As artificial intelligence (AI) tools like ChatGPT become more common, their role in computer science (CS) education continues to evolve, especially in large courses with time-limited assessments. This lightning talk presents an observation from a large-scale (over 300 students) introductory software design and engineering course at a university in the Southeastern United States. During a 30-minute, open-notes assessment where students were allowed to use generative AI, they were asked to extend one user story in an existing codebase they had previously built for the course. The task followed an all-or-nothing grading approach that required a fully functional user story implementation for credit. Although students often support using AI for learning, their reactions revealed a gap between what they expected AI to do and the actual thinking required to solve real problems quickly: Some students even expressed frustration, noting that AI tools offered little help under time pressure. To close the experience, we held a reflection lecture where students analyzed the role of AI as a tool and discussed its purpose in supporting augmented intelligence rather than replacing human reasoning. This case illustrates how assessment design can expose the limits of generative AI as a learning aid and highlights the importance of helping students build awareness of time, effort, and reflection when using these tools. The goal of this talk is to share these insights, invite discussion on AI use under time constraints, and explore how students adapt or resist adapting when AI cannot ''think fast enough'' for them.

As generative AI (Gen AI) tools reshape software engineering (SE) workflows, educators are exploring how to meaningfully integrate them into computing education. This experience report presents a structured benchmarking of widely used AI tools -- such as GitHub Copilot, GPT-4, Codeium, Claude 3.5, Gemini 1.5, Supermaven, TabNine, Testim, Postman, Eraser.io, and Lucidchart AI -- across key SE phases: design, implementation, debugging, and testing. Tools were selected based on industry relevance, accessibility for students, and alignment with common SE tasks. Through controlled experiments conducted by five AI-experienced evaluators with matched exposure levels, we assessed tool performance using standardized prompts, counterbalanced task roles, and a range of proxy metrics -- including prompt iterations, task completion time, human correction burden, hallucination frequency, output accuracy, and cross-file consistency -- to capture both cognitive load and tool limitations. While AI tools accelerated tasks such as boilerplate generation and UML sketching, they exhibited challenges in test coverage quality, cross-file coherence, and reliability under complex prompts. We discuss educational implications, including managing cognitive load, aligning tools with task types, and explicitly teaching prompt refinement and verification strategies. The paper offers actionable guidance for instructors, curriculum-ready artifacts, and a roadmap for scaling AI integration in SE classrooms, while also noting key limitations to support replication and contextual adoption.

Generative Artificial Intelligence (Gen-AI) has revolutionized software engineering (SE) by automating tasks across design, coding, and testing [1] [2].Tools like ChatGPT and GitHub Copilot streamline code generation, architectural modeling, debugging, and testcase creation [3] [4].Despite their rapid adoption in industry, the pedagogical implications of these tools in computing education have not been systematically examined.This study solves the existing gap by conducting a comprehensive benchmarking study of Gen-AI tools across four core SE phases-design documentation, feature implementation, debugging support, and testing -to address two research questions:RQ1: What strengths and limitations do Gen-AI tools exhibit in each phase?RQ2: How can insights from benchmarking inform effective integration of Gen-AI into SE curricula?To answer these questions, a diverse set of Gen-AI tools is evaluated, ranging from design-focused assistants such as Lucidchart, Mermaid.js and UIzard; implementation-oriented systems including GitHub Copilot, TabNine, Codeium and Supermaven; debugging supports like GPT-4 and Claude 3.5 Sonnet; and testing frameworks such as Testim, Mabl and Applitools-while also surveying emerging platforms (as of summer 2024) like Replit, Postman, Visily, Gemini, Eraser.io and others.For each tool and development phase, we applied phase-specific metrics: in design documentation, we assessed diagram accuracy, completeness, user effort, and IDE integration; in feature implementation, we measured pattern-based code generation quality, code-completion effectiveness, refactoring robustness, and UI/UX scaffolding; in debugging, we evaluated error-detection accuracy, hallucination rates, and clarity of explanatory feedback; and in testing, we examined test-case relevance and defect-detection coverage.Across all phases, we tracked prompt engineering complexity as a key mediating factor influencing tool performance.Our evaluation reveals speed-fidelity trade-offs: Code-completion assistants accelerate boilerplate generation but demand manual oversight to ensure cross-file consistency and manage higher-order abstractions; diagramming tools can produce precise UML models with minimal effort-but at the cost of iterative prompt refinement for complex cases; LLM debuggers deliver context-sensitive fixes

Artificial Intelligence (AI) tools have transformed software development, making it crucial to equip computer science (CS) students with the skills to leverage these technologies. This talk presents an innovative curriculum approach, integrating AI tools into an advanced CS capstone course at a stage where students possess foundational skills in software engineering. This strategic timing ensures students can critically engage with AI, recognizing biases and managing challenges like hallucinations in AI-generated outputs.

Manual processes for evaluating external course syllabi for transfer credit in higher education are time-consuming, inconsistent, and prone to bias. This project leverages Natural Language Processing (NLP) and large language models (LLMs) to automate the transfer credit evaluation process. The system processes external syllabi by embedding course content, conducting similarity searches, and providing structured reasoning for each match. Using techniques such as chain-of-thought reasoning and reflection agents, the system generates similarity scores and detailed explanations to support informed, data-driven decision-making by faculty. Validated against faculty decisions, the system promises to significantly improve the efficiency, consistency, and fairness of transfer credit evaluations. Future directions include expanding the system for advanced standing test evaluations and allowing faculty to query specific course components for more targeted analysis.

