Distribution shift occurs when the test distribution differs from the training distribution, and it can considerably degrade performance of machine learning models deployed in the real world. Temporal shifts -- distribution shifts arising from the passage of time -- often occur gradually and have the additional structure of timestamp metadata. By leveraging timestamp metadata, models can potentially learn from trends in past distribution shifts and extrapolate into the future. While recent works have studied distribution shifts, temporal shifts remain underexplored. To address this gap, we curate Wild-Time, a benchmark of 5 datasets that reflect temporal distribution shifts arising in a variety of real-world applications, including patient prognosis and news classification. On these datasets, we systematically benchmark 13 prior approaches, including methods in domain generalization, continual learning, self-supervised learning, and ensemble learning. We use two evaluation strategies: evaluation with a fixed time split (Eval-Fix) and evaluation with a data stream (Eval-Stream). Eval-Fix, our primary evaluation strategy, aims to provide a simple evaluation protocol, while Eval-Stream is more realistic for certain real-world applications. Under both evaluation strategies, we observe an average performance drop of 20% from in-distribution to out-of-distribution data. Existing methods are unable to close this gap. Code is available at https://wild-time.github.io/.

AI is buzzing again thanks to the recent release of ChatGPT, a natural language chatbot that people are using to write emails, poems, song lyrics and college essays. Early adopters have even used it to write Python code, as well as to reverse engineer shellcode and rewrite it in C. ChatGPT has sparked hope among people eager for the arrival of practical applications of AI, but it also begs the question of whether it will displace writers and developers in the same way robots and computers have replaced some cashiers, assembly-line workers and, perhaps in the future, taxi drivers. It's hard to say how sophisticated the AI text-creation capabilities will be in the future as the technology ingests more and more examples of our online writing. But I see it having very limited capabilities for programming. If anything, it could end up being just another tool in the developer's kit to handle tasks that don't take the critical thinking skills software engineers bring to the table.

We've all heard them: 24 carat gold, 100% Florida orange juice, 100% all natural, 100% made in the USA. Much of what we consume--from food to data--is qualified in some way to help us gain insights into what we're consuming. Sometimes, it's directly related to things like ingredients and other times, it's more about the social and political implications. But the rise of machine learning and natural language processing has led to the development of advanced language models such as GPT (Generative Pre-trained Transformer) and raised important questions about creativity and ownership, to name just a few. These models are amazing and are capable of generating human-like text, making it difficult to distinguish between text written by a human and text generated by a machine.

Creating autonomous robots that can actively explore the environment, acquire knowledge and learn skills continuously is the ultimate achievement envisioned in cognitive and developmental robotics. Their learning processes should be based on interactions with their physical and social world in the manner of human learning and cognitive development. Based on this context, in this paper, we focus on the two concepts of world models and predictive coding. Recently, world models have attracted renewed attention as a topic of considerable interest in artificial intelligence. Cognitive systems learn world models to better predict future sensory observations and optimize their policies, i.e., controllers. Alternatively, in neuroscience, predictive coding proposes that the brain continuously predicts its inputs and adapts to model its own dynamics and control behavior in its environment. Both ideas may be considered as underpinning the cognitive development of robots and humans capable of continual or lifelong learning. Although many studies have been conducted on predictive coding in cognitive robotics and neurorobotics, the relationship between world model-based approaches in AI and predictive coding in robotics has rarely been discussed. Therefore, in this paper, we clarify the definitions, relationships, and status of current research on these topics, as well as missing pieces of world models and predictive coding in conjunction with crucially related concepts such as the free-energy principle and active inference in the context of cognitive and developmental robotics. Furthermore, we outline the frontiers and challenges involved in world models and predictive coding toward the further integration of AI and robotics, as well as the creation of robots with real cognitive and developmental capabilities in the future.

