Volume 3 (2024)

Each volume of Journal of AI, Robotics and Workplace Automation consists of four 100-page issues published both in print and online.

Articles published in Volume 3 include:

  • Editorial
    Andreas Welsch, Chief AI Strategist, Intelligence Briefing and Editorial Board member, Journal of AI, Robotics & Workplace Automation
  • Generative AI Papers
    Enabling generative AI through use cases in a big enterprise
    Enrique Mora and Luca Dell’Orletta, Nestlé

    In the emerging field of generative artificial intelligence (GenAI), we possess the potential to significantly enhance our business operations and processes. Achieving this goal within a large corporation like Nestlé is challenging, however, given the immature stage of this technology. This paper outlines the approach to implementing GenAI at Nestlé, guided by the most influential use cases. It also underscores the importance of scaling people’s capabilities and establishing legal, ethical and compliance frameworks to support the deployment of this technology.
    Keywords: AI; generative AI; GenAI; LLM; enterprise

  • Building resilient SMEs: Harnessing large language models for cyber security in Australia
    Ben Kereopa-Yorke, Telco

    The escalating digitalisation of our lives and enterprises has led to a parallel growth in the complexity and frequency of cyberattacks. Small and medium-sized enterprises (SMEs), particularly in Australia, are experiencing increased vulnerability to cyber threats, posing a significant challenge to the nation’s cyber security landscape. Embracing transformative technologies such as artificial intelligence (AI), machine learning (ML) and large language models (LLMs) can potentially strengthen cyber security policies for Australian SMEs. Their practical application, advantages and limitations remain underexplored, however, with prior research mainly focusing on large corporations. This study aims to address this gap by providing a comprehensive understanding of the potential role of LLMs in enhancing cyber security policies for Australian SMEs. Employing a mixed-methods study design, this research includes a literature review, qualitative analysis of SME case studies and a quantitative assessment of LLM performance metrics in cyber security applications. The findings highlight the promising potential of LLMs across various performance criteria, including relevance, accuracy and applicability, although gaps remain in areas such as completeness and clarity. The study underlines the importance of integrating human expertise with LLM technology and refining model development to address these limitations. By proposing a robust conceptual framework guiding the effective adoption of LLMs, this research aims to contribute to a safer and more resilient cyber environment for Australian SMEs, enabling sustainable growth and competitiveness in the digital era.
    Keywords: cyber security; artificial intelligence; AI; large language models; LLM; AI in cyber security; technological innovation

  • Measuring the business value of generative AI
    Jim Sterne, Target Marketing of Santa Barbara

    Generative artificial intelligence (GenAI) can deliver tangible and intangible values that can be calculated to decide which projects benefit from GenAI and which do not. This paper is intended to be a guide for businesses just starting to build traction for their ideas. The focus is on evaluating and leveraging GenAI’s potential to innovate faster and compete effectively in a rapidly evolving digital economy. The paper specifies the many ways GenAI can have an impact on a business and considers how to measure that impact. It starts with standard business metrics (revenue, profit, customer satisfaction, etc.) and then turns to the more esoteric task of measuring the impact on creativity, inspiration and innovation, followed by business disruption and process metrics. It finishes with a look at improving process improvement.
    Keywords: generative AI; business metrics; economic impact; innovation; digital transformation.

  • Machine unlearning for generative AI
    Yashaswini Viswanath, Resident Researcher, Business School of AI, et al

    This paper introduces a new field of AI research called machine unlearning and examines the challenges and approaches to extend machine unlearning to generative AI (GenAI). Machine unlearning is a model-driven approach to make an existing artificial intelligence (AI) model unlearn a set of data from its learning. Machine unlearning is becoming important for businesses to comply with privacy laws such as General Data Protection Regulation (GDPR) customer’s right to be forgotten, to manage security and to remove bias that AI models learn from their training data, as it is expensive to retrain and deploy the models without the bias or security or privacy compromising data. This paper presents the state of the art in machine unlearning approaches such as exact unlearning, approximate unlearning, zero-shot learning (ZSL) and fast and efficient unlearning. The paper highlights the challenges in applying machine learning to GenAI which is built on a transformer architecture of neural networks and adds more opaqueness to how large language models (LLM) learn in pre-training, fine-turning, transfer learning to more languages and in inference. The paper elaborates on how models retain the learning in a neural network to guide the various machine unlearning approaches for GenAI that the authors hope can be built upon their work. The paper suggests possible futuristic directions of research to create transparency in LLM and particularly looks at hallucinations in LLMs when they are extended to do machine translation for new languages beyond their training with ZSL to shed light on how the model stores its learning of newer languages in its memory and how it draws upon it during inference in GenAI applications. Finally, the paper calls for collaborations for future research in machine unlearning for GenAI, particularly LLMs, to add transparency and inclusivity to language AI.
    Keywords: machine unlearning; privacy; right to be forgotten; generative AI; fine-tuning; large language models; LLM; zero shot learning; explainability

