Race After Technology opens with a brief personal history set in the Crenshaw neighborhood of Los Angeles, where sociologist Ruha Benjamin spent a portion of her childhood. Recalling the time she set up shop on her grandmother's porch with a chalkboard and invited other kids to do math problems, she writes, "For the few who would come, I would hand out little slips of paper…until someone would insist that we go play tag or hide-and-seek instead. Needless to say, I didn't have that many friends!" As she gazed out the back window during car rides, she saw "boys lined up for police pat-downs," and inside the house she heard "the nonstop rumble of police helicopters overhead, so close that the roof would shake." The omnipresent surveillance continued when she visited her grandmother years later as a mother, her homecomings blighted by "the frustration of trying to keep the kids asleep with the sound and light from the helicopter piercing the window's thin pane." Benjamin's personal beginning sets the tone for her book's approach, one that focuses on how modern invasive technologies--from facial recognition software to electronic ankle monitors to the metadata of photos taken at protests--further racial inequality.

The right to contest a decision with consequences on individuals or the society is a well-established democratic right. Despite this right also being explicitly included in GDPR in reference to automated decision-making, its study seems to have received much less attention in the AI literature compared, for example, to the right for explanation. This paper investigates the type of assurances that are needed in the contesting process when algorithmic black boxes are involved, opening new questions about the interplay of contestability and explainability. We argue that specialised complementary methodologies to evaluate automated decision-making in the case of a particular decision being contested need to be developed. Further, we propose a combination of well-established software engineering and rule-based approaches as a possible socio-technical solution to the issue of contestability, one of the new democratic challenges posed by the automation of decision making.

ML-based predictive systems are increasingly used to support decisions with a critical impact on individuals' lives such as college admission, job hiring, child custody, criminal risk assessment, etc. As a result, fairness emerged as an important requirement to guarantee that predictive systems do not discriminate against specific individuals or entire sub-populations, in particular, minorities. Given the inherent subjectivity of viewing the concept of fairness, several notions of fairness have been introduced in the literature. This paper is a survey of fairness notions that, unlike other surveys in the literature, addresses the question of "which notion of fairness is most suited to a given real-world scenario and why?". Our attempt to answer this question consists in (1) identifying the set of fairness-related characteristics of the real-world scenario at hand, (2) analyzing the behavior of each fairness notion, and then (3) fitting these two elements to recommend the most suitable fairness notion in every specific setup. The results are summarized in a decision diagram that can be used by practitioners and policy makers to navigate the relatively large catalogue of fairness notions.

Increasingly, organizations are pairing humans with AI systems to improve decision-making and reducing costs. Proponents of human-centered AI argue that team performance can even further improve when the AI model explains its recommendations. However, a careful analysis of existing literature reveals that prior studies observed improvements due to explanations only when the AI, alone, outperformed both the human and the best human-AI team. This raises an important question: can explanations lead to complementary performance, i.e., with accuracy higher than both the human and the AI working alone? We address this question by devising comprehensive studies on human-AI teaming, where participants solve a task with help from an AI system without explanations and from one with varying types of AI explanation support. We carefully controlled to ensure comparable human and AI accuracy across experiments on three NLP datasets (two for sentiment analysis and one for question answering). While we found complementary improvements from AI augmentation, they were not increased by state-of-the-art explanations compared to simpler strategies, such as displaying the AI's confidence. We show that explanations increase the chance that humans will accept the AI's recommendation regardless of whether the AI is correct. While this clarifies the gains in team performance from explanations in prior work, it poses new challenges for human-centered AI: how can we best design systems to produce complementary performance? Can we develop explanatory approaches that help humans decide whether and when to trust AI input?

