Lo-fi hip hop music has been my go-to study buddy ever since college. It creates a cozy and calming vibe with a relatively simple musical structure. Some jazzy chord progressions, groovy drum patterns, ambience sounds, and nostalgic movie quotes can give us a pretty decent sounding lo-fi hip hop track. On top of the musical side, animated visuals are also a crucial part of the lo-fi aesthetics, setting the ambience along with the nature sounds of water, wind, and fire. The idea to create my own lo-fi web player occurred to me one Sunday afternoon when I was learning about deep generative models.

Again, these are broad generalizations and there is a lot of diversity within each generation. It's important to remember that individuals are more than just their generation, and that people of all ages can have a wide range of characteristics and experiences. The format of the above response is typical for queries that require factual information. ChatGPT begins with a generic introduction, then offers several facts, and ends with a suggestion to consider additional information for a more comprehensive understanding.

Social media company and Snapchat maker Snap has for years defined itself as a "camera company," despite its failures to turn its photo-and-video recording glasses known as Spectacles into a mass-market product and, more recently, its decision to kill off its camera-equipped drone. But that hasn't stopped the company from envisioning a future where AR glasses are a commonly used device, and one, as the company revealed on Tuesday's fourth-quarter earnings call, that will eventually be powered by AI technology. Investors wanted to get a sense of how Snap was thinking about the latest developments in AI -- particularly in buzzy areas like generative AI. which has benefited from advances in algorithms, language models and the increased processing power available to run the necessary calculations. One pointed to the AI image generator Midjourney's bot for Discord as an example of how AI could lead to increased user engagement within an app. Snap CEO Evan Spiegel agreed that, in the near term, there were a lot of opportunities to use generative AI to make Snap's camera more powerful.

Have you ever seen the movie The Terminator? If you have, then you'll get where I'm going with this. If you haven't, then I've aged myself big time, and here's a quick summary: The Terminator is based on a cyborg assassin (played by Arnold Schwarzenegger) who is sent back in time from 2029 to 1984 to kill Sarah Connor, whose unborn son is destined to save mankind from extinction by a company called Skynet. Skynet is an Artificial Intelligence (A.I.) industry responsible for a post-apocalyptic future where robots take over the world. While this movie (and its many sequels) became a big hit all over the world, it's also one of the movies that have inspired the general public to not trust A.I. I can't tell you how many times a new piece of technology is created without hearing some form of warning or fear: "Watch out for the robots will come and take over the world!"

Modern deep learning systems are increasingly deployed in situations such as personalization and federated learning where it is necessary to support i) learning on small amounts of data, and ii) communication efficient distributed training protocols. In this work, we develop FiLM Transfer (FiT) which fulfills these requirements in the image classification setting by combining ideas from transfer learning (fixed pretrained backbones and fine-tuned FiLM adapter layers) and meta-learning (automatically configured Naive Bayes classifiers and episodic training) to yield parameter efficient models with superior classification accuracy at low-shot. The resulting parameter efficiency is key for enabling few-shot learning, inexpensive model updates for personalization, and communication efficient federated learning. We experiment with FiT on a wide range of downstream datasets and show that it achieves better classification accuracy than the leading Big Transfer (BiT) algorithm at low-shot and achieves state-of-the art accuracy on the challenging VTAB-1k benchmark, with fewer than 1% of the updateable parameters. Finally, we demonstrate the parameter efficiency and superior accuracy of FiT in distributed low-shot applications including model personalization and federated learning where model update size is an important performance metric.

This paper describes deep learning models based on convolutional neural networks applied to the problem of predicting EM wave propagation over rural terrain. A surface integral equation formulation, solved with the method of moments and accelerated using the Fast Far Field approximation, is used to generate synthetic training data which comprises path loss computed over randomly generated 1D terrain profiles. These are used to train two networks, one based on fractal profiles and one based on profiles generated using a Gaussian process. The models show excellent agreement when applied to test profiles generated using the same statistical process used to create the training data and very good accuracy when applied to real life problems.

While contextualized word embeddings have been a de-facto standard, learning contextualized phrase embeddings is less explored and being hindered by the lack of a human-annotated benchmark that tests machine understanding of phrase semantics given a context sentence or paragraph (instead of phrases alone). To fill this gap, we propose PiC -- a dataset of ~28K of noun phrases accompanied by their contextual Wikipedia pages and a suite of three tasks for training and evaluating phrase embeddings. Training on PiC improves ranking models' accuracy and remarkably pushes span-selection (SS) models (i.e., predicting the start and end index of the target phrase) near-human accuracy, which is 95% Exact Match (EM) on semantic search given a query phrase and a passage. Interestingly, we find evidence that such impressive performance is because the SS models learn to better capture the common meaning of a phrase regardless of its actual context. SotA models perform poorly in distinguishing two senses of the same phrase in two contexts (~60% EM) and in estimating the similarity between two different phrases in the same context (~70% EM).

In 2016, Studio Ghibli co-founder and director Hayao Miyazaki, responsible for beloved anime classics like Princess Mononoke and Kiki's Delivery Service, made headlines around the world for his reaction to an AI animation program. "I would never wish to incorporate this technology into my work at all," Miyazaki told the software engineers who came to show their creation to him. "I strongly feel that this is an insult to life itself." A half-decade later, artificial intelligence and the potential role it could play in anime productions is once again in the spotlight. This week, Netflix shared Dog and Boy, an animated short the streaming giant described as an "experimental effort" to address the anime industry's ongoing labor shortage.

Will artificial intelligence (AI) soon replace journalists? Many have been asking this question since the boom of generative AI tools such as ChatGPT, which can write a high school essay, a poem, or even pass a medical licensing exam in a matter of seconds. Now, AI tools are seeping into newsrooms. CNET, an American tech news outlet, has acknowledged using AI to write financial articles, seemingly as early as November 2022. When looking more closely at the articles on CNET, a disclaimer reads: "This article was assisted by an AI engine and reviewed, fact-checked and edited by our editorial staff".

Spain's regional elections are still nearly four months away, but Irene Larraz and her team at Newtral are already braced for impact. Each morning, half of Larraz's team at the Madrid-based media company sets a schedule of political speeches and debates, preparing to fact-check politicians' statements. The other half, which debunks disinformation, scans the web for viral falsehoods and works to infiltrate groups spreading lies. Once the May elections are out of the way, a national election has to be called before the end of the year, which will likely prompt a rush of online falsehoods. "It's going to be quite hard," Larraz says.