Adult content creators are having their performances used without consent. This is just one way that AI now threatens their rights and livelihoods. When Jennifer got a job doing research for a nonprofit in 2023, she ran her new professional headshot through a facial recognition program. She wanted to see if the tech would pull up the porn videos she'd made more than 10 years before, when she was in her early 20s. It did in fact return some of that content, and also something alarming that she'd never seen before: one of her old videos, but with someone else's face on her body. "At first, I thought it was just a different person," says Jennifer, who is being identified by a pseudonym to protect her privacy. But then she recognized a distinctly garish background from a video she'd shot around 2013, and she realized: "Somebody used me in a deepfake."

In recent years, the field of NLP has seen growing interest in modeling both semantic and pragmatic dimensions. Despite this progress, two key challenges persist: firstly, the complex task of mapping and analyzing the interactions between semantic and pragmatic features; secondly, the insufficient incorporation of relevant insights from related disciplines outside NLP. Addressing these issues, this study introduces a novel geometric metric that utilizes word co-occurrence patterns. This metric maps two fundamental properties - semantic typicality (cognitive) and pragmatic salience (socio-cultural) - for basic-level categories within a two-dimensional hyperbolic space. Our evaluations reveal that this semantic-pragmatic metric produces mappings for basic-level categories that not only surpass traditional cognitive semantics benchmarks but also demonstrate significant socio-cultural relevance. This finding proposes that basic-level categories, traditionally viewed as semantics-driven cognitive constructs, should be examined through the lens of both semantic and pragmatic dimensions, highlighting their role as a cognitive-cultural interface. The broad contribution of this paper lies in the development of medium-sized, interpretable, and human-centric language embedding models, which can effectively blend semantic and pragmatic dimensions to elucidate both the cognitive and socio-cultural significance of linguistic categories.

Drawing on contemporary pragmatist philosophy and linguistic theories on cognition, meaning, and communication, this paper presents a dynamic, metasemantic-metapragmatic taxonomy for grounding and conceptualizing human-like multimodal communicative alignment. The framework is rooted in contemporary developments of the three basic communicative capacities initially identified by American logician and pragmatist philosopher Charles Sanders Peirce: iconic (sensory and perceptual qualities), indexical (contextual and sociocultural associations), and rule-like (symbolic and intuitive reasoning). Expanding on these developments, I introduce the concept of indexical contextualization and propose the principle of "contextualization directionality" for characterizing the crucial metapragmatic capacity for maintaining, navigating, or transitioning between semantic and pragmatic modes of multimodal communication. I contend that current cognitive-social computational and engineering methodologies disproportionately emphasize the semantic/metasemantic domain, overlooking the pivotal role of metapragmatic indexicality in traversing the semantic-pragmatic spectrum of communication. The framework's broader implications for intentionality, identity, affect, and ethics in within-modal and cross-modal human-machine alignment are also discussed.

During one of my more desperate phases as a young novelist, I began to question whether I should actually be writing my own stories. I was deeply uninterested at the time in anything that resembled a plot, but I acknowledged that if I wanted to attain any sort of literary success I would need to tell a story that had a distinct beginning, middle, and end. This was about twenty years ago. My graduate-school friends and I were obsessed with a Web site called the Postmodernism Generator that spat out nonsensical but hilarious critical-theory papers. The site, which was created by a coder named Andrew C. Bulhak, who was building off Jamie Zawinski's Dada Engine, is still up today, and generates fake scholarly writing that reads like, "In the works of Tarantino, a predominant concept is the distinction between creation and destruction. Marx's essay on capitalist socialism holds that society has objective value. But an abundance of appropriations concerning not theory, but subtheory exist."

With the success of Peter Jackson's Get Back, the documentary streaming on Disney Plus, Beatlemania is back. Watching Paul McCartney create the eponymous song out of seemingly nothing, as George Harrison stands nearby yawning, is one of 2021's cinematic pleasures. The Beatles are arguably the most successful pop group in history and, in the years since their heyday, countless artists, producers and songwriters, not to mention record companies and now music streaming services, have tried to recreate the same magic. The latest tool for capturing elusive pop music gold is artificial intelligence. Usually when we think of artificial intelligence creating art, it's making something bizarre or unintentionally hilarious. Take Google's horrifying Deep Dream with its thousands of dog eyes, or Sunspring, a movie written by an AI that was fed hundreds of sci-fi scripts.

In November, the musician Grimes made a bold prediction. "I feel like we're in the end of art, human art," she said on Sean Carroll's Mindscape podcast. "Once there's actually AGI (Artificial General Intelligence), they're gonna be so much better at making art than us." Her comments sparked a meltdown on social media. The musician Zola Jesus called Grimes the "voice of silicon fascist privilege."

They may never be able to fill a stadium for a rock concert, but computers are making inroads in the music industry, capable of producing songs--and convincingly so--as illustrated at the South by Southwest festival in Texas. Already, an album featuring eight tracks has been produced entirely with artificial intelligence, an unprecedented feat. "I Am AI" was released last fall by YouTube star Taryn Southern, who doesn't know how to play any instruments. "For my first music video in 2017, I had a lot of friction as a non-musician," the young artist told a panel discussion on Sunday at the festival running from March 8-17. "I wrote lyrics, I had a melodic line but it was difficult to compose and record the actual music."

One day in March of 2000, six of Google's best engineers gathered in a makeshift war room. The company was in the midst of an unprecedented emergency. In October, its core systems, which crawled the Web to build an "index" of it, had stopped working. Although users could still type in queries at google.com, the results they received were five months out of date. More was at stake than the engineers realized. Google's co-founders, Larry Page and Sergey Brin, were negotiating a deal to power a search engine for Yahoo, and they'd promised to deliver an index ten times bigger than the one they had at the time--one capable of keeping up with the World Wide Web, which had doubled in size the previous year.

Modern massive datasets pose an enormous computational burden to practitioners. Distributed computation has emerged as a universal approach to ease the burden: Datasets are partitioned over machines, which compute locally, and communicate short messages. Distributed data also arises due to privacy reasons, such as in medicine. It is important to study how to do statistical inference and machine learning in a distributed setting. In this paper, we study one-step parameter averaging in statistical linear models under data parallelism. We do linear regression on each machine, and take a weighted average of the parameters. How much do we lose compared to doing linear regression on the full data? Here we study the performance loss in estimation error, test error, and confidence interval length in high dimensions, where the number of parameters is comparable to the training data size. We discover several key phenomena. First, averaging is not optimal, and we find the exact performance loss. Our results are simple to use in practice. Second, different problems are affected differently by the distributed framework. Estimation error and confidence interval length increases a lot, while prediction error increases much less. These results match simulations and a data analysis example. We rely on recent results from random matrix theory, where we develop a new calculus of deterministic equivalents as a tool of broader interest.

Understanding the learning dynamics of neural networks is one of the key issues for the improvement of optimization algorithms as well as for the theoretical comprehension of why deep neural nets work so well today. In this paper, we introduce a random matrix-based framework to analyze the learning dynamics of a single-layer linear network on a binary classification problem, for data of simultaneously large dimension and size, trained by gradient descent. Our results provide rich insights into common questions in neural nets, such as overfitting, early stopping and the initialization of training, thereby opening the door for future studies of more elaborate structures and models appearing in today's neural networks.