces skepticism in the medical care business

    Yahoo, 07/09/23. The healthcare industry remains skeptical about the readiness of artificial intelligence (AI) for medical use, according to a recent GE HealthCare survey. Over half of medical professionals surveyed believe that AI is not yet prepared for medical applications, with only 42% globally and 26% in the US believing that AI can be trusted. Concerns include potential patient harm, data privacy and security issues. To address these concerns, it is important to involve clinicians in the design process and provide adequate training. Despite these challenges, there are already successful applications of AI in areas such as medical imaging and health data analysis. The adoption of AI in healthcare holds promise but will require building trust and addressing regulatory frameworks. READ THE ARTICLE

  • Legal

    Authors sue ChatGPT creators for using their books to train AI without consent

    Dexerto, 07/09/23. Mona Awad and Paul Tremblay have filed a lawsuit against OpenAI, alleging copyright infringement. The authors claim that OpenAI used their books to train their AI chatbot, ChatGPT, without their consent. They argue that the AI’s ability to generate summaries of their works suggests that it was trained on their copyrighted materials. As ChatGPT is a software product sold by OpenAI, the authors assert that the company is profiting from their copyrighted materials without permission. This lawsuit brings attention to the concerns of creatives regarding the use of AI tools and the need to respect copyright laws. READ THE ARTICLE

  • Blockchain

    Crypto And AI Innovation: The London Attraction

    Forbes, 07/09/23. London is positioning itself as a global hub for cryptocurrency and AI innovation. It offers a robust network of regulatory support, institutional backing, business-friendly policies, and a supportive community, making it an appealing destination for crypto companies. London’s policymakers actively work with the industry to shape its future, creating an environment where startups can thrive and transition from centralized to decentralized networks. The city also benefits from a rich academic network, attracting talented individuals who drive innovation. With endorsements from influential venture capital firms and visionary leaders, London has the infrastructure, resources, and commitment to lead the world in shaping the future of crypto. However, striking a balance between innovation and consumer protection remains a challenge. The UK’s aspiration to solidify itself as a crypto hub could stimulate economic growth and foster innovation, making it an attractive option for crypto ventures. Ultimately, London’s balanced approach to regulation and its fusion of legacy and innovation position it as a driving force in transforming the world of finance and technology. READ THE ARTICLE

  • Use Case

    Unearthing Rare Earth Elements – Scientists Use AI To Find Rare Materials

    SciTechDaily, 07/09/23. A new machine-learning model developed by Shaunna Morrison and Anirudh Prabhu has the ability to predict the locations of minerals on Earth and potentially other planets. By analyzing patterns in mineral associations, this model provides valuable insights for scientists and industries involved in exploring mineral deposits. This advancement is crucial for understanding the planet’s history and extracting resources for practical applications such as rechargeable batteries. The model utilizes data from the Mineral Evolution Database and has successfully predicted the occurrence of previously unknown minerals. Mineral association analysis is a powerful tool for various scientific disciplines. READ THE ARTICLE