A'scorpion-milking' robot has been developed to extract venom from the arachnids faster and more safely for use in cancer research. Scorpion venom is normally milked by toxicologists manually, a dangerous procedure where one wrong move can prove deadly. The new machine allows researchers to strap scorpions into an extractor, reducing their contact time with the beasts and making venom extraction safer. A'scorpion-milking' robot (pictured) has been developed to extract venom from the arachnids faster and more safely for use in cancer research. Current scorpion-milking methods can be dangerous both for the animals, due to punctures made to the venom gland or damage to the abdomen, and to the researchers, due to electric shocks from the equipment.

SHORTLY after 9/11, the US deployed a new form of high-tech warfare: sending drones into foreign airspace to kill terror suspects. At first the strikes were restricted to Afghanistan, but soon they were extended into Pakistan, Yemen and Somalia. The strategy has been escalated by presidents Obama and Trump. Initially the US had a virtual monopoly on drone technology, but commentators pointed out that this would only be temporary. Legal scholars also warned that the strikes were of dubious international legality. The implication was clear: if the US could strike with impunity, what was there to stop others from doing the same?

Regression or classification? This is perhaps the most basic question faced when tackling a new supervised learning problem. We present an Evolutionary Deep Learning (EDL) algorithm that automatically solves this by identifying the question type with high accuracy, along with a proposed deep architecture. Typically, a significant amount of human insight and preparation is required prior to executing machine learning algorithms. For example, when creating deep neural networks, the number of parameters must be selected in advance and furthermore, a lot of these choices are made based upon pre-existing knowledge of the data such as the use of a categorical cross entropy loss function. Humans are able to study a dataset and decide whether it represents a classification or a regression problem, and consequently make decisions which will be applied to the execution of the neural network. We propose the Automated Problem Identification (API) algorithm, which uses an evolutionary algorithm interface to TensorFlow to manipulate a deep neural network to decide if a dataset represents a classification or a regression problem. We test API on 16 different classification, regression and sentiment analysis datasets with up to 10,000 features and up to 17,000 unique target values. API achieves an average accuracy of $96.3\%$ in identifying the problem type without hardcoding any insights about the general characteristics of regression or classification problems. For example, API successfully identifies classification problems even with 1000 target values. Furthermore, the algorithm recommends which loss function to use and also recommends a neural network architecture. Our work is therefore a step towards fully automated machine learning.

This week Google South Africa announced major upgrades to its Google Translate function – used in 103 languages by more than 500 million users worldwide. For the first time since Google began in 1998, the Google Translate app will now include full isiZulu, isiXhosa and Kiswahili translation functionality. "Machine learning", Google's artificial intelligence team explained, is also being used to drastically improve existing Google search services. The major announcement, slipped in quite casually at a press conference held at Google's local headquarters in Johannesburg, was included in an explanation of machine learning by Blaise Aguera y Arcas, principal scientist for machine learning and artificial intelligence at Google. Aguera y Arcas steers a research team of 300 who are exploring the possibilities of emerging artificial intelligences.

For about thirty years, data analysis methods have largely demonstrated their effectiveness in the processing of data in many fields. Data reduction is one of these methods and part of the descriptive (or exploratory) statistics. It tries to summarize a sample of data using graphs or numerical characteristics. The main interpretation of data reduction is reducing the number of dimensions. This implies that data reduction is part of the multivariate exploratory statistics which seek to reduce the number of data dimensions by extracting a number of factors, dimensions, clusters, etc., which explain the dispersion of (multidimensional) data.

Instagram believes artificial intelligence could help fight trolls and junk messages. The photo-sharing service debuted Thursday two new tools it said would help it reduce the amount of spam it receives as well block offensive comments that appear on posts and live video. People can choose to have the automatic comment filter on or off. If an out-of-line comment still appears on a certain post with the filter on, people can still report it to Instagram as they would typically do. Instagram said that the comment filter is only available in English, but will debut in other languages at later dates.

Data science and machine learning for domain specific insights are at the core of Alibaba's strategy for global expansions. Apple's iPhone made its debut 10 years ago. Here's a look at how it has impacted business and the enterprise. The war in retail has long ago gone technological. Amazon is the poster child of this transition, paving the way first by taking its business online, then embracing the cloud and offering ever more advanced services for compute and storage to thirrd parties via Amazon Web Services (AWS).

Data science and machine learning for domain specific insights are at the core of Alibaba's strategy for global expansions. Apple's iPhone made its debut 10 years ago. Here's a look at how it has impacted business and the enterprise. The war in retail has long ago gone technological. Amazon is the poster child of this transition, paving the way first by taking its business online, then embracing the cloud and offering ever more advanced services for compute and storage to thirrd parties via Amazon Web Services (AWS).

In recent years, AI-based applications have increasingly been used in real-world domains. For example, game theory-based decision aids have been successfully deployed in various security settings to protect ports, airports, and wildlife. This article describes our unique problem-to-project educational approach that used games rooted in real-world issues to teach AI concepts to diverse audiences. Specifically, our educational program began by presenting real-world security issues, and progressively introduced complex AI concepts using lectures, interactive exercises, and ultimately hands-on games to promote learning. We describe our experience in applying this approach to several audiences, including students of an urban public high school, university undergraduates, and security domain experts who protect wildlife. We evaluated our approach based on results from the games and participant surveys.

Delegates at Newsweek and International Business Times' data science in capital markets event were mesmerised by a video of shoe box-sized satellites, known as "cube sats" being released into the earth's atmosphere. They were then shown the speeded up footage that was being captured: giant oil tanks with floating lids, rising like tides; ships being built from scratch in dry docks; burning flares from steel mills; a picture of the agricultural yields of the entire planet. Professor David Hand, chief scientific advisor, Winton, introducing the event, pointed out that the current AI summer is characterised by what he called "automatic data capture". He said: "Perhaps even more important than big data, we can now observe what people are doing all the time, like shopping for instance. Heterogeneous data is being produced constantly, as you sit here data about you is being captured by computers."