Generation AI Research Articles

Calculation of virtual 3D subtraction angiographies using conditional generative adversarial networks (cGANs)

ObjectiveSubtraction angiographies are calculated using a native and a contrast-enhanced 3D angiography images. This minimizes both bone and metal artifacts and results in a pure image of the vessels. However, carrying out the examination twice means double the radiation dose for the patient. With the help of generative AI, it could be possible to simulate subtraction angiographies from contrast-enhanced 3D angiographies and thus reduce the need for another dose of radiation without a cutback in quality. We implemented this concept by using conditional generative adversarial networks.MethodsWe selected all 3D subtraction angiographies from our PACS system, which had performed between 01/01/2018 and 12/31/2022 and randomly divided them into training, validation, and test sets (66%:17%:17%). We adapted the pix2pix framework to work on 3D data and trained a conditional generative adversarial network with 621 data sets. Additionally, we used 158 data sets for validation and 164 for testing. We evaluated two test sets with (n = 72) and without artifacts (n = 92). Five (blinded) neuroradiologists compared these datasets with the original subtraction dataset. They assessed similarity, subjective image quality, and severity of artifacts.ResultsImage quality and subjective diagnostic accuracy of the virtual subtraction angiographies revealed no significant differences compared to the original 3D angiographies. While bone and movement artifact level were reduced, artifact level caused by metal implants differed from case to case between both angiographies without one group being significant superior to the other.ConclusionConditional generative adversarial networks can be used to simulate subtraction angiographies in clinical practice, however, new artifacts can also appear as a result of this technology.

A Study of Undergraduates’ Behavioral Intention to Use Generative AI Tools—Ernie Bot as An Example

With the rapid advancement of artificial intelligence technology, its influence on various sectors is increasing. Ernie Bot, a leading tool in generative AI, has garnered significant attention and application. This study utilizes the Unified Theory of Acceptance and Use of Technology (UTAUT) and introduces the concepts of Perceived Risk and Information Quality to construct a model that examines the factors influencing undergraduates’ behavioral intention to use generative AI tools like Ernie Bot. Using questionnaire data collected from 175 undergraduates, a structural equation model was employed to analyze the relationships between variables. The findings indicate that Social Influence, Information Quality, and Performance Expectancy positively impact the Behavioral Intention to use Ernie Bot. However, Effort Expectancy was not found to be a significant determinant of Behavioral Intention among undergraduates. Additionally, the study revealed that Perceived Risk did not negatively affect Behavioral Intention. These results offer valuable insights for the design and development of generative AI products, suggesting that enhancing Information Quality and leveraging Social Influence could improve user acceptance.

Teacher Use of Generative AI for Read‐Aloud Question Prompts

AbstractThe advent of Generative AI technologies, such as ChatGPT, in November 2022, necessitated immediate and critical attention from the educational research community. The impact of GenAI in education, though not yet clear, has the potential to be transformative. More specifically, the focus of this paper is on how to integrate GenAI into elementary literacy education. We, as teacher educators, aim to showcase how to prompt ChatGPT to generate high‐quality questions during a read‐aloud. We discuss the easy access teachers have to GenAI tools and stress the pivotal role they have in decision‐making. We encourage teachers to explore, learn, and understand how to work with GenAI tools to get the most out of it and thus facilitate their agency, teaching, and learning.

ChatGPT as Teacher Assistant for Physics Teaching

This study explores the integration of ChatGPT as a teaching assistant in physics education, emphasizing its potential to transform traditional pedagogical approaches. ChatGPT facilitates interactive and inquiry-based learning grounded in constructivist learning theory, allowing students to engage actively in experiments and better grasp abstract concepts through hands-on activities. The AI's adaptive dialogue systems promote socio-constructivist learning by encouraging social interaction and personalized feedback, which is essential for addressing individual learning gaps and enhancing student engagement. ChatGPT's ability to simulate real-world physics problems and provide immediate feedback fosters experiential learning, making complex concepts more accessible and promoting critical thinking skills. By offering tailored interventions and adapting to individual learning paces, ChatGPT supports a personalized educational experience that caters to the unique needs of each student. This adaptability is particularly beneficial in physics education, where students often struggle with abstract concepts and require immediate clarification to progress effectively. The paper concludes that the synthesis of generative AI and pedagogy has the potential to reshape science education, fostering deeper understanding and curiosity among students. By leveraging innovative methodologies and augmented data, ChatGPT enriches teacher-student interactions, creating a comprehensive educational experience that promotes a culture of curiosity and exploration, thereby nurturing future scientists and engineers. Ultimately, the integration of ChatGPT into physics education offers a valuable opportunity to enhance student engagement and understanding of scientific concepts through interactive and personalized support.

