To be qualified for the position, you must have a MSc degree in Electrical or Mechanical Engineering or Computer Science or related subject. Perfect scientific skills with excellence in real life experimentation, as well as very good communication skills are considered as a strong plus. You will represent the group in different occasions, both in Sweden and world-wide, hence very good knowledge in English is a must. As a PhD you will perform research with substantial theoretical and experimental components that should be published in peer-reviewed major international journals and at major conferences. The position will include supervision of MSc and to assist in grant applications from research funding agencies/councils, the EU framework program or the industry.

A Hungarian-born startup in Copenhagen can detect if a video is fake, but no one was a buyer of their technology. With a clever change of direction and building up of the Colossyan brand, they are now trying to take the lead in an emerging market: they are producing manipulated videos themselves, but for ethical purposes. The market for synthetic media is growing, but the risk is still huge. The consortium of investors, led by the Hungarian Dayone Capital, also knows this. Exclusive excerpts from a business story that says more than anything about the age we live in.

November 24, 2021 - The Federal Trade Commission has issued limited guidance in the area of artificial intelligence and machine learning (AI), but through its enforcement actions and press releases has made clear its view that AI may pose issues that run afoul of the FTC Act's prohibition against unfair and deceptive trade practices. In recent years it has pursued enforcement actions involving automated decision-making and results generated by computer algorithms and formulas, which are some common uses of AI in the financial sector but may also be relevant in other contexts such as health care. In FTC v. CompuCredit Corp., FTC Case No. 108-CV-1976 (2008), the FTC alleged that subprime credit marketer CompuCredit violated the FTC Act by deceptively failing to disclose that it used a behavioral scoring model to reduce consumers' credit limits. If cardholders used their credit cards for cash advances or to make payments at certain venues, such as bars, nightclubs and massage parlors, their credit limit might be reduced. The company, the FTC alleged, did not inform consumers that these purchases could reduce their credit limit, neither at the time they signed up nor at the time they reduced the credit limit.

The Department of Homeland Security issued a request for comment for 11 proposed research topics the agency intends to pursue, with hopes that eligible small business partners will become more aware and attuned to those areas of study. The research areas were announced in a Nov. 16 pre-solicitation through Homeland's Science and Technology Directorate Small Business Innovation Research (SBIR) Program. These proposed areas include automated artificial intelligence sensing technology, counterfeit microelectronic detection, a broadband interoperability platform, biological hazard detection, a mass fatality tracking system, a wearable detector for chemical threats, low cost diagnostic devices, and streamlined airport checkpoint technology for passengers with limited mobility. "The topics span a broad range of homeland security needs that give small businesses the opportunity to partner with DHS and turn their ideas into effective solutions," Dusty Lang, the DHS SBIR director, said in a press release. "I encourage all innovative small businesses to review the topics in the Pre-Solicitation to better understand our research and development needs."

Often described as austere and even robotic, Social Democrat Olaf Scholz nonetheless managed to inspire German voters in this year's election with a campaign that played on his reputation as a safe pair of hands. Scholz, 63, is on the brink of becoming the next German chancellor after leading his party to a surprise victory in September's vote, relieving Angela Merkel of her duties after 16 years. Together with the leaders of the ecologist Greens and the liberal FDP, he unveiled a deal for Germany's next coalition on Wednesday. The Social Democrats (SPD) had begun the election campaign at rock bottom in the polls, with many completely writing off Scholz's chances of becoming chancellor -- so much so that he doesn't even have an official biography. Olaf Scholz staged an upset poll win by positioning himself as the best candidate to continue Angela Merkel's legacy as German chancellor Photo: AFP / Odd ANDERSEN But Scholz managed to stage a stunning upset by positioning himself as the best candidate to continue Merkel's legacy, even adopting her famous "rhombus" hand gesture on a magazine cover.

