Robots are already making cashiers obsolete--now they're targeting the musicians who staff America's grocery stores. Sony has unveiled two songs written by artificial intelligence, by way of the company's Flow Machines software. To be sure, the final products--"Daddy's Car" and "Mister Shadow"--were produced and arranged by French composer Benoît Carré. "Daddy's Car" may be a hacky ripoff that falls somewhere between the Beatles and Tame Impala for its mid-'60s British rock stylings, but hey, blame the software. The Flow Machines software learns music from its massive database of songs.

We can all agree that teaching artificially-intelligent robots how to arm themselves with rocket launchers and create a murder carnival is just a fantastic idea. But these robots aren't ready to be let loose outside our front door (yet). They need to train first. So we're ready if Hitler comes back as a robot, of course. Or perhaps because the influential 1993 game has simple 3D maps and multiple gameplay styles, which makes it a handy touchstone for AI research.

If you trade stocks, you've probably made a few losing trades that still keep you up at night. The stock market moves quickly, and volatility can lead even the most seasoned trader to buy or sell in a panic. Whatever strategy you choose, the most important thing for your long-term returns is that you stick to it. You'd be better off if you stopped yourself before trading in a panic. Unfortunately, human emotions are powerful, and it's hard to think rationally when your nervous system has other plans.

"We will cooperate with each other and exercise leadership to support the early commercialization of automated and connected vehicle technologies," a declaration adopted at a meeting in the resort town of Karuizawa, Nagano Prefecture, said. Ministers and representatives from Britain, Canada, France, Germany, Italy, Japan and the United States plus the European Union also pledged in the declaration to set up a group to work on resolving problems pertaining to introducing self-driving cars. "We obtained a common understanding to make efforts in the same direction to create regulation frameworks that (will) tend to vary depending on region," transport minister Keiichi Ishii told a news conference after the meeting. The envisioned international safety regulations would enable automakers to sell the same car models worldwide, resulting in lower sales prices due to mass production, experts say. The declaration noted that the introduction of self-driving cars is expected to have such benefits as reducing traffic accidents, improving efficiencies including logistics and alleviating burdens on drivers.

If you visit the data science competitions website Kaggle, the example problem that they have for new starters is predicting the survivors and non-survivors from the Titanic from a sampled training dataset. This training dataset contains various attributes of the passengers, like the passenger's ticket class, the boarding gate, number of siblings, age, gender, etc. One can build a model using whatever tool you prefer and test your algorithm on a sample and see how you faired against the best of the best. Don't be disillusioned if your model only comes out at 76% accuracy against the winner's 100%. One can just Google the actual names and see who lived and who did not.

Today's paper is by Gao Huang, Yu Sun, et al. It introduces a new way to perturb networks during training in order to improve their performance. Before I continue, let me first state that this paper is a real pleasure to read; it is concise and extremely well written. It gives an excellent overview of the motivating problems, previous solutions, and Huang and Sun's new approach. I highly recommended giving it a read!

A little while ago, you might have read about batch normalization being the next coolest thing since ReLu's. Things have since moved on, but it's worth mentioning because it has been adopted in most networks today. The original paper is pretty dense, and it's all goodies inside of there. So, this blog post is devoted to explaining the more confusing portions of batch normalization. What follows are a few concepts that you may find interesting or may not have fully understood when reading over Ioffe and Szegedy's paper.

HYDERABAD: Apple quietly acquired a little-known Hyderabad-based startup a few months ago for an undisclosed amount and while little is known about what Tuplejump does, the word is that this is part of the Cupertino-based giant's exploration into artificial intelligence. Known in the trade as an acqui-hire, nearly all of Tuplejump's 16 employees are in the process of becoming Apple staffers. A back-of-the-envelope calculation by an expert put the valuation at about 20 million (Rs 133 crore). "When there is a talent hire, the cost of hiring is usually calculated based on the last few years' salary," the expert said. "So, for a 16-member team, I think it would probably come up to around 4 million a year -- making this deal for something around 20 million (including for possible IP evaluation)."

Apple Inc. is pressing ahead with the development of an Echo-like smart-home device based on the Siri voice assistant, according to people familiar with the matter. Started more than two years ago, the project has exited the research and development lab and is now in prototype testing, said the people, who asked not to be identified discussing unannounced Apple projects. Like Amazon Inc.'s Echo, the device is designed to control appliances, locks, lights and curtains via voice activation, the people said. Apple hasn't finalized plans for the device and could still scrap the project. If a product reaches the market, it would be Apple's most significant piece of new hardware since the company announced the Apple Watch in 2014.

Apple has bought Tuplejump, a machine learning technology company with operations in both India and the United States, TechCrunch reports. Apple continues to buy companies with machine learning expertise. The reason the company had its eyes on Tuplejump was for its "FiloDB" project, according to the report. The lead engineer on FiloDB has been working at Apple since this May, according to his LinkedIn. The purchase price is unknown, although it's almost certainly lower than Apple's last machine learning purchase, Turi, which was bought for around 200 million earlier this year.