About a month into Japan's professional baseball season, the Fukuoka SoftBank Hawks, the 2014 and 2015 national champions, are doing no worse. Many say the team's strong lineup is underpinned by the cash-rich SoftBank Group, a big telecom and technology group. The reality is quite the reverse. Unlike their rivals, the Hawks are a stand-alone club, though one with the financial leeway to allocate profits to areas where the front office sees fit, such as player development and information technology. With the help of its tech-savvy parent, the club may be about to rewrite "Moneyball," the 2003 bestseller about how a Major League Baseball team in the U.S. used statistical analysis to beat high-spending opponents.

Movidius, machine intelligence partner to DJI, FLIR, Google and others, is introducing the first ever powerful deep learning processing accelerator that fits into a tiny USB Stick. It connects to existing systems and increases the performance of neural networking tasks by 20-30X. It performs at over 150GFLOPS while consuming under 1.2W. Called the Fathom Neural Compute Stick, It's basically the world's first supercomputer on a USB device. Developers, researchers, hobbyists (think raspberry pie) and anyone developing deep learning applications will benefit from Fathom.

Apple bought Emotient, an artificial intelligence startup that reads people's emotions by analyzing facial expressions, the Wall Street Journal reported. The report did not specify the financial terms of the deal. The tech giant's plans for Emotient were not immediately clear, the Journal reported, confirming the news with an Apple spokeswoman. Emotient's software reads the expressions of individuals and crowds to gain insights that can be used by advertisers to assess viewer reaction or a medical practitioner to better understand signs of pain in patients. San Diego-based Emotient had previously raised 8 million from investors including Intel Capital, the Journal said.

The last few years have seen an explosion of interest in Machine Learning (ML) technology and potential applications. Machine Learning is the unsung hero that powers many applications, systems, sensors, devices, and products. Today, Machine Learning is so pervasive that we can often assume its presence in most of the applications and systems without having to specifically call it out. In simple terms, machine learning is a computer's ability to learn from data, and it is one of the most useful tools we have to develop intelligent systems and applications. Machine learning is used widely today for all kinds of tasks, from churn prediction in large companies, to web search, to medical diagnostics, to robotics.

In my last post on the subject of AI, values,and socially beneficial simulation, I looked at one way of approaching the problem of AI values, society, and institutions through simulation and modeling. It occurred to me while writing it that there was also another way that might be interesting to think about, and it stems from a rejected piece I wrote last year on the subject of the computer as a "moral mirror." Though that piece was rejected for a reason (it wasn't that well-written), it has come into my mind again over the last week as the Internet has collectively reflected on my Slate article about AI values. I have some commonalities and differences with Stuart Russell's view of AI value alignment. I've looked at them here, here, here, and here.

Bossa Nova Robotics, a company specializing in building robotic technology for retailers, has raised 14 million to expedite the rollout of its robots in stores. Founded out of Pittsburgh, with offices in San Francisco as well, Bossa Nova has been developing its robot technology over the past few years, setting out to serve retailers through automation and analytics. Its machines analyze stock on shelves and collect data to optimize inventory, with fully autonomous robots unleashed in stores among shoppers. The company says it is now testing its "retail robots" with "five of the world's leading retail chains," though it wouldn't divulge any names. What we're effectively talking about here is automating a task that would ordinarily be carried out by humans, with a view toward improving efficiency and productivity and cutting costs.

As we go about our daily lives, we tend to assume that our perceptions -- sights, sounds, textures, tastes -- are an accurate portrayal of the real world. Sure, when we stop and think about it -- or when we find ourselves fooled by a perceptual illusion -- we realize with a jolt that what we perceive is never the world directly, but rather our brain's best guess at what that world is like, a kind of internal simulation of an external reality. Still, we bank on the fact that our simulation is a reasonably decent one. If it wasn't, wouldn't evolution have weeded us out by now? The true reality might be forever beyond our reach, but surely our senses give us at least an inkling of what it's really like. Not so, says Donald D. Hoffman, a professor of cognitive science at the University of California, Irvine. Hoffman has spent the past three decades studying perception, artificial intelligence, evolutionary game theory and the brain, and his conclusion is a dramatic one: The world presented to us by our perceptions is nothing like reality.

When teenage carpet salesman Lee Biggins decided to set up a jobseekers website, he wasn't going to let the fact he didn't have any computer skills hold him back. This was back in 1999, and the then 19-year-old had big ambitions for his business idea. So he went out and spent 899 on a computer, and an internet how-to book. Mr Biggins, who had left school at 15 "with some terrible grades - Es, Fs, and Gs", also enrolled on a computer literacy course in his hometown of Fleet, in Hampshire, 45 miles south west of London. But realising he could still do with some technical assistance, he says he went down his local pub one evening, and asked everyone: "Does anyone know someone who can build a website?"

In many technical fields, single-objective optimization procedures in continuous domains involve expensive numerical simulations. In this context, an improvement of the Artificial Bee Colony (ABC) algorithm, called the Artificial super-Bee enhanced Colony (AsBeC), is presented. AsBeC is designed to provide fast convergence speed, high solution accuracy and robust performance over a wide range of problems. It implements enhancements of the ABC structure and hybridizations with interpolation strategies. The latter are inspired by the quadratic trust region approach for local investigation and by an efficient global optimizer for separable problems. Each modification and their combined effects are studied with appropriate metrics on a numerical benchmark, which is also used for comparing AsBeC with some effective ABC variants and other derivative-free algorithms. In addition, the presented algorithm is validated on two recent benchmarks adopted for competitions in international conferences. Results show remarkable competitiveness and robustness for AsBeC.

Bandits are becoming an essential tool in modern recommenders systems [9, 12]. Most recommendation setting involve an ever changing dynamic set of items, in many domains such as news and ads recommendation the item set is changing so rapidly that is impossible to use standard collaborative filtering techniques. In these settings bandit algorithms such as contextual bandits have been proven to work well [10] since they provide a principled way to gauge the appeal of the new items. Yet, one drawback of contextual bandits is that they mainly work in a content-dependent regime, the user and item content features determine the preference scores so that any collaborative effects (joint user preferences over groups of items) that arise are being ignored. Incorporating collaborative effects into bandit algorithms can lead to a dramatic increase in the quality of recommendations. In bandit algorithms this has been mainly done by clustering the user. For instance, we may want to serve content to a group of users by taking advantage of an underlying network of preference relationships among them. These preference relationships can either be explicitly encoded in a graph, where adjacent nodes/users are deemed similar to one another, or implicitly contained in the data, and given as the outcome of an inference process that recognizes similarities across users based on their past behavior. To deal with this issue a new type of bandit algorithms has been developed which work under the assumption that users can be grouped (or clustered) based on their selection of items e.g.