Prediction-based decision-making systems are becoming increasingly prevalent in various domains. Previous studies have demonstrated that such systems are vulnerable to runaway feedback loops, e.g., when police are repeatedly sent back to the same neighborhoods regardless of the actual rate of criminal activity, which exacerbate existing biases. In practice, the automated decisions have dynamic feedback effects on the system itself that can perpetuate over time, making it difficult for short-sighted design choices to control the system's evolution. While researchers started proposing longer-term solutions to prevent adverse outcomes (such as bias towards certain groups), these interventions largely depend on ad hoc modeling assumptions and a rigorous theoretical understanding of the feedback dynamics in ML-based decision-making systems is currently missing. In this paper, we use the language of dynamical systems theory, a branch of applied mathematics that deals with the analysis of the interconnection of systems with dynamic behaviors, to rigorously classify the different types of feedback loops in the ML-based decision-making pipeline. By reviewing existing scholarly work, we show that this classification covers many examples discussed in the algorithmic fairness community, thereby providing a unifying and principled framework to study feedback loops. By qualitative analysis, and through a simulation example of recommender systems, we show which specific types of ML biases are affected by each type of feedback loop. We find that the existence of feedback loops in the ML-based decision-making pipeline can perpetuate, reinforce, or even reduce ML biases.

This paper presents a holistic approach to gradient leakage resilient distributed Stochastic Gradient Descent (SGD). First, we analyze two types of strategies for privacy-enhanced federated learning: (i) gradient pruning with random selection or low-rank filtering and (ii) gradient perturbation with additive random noise or differential privacy noise. We analyze the inherent limitations of these approaches and their underlying impact on privacy guarantee, model accuracy, and attack resilience. Next, we present a gradient leakage resilient approach to securing distributed SGD in federated learning, with differential privacy controlled noise as the tool. Unlike conventional methods with the per-client federated noise injection and fixed noise parameter strategy, our approach keeps track of the trend of per-example gradient updates. It makes adaptive noise injection closely aligned throughout the federated model training. Finally, we provide an empirical privacy analysis on the privacy guarantee, model utility, and attack resilience of the proposed approach. Extensive evaluation using five benchmark datasets demonstrates that our gradient leakage resilient approach can outperform the state-of-the-art methods with competitive accuracy performance, strong differential privacy guarantee, and high resilience against gradient leakage attacks. The code associated with this paper can be found: https://github.com/git-disl/Fed-alphaCDP.

As artificial intelligence systems play a bigger role in everyday life, they're changing the world of education, too. I am a literacy educator and researcher, and here are four ways I believe these kinds of systems can be used to help students learn. Teachers are taught to identify the learning goals of all students in a class and adapt instruction for the specific needs of individual students. But with 20 or more students in a classroom, fully customized lessons aren't always realistic. An AI system can observe how a student proceeds through an assigned task, how much time they take and whether they are successful.

Journalists, screenwriters and college professors are among widening groups of people who are concerned about eventually losing their livelihoods to artificial intelligence programs like ChatGPT, which can produce copy faster and possibly better than humans. But one entrepreneur is pursuing technology to make it easier to distinguish between text written by people and that composed by a machine. Edward Tian, a 22-year-old Princeton University student studying computer science and journalism, developed an app called GPTZero to deter the misuse of the viral chatbot ChatGPT in classrooms. The app has racked up 1.2 million registered users since January. He's now launching a new program called Origin aimed at "saving journalism," by distinguishing AI-generated disinformation from fact in online media.

New Jersey parents Christina Balestriere and Kristen Cobo discuss being sued by a school librarian for speaking out against'inappropriate books' on'Jesse Watters Primetime.' The University of Florida offers a class that examines race in the "genre of horror and its trends with a particular focus on representations of racial Otherness and racism," including "white terror" in literary classics, like Frankenstein. As part of the African American Studies class, titled "Black Horror, White Terror," students are instructed to analyze horror books and movies through the lens of "racial identity and oppression" using materials about "the power and horror of whiteness," "black feminism" and "queering personhood," according to a fall 2022 syllabus obtained by The College Fix. "We will also consider the relationship between horror and Black literary modes and traditions focusing on key moments that depict fears of Blackness and/or the terror associated with being Black in America," the syllabus reads. "This course will study the works of Black authors and producers as a way to explore racial identity and oppression."

