What do companies like Netflix, Airbnb, Spotify and Lyft know that you don't know? What do these companies have that you might be lacking? Sure, they had an initial idea. But we've all had good ideas. How did each of these new market heroes move from the seed of an idea to the disruption of an entire industry?

One way to analyse the behaviour of machine learning models is through local explanations that highlight input features that maximally influence model predictions. Sensitivity analysis, which involves analysing the effect of input perturbations on model predictions, is one of the methods to generate local explanations. Meaningful input perturbations are essential for generating reliable explanations, but there exists limited work on what such perturbations are and how to perform them. This work investigates these questions in the context of machine listening models that analyse audio. Specifically, we use a state-of-the-art deep singing voice detection (SVD) model to analyse whether explanations from SoundLIME (a local explanation method) are sensitive to how the method perturbs model inputs. The results demonstrate that SoundLIME explanations are sensitive to the content in the occluded input regions. We further propose and demonstrate a novel method for quantitatively identifying suitable content type(s) for reliably occluding inputs of machine listening models. The results for the SVD model suggest that the average magnitude of input mel-spectrogram bins is the most suitable content type for temporal explanations.

Internet speeds are still running largely as normal despite the increased pressure of lockdown, according to research from regulator Ofcom. Download speeds have only dropped by an average of 2 per cent, according to the research, even with the extra load. That is despite some high-profile outages, including Virgin Media problems that took the internet offline for users across the country. Networks have been under increased strain with more people across the country working from home, children using online platforms to carry on school work, and greater gaming and streaming as a source of entertainment. The communications regulator measured broadband performance for 3,481 users at the beginning and end of March, to compare results before and after lockdown started.

The digital age, like the age of the printing press, has transformed our modes of social interaction -- our reading and writing habits and methods of communication and consumption. It is also disrupting long-standing institutions, threatening old hierarchies of knowledge and power. The benefits of this transformation are manifest. But the near universal availability of information also brings certain dangers. One is the prevalence of disinformation online, and the attendant challenge of discriminating between reliable and unreliable information. Both the need for reliable information and the dangers of disinformation have been thrown into sharp relief by the novel coronavirus pandemic.

The "tourist guy" pic, allegedly taken on top of the World Trade Center in NYC just seconds before the 9/11 attacks, is a classic example of a sensational, fabricated image circulated to millions of internet users, then via email. Misleading news, doctored visuals, disguised propaganda: The spread of digital mis- and disinformation has become a major problem over the last years, and it looks like the Corona pandemic has led to yet another peak. What better time to start a new verification project: KID, kicked-off in April 2020, will help counter the "infodemic" by fostering media and information literacy (MIL) and exploring artificial intelligence (AI) tools to improve verification processes. The acronym KID resolves to the German "Künstliche Intelligenz gegen Desinformation" (Artificial Intelligence against Mis-/Disinformation) and is funded by the BKM (Beauftragte der Bundesregierung für Kultur und Medien the German Government's Commissioner for Culture and Media). KID follows in the footsteps of and is partly linked to other DW verification/MIL projects and services such as Digger, Truly Media, or WeVerify.

WEST LAFAYETTE, Ind. -- Just to solve a puzzle or play a game, artificial intelligence can require software running on thousands of computers. That could be the energy that three nuclear plants produce in one hour. A team of engineers has created hardware that can learn skills using a type of AI that currently runs on software platforms. Sharing intelligence features between hardware and software would offset the energy needed for using AI in more advanced applications such as self-driving cars or discovering drugs. "Software is taking on most of the challenges in AI. If you could incorporate intelligence into the circuit components in addition to what is happening in software, you could do things that simply cannot be done today," said Shriram Ramanathan, a professor of materials engineering at Purdue University.

Turning night into day, clearing up fog, and removing cloud cover in image and video sources, in real-time, Digital Harmonic's PurePixel solution will help the Federal mission secure critical areas and communities, enabling our military to make better-informed command and control decisions. Digital Harmonic LLC, a leading image, video, and signal processing technology company, announced today they have chosen to collaborate with Dell Technologies OEM Embedded & Edge Solutions to bring PurePixel, designed on a scalable suite of hardware solutions from edge devices to rack mounted servers, to federal agencies. PurePixel is an upstream pre-processing component to enhance the quality and efficiency of machine learning (ML) and computer vision (CV) algorithms increasing the success of the output for real-time video. PurePixel also improves advanced analytics software with cloud-based ML algorithms for object recognition, object detection, image annotation/labeling, semantic image segmentation analysis, and edge computing capabilities. One of the options for delivering PurePixel to customers is on the Dell EMC PowerEdge C4140 server, which is an ultra-dense, accelerator optimized rack server system purpose-built for artificial intelligence (AI) solutions with a leading GPU-accelerated infrastructure.

Eurovision 2020 was unsurprisingly cancelled due to the pandemic, but AI has stepped in to fill its glittery shoes. Dutch broadcaster VPRO has just wrapped up a Eurovision-inspired AI Song Contest, with 13 teams from Europe and Australia training algorithms to become budding pop stars while experts judge their output. As BBC and Bloomberg point out, the results are a mix of surprisingly well-done and frighteningly dystopic tunes... a bit like the real thing, really. The winning entry came from Australian team Uncanny Valley, whose song "Beautiful the World" was built by an AI trained on a mix of Eurovision hits and local animals affected by wildfires, including koalas, kookaburras and Tasmanian devils. It has the same catchy dance pop riffs you'd expect to get the full douze points from a Eurovision vote, just with nonsensical lyrics.

If you ask someone what comes to their mind after hearing the words'artificial intelligence', they would probably answer something like'robots, self-driving cars' etc. However, AI goes way beyond that. Furthermore, you have probably already encountered it in your life but didn't even know it. The perfect example is Siri. AI is more than just robots.

The news: In its latest Community Standards Enforcement Report, released today, Facebook detailed the updates it has made to its AI systems for detecting hate speech and disinformation. The tech giant says 88.8% of all the hate speech it removed this quarter was detected by AI, up from 80.2% in the previous quarter. The AI can remove content automatically if the system has high confidence that it is hate speech, but most is still checked by a human being first. Behind the scenes: The improvement is largely driven by two updates to Facebook's AI systems. First, the company is now using massive natural-language models that can better decipher the nuance and meaning of a post.