An artificial intelligence algorithm has defeated a human F-16 fighter pilot in a virtual dogfight simulation. The Aug. 20 event was the finale of the Pentagon research agency's AI air combat competition. The algorithm, developed by Heron Systems, easily defeated the fighter pilot in all five rounds that capped off a yearlong competition hosted by the Defense Advanced Research Projects Agency. The competition, called the AlphaDogfight Trials, was part of DARPA's Air Combat Evolution program, which is exploring automation in air-to-air combat and looking to improve human trust in AI systems. "It's easy to go down the wrong path of thinking that that is either A) definitive in some way as to what the future of [basic fighter maneuvers will be]; or B) that it is a bad outcome," said Justin Mock of DARPA, a fighter pilot and commentator for the trials.

Jaynarayan H. Lala (jay.lala@rtx.com) is a Senior Principal Engineering Fellow at Raytheon Technologies, San Diego, CA, USA. Carl E. Landwehr (carl.landwehr@gmail.com) is a Research Scientist at George Washington University and a Visiting Professor at University of Michigan, Ann Arbor, MI, USA. John F. Meyer (jfm@umich.edu) is a Professor Emeritus of Computer Science and Engineering at University of Michigan, Ann Arbor, MI, USA. This Viewpoint is derived from material produced as part of the Intelligent Vehicle Dependability and Security (IVDS) project of IFIP Working Group 10.4.

Shifting artificial intelligence to the "edge" of the network could transform computing . . . and everyday life.

Automated machine learning is likely to introduce two critical problems. Fortunately, vendors are introducing tools to tackle both of them. Adoption of automated machine learning -- tools that help data scientists and business analysts (and even business users) automate the construction of machine learning models -- is expected to increase over the next few years because these tools simplify model building. For example, in some of the tools, all the user needs to do is specify the outcome or target variable of interest along with the attributes believed to be predictive. The automated machine learning (autoML) platform picks the best model.

The algorithms build on FAIR's work in January of this year, when an agent was trained in Habitat to navigate unfamiliar environments without a map. Using just a depth-sensing camera, GPS, and compass data, it learned to enter a space much as a human would, and find the shortest possible path to its destination without wrong turns, backtracking, or exploration. The first of these new algorithms can now build a map of the space at the same time, allowing it to remember the environment and navigate through it faster if it returns. The second improves the agent's ability to map the space without needing to visit every part of it. Having been trained on enough virtual environments, it is able to anticipate certain features in a new one; it can know, for example, that there is likely to be empty floor space behind a kitchen island without navigating to the other side to look.

It is an increasingly common modern annoyance: arriving at the front of the queue to pay in a shop, pulling out a smartphone for a hygienic contact-free payment, and staring down at an error message because your phone fails to recognise your masked face. As more and more nations mandate masks to prevent the spread of coronavirus, technology companies are scrambling to keep up with the changing world. But some experts are warning that the change may have to start with users themselves. Apple's Face ID is the most well-known example of a consumer facial verification system. The technology, which uses a grid of infrared dots to measure the physical shape of a user's face, secures access to the company's iPhones and iPads, as well as other features such as Apple Pay.

Variability is a problem for the scalability of semiconductor quantum devices. The parameter space is large, and the operating range is small. Our statistical tuning algorithm searches for specific electron transport features in gate-defined quantum dot devices with a gate voltage space of up to eight dimensions. Starting from the full range of each gate voltage, our machine learning algorithm can tune each device to optimal performance in a median time of under 70 minutes. This performance surpassed our best human benchmark (although both human and machine performance can be improved). The algorithm is approximately 180 times faster than an automated random search of the parameter space, and is suitable for different material systems and device architectures. Our results yield a quantitative measurement of device variability, from one device to another and after thermal cycling. Our machine learning algorithm can be extended to higher dimensions and other technologies. To optimize operating conditions of large scale semiconductor quantum devices, a large parameter space has to be explored. Here, the authors report a machine learning algorithm to navigate the entire parameter space of gate-defined quantum dot devices, showing about 180 times faster than a pure random search.

Researchers at the Research Center for Molecular Medicine of the Austrian Academy of Sciences have created knowledge-primed neural networks (KPNNs) which utilize signaling pathways and gene-regulatory networks. Each node in a KPNN corresponds to a protein or gene, while each edge possesses a mechanistic biological interpretation. By requiring this closer correspondence, KPNNs integrate deep learning with the interpretability of biological network models, yielding tangible insights into biological systems with high prediction performance. KPNNs are especially applicable to single-cell RNA-seq data, which is produced at massive scale with single-cell sequencing assays. The findings illustrate the future impact that artificial intelligence (AI) and deep learning will have on mechanistic biology as the scientific community learns to add biologically interpretability to AI outcomes, the researchers say.

One of the world's smallest microrobots is able to carry 2.6 times its own body weight thanks to a muscular system powered by alcohol. Conventionally, the "muscles" of small robots have been tethered to an external power source. Alternatively, they have been powered by batteries, the weight and size of which have limited efficiency and how small the robots can be. Top-of-the-range batteries have an energy density of around 1.8 megajoules per kilogram, a fraction of what you get from animal fat, which is about 38 MJ/kg. The methanol-powered muscles used by RoBeetle, an 88-milligram-long microrobot, can use catalytic combustion to reach energy levels up to 20 MJ/kg.

Washington, DC (CNN)Michigan announced Thursday that it's teaming with tech and auto companies to attempt to retrofit a roughly 40-mile stretch of two roads outside Detroit exclusively for self-driving vehicles. Michigan's partners include Ford and Sidewalk Infrastructure Partners, a company that Alphabet has invested in. Alphabet owns Google (GOOG) and Waymo, one of the companies at the forefront of developing self-driving vehicles. Both Interstate 94 and Michigan Avenue between Detroit and Ann Arbor, Michigan, would be retrofitted to include a dedicated lane for self-driving vehicles. Sensors and cameras added to the roads would help the vehicles better understand their surroundings.