Andrew Flemming and Geoff Fowler, both 29, along with their friend and business partner, Will Hamilton, 37, were pouring their creative energies into a high-tech training device the likes of which the sporting world had never seen. They were building a better broom. From a report: Not just any broom, but one that they thought could be essential to the sport of curling, which relies on the best broom handling out there as teams strategically cajole a polished granite rock across a sheet of ice. They wound up calling it the SmartBroom, and in a sport that can come across as vaguely primordial, their piece of 21st-century gadgetry could play a role in determining who wins gold at the 2018 Winter Olympics in Pyeongchang, South Korea.

Each SmartBroom has four sensors in the broom head that relay data to a small display unit. Hamilton took one for a spin down the ice, and the data was instantaneous -- line graphs along with a slew of numbers that showed his force in pounds and his stroke rate in hertz. Hamilton also pointed to a figure that he described as his "sweeping performance index," or S.P.I., a metric that combines power and speed in one easy-to-digest figure. Patrick Janssen, a world-class curler from Canada, has consistently registered an S.P.I. in the 2,800 range. The numbers by themselves might not mean much, Flemming said, but subtle changes in technique can lead to big differences in the quality of each stroke. And now curlers have that information at their disposal. They can experiment to see which stroke works best for them.

A report released on Thursday by the job marketplace Hired reveals that salaries for workers in technology roles, including software engineers, product managers, and data analysts, peak around age 45. After that, earnings level off or drop until retirement. According to the report, the average salary of US technology workers is about $135,000, with the highest pay in the San Francisco area.

Jason Pontin, writing for Wired: Sundar Pichai, the chief executive of Google, has said that AI "is more profound than ... electricity or fire." Andrew Ng, who founded Google Brain and now invests in AI startups, wrote that "If a typical person can do a mental task with less than one second of thought, we can probably automate it using AI either now or in the near future." Their enthusiasm is pardonable.

[...] But there are many things that people can do quickly that smart machines cannot. Natural language is beyond deep learning; new situations baffle artificial intelligences, like cows brought up short at a cattle grid. None of these shortcomings is likely to be solved soon. Once you've seen you've seen it, you can't un-see it: deep learning, now the dominant technique in artificial intelligence, will not lead to an AI that abstractly reasons and generalizes about the world. By itself, it is unlikely to automate ordinary human activities.

To see why modern AI is good at a few things but bad at everything else, it helps to understand how deep learning works. Deep learning is math: a statistical method where computers learn to classify patterns using neural networks. [...] Deep learning's advances are the product of pattern recognition: neural networks memorize classes of things and more-or-less reliably know when they encounter them again. But almost all the interesting problems in cognition aren't classification problems at all.

Intel has released new microcode to address the stability and reboot issues on systems after installing its initial mitigations for Variant 2 of the Meltdown and Spectre attacks. From a report: The stability issues caused by Intel's microcode updates resulted in Lenovo, HP, and Dell halting their deployment of BIOS updates last month as Intel worked to resolve the problems. Intel initially said unexpected reboots were only seen on Broadwell and Haswell chips, but later admitted newer Skylake architecture chips were also affected. Microsoft also said it had also seen Intel's updates cause data loss or corruption in some cases.

Google is facing a lawsuit over the original Pixel. From a report: In a class action complaint filed this week, plaintiffs allege that the microphones in their Pixel and Pixel XL phones were defective from the start, and that Google knowingly sold defective phones amid widespread complaints immediately after launch. The lawsuit also claims that some warranty replacement phones continued to have problems, though neither of the named plaintiffs in the lawsuit had their phones repaired within Google's standard warranty period. Google acknowledged the Pixel phones' microphone issues in March 2017. An employee on Google's support forums attributed the problems to "a hairline crack in the solder connection on the audio codec," and said the problem can come and go depending on the temperature of the phone or the way it's being held.

Fake news, bots, and propaganda were hot topics at the World Economic Forum meeting in Davos last month, and Google executives there floated an intriguing idea to some fellow attendees -- what if the company could tell users whether information is trustworthy before they shared it on social networks like Facebook and Twitter? From a report: Representatives from Google and its parent company Alphabet eagerly discussed how the company can play a greater role in reducing misleading information online, several Davos attendees involved in and briefed on these conversations told Quartz. A notification system, perhaps via an optional extension for Google's Chrome browser, was an idea that these people said was broached more than once. Such a browser-based system controlled by Google could alert users on Facebook's or Twitter's websites when they're seeing or sharing a link deemed to be false or untrustworthy. Right now, this appears to be merely an idea company executives are discussing, not a product in development.

