A landmark trial currently under way in Washington may well decide the future of the internet. In the dock is Google, the world's largest search engine. The United States Department of Justice has accused the search giant of muscling its way to dominance by paying other companies like Apple to be the default search engine on their devices. "Google pays billions of dollars each year to distributors -- including popular-device manufacturers such as Apple, LG, Motorola, and Samsung … to secure default status for its general search engine," the Justice Department's complaint says. This, the DOJ thinks, chokes off competition that includes other search engines like Microsoft's Bing, and privately held DuckDuckGo.

When The Atlantic revealed last month that tens of thousands of books published in the past 20 years had been used without permission to train Meta's AI language model, well-known authors were outraged, calling it a "smoking gun" for mega-corporate misbehavior. Now that the magazine has put out a searchable database of affected books, the outrage is redoubled: "I would never have consented for Meta to train AI on any of my books, let alone five of them," wrote the novelist Lauren Groff. The original Atlantic story gestured at this sense of violation and affront: "The future promised by AI is written with stolen words," it said. Still I was mystified, at first, by the Sturm und Drang response, and by the claim that generative AI is "powered by mass theft." Perhaps I was just jealous of the famous writers who were being singled out as victims--Stephen King, Zadie Smith, Michael Pollan, and others who command huge speaking fees and lucrative secondary-rights deals.

Semi-supervised learning (SSL) is a promising approach for training deep classification models using labeled and unlabeled datasets. However, existing SSL methods rely on a large unlabeled dataset, which may not always be available in many real-world applications due to legal constraints (e.g., GDPR). In this paper, we investigate the research question: Can we train SSL models without real unlabeled datasets? Instead of using real unlabeled datasets, we propose an SSL method using synthetic datasets generated from generative foundation models trained on datasets containing millions of samples in diverse domains (e.g., ImageNet). Our main concepts are identifying synthetic samples that emulate unlabeled samples from generative foundation models and training classifiers using these synthetic samples. To achieve this, our method is formulated as an alternating optimization problem: (i) meta-learning of generative foundation models and (ii) SSL of classifiers using real labeled and synthetic unlabeled samples. For (i), we propose a meta-learning objective that optimizes latent variables to generate samples that resemble real labeled samples and minimize the validation loss. For (ii), we propose a simple unsupervised loss function that regularizes the feature extractors of classifiers to maximize the performance improvement obtained from synthetic samples. We confirm that our method outperforms baselines using generative foundation models on SSL. We also demonstrate that our methods outperform SSL using real unlabeled datasets in scenarios with extremely small amounts of labeled datasets. This suggests that synthetic samples have the potential to provide improvement gains more efficiently than real unlabeled data.

Survival analysis plays a crucial role in many healthcare decisions, where the risk prediction for the events of interest can support an informative outlook for a patient's medical journey. Given the existence of data censoring, an effective way of survival analysis is to enforce the pairwise temporal concordance between censored and observed data, aiming to utilize the time interval before censoring as partially observed time-to-event labels for supervised learning. Although existing studies mostly employed ranking methods to pursue an ordering objective, contrastive methods which learn a discriminative embedding by having data contrast against each other, have not been explored thoroughly for survival analysis. Therefore, in this paper, we propose a novel Ontology-aware Temporality-based Contrastive Survival (OTCSurv) analysis framework that utilizes survival durations from both censored and observed data to define temporal distinctiveness and construct negative sample pairs with adjustable hardness for contrastive learning. Specifically, we first use an ontological encoder and a sequential self-attention encoder to represent the longitudinal EHR data with rich contexts. Second, we design a temporal contrastive loss to capture varying survival durations in a supervised setting through a hardness-aware negative sampling mechanism. Last, we incorporate the contrastive task into the time-to-event predictive task with multiple loss components. We conduct extensive experiments using a large EHR dataset to forecast the risk of hospitalized patients who are in danger of developing acute kidney injury (AKI), a critical and urgent medical condition. The effectiveness and explainability of the proposed model are validated through comprehensive quantitative and qualitative studies.

