Making fair decisions is crucial to ethically implementing machine learning algorithms in social settings. In this work, we consider the celebrated definition of counterfactual fairness [Kusner et al., NeurIPS, 2017]. We begin by showing that an algorithm which satisfies counterfactual fairness also satisfies demographic parity, a far simpler fairness constraint. Similarly, we show that all algorithms satisfying demographic parity can be trivially modified to satisfy counterfactual fairness. Together, our results indicate that counterfactual fairness is basically equivalent to demographic parity, which has important implications for the growing body of work on counterfactual fairness. We then validate our theoretical findings empirically, analyzing three existing algorithms for counterfactual fairness against three simple benchmarks. We find that two simple benchmark algorithms outperform all three existing algorithms -- in terms of fairness, accuracy, and efficiency -- on several data sets. Our analysis leads us to formalize a concrete fairness goal: to preserve the order of individuals within protected groups. We believe transparency around the ordering of individuals within protected groups makes fair algorithms more trustworthy. By design, the two simple benchmark algorithms satisfy this goal while the existing algorithms for counterfactual fairness do not.

We consider representation learning for proteins with 3D structures. We build 3D graphs based on protein structures and develop graph networks to learn their representations. Depending on the levels of details that we wish to capture, protein representations can be computed at different levels, e.g., the amino acid, backbone, or all-atom levels. Importantly, there exist hierarchical relations among different levels. In this work, we propose to develop a novel hierarchical graph network, known as ProNet, to capture the relations. Our ProNet is very flexible and can be used to compute protein representations at different levels of granularity. By treating each amino acid as a node in graph modeling as well as harnessing the inherent hierarchies, our ProNet is more effective and efficient than existing methods. We also show that, given a base 3D graph network that is complete, our ProNet representations are also complete at all levels. Experimental results show that ProNet outperforms recent methods on most datasets. In addition, results indicate that different downstream tasks may require representations at different levels. Our code is publicly available as part of the DIG library (https://github.com/divelab/DIG). Proteins consist of one or more amino acid chains and perform various functions by folding into 3D conformations.

Despite recent advances in modern machine learning algorithms, the opaqueness of their underlying mechanisms continues to be an obstacle in adoption. To instill confidence and trust in artificial intelligence systems, Explainable Artificial Intelligence has emerged as a response to improving modern machine learning algorithms' explainability. Inductive Logic Programming (ILP), a subfield of symbolic artificial intelligence, plays a promising role in generating interpretable explanations because of its intuitive logic-driven framework. ILP effectively leverages abductive reasoning to generate explainable first-order clausal theories from examples and background knowledge. However, several challenges in developing methods inspired by ILP need to be addressed for their successful application in practice. For example, existing ILP systems often have a vast solution space, and the induced solutions are very sensitive to noises and disturbances. This survey paper summarizes the recent advances in ILP and a discussion of statistical relational learning and neural-symbolic algorithms, which offer synergistic views to ILP. Following a critical review of the recent advances, we delineate observed challenges and highlight potential avenues of further ILP-motivated research toward developing self-explanatory artificial intelligence systems.

The ability to acquire abstract knowledge is a hallmark of human intelligence and is believed by many to be one of the core differences between humans and neural network models. Agents can be endowed with an inductive bias towards abstraction through meta-learning, where they are trained on a distribution of tasks that share some abstract structure that can be learned and applied. However, because neural networks are hard to interpret, it can be difficult to tell whether agents have learned the underlying abstraction, or alternatively statistical patterns that are characteristic of that abstraction. In this work, we compare the performance of humans and agents in a meta-reinforcement learning paradigm in which tasks are generated from abstract rules. We define a novel methodology for building "task metamers" that closely match the statistics of the abstract tasks but use a different underlying generative process, and evaluate performance on both abstract and metamer tasks. We find that humans perform better at abstract tasks than metamer tasks whereas common neural network architectures typically perform worse on the abstract tasks than the matched metamers. This work provides a foundation for characterizing differences between humans and machine learning that can be used in future work towards developing machines with more human-like behavior.