Proposed legislation in Congress aimed at preventing China from acquiring sensitive technology is being softened after protests by big U.S. companies who fear a loss in sales, Reuters reported on Thursday, citing people with knowledge of the matter. From the report: Two bills in the House of Representatives and Senate would broaden the powers of the inter-agency Committee on Foreign Investment in the United States (CFIUS) in hopes of stopping Chinese efforts to acquire sophisticated U.S. technology. The bipartisan legislation has the support of President Donald Trump's administration. "We are concerned that it vastly expands the scope and jurisdiction (of CFIUS)," said Nancy McLernon, chief executive of the Organization for International Investment, a group that represents global companies with U.S. operations. Given the alarm that the legislation has caused, Senator John Cornyn's staff is drafting changes to address industry concerns, according to three sources. Cornyn's office did not respond to a request for comment.

Bethany Brookshire, writing for Science News: Wikipedia is a gold mine for science fans, science bloggers and scientists alike. But even though scientists use Wikipedia, they don't tend to admit it. The site rarely ends up in a paper's citations as the source of, say, the history of the gut-brain axis or the chemical formula for polyvinyl chloride. But scientists are browsing Wikipedia just like everyone else. A recent analysis found that Wikipedia stays up-to-date on the latest research -- and vocabulary from those Wikipedia articles finds its way into scientific papers. The results don't just reveal the Wiki-habits of the ivory tower. They also show that the free, widely available information source is playing a role in research progress, especially in poorer countries.

darthcamaro writes: Apparently YouTube isn't the only site that is draining CPU power with unauthorized cryptocurrency miners. A water utility provider in Europe is literally being drained of its CPU power via an cryptojacking attack that was undetected for three weeks. eWeek reports: "At this point, Radiflow's (the security firm that discovered the cryptocurrency mining malware) investigation indicates that the cryptocurrency mining malware was likely downloaded from a malicious advertising site. As such, the theory that Radiflow CTO Yehonatan Kfir has is that an operator at the water utility was able to open a web browser and clicked on an advertising link that led the mining code being installed on the system. The actual system that first got infected is what is known as a Human Machine Interface (HMI) to the SCADA network and it was running the Microsoft Windows XP operating system. Radiflow's CEO, Ilan Barda, noted that many SCADA environments still have Windows XP systems deployed as operators tend to be very slow to update their operating systems." Radiflow doesn't know how much Monero (XMR) cryptocurrency was mined by the malware, but a recent report from Cisco's Talos research group revealed that some of the top un-authorized cryptocurrency campaigns generate over a million dollars per year. The average system would generate nearly $200,000 per year.

Baron_Yam shares a report from Engadget: Police in China are now sporting glasses equipped with facial recognition devices and they're using them to scan train riders and plane passengers for individuals who may be trying to avoid law enforcement or are using fake IDs. So far, police have caught seven people connected to major criminal cases and 26 who were using false IDs while traveling, according to People's Daily. The Wall Street Journal reports that Beijing-based LLVision Technology Co. developed the devices. The company produces wearable video cameras as well and while it sells those to anyone, it's vetting buyers for its facial recognition devices. And, for now, it isn't selling them to consumers. LLVision says that in tests, the system was able to pick out individuals from a database of 10,000 people and it could do so in 100 milliseconds. However, CEO Wu Fei told the Wall Street Journal that in the real world, accuracy would probably drop due to "environmental noise." Additionally, aside from being portable, another difference between these devices and typical facial recognition systems is that the database used for comparing images is contained in a hand-held device rather than the cloud."

dryriver shares a report from the BBC: U.S. suicide rates spiked in the months after Robin Williams killed himself in 2014, according to researchers. In the five months after the actor's death there were 10% more suicides than might be expected, or 1,841 extra cases, PLOS One journal reports. The potential risk of copycat incidents after celebrity cases is known to public health bodies. It cannot be known for certain if his death led to the spike but it appeared to be connected, the new study said. Experts say "irresponsible" media coverage of suicides can play a big part in copycat cases. At the time of his death, the Samaritans warned about a large number of news articles giving too much detail about the nature of his suicide, against media guidelines. Guidance from the World Health Organization, the Independent Press Standards Organization's editors' code of practice, the Ofcom broadcasting code and the BBC's editorial guidelines all advise against going into explicit detail about the methods used. However, researchers said there was "substantial evidence" that many media outlets had tended to deviate from these guidelines.