In a newly published paper on the preprint server Arxiv.org, Through techniques like compute-efficient machine learning, federated learning, and data sovereignty, the coauthors assert scientists and practitioners have the power to cut contributions to the carbon footprint while restoring trust in historically opaque systems. Sustainability, privacy, and transparency remain underaddressed and unsolved challenges in AI. In June 2019, researchers at the University of Massachusetts at Amherst released a study estimating that the amount of power required for training and searching a given model involves the emission of roughly 626,000 pounds of carbon dioxide -- equivalent to nearly 5 times the lifetime emissions of the average U.S. car. Partnerships like those pursued by DeepMind and the U.K.'s National Health Service conceal the true nature of AI systems being developed and piloted.

In 1906, in response to shocking reports about the disgusting conditions in U.S. meat-packing facilities, Congress created the Food and Drug Administration (FDA) to ensure safe and sanitary food production. In 1934, in the wake of the worst stock market crash in U.S. history, Congress created the Securities and Exchange Commission (SEC) to regulate capital markets. In 1970, as the nation became increasingly alarmed about the deterioration of the natural environment, Congress created the Environmental Protection Agency (EPA) to ensure cleaner skies and waters. When an entire field begins to create a broad set of challenges for the public, demanding thoughtful regulation, a proven governmental approach is to create a federal agency focused specifically on engaging with and managing that field. The time has come to create a federal agency for artificial intelligence.

According to the resolution, the where and how Level 3 will be allowed is spelled out in detailed specificity. Any self-driving car will require a data-collecting "black box" for the ALKS system. They will operate only in lanes without pedestrians or cyclists, there will have to be a physical "separation" between traffic moving in opposite directions and the cars will be limited to a maximum of 60 kilometres an hour while in self-driving mode. These last two might require some deft stickhandling since low speed often means urban centres with both pedestrian and cyclists. The cars will also have to be equipped with Driver Availability Recognition Systems which test whether the human seated at the wheel could take over if need be.

A number of AI researchers, data scientists, sociologists, and historians have written an open letter to end the publishing of research that claims artificial intelligence or facial recognition can predict whether a person is likely to be a criminal. The letter, signed by over 1000 experts, argues that data generated by the criminal justice system cannot be used to "identify criminals" or predict behaviour. Historical court and arrest data reflect the policies and practises of the criminal justice system and are therefore biased, the experts say. "These data reflect who police choose to arrest, how judges choose to rule, and which people are granted longer or more lenient sentences," the letter reads. Moreover, by continuing these studies, "'criminality' operates as a proxy for race due to racially discriminatory practices in law enforcement and criminal justice, research of this nature creates dangerous feedback loops" the letter says.

Two Google executives said Friday that bias in artificial intelligence is hurting already marginalized communities in America, and that more needs to be done to ensure that this does not happen. X. Eyeé, outreach lead for responsible innovation at Google, and Angela Williams, policy manager at Google, spoke at (Not IRL) Pride Summit, an event organized by Lesbians Who Tech & Allies, the world's largest technology-focused LGBTQ organization for women, non-binary and trans people around the world. In separate talks, they addressed the ways in which machine learning technology can be used to harm the black community and other communities in America -- and more widely around the world. Bias in algorithms IS NOT JUST A DATA PROBLEM. The choice to use AI can be biased, the way the algorithm learns can be biased, and the way users are impacted/interact with/perceive a system can reinforce bias! checkout @timnitGebru's work to learn more!

Retrosynthetic planning is a critical task in organic chemistry which identifies a series of reactions that can lead to the synthesis of a target product. The vast number of possible chemical transformations makes the size of the search space very big, and retrosynthetic planning is challenging even for experienced chemists. However, existing methods either require expensive return estimation by rollout with high variance, or optimize for search speed rather than the quality. In this paper, we propose Retro*, a neural-based A*-like algorithm that finds high-quality synthetic routes efficiently. It maintains the search as an AND-OR tree, and learns a neural search bias with off-policy data. Then guided by this neural network, it performs best-first search efficiently during new planning episodes. Experiments on benchmark USPTO datasets show that, our proposed method outperforms existing state-of-the-art with respect to both the success rate and solution quality, while being more efficient at the same time.