If cheating is optimisation then assessment must not be pure

The emergence of generative AI has broadened the question of how to ensure academic integrity. Where, in the past, many tools sought to counter specific threats to integrity (for example, detecting plagiarism or outsourcing), we now need to consider integrity more fundamentally: What work can we be assured the student did? One way of viewing cheating is from the perspective of optimisation: producing a result without performing the work. This perspective creates an interesting parallel with computer programming. A programmer often wants their compiler to optimise their program, altering its instructions to make it run more efficiently, without affecting its observable behaviour. Some characteristics of programs make this easier or harder to do. For example, a function is said to be “pure” if it always produces the same outputs for the same inputs and has no other effects. If the result is known, then performing the function can be invisibly replaced by inserting its result. If, however, a function has an observable effect, then the work of performing the function cannot be optimised away. This paper explores academic integrity assurance through the lens of “effect types”. In functional languages, functions are often written in such a manner that the types of effects they have are explicit and tracked by the compiler. The paper likens this to tracking the types of academic integrity measures that are present in each assessment, how they compose across a course, and therefore what observable effects are present to assure each learning outcome.

Technology-agnostic framework for designing assessments in the era of artificial intelligence

There is an urgent need to understand and benefit from artificial intelligence within schools. However, government policies that focus on academic integrity and duty of care do not address how students can leverage AI to enhance their learning nor how teachers can intentionally design assessments to account for AI. The Australian Framework for Generative Artificial Intelligence in Schools suggests that assessments need to clearly state “how generative AI tools should or should not be used” while also permitting a “clear and unbiased evaluation of student ability”. This is a worthy aspiration, yet there are presently few tools and examples to guide teachers in creating assessments. This conceptual paper draws from the field of design to articulate a framework for developing assessments that focuses on AI dialogue and trace-augmented critical reflection.

Investigating translation for Indic languages with BLOOMZ-3b through prompting and LoRA fine-tuning

In the domain of natural language processing, the rise of Large Language Models and Generative AI represents a noteworthy transition, enabling machines to understand and generate text resembling that produced by humans. This research conducts a thorough examination of this transformative technology, with a focus on its influence on machine translation. The study explores the translation landscape between English and Indic languages, which include Hindi, Kannada, Malayalam, Tamil, and Telugu. To address this, the Large Language Model, BLOOMZ-3b, is utilized, which has been primarily developed for a text generation task. Multiple prompting engineering techniques for machine translation are prominently explored. The study further traverse fine-tuning the BLOOMZ-3b model using a Parameter Efficient Fine-Tuning technique called Low Rank Adaptation, aiming to reduce computational complexity. Hence, by combining innovative prompting approaches using BLOOMZ-3b model and fine-tuning the model, it contributes to continuous development of machine translation technologies beyond traditional borders of what can be done with respect to language processing. In this regard, not only does this research shed light on the intricacy of translation problems but it also sets a precedence for optimizing or adapting big language models to various languages which end up advancing Artificial Intelligence and Natural Language Processing at large.

Let Us Not Squander the Affordances of LLMs for the Sake of Expedience: Using Retrieval Augmented Generative AI Chatbots to Support and Evaluate Student Reasoning

Let Us Not Squander the Affordances of LLMs for the Sake of Expedience: Using Retrieval Augmented Generative AI Chatbots to Support and Evaluate Student Reasoning

Can AI Get a Degree in Geoscience? Performance Analysis of a GPT-Based Artificial Intelligence System Trained for Earth Science (GeologyOracle)

