Media bias can significantly impact the formation and development of opinions and sentiments in a population. It is thus important to study the emergence and development of partisan media and political polarization. However, it is challenging to quantitatively infer the ideological positions of media outlets. In this paper, we present a quantitative framework to infer both political bias and content quality of media outlets from text, and we illustrate this framework with empirical experiments with real-world data. We apply a bidirectional long short-term memory (LSTM) neural network to a data set of more than 1 million tweets to generate a two-dimensional ideological-bias and content-quality measurement for each tweet. We then infer a ``media-bias chart'' of (bias, quality) coordinates for the media outlets by integrating the (bias, quality) measurements of the tweets of the media outlets. We also apply a variety of baseline machine-learning methods, such as a naive-Bayes method and a support-vector machine (SVM), to infer the bias and quality values for each tweet. All of these baseline approaches are based on a bag-of-words approach. We find that the LSTM-network approach has the best performance of the examined methods. Our results illustrate the importance of leveraging word order into machine-learning methods in text analysis.

Voting is used widely to identify a collective decision for a group of agents, based on their preferences. In this paper, we focus on evaluating and designing voting rules that support both the privacy of the voting agents and a notion of fairness over such agents. To do this, we introduce a novel notion of group fairness and adopt the existing notion of local differential privacy. We then evaluate the level of group fairness in several existing voting rules, as well as the trade-offs between fairness and privacy, showing that it is not possible to always obtain maximal economic efficiency with high fairness or high privacy levels. Then, we present both a machine learning and a constrained optimization approach to design new voting rules that are fair while maintaining a high level of economic efficiency. Finally, we empirically examine the effect of adding noise to create local differentially private voting rules and discuss the three-way trade-off between economic efficiency, fairness, and privacy. This paper appears in the special track on AI & Society.

Check out the on-demand sessions from the Low-Code/No-Code Summit to learn how to successfully innovate and achieve efficiency by upskilling and scaling citizen developers. Using AI in the real world remains challenging in many ways. Organizations are struggling to attract and retain talent, build and deploy AI models, define and apply responsible AI practices, and understand and prepare for regulatory framework compliance. At the same time, the DeepMinds, Googles and Metas of the world are pushing ahead with their AI research. Their talent pool, experience and processes around operationalizing AI research rapidly and at scale puts them on a different level from the rest of the world, creating a de facto AI divide. These are 4 AI research trends that the tech giants are leading on, but everyone else will be talking about and using in the near future.

This is not Beamery's first audit of its AI tools. It conducted internal audits to test for compliance with General Data Protection Regulation, the 2016 European Union law that protects consumer identity and privacy. For AI anti-bias audits that fall under the New York City law, Beamery sought to test how its talent acquisition tools handle a potential job candidate's gender and ethnicity during the recruitment process. The first audit took place in the summer followed by a month-long audit in October.

A deepfake version of Meta CEO Mark Zuckerberg thanked Congressional Democrats for their'service and inaction' on antitrust legislation. The eerie, convincing clip is the work of advocacy group Demand Progress Action, which used deepfake technology to turn an actor into Zuckerberg - who thanks Democratic leaders Nancy Pelosi and Chuck Schumer for holding up two major pieces of antitrust legislation this year. 'Over the past five years, Congress has held over 30 hearings designed to hold Big Tech accountable,' fake Zuckerberg says in the ad, which the liberal group plans to use for television ads in New York and Washington, D.C. 'Sometimes you land a punch.' 'Most of the time, it felt like playing paddy cake,' fake Zuckerberg boasts. 'So I'd like to propose a toast to Speaker of the House Nancy Pelosi and Senate Majority Leader Chuck Schumer (above),' fake Zuckerberg says, holding a glass of champagne We then see a clip of Republican Sen. Hatch asking him how he sustains a business model where users don't pay for services. A grinning Zuckerberg says: 'Senator, we run ads.' 'Either way, it looks like the most consequential action that Congress is poised to take, a bipartisan bill to prevent companies like mine from self-dealing, is about to fade away like so many efforts to rein in big tech in the past.'

The Bureau of Multiversal Arbitration is an unusual workplace. Maude Fletcher's alright, though she needs to learn how to turn off caps lock in the company chat. But trying to deal with Byron G Snodgrass is like handling an energetic poodle, and Phil is a bit stiff. Byron G Snodgrass is an energetic poodle. A peace lily, I think.

When "virtual" became the standard medium in early 2020 for business communications from board meetings to office happy hours, companies like Zoom found themselves hot in demand. They also became prime targets for the next big cyberattack. In the case of Zoom, hackers succeeded in April 2020, with a large data breach that exposed an estimated 500,000 user passwords. This breach, along with the thousands of daily reported cyber attacks in that year show that of the major threats facing businesses today, cybercrime nears the top of the list. Due to a combination of factors, including operations that have moved online and sensitive information crossing remote networks, opportunities for cybercrime are rife and the attacks are becoming more advanced and difficult to contain.

A lobby group backed by Elon Musk and associated with a controversial ideology popular among tech billionaires is fighting to prevent killer robots from terminating humanity, and it's taken hold of Europe's Artificial Intelligence Act to do so. The Future of Life Institute (FLI) has over the past year made itself a force of influence on some of the AI Act's most contentious elements. Despite the group's links to Silicon Valley, Big Tech giants like Google and Microsoft have found themselves on the losing side of FLI's arguments. In the EU bubble, the arrival of a group whose actions are colored by fear of AI-triggered catastrophe rather than run-of-the-mill consumer protection concerns was received like a spaceship alighting in the Schuman roundabout. Some worry that the institute embodies a techbro-ish anxiety about low-probability threats that could divert attention from more immediate problems.

While most of the robots listed in the SFPD's inventory are primarily used for defusing bombs or dealing with hazardous materials, newer Remotec models have an optional weapons system, and the department's existing F5A has a tool called the PAN disruptor that can load 12-gauge shotgun shells. It's typically used to detonate bombs from a distance. The department's QinetiQ Talon can also be modified to hold various weapons -- a weaponized version of the robot is currently used by the US Army and can equip grenade launchers, machine guns, or even a .50-caliber

Noninvasive X-ray imaging of nanoscale three-dimensional objects, e.g. integrated circuits (ICs), generally requires two types of scanning: ptychographic, which is translational and returns estimates of complex electromagnetic field through ICs; and tomographic scanning, which collects complex field projections from multiple angles. Here, we present Attentional Ptycho-Tomography (APT), an approach trained to provide accurate reconstructions of ICs despite incomplete measurements, using a dramatically reduced amount of angular scanning. Training process includes regularizing priors based on typical IC patterns and the physics of X-ray propagation. We demonstrate that APT with 12-time reduced angles achieves fidelity comparable to the gold standard with the original set of angles. With the same set of reduced angles, APT also outperforms baseline reconstruction methods. In our experiments, APT achieves 108-time aggregate reduction in data acquisition and computation without compromising quality. We expect our physics-assisted machine learning framework could also be applied to other branches of nanoscale imaging.