Doom-and-gloom headlines about the shrinking number of stores in malls, closing department stores, and thinning foot traffic have been everywhere lately, and not without good reason. If one were to venture a guess based on brick-and-mortar store performance, it would appear the U.S. consumer has stopped spending. But in fact, consumer spending has been rising steadily every year since 2009. It's just that an ever-greater number of people are choosing to do their shopping online. According to the U.S. Census Bureau's monthly retail report, retail sales were up 4.5% year over year in April, but online sales grew 12%.

Elon Musk's company SpaceX will attempt to launch a Falcon 9 rocket Saturday evening, in a bid to send a refurbished Dragon cargo spacecraft to the International Space Station (ISS). The CRS-11 mission is the eleventh for the commercial space company and was originally scheduled for Thursday evening, but had to be postponed due to weather conditions. The weather could play spoilsport once again Saturday, which has an instantaneous launch window at 5:07 p.m. EDT. According to forecast by meteorologists with the U.S. Air Force 45th Weather Squadron, announced Friday morning, there was a 60 percent chance of favorable weather at the time of the planned launch. That forecast was worse than the 70 percent favorable forecast announced Thursday night. "The primary weather concerns for Saturday's launch are anvil clouds, cumulus clouds and flight through precipitation," the forecast announcement on NASA's website said.

These days there are so many self-driving cars coming down the pipeline it seems inevitable they'll soon be stuck in a robot traffic jam – just like the human-piloted cars of today. Well, not if Anthony Barrs and Baiyu Chen get their way. The two graduate students at the University of California, Berkeley, have devised a system that would have tightly-packed clusters of autonomous vehicles zipping past local traffic at speeds of more than 100mph, all on existing roadways. They call it Hyperlane, and it works a lot like high-speed toll lanes already do, only with a central computer controlling everything. Although fully autonomous cars are not yet legal on most public roads, manufacturers like Volvo and Tesla already offer autonomous features on their vehicles – adaptive cruise control and, in some cases, systems that steer the car with limited driver input.

Editor's note: This post originally appeared as an answer to a Quora question, which also included the following: "As I understand, the chance of having a derivative zero in each of the thousands of direction is low. Is there some other reason besides this?" Yes, there is a'theoretical justification', and has taken a couple decades to flush it out. I will first point out, however, it has been observed in practice. This was pointed out by LeCun in his early work on LeNet, and is actually discussed in the'orange book', "Pattern Classification" by David G. Stork, Peter E. Hart, and Richard O. Duda. The problem has been addressed in condensed matter physics 20 years ago in the study of spin glasses.

Web crawling, snowball sampling, and respondent-driven sampling (RDS) are three types of network sampling techniques used to contact individuals in hard-to-reach populations. This paper studies these procedures as a Markov process on the social network that is indexed by a tree. Each node in this tree corresponds to an observation and each edge in the tree corresponds to a referral. Indexing with a tree (instead of a chain) allows for the sampled units to refer multiple future units into the sample. In survey sampling, the design effect characterizes the additional variance induced by a novel sampling strategy. If the design effect is some value $DE$, then constructing an estimator from the novel design makes the variance of the estimator $DE$ times greater than it would be under a simple random sample with the same sample size $n$. Under certain assumptions on the referral tree, the design effect of network sampling has a critical threshold that is a function of the referral rate $m$ and the clustering structure in the social network, represented by the second eigenvalue of the Markov transition matrix, $\lambda_2$. If $m < 1/\lambda_2^2$, then the design effect is finite (i.e. the standard estimator is $\sqrt{n}$-consistent). However, if $m > 1/\lambda_2^2$, then the design effect grows with $n$ (i.e. the standard estimator is no longer $\sqrt{n}$-consistent). Past this critical threshold, the standard error of the estimator converges at the slower rate of $n^{\log_m \lambda_2}$. The Markov model allows for nodes to be resampled; computational results show that the findings hold in without-replacement sampling. To estimate confidence intervals that adapt to the correct level of uncertainty, a novel resampling procedure is proposed. Computational experiments compare this procedure to previous techniques.

Might such things as school examinations, tests, marked classwork and progress checks soon all be a thing of the past? They will be if Rose Luckin, Professor of Learner-Centred Design at the Knowledge Lab at the University College, London (UCL) Institute of Education, has her way. She argues that the way school pupils are assessed today is unsatisfactory. "Decades of research have shown that knowledge and understanding cannot be rigorously evaluated through a series of 90-minute exams. The prevailing exam paradigm is stressful, unpleasant, can turn students away from education, and requires that both students and teachers take time away from learning. And yet we persist in relying on these blunt instruments, sending students off to universities and the workplace ill-equipped for their futures," writes Professor Luckin in a research paper.

A disembodied voice sounded over a loudspeaker. Take cover," it warned to anyone within earshot. Then, the sirens began to wail. Erin Delaney assumed it was a drill. She peeked down the hallway to see how other people were responding. Then she hit the deck. The NATO base in Kabul where Delaney had been working for weeks was being attacked. Delaney, 24, had never had any military training. She grew up in San Diego, traveled up the coast for college at UC Berkeley, and spent the next two years nestled in the safe, Tesla-filled San Francisco bubble, working in the compliance department at Dropbox. Now, with her nose to the ground, she was getting a taste--however brief--of life in a war zone. She flipped over the visitor's badge she'd received when she first arrived at the base. In case of attack, it said, she should stay on the ground for two minutes. Assuming nothing dire happened, she was to shelter in place until the shelling stopped. So, for about an hour, that's what she did.

Robots and automation are already commonplace in food warehouses, but they could soon be coming to supermarket aisles. Fujitsu's Matey, shown at the company's annual showcase in Tokyo, move autonomously through shops to monitor produce and help retailers stack shelves more effectively. The robot is about five-feet tall and moves on wheels with a screen on the front used to communicate with any humans it comes across. A Fujitsu spokesperson explained that initially it will only be used when a shop is closed, trundling up and down aisles to check stock levels. Using a cloud-based AI, the robot could also analyse the placement of products to work out if they could be arranged more effectively.

TOKYO (Reuters) - Japan on Thursday lodged a protest with China after four Chinese coastguard vessels entered what Tokyo considers its territorial waters near disputed East China Sea islets and a drone-like object flew near one ship, the Japanese government said.

Uber is hiring Raquel Urtasun, a prominent artificial intelligence researcher at the University of Toronto, as the ride-hailing company aims to build a lab for driverless car research in the Canadian city, a hotbed for AI talent. Urtasun--an associate professor at the university who specializes in the computer vision software that allows driverless cars to view the world around them--will oversee the new venture. "We hope to draw from the region's impressive talent pool as we grow, helping the dozens of researchers we plan to hire stay connected to the Toronto-Waterloo Corridor," Travis Kalanick, Uber's embattled CEO, wrote in a blog post published this morning. The move resonates on multiple levels, given the ongoing legal attack against Uber's existing computer vision technology by Waymo--the driverless car company that grew out of Google--and the widespread controversy over Uber's allegedly misogynistic internal culture. Urtasun could help the company forge another much-needed path to the kind of AI that driverless cars will require.