Generative artificial intelligence systems based on the GPT model have shown groundbreaking capabilities in producing scientific texts, revolutionising how geoscientists research and teach. However, there is a noticeable absence of generative AI systems tailored specifically for geoscience, and the efficacy of GPT technology has not been examined within the Earth Science realm. To fill these gaps, we propose a new artificial intelligence system (GeologyOracle) built upon the GPT-4o model and trained on Earth Science data. It is designed to simulate a conversation with a geoscientist, having the capabilities to analyse geologic datasets, suggest new geoscience hypotheses, explain Earth-Science concepts, and interpret geosites. To evaluate the performance of GeologyOracle, the system was tested with 152 geoscience questions provided and evaluated by a panel of geoscience academics on a scale of 0–10. Performance analysis revealed that 79.6% of the answers scored equal to or above the passing mark of 5, with a tendency to deliver high-quality responses (mean: 6.5; median = 7; interquartile range: 5–8). The results demonstrate that GeologyOracle is effective in performing complex geoscience tasks, such as identifying rocks, fossils, and minerals, and interpreting outcrop, core data, and quantitative datasets. The new AI system has exhibited noteworthy ability in the interpretation of internationally-acknowledged geosites situated within geoparks and nature reserves. The performance of the AI system is comparable to that of trained geoscience specialists, suggesting its potential as an assistant in various geoscience fields, including structural geology, palaeontology, geomorphology, sedimentology, and economic geology. While AI does not render geoscientists obsolete, it offers significant potential for accelerating scientific discovery, automating geoscience research, and assisting educators, students, and geotourists.

Generative AI is already widespread in the public sector: evidence from a survey of UK public sector professionals

Generative AI has the potential to transform how public services are delivered by enhancing productivity and reducing time spent on bureaucracy. But to what extent is the technology already in use? Our survey of UK public service professionals (in education, health, social work, and emergency services) seeks to answer this question. We find that use of generative AI is widespread: 45% of respondents were aware of colleagues using generative AI, while 22% use it themselves. Respondents were positive about its potential to enhance their efficiency and reduce their bureaucratic workload. For example, those working in the health service thought that time spent on bureaucracy could drop by the equivalent of one day per week, an enormous potential impact. Our survey also found a high amount of trust (61%) in generative AI and a low fear of replacement (16%). However, only a minority of respondents (32%) felt like there was clear guidance on generative AI usage in their workplaces. In other words, it is clear that generative AI is already coming into the public sector, but uptake is happening in a disorganised fashion without clear guidelines. The UK’s public sector urgently needs to develop more systematic methods for taking advantage of the technology.

"I'm a Solo Developer but AI is My New Ill-Informed Co-Worker": Envisioning and Designing Generative AI to Support Indie Game Development

Indie game developers are often defined as game developers who are typically not employed by or affiliated with tech giants or large gaming companies/publishers. Although people may decide to "go indie" for various purposes, indie game development has become a crucial part of the global gaming culture. However, this community is now facing unprecedented tensions as generative AI technologies are shifting how games can be designed, produced, and experienced. Through a qualitative analysis of 3,091 online posts and comments from subreddits and Facebook groups for indie game developers, we offer an in-depth investigation of how indie game developers perceive and envision the multifaceted role of generative AI in their creative practices. Our empirical investigation reveals that generative AI both promotes and harms indie game developers' endeavors to innovate the traditional game production model, which further influences the nature and workflow of creativity in game development. We also propose three principles for designing future generative AI technologies to improve indie developers' work while mitigating potential risks, harm, and negative impacts of AI. We hope that this study can help design and develop future generative AI technologies to foster and sustain more democratic and inclusive practices in game development rather than replacing human creators.

Ironies of Generative AI: Understanding and Mitigating Productivity Loss in Human-AI Interaction

Generative AI (GenAI) systems offer opportunities to increase user productivity in many tasks, such as programming and writing. However, while they boost productivity in some studies, many others show that users are working ineffectively with GenAI systems and losing productivity. Despite the apparent novelty of these usability challenges, these ‘ironies of automation’ have been observed for over three decades in Human Factors research on the introduction of automation in domains such as aviation, automated driving, and intelligence. We draw on this extensive research alongside recent GenAI user studies to outline four key reasons for productivity loss with GenAI systems: a shift in users’ roles from production to evaluation, unhelpful restructuring of workflows, interruptions, and a tendency for automation to make easy tasks easier and hard tasks harder. We then suggest how Human Factors research can also inform GenAI system design to mitigate productivity loss by using approaches such as continuous feedback, system personalization, ecological interface design, task stabilization, and clear task allocation. Thus, we ground developments in GenAI system usability in decades of Human Factors research, ensuring that the design of human-AI interactions in this rapidly moving field learns from history instead of repeating it.

