Stories
Slash Boxes
Comments

SoylentNews is people

Log In

Log In

Create Account  |  Retrieve Password


Site News

Join our Folding@Home team:
Main F@H site
Our team page


Funding Goal
For 6-month period:
2022-07-01 to 2022-12-31
(All amounts are estimated)
Base Goal:
$3500.00

Currently:
$438.92

12.5%

Covers transactions:
2022-07-02 10:17:28 ..
2022-10-05 12:33:58 UTC
(SPIDs: [1838..1866])
Last Update:
2022-10-05 14:04:11 UTC --fnord666

Support us: Subscribe Here
and buy SoylentNews Swag


We always have a place for talented people, visit the Get Involved section on the wiki to see how you can make SoylentNews better.

The Best Star Trek

  • The Original Series (TOS) or The Animated Series (TAS)
  • The Next Generation (TNG) or Deep Space 9 (DS9)
  • Voyager (VOY) or Enterprise (ENT)
  • Discovery (DSC) or Picard (PIC)
  • Lower Decks or Prodigy
  • Strange New Worlds
  • Orville
  • Other (please specify in comments)

[ Results | Polls ]
Comments:83 | Votes:90

posted by CoolHand on Monday July 09 2018, @10:37PM   Printer-friendly
from the the-road-to-skynet dept.

Submitted via IRC for BoyceMagooglyMonkey

AI agents continue to rack up wins in the video game world. Last week, OpenAI's bots were playing Dota 2; this week, it's Quake III, with a team of researchers from Google's DeepMind subsidiary successfully training agents that can beat humans at a game of capture the flag.

As we've seen with previous examples of AI playing video games, the challenge here is training an agent that can navigate a complex 3D environment with imperfect information. DeepMind's researchers used a method of AI training that's also becoming standard: reinforcement learning, which is basically training by trial and error at a huge scale.

Agents are given no instructions on how to play the game, but simply compete against themselves until they work out the strategies needed to win. Usually this means one version of the AI agent playing against an identical clone. DeepMind gave extra depth to this formula by training a whole cohort of 30 agents to introduce a "diversity" of play styles. How many games does it take to train an AI this way? Nearly half a million, each lasting five minutes.

Source: https://www.theverge.com/2018/7/4/17533898/deepmind-ai-agent-video-game-quake-iii-capture-the-flag


Original Submission

posted by martyb on Monday July 09 2018, @09:01PM   Printer-friendly
from the whole-is-greater-than-the-sum-of-its-parts dept.

Instead of Building Single Monster Scopes like James Webb, What About Swarms of Space Telescopes Working Together?

... Jayce Dowell and Gregory B. Taylor, a research assistant professor and professor (respectively) with the Department of Physics and Astronomy at the University of New Mexico [...] outlined their idea in a study titled "The Swarm Telescope Concept [pdf]", which recently appeared online and was accepted for publication by the Journal of Astronomical Instrumentation.

[...] Instead of a single instrument, the telescope would consist of a distributed array where many autonomous elements come together through a data transport system to function as a single facility. This approach, they claim, would be especially useful when it comes to the Next Generation Very Large Array (NGVLA) – a future interferometer that will build on the legacy of the Karl G. [J]ansky Very Large Array and Atacama Large Millimeter/submillimeter Array (ALMA). As they state in their study:

At the core of the swarm telescope is a shift away from thinking about an observatory as a monolithic entity. Rather, an observatory is viewed as many independent parts that work together to accomplish scientific observations. This shift requires moving part of the decision making about the facility away from the human schedulers and operators and transitioning it to "software defined operators" that run on each part of the facility. These software agents then communicate with each other and build dynamic arrays to accomplish the goals of multiple observers, while also adjusting for varying observing conditions and array element states across the facility.


Original Submission

posted by martyb on Monday July 09 2018, @07:24PM   Printer-friendly
from the why-pay-twice-when-you-can-pay-three-times-for-the-same-data dept.

