Last month, dermatologists were told they had narrowly lost a competition. "Man against machine," a study by Holger Haenssle and colleagues, found that artificial intelligence known as deep learning convolutional neural network edged out 58 dermatologists in the photographic diagnosis of melanoma. They were charged with differentiating melanoma from benign moles using images obtained via dermoscopy, a technique that allows dermatologists to view the skin through a high-quality magnifying lens with a powerful lighting system. This story made headlines: "AI beats doctors at cancer diagnoses." As dermatologists, we read the headlines with surprise and thought, "Aren't we on the same team?"

Neural networks are the crown jewel of the AI boom. They gorge on data and do things like transcribe speech or describe images with near-perfect accuracy (see "10 breakthrough technologies 2013: Deep learning"). The catch is that neural nets, which are modeled loosely on the structure of the human brain, are typically constructed in software rather than hardware, and the software runs on conventional computer chips. IBM has now shown that building key features of a neural net directly in silicon can make it 100 times more efficient. Chips built this way might turbocharge machine learning in coming years.

This paper aims to detect linguistic characteristics and grammatical patterns from speech transcriptions generated by Alzheimer's disease (AD) patients. This is considered an important application of natural language processing and deep learning techniques for computational health. The authors propose several neural models such as CNNs and LSTM-RNNs -- and combinations of them -- to enhance an AD classification task. The trained neural models are used to interpret linguistic characteristics of AD patients (including gender variation) via activation clustering and first-derivative saliency techniques. Language variation can serve as a proxy that monitors how patients' cognitive functions have been affected (e.g., issues with word finding and impaired reasoning).

Researchers at a British university are working on an artificial intelligence (AI) engine that can diagnose illness simply by smelling the breath of a person. Andrea Soltoggio, a member of the data science team at Loughborough University, said the engine is being taught how to identify a range of illness-revealing substances that humans might exhale. "Compared to that of animals, the human sense of smell is far less developed and certainly not used to carry out daily activities. For this reason, humans aren't particularly aware of the richness of information that can be transmitted through the air, and can be perceived by a highly sensitive olfactory system. AI may be about to change that," Soltoggio wrote in an article for online publication Smithsonian.com.

Google CEO Sundar Pichai believes artificial intelligence could have "more profound" implications for humanity than electricity or fire, according to recent comments. Pichai also warned that the development of artificial intelligence could pose as much risk as that of fire if its potential is not harnessed correctly. Pichai went on to warn of the potential dangers associated with developing advanced AI, saying that developers need to learn to harness its benefits in the same way humans did with fire. Google has invested heavily in artificial intelligence research, having acquired the London-based startup DeepMind for £300 million in 2014. DeepMind is often cited by AI experts and academics as the leading pioneer in AI research for its work in developing an algorithm capable of beating human champions at the ancient board game Go, as well as its work with the National Health Service (NHS) in the UK.

An undergraduate student from China's Fudan University showcased a new artificial intelligence (AI) software that can turn regular human photos into an anime masterpiece using the Generative Adversarial Network (GAN) and Deep Learning method. Yanghua Jin is attempting to create a computer program that can learn from its own mistakes the longer it works. This is all done using GAN's two networks: the generator and the discriminator, according to SoraNews24. The generator is in charge of producing the anime picture that runs through the software. It uses attributes taken from anime images, such as hair and eye color, whether the hair is long or short, and whether the mouth is open or not, and studies them.

It's already been about a week and the SDS organising committee members are still feeling the exhaustion that stays after the completion of a grand event. What had started as a half day workshop in 2014 with less than a hundred participants and born out of the Datalab in InIT ZHAW, is now a day conference gathering around 400 people. The conference has been a flagship event of the Swiss Alliance for Data-Intensive Services (Data Service Alliance), a Switzerland wide network of Data and Service scientists and professionals, for the past 2 years with a strive to get bigger and better each year in terms of content and participation. The main conference was held on the 7th and we had a pre-conference event on Deep Learning co-organized by Fernfachhochschule (FFHS) the day before. There were 3 talks followed by an apero and thereafter the Data Service Alliance General Assembly.

Deep learning is changing the world around us. To make the greatest impact, Hailo believes deep learning should be physically close to us. To make that happen, we need a new type of computer which can bring intelligence to any product – and this is exactly what we're building. Imagine intelligent devices that are empowered with the performance of a datacenter class computer, operating in real time at reduced power consumption, size and cost. Hailo processors enable edge devices to go beyond handling sensors and streaming volumes of data for remote processing–they can do the processing themselves.

Two agents, Red and Blue, collect apples (green) and occasionally tag the other agent (yellow lines).

Count Wally Rhines, semiconductor industry veteran and long-time CEO of Mentor Graphics, among the many who believe that deep-learning hardware will drive the next wave of growth for the semiconductor industry. Speaking at the GSA European Executive Forum here this week, Rhines added that memory will continue to be a key driver of the chip industry going forward. Despite the volatility of the semiconductor industry, R&D investment continues to be around 14% of revenue as it has been for the last 36 years, Rhines said, dismissing arguments put forth by some that there isn't enough being ploughed back into R&D to maintain sustained growth. On growth, we should be watching out for China, and a lot of growth will come from visual processing for AI applications, Rhines said. He offered his perspective on the future of the semiconductor industry and why revenue forecasts from research and analyst firms have been consistently off target.