Responsible Crowdsourcing for Responsible Generative AI: Engaging Crowds in AI Auditing and Evaluation

With the rise of generative AI (GenAI), there has been an increased need for participation by large and diverse user bases in AI evaluation and auditing. GenAI developers are increasingly adopting crowdsourcing approaches to test and audit their AI products and services. However, it remains an open question how to design and deploy responsible and effective crowdsourcing pipelines for AI auditing and evaluation. This workshop aims to take a step towards bridging this gap. Our interdisciplinary team of organizers will work with workshop participants to explore several key questions, such as how to improve the output quality and workers' productivity for GenAI evaluation crowdsourcing tasks compared to discriminative AI systems, how to guide crowds in auditing problematic AI-generated content while managing their psychological impact, ensuring marginalized voices are heard, and setting up responsible and effective crowdsourcing pipelines for real-world GenAI evaluation. We hope this workshop will produce a research agenda and best practices for designing responsible crowd-based approaches to AI auditing and evaluation.

Sketchar: Supporting Character Design and Illustration Prototyping Using Generative AI

Character design in games involves interdisciplinary collaborations, typically between designers who create the narrative content, and illustrators who realize the design vision. However, traditional workflows face challenges in communication due to the differing backgrounds of illustrators and designers, the latter with limited artistic abilities. To overcome these challenges, we created Sketchar, a Generative AI (GenAI) tool that allows designers to prototype game characters and generate images based on conceptual input, providing visual outcomes that can give immediate feedback and enhance communication with illustrators' next step in the design cycle. We conducted a mixed-method study to evaluate the interaction between game designers and Sketchar. We showed that the reference images generated in co-creating with Sketchar fostered refinement of design details and can be incorporated into real-world workflows. Moreover, designers without artistic backgrounds found the Sketchar workflow to be more expressive and worthwhile. This research demonstrates the potential of GenAI in enhancing interdisciplinary collaboration in the game industry, enabling designers to interact beyond their own limited expertise.

Human-AI Collaboration in Cooperative Games: A Study of Playing Codenames with an LLM Assistant

Playing partial information, restricted communication, cooperative (PIRCC) games with humans have proven challenging for AI, due to our reliance on social dynamics and sophisticated cognitive techniques. Yet, recent advances in generative AI may change this situation through new forms of human-AI collaboration. This paper investigates how teams of players interact with an AI assistant in the PIRCC game Codenames and the impact this has on cognition, social dynamics, and player experience. We observed gameplay and conducted post-game focus groups with 54 participants across ten groups. Each group played three rounds of Codenames, with an AI assistant supporting Cluegivers. We found the AI assistant enhanced players' convergent and divergent thinking, but interfered with formation of team mental models, highlighting a tension in the use of AI in creative team scenarios. The presence of the AI challenged many players' understanding of the 'spirit of the game'. Furthermore, the presence of the AI assistants weakened social connections between human teammates, but strengthened connections across teams. This paper provides an empirical account of an AI assistant's effect on cognition, social dynamics, and player experience in Codenames. We highlight the opportunities and challenges that arise when designing hybrid digital boardgames that include AI assistants.

Addressing student non-compliance in AI use declarations: implications for academic integrity and assessment in higher education

This study examines the factors driving student non-compliance with AI use declarations in academic assessments at King’s Business School, where 74% of students failed to declare AI usage despite declaration being a requirement of a mandatory coursework coversheet. Utilising the Theory of Planned Behaviour (TPB) as a framework, the research combines service evaluation survey data and semi-structured interviews to explore how attitudes, subjective norms, and perceived behavioural control influence compliance. Findings reveal that fear of academic repercussions, ambiguous guidelines, inconsistent enforcement, and peer influence are key barriers to AI use declaration. These factors complicate the declaration process, undermine transparency, and challenge academic integrity. The study extends the TPB model by highlighting the ethical and practical dilemmas posed by generative AI, which blur traditional norms of academic integrity. This research offers critical insights for policymakers, suggesting that clear, consistent, and trust-based policies are crucial in fostering ethical AI use. The findings underscore the importance of transparent communication and supportive institutional cultures to improve compliance. Ultimately, this study informs policy development by evaluating the effectiveness of declaration mechanisms and providing actionable recommendations to promote a culture of academic integrity in the evolving landscape of AI technologies.