The bad blood and high prices with academic publishing houses go back many years. Now the German Rectors' Conference (HRK) has issued a press release regarding the publisher Elesevier's unacceptable demands on the academic community, forcing the community's hand to suspend even negotiations. The HRK is the association of public and government-recognised universities in Germany consisting of 268 member institutions, in which around 94 percent of all students in Germany are enrolled. The German universities, like those in other countries, have been wishing to move to Open Access but have been stymied for decades by the big publishing houses.

“As far as we’re concerned, the aim of the ongoing negotiations with the three biggest academic publishers is to develop a future-oriented model for the publishing and reading of scientific literature. What we want is to bring an end to the pricing trend for academic journals that has the potential to prove disastrous for libraries as it stands. We are also working to promote open access, with a view to essentially making the results of publicly funded research freely accessible. The publishers should play a crucial role in achieving this. We have our sights set on a sustainable publish and read model, which means fair payment for publication and unrestricted availability for readers afterwards. Elsevier, however, is still not willing to offer a deal in the form of a nationwide agreement in Germany that responds to the needs of the academic community in line with the principles of open access and that is financially sustainable,” said Hippler.

The trouble shows no signs of abating. Even now, in a case of the fox watching the hen house, these problematic publishers have inserted themselves between the EU money and the universities even in the matter of advancing open access.

From HRK's web site: DEAL and Elsevier negotiations: Elsevier demands unacceptable for the academic community


Original Submission

posted by martyb on Monday July 09 2018, @05:47PM   Printer-friendly
from the Better-ask-Betteridge? dept.

Andres Guadamuz has written a blog post analyzing why last Thursday's vote by the JURI Committee to reject fast-tracking the proposal concerning "harmonization" of copyright in the EU went as it did. The rejection of fast-tracking means that the issue will still come up for a general vote in parliament in September but the interesting part is that for the first time in Europe a wide coalition has managed to defeat powerful media lobbies, at least for now. He goes into how this was possible and what needs to happen in September.

The main result of this change from a political standpoint is that now we have two lobbying sides in the debate, which makes all the difference when it comes to this type of legislation. In the past, policymakers could ignore experts and digital rights advocates because they never had the potential to reach them, letters and articles by academics were not taken into account, or given lip service during some obscure committee discussion just to be hidden away. Tech giants such as Google have provided lobbying access in Brussels, which has at least levelled the playing field when it comes to presenting evidence to legislators.

Earlier on SN:
The EU's Dodgy Article 13 Copyright Directive has Been Rejected (2018)
EU Committee Approves Controversial Copyright Directive (2018)
Censorship Machines Are Coming: It’s Time for the Free Software Community to Use its Political Clout (2018)
Mulled EU Copyright Shakeup Will Turn Us Into Robo-Censors (2018)
EU Study Finds Even Publishers Oppose the "Link Tax" (2017)


Original Submission

posted by martyb on Monday July 09 2018, @04:10PM   Printer-friendly
from the windfall? dept.

Judge Orders Chinese Wind-Turbine Maker To Pay $59 Million For Stealing Trade Secrets

A federal judge has ordered China's largest wind-turbine firm, Sinovel, to pay $59 million for stealing trade secrets from a Massachusetts-based technology company.

Last January, Sinovel was found guilty of stealing trade secrets in federal criminal court in Madison, Wis. The company paid an Austria-based employee of American Superconductor Corp. to steal its source code for software that powered wind turbines.

[...] Sinovel was the largest customer of American Superconductor Corp. And then the Chinese company suddenly began rejecting shipments of American Superconductor's electronic components in 2011. The Massachusetts tech company learned that Sinovel was using a pirated version of the software it made in the wind turbines it installed. The ordeal left American Superconductor in perilous financial shape, and Wall Street analysts wrote it off as dead. The U.S. Department of Justice said that the company lost more than $1 billion in shareholder equity and 700 jobs.

Previously: China's 'Great Brain Robbery' Hacking of US Companies a National Security Emergency

Owner of American Superconductor, Daniel McGahn, discovered a version of its latest, unreleased software being used on a turbine in China. Despite doing everything possible to keep its source code off the internet it was discovered that the Chinese company turned one of McGahn's employees by offering him money, women and an apartment in return for the full operating code.