Collision avoidance is a vital capability of any marine vessel navigating in public waterways; this is particularly true for autonomous surface vehicles (ASVs), which cannot benefit by the real-time guidance of a human operator. Safe maritime navigation remains a challenge due to the fact that it requires the seamless coordination of multiple complex subsystems. First, vessels must be able to perceive their surroundings under a wide range of environmental conditions. This is typically accomplished using one or more line-of-sight sensors, which emit electromagnetic or acoustic signals, and detect the reflections produced by nearby obstacles (Robinette et al., 2019). However, in the marine environment, vessels can also utilize the Automatic Information System (AIS) protocol to track nearby vessels. The merits and drawbacks of these sensing modalities will be discussed in Section 1.1. Once an obstacle is detected, the ASV must react quickly and intelligently to avoid it, in accordance with the "rules of the road" set forth by the 1972 International Regulations for Prevention of Collisions at Sea (COLREGs) (International Maritime Organization, 2003). Many ASVs remain unable to perform one or more of these crucial tasks, limiting their adoption beyond the oceanographic research community. B. Cole is with the Laboratory for Autonomous Marine Sensing Systems, Department of Mechanical Engineering.

Simulation-based inference with conditional neural density estimators is a powerful approach to solving inverse problems in science. However, these methods typically treat the underlying forward model as a black box, with no way to exploit geometric properties such as equivariances. Equivariances are common in scientific models, however integrating them directly into expressive inference networks (such as normalizing flows) is not straightforward. We here describe an alternative method to incorporate equivariances under joint transformations of parameters and data. Our method -- called group equivariant neural posterior estimation (GNPE) -- is based on self-consistently standardizing the "pose" of the data while estimating the posterior over parameters. It is architecture-independent, and applies both to exact and approximate equivariances. As a real-world application, we use GNPE for amortized inference of astrophysical binary black hole systems from gravitational-wave observations. We show that GNPE achieves state-of-the-art accuracy while reducing inference times by three orders of magnitude.

Among the several paradigms of artificial intelligence (AI) or machine learning (ML), a remarkably successful paradigm is deep learning. Deep learning's phenomenal success has been hoped to be interpreted via fundamental research on the theory of deep learning. Accordingly, applied research on deep learning has spurred the theory of deep learning-oriented depth and breadth of developments. Inspired by such developments, we pose these fundamental questions: can we accurately approximate an arbitrary matrix-vector product using deep rectified linear unit (ReLU) feedforward neural networks (FNNs)? If so, can we bound the resulting approximation error? In light of these questions, we derive error bounds in Lebesgue and Sobolev norms that comprise our developed deep approximation theory. Guided by this theory, we have successfully trained deep ReLU FNNs whose test results justify our developed theory. The developed theory is also applicable for guiding and easing the training of teacher deep ReLU FNNs in view of the emerging teacher-student AI or ML paradigms that are essential for solving several AI or ML problems in wireless communications and signal processing; network science and graph signal processing; and network neuroscience and brain physics.

The concept of meaningful human control has been proposed to address responsibility gaps and mitigate them by establishing conditions that enable a proper attribution of responsibility for humans (e.g., users, designers and developers, manufacturers, legislators). However, the relevant discussions around meaningful human control have so far not resulted in clear requirements for researchers, designers, and engineers. As a result, there is no consensus on how to assess whether a designed AI system is under meaningful human control, making the practical development of AI-based systems that remain under meaningful human control challenging. In this paper, we address the gap between philosophical theory and engineering practice by identifying four actionable properties which AI-based systems must have to be under meaningful human control. First, a system in which humans and AI algorithms interact should have an explicitly defined domain of morally loaded situations within which the system ought to operate. Second, humans and AI agents within the system should have appropriate and mutually compatible representations. Third, responsibility attributed to a human should be commensurate with that human's ability and authority to control the system. Fourth, there should be explicit links between the actions of the AI agents and actions of humans who are aware of their moral responsibility. We argue these four properties are necessary for AI systems under meaningful human control, and provide possible directions to incorporate them into practice. We illustrate these properties with two use cases, automated vehicle and AI-based hiring. We believe these four properties will support practically-minded professionals to take concrete steps toward designing and engineering for AI systems that facilitate meaningful human control and responsibility.

We focus on some shortcomings in current data protection regulation's ability to adequately address the ramifications of AI-driven data processing practices, in particular those of combining data sets. We propose that privacy regulation relies less on individuals' privacy expectations and recommend regulatory reform in two directions: (1) abolishing the distinction between personal and anonymized data for the purposes of triggering the application of data protection laws and (2) developing methods to prioritize regulatory intervention based on the level of privacy risk posed by individual data processing actions. This is an interdisciplinary paper that intends to build a bridge between the various communities involved in privacy research. We put special emphasis on linking technical notions with their regulatory implications and introducing the relevant technical and legal terminology in use to foster more efficient coordination between the policymaking and technical communities and enable a timely solution of the problems raised.