Lucas Rizzotto, a YouTuber from Brazil, had no idea what to expect when he gave his Alexa powered smart microwave a brain transplant, replacing the Amazon... 21.04.2022, What he created is a frightening abomination of a poet with an affinity towards Hitler, the British crown and the ending of what it calls the parasitic American empire. Oh, and it wants to kill its creator. There is that, too.Rizzotto, who makes humorous videos about technology projects he builds, used an imaginary friend he had as a child who happened to be embodied in his family's microwave as inspiration. "Magnetron" was a turn-of-the-century British poet who served in World War I, lost his family to the war and later became an expert StarCraft player.

The remarkable progress in Natural Language Processing (NLP) brought about by deep learning, particularly with the recent advent of large pre-trained neural language models, is brought into scrutiny as several studies began to discuss and report potential biases in NLP applications. Bias in NLP is found to originate from latent historical biases encoded by humans into textual data which gets perpetuated or even amplified by NLP algorithm. We present a survey to comprehend bias in large pre-trained language models, analyze the stages at which they occur in these models, and various ways in which these biases could be quantified and mitigated. Considering wide applicability of textual affective computing based downstream tasks in real-world systems such as business, healthcare, education, etc., we give a special emphasis on investigating bias in the context of affect (emotion) i.e., Affective Bias, in large pre-trained language models. We present a summary of various bias evaluation corpora that help to aid future research and discuss challenges in the research on bias in pre-trained language models. We believe that our attempt to draw a comprehensive view of bias in pre-trained language models, and especially the exploration of affective bias will be highly beneficial to researchers interested in this evolving field. The examples provided in this paper may be offensive in nature and may hurt your moral beliefs.

In Barcelona especially, physical remnants of this past abound. The city is known for its Catalan modernism, an iconic aesthetic popularized by Antoni Gaudí, the mastermind behind the Sagrada Familia. The architectural movement was born in part from the investments of wealthy Spanish families who amassed riches from their colonial businesses and funneled the money into lavish mansions. One of the most famous, known as the Casa Lleó Morera, was built early in the 20th century with profits made from the sugar trade in Puerto Rico. While tourists from around the world today visit the mansion for its beauty, Puerto Rico still suffers from food insecurity because for so long its fertile land produced cash crops for Spanish merchants instead of sustenance for the local people.

It was meant to be a temporary side job--a way to earn some extra money. Oskarina Fuentes Anaya signed up for Appen, an AI data-labeling platform, when she was still in college studying to land a well-paid position in the oil industry. But then the economy tanked in Venezuela. Inflation skyrocketed, and a stable job, once guaranteed, was no longer an option. Her side gig was now full time; the temporary now the foreseeable future.

Welcome to episode 4 of New voices in AI. In this episode Maria De-Arteaga shares her work and journey into algorithmic fairness and human algorithm collaboration. You can find out more on Maria's website and follow her on Twitter, @mariadearteaga. Daly: Hello and welcome to New Voices in AI the series from Ai hub where we celebrate the voices of Masters and PhD students, early career researchers and those with a new perspective on AI. I am Joe Daly, engagement manager for AI hub and this week I am talking to Maria De-Arteaga about some of her research.

Appen is among dozens of companies that offer data-labeling services for the AI industry. If you've bought groceries on Instacart or looked up an employer on Glassdoor, you've benefited from such labeling behind the scenes. Most profit-maximizing algorithms, which underpin e-commerce sites, voice assistants, and self-driving cars, are based on deep learning, an AI technique that relies on scores of labeled examples to expand its capabilities. The insatiable demand has created a need for a broad base of cheap labor to manually tag videos, sort photos, and transcribe audio. The market value of sourcing and coordinating that "ghost work," as it was memorably dubbed by anthropologist Mary Gray and computational social scientist Siddharth Suri, is projected to reach $13.7 billion by 2030.

Recurrent neural networks have seen widespread use in modeling dynamical systems in varied domains such as weather prediction, text prediction and several others. Often one wishes to supplement the experimentally observed dynamics with prior knowledge or intuition about the system. While the recurrent nature of these networks allows them to model arbitrarily long memories in the time series used in training, it makes it harder to impose prior knowledge or intuition through generic constraints. In this work, we present a path sampling approach based on principle of Maximum Caliber that allows us to include generic thermodynamic or kinetic constraints into recurrent neural networks. We show the method here for a widely used type of recurrent neural network known as long short-term memory network in the context of supplementing time series collected from different application domains. These include classical Molecular Dynamics of a protein and Monte Carlo simulations of an open quantum system continuously losing photons to the environment and displaying Rabi oscillations. Our method can be easily generalized to other generative artificial intelligence models and to generic time series in different areas of physical and social sciences, where one wishes to supplement limited data with intuition or theory based corrections.

In part two, we head to Venezuela, where AI data-labeling firms found cheap and desperate workers amid a devastating economic crisis, creating a new model of labor exploitation. The series also looks at ways to move away from these dynamics. In part three, we visit ride-hailing drivers in Indonesia who, by building power through community, are learning to resist algorithmic control and fragmentation. In part four, we end in Aotearoa, the Maori name for New Zealand, where an Indigenous couple are wresting back control of their community's data to revitalize its language. Together, the stories reveal how AI is impoverishing the communities and countries that don't have a say in its development--the same communities and countries already impoverished by former colonial empires.

Intelligent systems, such as Multi-Agent Systems (MAS), can be seen as a set of intelligent entities capable of proactively decide how to act to fulfill their own goals. These entities, called generally agents, are notoriously autonomous, i.e., they do not expect input from an user to act, and social, i.e., they usually communicate amongst each other to achieve common goals. Software systems are not easy to trust in general. This is especially true in the case of complex and distributed systems, such as MAS. Because of this, we need verification techniques to verify that such systems behave as expected. More specifically, in the case of MAS, it is relevant to know whether the agents are capable of achieving their own goals, by themselves or by collaborating with other agents by forming a coalition. This is usually referred to as the process of finding a strategy for the agent(s). A well-known formalism for reasoning about strategic behaviours in MAS is Alternating-time Temporal Logic (AT L) [1]. Before verifying AT L specifications, two questions need to be answered: (i) does each agent know everything about the system?

Since frequent severe droughts are lengthening the dry season in the Amazon Rainforest, it is important to detect wildfires promptly and forecast possible spread for effective suppression response. Current wildfire detection models are not versatile enough for the low-technology conditions of South American hot spots. This deep learning study first trains a Fully Convolutional Neural Network on Landsat 8 images of Ecuador and the Galapagos, using Green and Short-wave Infrared bands to predict pixel-level binary fire masks. This model achieves a 0.962 validation F2 score and a 0.932 F2 score on test data from Guyana and Suriname. Afterward, image segmentation is conducted on the Cirrus band using K-Means Clustering to simplify continuous pixel values into three discrete classes representing differing degrees of cirrus cloud contamination. Three additional Convolutional Neural Networks are trained to conduct a sensitivity analysis measuring the effect of simplified features on model accuracy and train time. The Experimental model trained on the segmented cirrus images provides a statistically significant decrease in train time compared to the Control model trained on raw cirrus images, without compromising binary accuracy. This proof of concept reveals that feature engineering can improve the performance of wildfire detection models by lowering computational expense.