The emergence of large language models has led to the development of powerful tools such as ChatGPT that can produce text indistinguishable from human-generated work. With the increasing accessibility of such technology, students across the globe may utilize it to help with their school work—a possibility that has sparked ample discussion on the integrity of student evaluation processes in the age of artificial intelligence (AI). To date, it is unclear how such tools perform compared to students on university-level courses across various disciplines. Further, students’ perspectives regarding the use of such tools in school work, and educators’ perspectives on treating their use as plagiarism, remain unknown. Here, we compare the performance of the state-of-the-art tool, ChatGPT, against that of students on 32 university-level courses. We also assess the degree to which its use can be detected by two classifiers designed specifically for this purpose. Additionally, we conduct a global survey across five countries, as well as a more in-depth survey at the authors’ institution, to discern students’ and educators’ perceptions of ChatGPT’s use in school work. We find that ChatGPT’s performance is comparable, if not superior, to that of students in a multitude of courses. Moreover, current AI-text classifiers cannot reliably detect ChatGPT’s use in school work, due to both their propensity to classify human-written answers as AI-generated, as well as the relative ease with which AI-generated text can be edited to evade detection. Finally, there seems to be an emerging consensus among students to use the tool, and among educators to treat its use as plagiarism. Our findings offer insights that could guide policy discussions addressing the integration of artificial intelligence into educational frameworks.
The evolution of natural language processing techniques has led to the development of advanced conversational tools such as ChatGPT, capable of assisting users with a variety of activities. Media attention has centered on ChatGPT’s potential impact, policy implications, and ethical ramifications, particularly in the context of education. As such tools become more accessible, students across the globe may use them to assist with their homework. However, it is still unclear whether ChatGPT’s performance is advanced enough to pose a serious risk of plagiarism. We fill this gap by evaluating ChatGPT on two introductory and two advanced university-level courses. We find that ChatGPT receives near-perfect grades on the majority of questions in the introductory courses but has not yet reached the level of sophistication required to pass in advanced courses. Moreover, adding a few full stops or typos may fool a machine learning algorithm designed to detect ChatGPT-generated text. These findings suggest that, at least for some courses, current artificial intelligence tools pose a real threat that can no longer be overlooked by educational institutions.
Text-to-image generative AI models such as Stable Diffusion are used daily by millions worldwide. However, the extent to which these models exhibit racial and gender stereotypes is not yet fully understood. Here, we document significant biases in Stable Diffusion across six races, two genders, 32 professions, and eight attributes. Additionally, we examine the degree to which Stable Diffusion depicts individuals of the same race as being similar to one another. This analysis reveals significant racial homogenization, e.g., depicting nearly all Middle Eastern men as bearded, brown-skinned, and wearing traditional attire. We then propose debiasing solutions that allow users to specify the desired distributions of race and gender when generating images while minimizing racial homogenization. Finally, using a preregistered survey experiment, we find evidence that being presented with inclusive AI-generated faces reduces people’s racial and gender biases, while being presented with non-inclusive ones increases such biases, regardless of whether the images are labeled as AI-generated. Taken together, our findings emphasize the need to address biases and stereotypes in text-to-image models.
27 November 2023 in Proceedings of the ACM on Networking
The rapid growth of satellite network operators (SNOs) has revolutionized broadband communications, enabling global connectivity and bridging the digital divide. As these networks expand, it is important to evaluate their performance and efficiency. This paper presents the first comprehensive study of SNOs. We take an opportunistic approach and devise a methodology which allows to identify public network measurements performed via SNOs. We apply this methodology to both M-Lab and RIPE public datasets which allowed us to characterize low level performance and footprint of up to 18 SNOs operating in different orbits. Finally, we identify and recruit paid testers on three popular SNOs (Starlink, HughesNet, and ViaSat) to evaluate the performance of popular applications like web browsing and video streaming.
A significant fraction of the World Wide Web suffers from the excessive usage of JavaScript (JS). Based on an analysis of popular webpages, we observed that a considerable number of JS elements utilized by these pages are not essential for their visual and functional features. In this paper, we propose JSCleaner, a JavaScript de-cluttering engine that aims at simplifying webpages without compromising their content or functionality. JSCleaner relies on a rule-based classification algorithm that classifies JS into three main categories: non-critical, replaceable, and critical. JSCleaner removes non-critical JS from a webpage, translates replaceable JS elements with their HTML outcomes, and preserves critical JS. Our quantitative evaluation of 500 popular webpages shows that JSCleaner achieves around 30% reduction in page load times coupled with a 50% reduction in the number of requests and the page size. In addition, our qualitative user study of 103 evaluators shows that JSCleaner preserves 95% of the page content similarity, while maintaining nearly 88% of the page functionality (the remaining 12% did not have a major impact on the user browsing experience).
Citations are widely considered in scientists’ evaluation. As such, scientists may be incentivized to inflate their citation counts. While previous literature has examined self-citations and citation cartels, it remains unclear whether scientists can purchase citations. Here, we compile a dataset of ~1.6 million profiles on Google Scholar to examine instances of citation fraud on the platform. We survey faculty at highly-ranked universities, and confirm that Google Scholar is widely used when evaluating scientists. We then engage with a citation-boosting service, and manage to purchase 50 citations while assuming the identity of a fictional author. Taken as a whole, our findings bring to light new forms of citation manipulation, and emphasize the need to look beyond citation counts.