When I learned that Meta's programmers downloaded 183,000 books for a database to teach the company's generative A.I. machines how to write, I was curious whether any of my own books had been fed into the crusher. Alex Reisner of the Atlantic has provided a handy search tool--type in an author's name, out comes all of his or her books that the LLaMA used. I typed "Fred Kaplan" and found that three of my six books (1959, Dark Territory, and The Insurgents) had been assimilated into the digital Borg. My first reaction, like that of many other authors, was outrage at the violation. However, my second reaction--also, I assume, like that of many other authors--was outrage that the program didn't include my other three books (The Bomb, Daydream Believers, and The Wizards of Armageddon). Were there really 182,997 books that were better than those three?

OpenAI is bringing audio and image capabilities to ChatGPT. The platform, which has long been limited to written prompts, will be adding the new features over the next two weeks to paid versions of the app, OpenAI announced in a blog post on Monday. Everyone else will be receiving the features "soon after". Users can have voice conversations with the chatbot, bringing it closer to popular AI assistants such as Apple's Siri and Amazon's Alexa. ChatGPT's new voice feature can also narrate bedtime stories, settle debates at the dinner table and speak out loud text input from users.

Allen explained that he "input numerous revisions and text prompts at least 624 times to arrive at the initial version of the image," which created an initial version of the image that Allen edited with Photoshop.

Among creative professionals, Generative Artificial Intelligence (GenAI) has sparked excitement over its capabilities and fear over unanticipated consequences. How does GenAI impact User Experience Design (UXD) practice, and are fears warranted? We interviewed 20 UX Designers, with diverse experience and across companies (startups to large enterprises). We probed them to characterize their practices, and sample their attitudes, concerns, and expectations. We found that experienced designers are confident in their originality, creativity, and empathic skills, and find GenAI's role as assistive. They emphasized the unique human factors of "enjoyment" and "agency", where humans remain the arbiters of "AI alignment". However, skill degradation, job replacement, and creativity exhaustion can adversely impact junior designers. We discuss implications for human-GenAI collaboration, specifically copyright and ownership, human creativity and agency, and AI literacy and access. Through the lens of responsible and participatory AI, we contribute a deeper understanding of GenAI fears and opportunities for UXD.

Recent years have witnessed remarkable progress made in large language models (LLMs). Such advancements, while garnering significant attention, have concurrently elicited various concerns. The potential of these models is undeniably vast; however, they may yield texts that are imprecise, misleading, or even detrimental. Consequently, it becomes paramount to employ alignment techniques to ensure these models to exhibit behaviors consistent with human values. This survey endeavors to furnish an extensive exploration of alignment methodologies designed for LLMs, in conjunction with the extant capability research in this domain. Adopting the lens of AI alignment, we categorize the prevailing methods and emergent proposals for the alignment of LLMs into outer and inner alignment. We also probe into salient issues including the models' interpretability, and potential vulnerabilities to adversarial attacks. To assess LLM alignment, we present a wide variety of benchmarks and evaluation methodologies. After discussing the state of alignment research for LLMs, we finally cast a vision toward the future, contemplating the promising avenues of research that lie ahead. Our aspiration for this survey extends beyond merely spurring research interests in this realm. We also envision bridging the gap between the AI alignment research community and the researchers engrossed in the capability exploration of LLMs for both capable and safe LLMs.

Knowledge graph completion is a task that revolves around filling in missing triples based on the information available in a knowledge graph. Among the current studies, text-based methods complete the task by utilizing textual descriptions of triples. However, this modeling approach may encounter limitations, particularly when the description fails to accurately and adequately express the intended meaning. To overcome these challenges, we propose the augmentation of data through two additional mechanisms. Firstly, we employ ChatGPT as an external knowledge base to generate coherent descriptions to bridge the semantic gap between the queries and answers. Secondly, we leverage inverse relations to create a symmetric graph, thereby creating extra labeling and providing supplementary information for link prediction. This approach offers additional insights into the relationships between entities. Through these efforts, we have observed significant improvements in knowledge graph completion, as these mechanisms enhance the richness and diversity of the available data, leading to more accurate results.