Updating a Code for Teaching Ethical Visualizations

The prevalent paradigm for understanding what constitutes a ‘good’ data visualization, and what we are likely teaching business students, relates to a conventional wisdom of efficiency, clarity, transparency, and faithful representation of truth. Teaching about the ethics of visualizations seems to be largely absent from business school curricula. This paper suggests that business students who make and use visualizations should be taught to use a code of ethics for visualizations based not only on conventional wisdom, but also on ideas from the Machine Learning/AI and ‘ethical visualizations’ communities. Elaborating on Sheppard’s code, the code suggested here incorporates ideas about bias, transparency, user agency, and privacy. Ethical awareness and practice become even more important as more data collection, processing, and visualization is shifted to Machine Learning and Generative AI. A threshold model is adapted to understand when visualization decisions require more scrutiny. Ideas from the code and the threshold model are used to analyze choices made by the CDC’s COVID-19 Rapid Response Team for an early pandemic visualization that circulated widely beyond the initial report in which it appeared. This and other examples underscore the need for business professionals to consider their visualizations from the perspective of the code.

EPortfolio traditions that are ready to be broken and new ones ready to be established

Although ePortfolio practices have been with us for a good 20+ years, in some form or another, the options that students and staff now have to represent their professional experiences have expanded over recent years. This is largely due to the plethora of online platforms that allow a user-friendly and templated experience, giving students freedoms they have not previously had, unless they had website development skills. This paper will report on a study of current ePortfolio platform options provided by institutions in Australasia, highlighting a divergence in, and opinions around, what constitutes contemporary good practice. It is fair to say that what was once seen as the responsibility of an institution to provide a dedicated ePortfolio platform for their students has started to fracture and that many institutions are now opting to mediate third-party and/or discipline-relevant options for their students. This paper will look at the pros and cons of these options, but also some emerging trends brought on by the advent of Generative AI, that institutions can now take advantage of.

Effectiveness of Generative AI for Post-Earthquake Damage Assessment

After an earthquake, rapid assessment of building damage is crucial for emergency response, reconstruction planning, and public safety. This study evaluates the performance of various Generative Artificial Intelligence (GAI) models in analyzing post-earthquake images to classify structural damage according to the EMS-98 scale, ranging from minor damage to total destruction. Correct classification rates for masonry buildings varied from 28.6% to 64.3%, with mean damage grade errors between 0.50 and 0.79, while for reinforced concrete buildings, rates ranged from 37.5% to 75.0%, with errors between 0.50 and 0.88. Fine-tuning these models could substantially improve accuracy. The practical implications are significant: integrating accurate GAI models into disaster response protocols can drastically reduce the time and resources required for damage assessment compared to traditional methods. This acceleration enables emergency services to make faster, data-driven decisions, optimize resource allocation, and potentially save lives. Furthermore, the widespread adoption of GAI models can enhance resilience planning by providing valuable data for future infrastructure improvements. The results of this work demonstrate the promise of GAI models for rapid, automated, and precise damage evaluation, underscoring their potential as invaluable tools for engineers, policymakers, and emergency responders in post-earthquake scenarios.

Using Generative AI to Promote Psychological, Feedback, and Artificial Intelligence Literacies in Undergraduate Psychology

Background With the arrival of generative artificial intelligence (genAI) tools, psychology educators are rethinking their assessment practices. Objective This paper describes one approach to integrating genAI into an assessment designed to promote psychological literacy. Method Students used ChatGPT to generate a media release about a published article and then wrote a critique. We evaluated whether students were able to use the marking rubric to assess the ChatGPT output, and whether working with the rubric early in the assessment process had benefits for their grades on subsequent tasks. Results The results show that students accurately assessed the ChatGPT output against the marking rubric, judging the output to be stylistically good but lacking in accurate coverage of the aims, methods, and results of the research. Working with genAI and the marking rubric early in the assessment process had benefits for performance, relative to cohorts that had engaged in peer review. Conclusion By allowing students to use genAI and scaffolding the process of critiquing and revising, students gained competencies in psychological, feedback, and AI literacies. Teaching implications Integrating genAI presents opportunities for learning, if educators can think beyond the artifact and design assessment that allows our students to showcase their learning process.