  • Business value of generative AI use cases
    Fuad Hendricks, Accenture

    This paper discusses the significant impact of artificial intelligence (AI), specifically generative AI (GenAI), on various industries and business processes. It highlights the rapid adoption of AI technologies and their profound influence on global business operations. The paper shares views on the substantial growth rate for the technology, with many businesses already in piloting phases or production stages. It emphasises the transformative role of generative AI in creating new services and products, necessitating changes in operating models, technology stacks and workforce skills. The paper also touches on various industries affected by AI, the potential for automation and augmentation and the strategic planning required for businesses to effectively implement and benefit from these technologies. Additionally, it discusses the importance of responsible AI, addressing risks such as bias and privacy, and complying with emerging regulations. The paper highlights the crucial role of AI in modernising business practices and creating competitive advantages in various sectors.
    Keywords: Generative AI; data strategy; business transformation; efficiency; AI at scale; responsible AI; regulation

  • Minimise model risk management oversight for cyber security solutions
    Liming Brotcke, Ally Financial

    Adoption of artificial intelligence (AI) and machine learning (ML)-powered cyber security tools and models by financial institutions has received considerable attention in the model risk management community. In parallel, developing trustworthy AI that is more explainable, fair, robust, private and transparent has also received considerable research and regulatory attention. Appropriate governing of cyber security models is inevitable. The prevailing thought at present is to have the model risk management function to oversee the development, implementation and use of such cyber security tools and models. This study first demonstrates two primary challenges of executing this oversight and then offers a few practical suggestions to ensure a reasonable application.
    Keywords: cyber security; data science; model validation; machine learning

  • Analysis of ChatGPT and the future of artificial intelligence: Its effect on teaching and learning
    Gunja Kumari Sah, Dipak Kumar Gupta, Tribhuvan University, and Amar Prasad Yadav,Tribhuvan University and Rajarshi Janak University

    This research paper aims to analyse the use of ChatGPT and its future in the teaching and learning process. The research was based on a descriptive research design and model that examined the use of ChatGPT among university students and leveraged the technology acceptance model (TAM). Samples were obtained from Nepal’s oldest universities, Tribhuvan University, Purwanchal University, Kathmandu University and Pokhara University, by purposive sampling techniques. Data was directly observed by field and Internet surveys at sample universities by structured questionnaires. The research included respondents based on their use of ChatGPT. Information related to ChatGPT was collected from 400 respondents from the science and management faculty. This application was only used by 280 responders, however; therefore, these 280 replies were examined to reach a decision. Data was entered into the SPSS and AMOS software for structural equation modelling. The measurement and structural model was found to be reliable and valid. According to the study, 54.3 per cent of users felt ChatGPT was crucial and 53.9 per cent of users believed it provided reliable information; however, 47.1 per cent of respondents said it is neither secure nor unsafe and 71.4 per cent said ChatGPT has an advantageous effect on users’ performance. ChatGPT is used by 42.9 per cent of users for help, mostly to find answers to questions. The environment and conduct of users of facilities are significantly influenced by performance expectations. The relationship between the performance expectations of teachers and students toward ChatGPT and their behaviour is also mediated by the facility circumstances.
    Keywords: artificial intelligence; AI; generative pre-trained transformer; GPT; ChatGPT; learning; technology acceptance model; TAM; teaching

  • Automation and AI in the workplace: The future of work is more complex than ever
    Chelsea Perino, The Executive Centre

    Artificial intelligence (AI) and automation are revolutionising the workplace, transforming the way businesses operate, how people interact and having an impact on the future of work. These technologies have the potential to enhance productivity, address societal challenges and contribute to economic growth. With machines becoming increasingly capable, they can now perform tasks previously done by humans, complement human work and even surpass human capabilities in certain areas. As a result, the nature of work is changing, with some occupations declining, others growing and many more undergoing significant transformations. This paper explores the promise and challenges of AI automation in the workplace, highlighting key workforce transitions and presenting several critical issues that need to be solved.
    Keywords: AI; automation; future of work; coworking; real estate; technology; flexible workspace; hybrid work

  • Practice Paper
    What is in the black box: The ethical implications of algorithms and transparency in the age of the GDPR
    Senna Mougdir, Dutch Data Protection Authority

    Algorithms silently construct our lives. They can determine whether someone is hired or promoted, provide loans or housing, and decide which political advertisements and news articles consumers see. The General Data Protection Regulation (GDPR) which came into force on 25th May, 2018 regulates the protection of personal data of individuals within the European Union (EU) member states. While innovation is important to our society, it is also important that organisations using artificial intelligence technologies and big data comply with the GDPR to ensure our privacy and data is protected. The ethical implications and responsibility of the algorithm in these important decisions are, however, unclear. Algorithms are important participants in ethical decision making and influence the delegation of roles and responsibilities in these decisions. This paper focuses on the bias in algorithms and determines whether developers are responsible for the algorithms they use in the future, what these organisations are responsible for, and the normative basis of this responsibility. Finally, it will give recommendations that could be useful for organisations using algorithms in automated decision making.
    Keywords: General Data Protection Regulation (GDPR); artificial intelligence (AI); algorithms; machine learning (ML); big data; ethics