This research paper explores the prevalence of the Impostor Phenomenon (IP) among undergraduate college students in computer science (CS) courses, addressing its variations concerning different student demographics such as gender, ethnicity, and institutional background. We surveyed 502 students from two Southeastern U.S. institutions-one in Georgia and one in Florida-using the Clance Impostor Phenomenon Scale (CIPS). We found that 63% of our participants reported experiencing IP, with higher rates among female students (68%) compared to males (62 %) and those of other gender identities (46 %). Our findings also suggest that IP experiences are significantly influenced by the interplay of students' gender and their institutional context. At the first institution (Georgia Institute of Technology), IP was found in 58% of female students and 65% of male students, compared to 29% among students with other gender identities. At the second institution (University of Florida), these figures were 80% for females, 57% for males, and 67% for students with other gender identities. Our study underscores the importance of developing targeted support strategies within the CS education community to address the high prevalence of IP, particularly considering its varying impact across different demographics and institutional contexts.

Designing virtual humans to support mental health scenarios is challenging, as it requires designers to think of characteristics and strategies to promote rapport, increase trust, and favor the agents’ effectiveness during the intended interactions. This paper explores the impact of featuring the designer as a product cue in virtual human interactions, with a particular focus on how gender synchrony among users, virtual humans, and their designers influences trust and intentions towards well-being practices like gratitude journaling. We conducted an online, asynchronous, between-subjects user study and analyzed responses from 159 adult participants from the United States: Participants were female (n=80) or male (n=79). We assessed how gender synchrony could affect users’ intentions and the perceived trustworthiness of the virtual agents. Our findings suggest that gender synchrony does not significantly impact users’ intentions towards gratitude journaling. However, significant differences were observed concerning source credibility linked to agent trustworthiness when there was an age gap of less than ten years between users and virtual humans. These findings underscore the importance of considering both gender and age when designing virtual humans, as these factors can significantly enhance user acceptance and could foster the overall effectiveness of virtual human interactions in sensitive contexts such as mental health. Our study contributes to understanding how demographic characteristics and the use of the designer as a product cue can be strategically adapted to improve user experience in mental health scenarios.

Promoting active learning is challenging in large computing courses, often with hundreds of students. We present insights from a pedagogical strategy we designed to foster college Computer Science (CS) students' learning in a large introductory course on software design and engineering (SWE). Students first created an instructional video explaining SWE concepts. Next, they peer-reviewed their classmates' explanations. Our findings suggest that using student-created instructional videos can support students' learning processes and help promote active learning in large computing courses.

When it comes to computer science capstone courses, quality of projects is often a determining factor in learning outcomes and overall student experience. However, sourcing appropriate and interesting projects-whether devised by instructors or submitted by faculty or industry clients-can be challenging. This birds of a feather session invites educators, administrators, and even former students involved with CS capstones and other project-based courses to share some of the challenges they have encountered (e.g. client recruitment, project quality and variety, IP concerns) and potential strategies, processes, and tools for addressing these challenges.

The rise of digital music has led to a parallel rise in the need to manage music collections of several thousands of songs on a single device. Manual selection of songs for a music listening experience is a cumbersome task. In this paper, we present an initial exploration of the feasibility of using song signal properties and user context information to assist in automatic song selection. Users listened to music over the course of a month while their context and song selections were tracked. Initial results suggest the use of context information can improve automated song selection when patterns are learned for each individual.

Audio signal properties can provide a media player with highly descriptive feature sets in order to intelligently select similar songs for a music stream. A well-known problem among researchers in music information retrieval, however, is that extracting signal properties requires a significant amount of computational resources, thus making it impractical for even the most advanced mobile media players. Although other approaches to retrieving data are possible, local extraction still has unique benefits. Using a combination of machine learning and profiling techniques, this paper presents an initial evaluation of partial signal extraction, which reduces resource requirements by locally collecting signals from parts of a song rather than all. Our preliminary experiments suggest that this idea can offer significantly lower resource requirements while losing marginal song information.

Many pervasive computing systems have been built for using mobile computers to interact with networked devices. To deploy a device's user-interface, several systems dynamically generate the user-interface on a mobile computer. While this approach has several advantages, empirical results from different generators show that it takes a relatively long time for a mobile computer to create a user-interface from scratch. This paper shows that it is possible to overcome this limitation by efficiently mapping or retargeting a previously generated user-interface of one (source) device to another (target) device of the same or different type. Using the implementation of an existing generator and a set of real-world scenarios, we show that user-interface retargeting can yield deployment times that are often as good as or noticeably better than the approach of locally loading pre-existing manually-written user-interface code.

One of the visions of pervasive computing is using mobile computers to interact with networked devices. A question raised by this vision is: should the user-interfaces of these devices be handcrafted manually or generated automatically? Based on experience within the domain of desktop computing, the answer seems to be that automatic generation is not flexible enough to support a significant number of useful interfaces but requires substantially less coding effort for the interfaces it can create. We show that the answer is much more complicated when we consider networking of traditional appliances such as stereos and TVs. Using qualitative arguments and quantitative experimental data, we show that the manual vs. generated issue must be resolved based on: (a) not only user-interface programming and flexibility but also several other metrics such as space and time costs, binding time, and reliability (b) whether it is a graphical or speech based user-interface, (c) the size of the device user-interface, (d) whether the manually written user-interface code is available at the mobile computer or at a remote machine, and (e) the network bandwidth between the mobile computer and remote factory