Suddenly, the world is abuzz with chatter about chatbots. Artificially intelligent agents, like ChatGPT, have shown themselves to be remarkably adept at conversing in a very human-like fashion. ChatGPT, for instance, recently passed written exams at top business and law schools, among other feats both awe-inspiring and alarming. Researchers at Stanford University's Polarization and Social Change Lab and the Institute for Human-Centered Artificial Intelligence (HAI) wanted to probe the boundaries of AI's political persuasiveness by testing its ability to sway real humans on some of the hottest social issues of the day--an assault weapon ban, the carbon tax, and paid parental leave, among others. Indeed, AI-generated persuasive appeals were as effective as ones written by humans in persuading human audiences on several political issues," said Hui "Max" Bai, a postdoctoral researcher in the Polarization and Social Change Lab and first author on a new paper about the experiment in pre-print.

At Cadmus, we believe every student should have equal opportunity to achieve academic excellence; that's why we're changing how the world learns! Cadmus is a global EdTech company purposefully built for the higher education sector to break down global learning barriers by providing educators and diverse student cohorts with access to high-quality learning and assessment. Built on rich, student-centric values, Cadmus empowers individuals to achieve their highest potential and graduate with the critical skills and knowledge to profoundly impact our workforces, communities, and evolving world. At Cadmus, we're working hard in pursuit of an ambitious, world-shaping goal: to provide 1 billion students with access to high-quality education by 2050. That's why we need a Data Engineering superstar to join our team and help us maintain a high-quality experience for our users worldwide -- ensuring our cloud platform is reliable, performant and set up to scale.

Artificial intelligence has become a prominent issue in education circles after the unveiling of ChatGPT, a large-scale learning model that scours the internet for information it can use to produce text in a conversational format. Many academics have raised concerns that students will use ChatGPT or similar tools to generate written assignments instead of doing the work themselves. In February, the university posted an explainer on its website describing for faculty what ChatGPT and other artificial intelligence tools are and what their limitations can be. "You've likely seen a lot of panic and concerns about how to best adapt," the explainer says. "But as with any technology, this is an ideal opportunity to reflect on our current teaching practices, experiment with new opportunities, and brainstorm ways they could be utilized effectively in a classroom." It is unclear how much Dunne made from her TikTok post about Caktus AI, but the junior gymnast from Hillsdale, New Jersey has captured college athletics' new world of name, image and likeness (NIL) profits unlike any other.

ChatGPT, the artificial intelligence chatbot, continues to have internet users abuzz, given its ability to answer prompts on a stunning variety of subjects, to create songs, recipes, and jokes, to draft emails, and more. "It's amazing to have this technology do in seconds what it takes many of us hours to do," says Deborah Rossen-Knill, executive director of the University of Rochester's Writing, Speaking, and Argument Program and a professor of writing studies. Those endless possibilities, however, have faculty and administrators in higher education expressing anxiety as well as awe, because ChatGPT also can write essays and code, answer homework questions, and solve math problems. "We're all trying to figure out how it fits into the existing landscape of higher education," says Rachel Remmel, assistant dean and director of the University's Teaching Center. "Everyone is talking about it."

BEGiN has an exciting opportunity for a Senior Data Engineer to join our growing team! This role will be remote in Ontario, Canada. BEGiN is an award-winning educational technology company with world-wide impact. With products that are as effective as they are fun, BEGiN's family of brands builds critical skills for school and life. We're a diverse team of talented people passionate about creating educational content kids love.

As the buzz about ChatGPT intensifies, the Caribbean Examinations Council (CXC) has added its voice to the conversation, underscoring that "any form of cheating is a big concern" for the regional exam body. The artificial intelligence chatbot, which was launched in November 2022, is capable of generating impressively detailed humanlike written text, and has passed law and business school exams. But CXC Director of Operations Dr Nicole Manning says, based on the design of the School Based Assessment (SBA), a candidate's competence should not be determined from one assignment. "The SBA has both a formative and summative dimension. In relation to the formative dimension, it is expected that the teacher's focus would really be on assessment for learning, as such, facilitating the assessment of skills and competencies not otherwise easily assessed in the final examinations administered in our January or May/June sittings," Manning told The Sunday Gleaner. She explained that the SBA process is expected to assess skills such as complex decision making, communication, collaboration, creativity and innovation.