Explaining the predictions of AI models is paramount in safety-critical applications, such as in legal or medical domains. One form of explanation for a prediction is an extractive rationale, i.e., a subset of features of an instance that lead the model to give its prediction on that instance. For example, the subphrase ``he stole the mobile phone'' can be an extractive rationale for the prediction of ``Theft''. Previous works on generating extractive rationales usually employ a two-phase model: a selector that selects the most important features (i.e., the rationale) followed by a predictor that makes the prediction based exclusively on the selected features. One disadvantage of these works is that the main signal for learning to select features comes from the comparison of the answers given by the predictor to the ground-truth answers. In this work, we propose to squeeze more information from the predictor via an information calibration method. More precisely, we train two models jointly: one is a typical neural model that solves the task at hand in an accurate but black-box manner, and the other is a selector-predictor model that additionally produces a rationale for its prediction. The first model is used as a guide for the second model. We use an adversarial technique to calibrate the information extracted by the two models such that the difference between them is an indicator of the missed or over-selected features. In addition, for natural language tasks, we propose a language-model-based regularizer to encourage the extraction of fluent rationales. Experimental results on a sentiment analysis task, a hate speech recognition task as well as on three tasks from the legal domain show the effectiveness of our approach to rationale extraction.

Blockchain, also coined as decentralized AI, has the potential to empower AI to be more trustworthy by creating a decentralized trust of privacy, security, and audibility. However, systematic studies on the design principle of blockchain as a trust engine for an integrated society of cyber-physical-social-system (CPSS) are still absent. In this article, we provide an initiative for seeking the design principle of blockchain for a better digital world. Using a hybrid method of qualitative and quantitative studies, we examine the past origin, the current development, and the future directions of blockchain design principles. We have three findings. First, the answer to whether blockchain lives up to its original design principle as a distributed database is controversial. Second, the current development of the blockchain community reveals a taxonomy of 7 categories, namely, privacy and security, scalability, decentralization, applicability, governance and regulation, system design, and cross-chain interoperability. Both research and practice are more centered around the first category of privacy and security and the fourth category of applicability. Future scholars, practitioners, and policy-makers have vast opportunities in other, much less exploited facets and the synthesis at the interface of multiple aspects. Finally, in counter-examples, we conclude that a synthetic solution that crosses discipline boundaries is necessary to close the gaps between the current design of blockchain and the design principle of a trust engine for a truly intelligent world.

Central Michigan University faculty in partnership with Black Women In Artificial Intelligence is offering a STEM cohort that features a series of STEM talks related to Environmental Justice and the United Nations sustainability goals (SDG). Talks will occur each Tuesday night in February from 7-7:45 PM EST. The first three Tuesday evenings will be faculty presentations and the fourth meeting will feature participant poster presentations focused on environmental justice. Join us to learn learn how Dr. Mahon is contributing to SDG 14 – Life below water - by using molecular and genomic tools to explore biodiversity, evolution and phylogeography. Dr. Marquez is working to help ensure clean water and sanitation (SDG 6) by studying how contaminants behave in the environment and how we can develop technologies to remove contaminants from water.

Bixby is an intelligent personal assistant which is only available as a built-in application on Samsung flagship devices and wearables. This application uses Natural Language Understanding to perform tasks on these devices using voice/ text, including but not limited to making phone calls, sending text messages, setting up meetings, opening apps, setting alarms and timers, getting directions, answering general questions, providing information about restaurants and other businesses, etc. The Natural Language understanding team aims to create a delightful experience for Bixby customers by making Bixby understand the intent behind any spoken request quickly and accurately. You will collaborate closely with experts in Machine Learning and Natural Language Processing, and contribute to advancing the state of the art in human language understanding systems. As an NLP Engineer you will primarily focus on building the NLU platform for Bixby by working with Product Managers / Subject Matter Experts, Lab Leaders, Linguistic Experts, brainstorm different ideas, research, build POCs and propose solutions that cater to the broader business needs.

A chatbot powered by artificial intelligence will appear in court next month to help a defendant fight a traffic ticket, CBS News reports. The "robot lawyer," the first of its kind, is an experimental step toward exploring the capabilities of increasingly sophisticated AI tools. Consumer-focused tech firm DoNotPay is behind the AI-powered legal assistant. The company's CEO, Joshua Browder, says the company's creation runs on a smartphone that listens to court arguments. The information is then fed through an AI program that outputs legal arguments to the defendant through wireless headphones in real time.

It's no secret that the commercial application of NLP technologies has exploded in recent years. From chatbots and virtual assistants to machine translation and sentiment analysis, NLP technologies are now being used in a wide variety of applications across a range of industries. With the increasing demand for technologies that can process human language, investors have been eager to get a piece of the action. In this article, we look at NLP start-up funding over the past year, identifying the applications and domains that have received investment. A version of this article will appear in the Journal of Natural Language Engineering in early 2023.