The most famous virtual valets around today--Siri, Alexa, and Google Assistant--are a lot less impressive than the latest AI-powered chatbots like ChatGPT or Google Bard. When the fruits of the recent generative AI boom get properly integrated into those legacy assistant bots, they will surely get much more interesting. To get a preview of what's next, I took an experimental AI voice helper called vimGPT for a test run. When I asked it to "subscribe to WIRED," it got to work with impressive skill, finding the correct web page and accessing the online form. If it had access to my credit card details I'm pretty sure it would have nailed it.

Google is rolling out a swathe of updates on the generative AI front, including a new text-to-image tool. What's different about ImageFX is that it has an interface that features "expressive chips." The idea here is that these will help you "quickly experiment with adjacent dimensions of your creation and ideas." Alongside the debut of ImageFX, Google says it has improved MusicFX and TextFX. The company's claims that it's made upgrades to the MusicLM model that include faster generation of music and higher-quality audio, along with new features.

Challenges and opportunities faced by computing educators and students adapting to LLMs capable of generating accurate source code from natural-language problem descriptions.

As owner of GitHub and lead investor in OpenAI, the developer of the GPT-x series of large language models (LLMs), it did not take long for Microsoft to see the potential for collaboration between the two. Three years ago, GitHub partnered with OpenAI to develop Codex as an automated assistant for programmers, quickly followed by the Copilot code-completion tool. The public release of ChatGPT by OpenAI toward the end of 2022 made the technology even more widely available to software developers and people learning to program, with other vendors joining in the effort to automate the job of writing software using LLMs. Rapid scaling has enabled major improvements in the ability of artificial intelligence (AI) to turn natural-language requests into working code. Workplace studies have claimed LLMs boost productivity on real-world projects.

Warning: The content of this paper as well as some blurred images shown may include references to nudity, sexualization, violence, and gore. Text-to-image (T2I) models have emerged as a significant advancement in generative AI; however, there exist safety concerns regarding their potential to produce harmful image outputs even when users input seemingly safe prompts. This phenomenon, where T2I models generate harmful representations that were not explicit in the input, poses a potentially greater risk than adversarial prompts, leaving users unintentionally exposed to harms. Our paper addresses this issue by first introducing a formal definition for this phenomenon, termed harm amplification. We further contribute to the field by developing methodologies to quantify harm amplification in which we consider the harm of the model output in the context of user input. We then empirically examine how to apply these different methodologies to simulate real-world deployment scenarios including a quantification of disparate impacts across genders resulting from harm amplification. Together, our work aims to offer researchers tools to comprehensively address safety challenges in T2I systems and contribute to the responsible deployment of generative AI models.

This study explores the use of Generative AI chatbots for transforming public science exhibits into virtual experiences that can extend the engagement of exhibits into the classroom. The broader goal is to increase accessibility of science exhibits, especially for those marginalized in STEM due to various factors, including cultural barriers. We hypothesize that turning exhibits into first-person anthropomorphized chatbots with a personality, like quirky-talking asteroids or comets, can increase engagement and learning. The paper mainly explores if such techniques are possible using Generative AI (e.g. GPT) via prompt engineering alone. The research includes an investigation into the possibility of integrating interactive assessment via question-generation using Bloom's Taxonomy. Initial results indicate that it is possible to combine these techniques. As such, it lays a foundation for future classroom evaluations of such chatbots to gauge their overall efficacy in extending the reach of science exhibitions. The paper concludes by discussing extensions of the research to fully evaluate effectiveness in virtual field-trips. We also include a brief examination of additional ways to enhance student motivation towards learning via chatbots.

The potential harms of the under-representation of minorities in training data, particularly in multi-modal settings, is a well-recognized concern. While there has been extensive effort in detecting such under-representation, resolution has remained a challenge. With recent advancements in generative AI, large language models and foundation models have emerged as versatile tools across various domains. In this paper, we propose Chameleon, a system that efficiently utilizes these tools to augment a data set with a minimal addition of synthetically generated tuples, in order to enhance the coverage of the under-represented groups. Our system follows a rejection sampling approach to ensure the generated tuples have a high quality and follow the underlying distribution. In order to minimize the rejection chance of the generated tuples, we propose multiple strategies for providing a guide for the foundation model. Our experiment results, in addition to confirming the efficiency of our proposed algorithms, illustrate the effectiveness of our approach, as the unfairness of the model in a downstream task significantly dropped after data repair using Chameleon.

Text-to-image generative AI models such as Stable Diffusion are used daily by millions worldwide. However, many have raised concerns regarding how these models amplify racial and gender stereotypes. To study this phenomenon, we develop a classifier to predict the race, gender, and age group of any given face image, and show that it achieves state-of-the-art performance. Using this classifier, we quantify biases in Stable Diffusion across six races, two genders, five age groups, 32 professions, and eight attributes. We then propose novel debiasing solutions that outperform state-of-the-art alternatives. Additionally, we examine the degree to which Stable Diffusion depicts individuals of the same race as being similar to one another. This analysis reveals a high degree of stereotyping, e.g., depicting most middle eastern males as being dark-skinned, bearded, and wearing a traditional headdress. We address these limitations by proposing yet another novel solution that increases facial diversity across genders and racial groups. Our solutions are open-sourced and made publicly available.

Large language models demonstrate remarkable capabilities after learning from enormous volumes of text data (Anil et al., 2023; Hoffmann et al., 2022; OpenAI, 2023). Yet, reinforcement learning from human feedback (RLHF) greatly improves their behavior even after only tens of thousands of interactions (Bai et al., 2022; Glaese et al., 2022; Ouyang et al., 2022; Stiennon et al., 2020). The uptake of chatbots affords opportunities to gather increasing volumes of human feedback, with each engagement eliciting expressions of satisfaction or preference (OpenAI, 2022). It is natural to wonder what new capabilities may emerge with this growing source of data. Superhuman ingenuity remains an alluring possibility. With increasing volumes, more can be inferred from human feedback.

Hackathons and software competitions, increasingly pivotal in the software industry, serve as vital catalysts for innovation and skill development for both organizations and students. These platforms enable companies to prototype ideas swiftly, while students gain enriched learning experiences, enhancing their practical skills. Over the years, hackathons have transitioned from mere competitive events to significant educational tools, fusing theoretical knowledge with real-world problem-solving. The integration of hackathons into computer science and software engineering curricula aims to align educational proficiencies within a collaborative context, promoting peer connectivity and enriched learning via industry-academia collaborations. However, the infusion of advanced technologies, notably artificial intelligence (AI), and machine learning, into hackathons is revolutionizing their structure and outcomes. This evolution brings forth both opportunities, like enhanced learning experiences, and challenges, such as ethical concerns. This study delves into the impact of generative AI, examining its influence on student's technological choices based on a case study on the University of Iowa 2023 event. The exploration provides insights into AI's role in hackathons, and its educational implications, and offers a roadmap for the integration of such technologies in future events, ensuring innovation is balanced with ethical and educational considerations.