For the latest study, they looked at the monthly suicide rates from the U.S. government Centers for Disease Control and Prevention between January 1999 and December 2015 to see if there had been a spike. They found there were 18,690 suicides between August and December 2014 compared with the 16,849 cases they would have expected. In the weeks after Williams's death, there was a "drastic" increase in references to suicide and death in news media reports, as well as more posts on an internet suicide forum researchers monitored, the study found. David Fink, one of the study's authors, from Columbia University Mailman School of Public Health, said research had previously shown that suicide rates increased following a high-profile celebrity suicide, but this was a first time such a study had been done within the era of the 24-hour news cycle. Lorna Fraser, from the Samaritans' media advisory service, said: "This study builds on a strong body of research evidence that shows that irresponsible or overly detailed depictions of suicide can have a devastating impact. In the case of celebrities, the potential for someone at risk to make an emotional connection and over-identify with them is greater, in some cases even to interpret their death as affirmation that they could take their own life."

An anonymous reader quotes a report from Consumer Reports: Consumer Reports has found that millions of smart TVs can be controlled by hackers exploiting easy-to-find security flaws. The problems affect Samsung televisions, along with models made by TCL and other brands that use the Roku TV smart-TV platform, as well as streaming devices such as the Roku Ultra. We found that a relatively unsophisticated hacker could change channels, play offensive content, or crank up the volume, which might be deeply unsettling to someone who didn't understand what was happening. This could be done over the web, from thousands of miles away. (These vulnerabilities would not allow a hacker to spy on the user or steal information.) The findings were part of a broad privacy and security evaluation, led by Consumer Reports, of smart TVs from top brands that also included LG, Sony, and Vizio. The testing also found that all these TVs raised privacy concerns by collecting very detailed information on their users. Consumers can limit the data collection. But they have to give up a lot of the TVs' functionality -- and know the right buttons to click and settings to look for.

Jason Koebler shares a report from Motherboard: An anonymous person posted what experts say is the source code for a core component of the iPhone's operating system on GitHub, which could pave the way for hackers and security researchers to find vulnerabilities in iOS and make iPhone jailbreaks easier to achieve. The code is for "iBoot," which is the part of iOS that is responsible for ensuring a trusted boot of the operating system. It's the program that loads iOS, the very first process that runs when you turn on your iPhone. The code says it's for iOS 9, an older version of the operating system, but portions of it are likely to still be used in iOS 11. Bugs in the boot process are the most valuable ones if reported to Apple through its bounty program, which values them at a max payment of $200,000. "This is the biggest leak in history," Jonathan Levin, the author of a series of books on iOS and Mac OSX internals, told Motherboard in an online chat. "It's a huge deal." Levin, along with a second security researcher familiar with iOS, says the code appears to be the real iBoot code because it aligns with the code he reverse engineered himself.

According to Nikkei Asian Review, "Foxconn's panel arm Innolux is planning to slash more than 10,000 jobs this year as part of the company's aggressive efforts to increase the use of automation in manufacturing." Honorary Chairman Tuan Hsing-Chien said in a press conference on Tuesday: "We will reduce our total workforce to less than 50,000 people by the end of this year, from some 60,000 staff at the end of 2017." From the report: Innolux is a liquid crystal display-making affiliate of major iPhone assembler Hon Hai Precision Industry, better known as Foxconn Technology Group. Tuan is also a technology adviser to Foxconn, Sharp and Innolux. Tuan said up to 75% of production will be fully automated by the end of 2018. Most of Innolux's factories are in Taiwan. Tuan's pledge came a few days after Foxconn Chairman Terry Gou said the company would pour in some $342 million to overhaul its manufacturing process by using artificial intelligence.

An anonymous reader quotes a report from Ars Technica: There's a shakeup at Nest today. Following previous rumors back in November, Google just announced Nest will no longer be a standalone Alphabet company; instead, it will merge with the Google hardware team. The current Nest CEO, Marwan Fawaz, will report to Google Hardware SVP Rick Osterloh. Google's blog post says the merger will allow it to "combine hardware, software, and services" between the two companies, which are all "built with Google's artificial intelligence and the Assistant at the core." Nest and Google have been growing closer together even without this merger, with Nest getting a spot at the "Made By Google" Pixel 2 launch event to tout Nest and Google Assistant integration. An earlier report from The Wall Street Journal said that Google and Nest already combined their supply chain teams in 2016. While Google has focused on making the "Google" brand well known in the hardware world with the Pixel phones and Google Home, CNET reports that Google won't be dumping the Nest brand.