Large AI Models (LAIMs), of which large language models are the most prominent recent example, showcase some impressive performance. However they have been empirically found to pose serious security issues. This paper systematizes our knowledge about the fundamental impossibility of building arbitrarily accurate and secure machine learning models. More precisely, we identify key challenging features of many of today's machine learning settings. Namely, high accuracy seems to require memorizing large training datasets, which are often user-generated and highly heterogeneous, with both sensitive information and fake users. We then survey statistical lower bounds that, we argue, constitute a compelling case against the possibility of designing high-accuracy LAIMs with strong security guarantees.

As voice-based Conversational Assistants (CAs), including Alexa, Siri, Google Home, have become commonly embedded in households, many children now routinely interact with Artificial Intelligence (AI) systems. It is important to research children's experiences with consumer devices which use AI techniques because these shape their understanding of AI and its capabilities. We conducted a mixed-methods study (questionnaires and interviews) with primary-school children aged 6-11 in Scotland to establish children's understanding of how voice-based CAs work, how they perceive their cognitive abilities, agency and other human-like qualities, their awareness and trust of privacy aspects when using CAs and what they perceive as appropriate verbal interactions with CAs. Most children overestimated the CAs' intelligence and were uncertain about the systems' feelings or agency. They also lacked accurate understanding of data privacy and security aspects, and believed it was wrong to be rude to conversational assistants. Exploring children's current understanding of AI-supported technology has educational implications; such findings will enable educators to develop appropriate materials to address the pressing need for AI literacy.

In recent years neural networks have achieved impressive results on many technological and scientific tasks. Yet, the mechanism through which these models automatically select features, or patterns in data, for prediction remains unclear. Identifying such a mechanism is key to advancing performance and interpretability of neural networks and promoting reliable adoption of these models in scientific applications. In this paper, we identify and characterize the mechanism through which deep fully connected neural networks learn features. We posit the Deep Neural Feature Ansatz, which states that neural feature learning occurs by implementing the average gradient outer product to up-weight features strongly related to model output. Our ansatz sheds light on various deep learning phenomena including emergence of spurious features and simplicity biases and how pruning networks can increase performance, the "lottery ticket hypothesis." Moreover, the mechanism identified in our work leads to a backpropagation-free method for feature learning with any machine learning model. To demonstrate the effectiveness of this feature learning mechanism, we use it to enable feature learning in classical, non-feature learning models known as kernel machines and show that the resulting models, which we refer to as Recursive Feature Machines, achieve state-of-the-art performance on tabular data.

This paper introduces a novel Arabic pronunciation learning application QVoice, powered with end-to-end mispronunciation detection and feedback generator module. The application is designed to support non-native Arabic speakers in enhancing their pronunciation skills, while also helping native speakers mitigate any potential influence from regional dialects on their Modern Standard Arabic (MSA) pronunciation. QVoice employs various learning cues to aid learners in comprehending meaning, drawing connections with their existing knowledge of English language, and offers detailed feedback for pronunciation correction, along with contextual examples showcasing word usage. The learning cues featured in QVoice encompass a wide range of meaningful information, such as visualizations of phrases/words and their translations, as well as phonetic transcriptions and transliterations. QVoice provides pronunciation feedback at the character level and assesses performance at the word level.

Theory of mind (ToM) refers to humans' ability to understand and infer the desires, beliefs, and intentions of others. The acquisition of ToM plays a key role in humans' social cognition and interpersonal relations. Though indispensable for social intelligence, ToM is still lacking for modern AI and NLP systems since they cannot access the human mental state and cognitive process beneath the training corpus. To empower AI systems with the ToM ability and narrow the gap between them and humans, in this paper, we propose COKE: the first cognitive knowledge graph for machine theory of mind. Specifically, COKE formalizes ToM as a collection of 45k+ manually verified cognitive chains that characterize human mental activities and subsequent behavioral/affective responses when facing specific social circumstances. Beyond that, we further generalize COKE using pre-trained language models and build a powerful cognitive generation model COKE+. Experimental results in both automatic and human evaluation demonstrate the high quality of COKE and the superior ToM ability of COKE+.