To make matters worse, when McGahn decided to sue Sinovel for $1.2bn (£840m) and hire a computer security firm to investigate the case, his firm claims the Chinese company hacked the company's system to see what course of legal action it was taking in order to get a leg-up in proceedings.

"Whenever there's a big lawsuit we'll see the Chinese government actually break into that company, break into the legal department and figure out what's going on behind the scenes so they can better deal with that lawsuit," said George Kurtz, co-founder of computer security company CrowdStrike.


Original Submission

posted by martyb on Monday July 09 2018, @02:33PM   Printer-friendly
from the lending-a-hand dept.

Elon Musk built a "submarine" to rescue Thailand kids—here's what it looks like

Scuba divers in Thailand have already rescued four of the 12 boys who have been stranded, along with their coach, in a flooded cave. And they are hoping to rescue the rest in the next couple of days. But in the meantime, Elon Musk has continued working on alternative strategies divers could use if conventional diving proves too difficult for the remaining boys.

On Saturday, Musk settled on the idea of building "a tiny, kid-size submarine using the liquid oxygen transfer tube of Falcon rocket as hull." He ordered SpaceX engineers to begin building the device, saying that it could be ready by the end of the day on Saturday.

By the time it was finished, Thai officials had already begun their rescue operation without Musk's help, so the technology likely won't be needed. But Musk decided to press forward anyway, tweeting that "even if not useful here, perhaps it will be in a future situation."

[Out of the 12 boys and the coach who were trapped in the cave, reports are that 8 boys have been rescued so far. Live reporting is available from: BBC, CBS News, NYT, and The Guardian; weather report via Weather Underground.

Wikipedia has a page on the rescue (be aware there is discussion to rename the article). The summary:

Twelve boys aged 11 to 17 and a 25-year-old man became stranded in Tham Luang Nang Non (Thai: ถ้ำหลวงนางนอน), a cave in Thailand's Chiang Rai Province on 23 June 2018. Heavy rains partially flooded the cave during their visit. The boys – all members of a local association football team – and their assistant coach were reported missing a few hours later, and search operations began immediately.

Efforts to locate them were hampered by rising water levels, and no contact was made for over a week. The rescue effort expanded into a massive operation amid intense worldwide media coverage and public interest. After struggles through narrow passages and muddy waters, British divers discovered the missing people, all alive, on an elevated rock about 3.2 kilometres (2.0 mi)[12] from the cave mouth, on 2 July, over nine days after they went missing. Rescuer organizers discussed whether to teach the boys and their coach basic dive techniques to enable their early rescue or wait months for the floodwaters to subside at the end of the monsoon season. After days of pumping water from the cave system and a respite from rain, four of the boys were rescued on 8 July, four more on 9 July,[13] with further rescue operations expected to take an additional few days.

Over 1,000 people have been involved in the rescue operation, including Thai Navy SEALs as well as volunteers, teams and technical assistance from multiple countries. They are providing food and water. A 38-year-old former SEAL died while attempting to return after delivering supplies of air to the cave on 5 July; he was unable to breathe and could not be revived.

There are reports that the level of oxygen in the air has dropped to 15% (from the normal 21%) which may cause breathing problems for the team and rescuers in the cave.

Please reply in the comments with any on-line resources that you have found helpful in following this story. --martyb]


Original Submission

posted by martyb on Monday July 09 2018, @12:56PM   Printer-friendly
from the is-that-a-Cray-in-your-pocket? dept.

Samsung is preparing to manufacture 7LPP and 5LPE process ARM chips:

Samsung has said its chip foundry building Arm Cortex-A76-based processors will use 7nm process tech in the second half of the year, with 5nm product expected mid-2019 using the extreme ultra violet (EUV) lithography process.

The A76 64-bit chips will be able to pass 3GHz in clock speed. Back in May we wrote: "Arm reckoned a 3GHz 7nm A76 single core is up to 35 per cent faster than a 2.8GHz 10nm Cortex-A75, as found in Qualcomm's Snapdragon 845, when running mixed integer and floating-point math benchmarks albeit in a simulator."

[...] Samsung eventually envisages moving to a 3nm Gate-All-Round-Early (3AAE) on its process technology roadmap. Catch up, Intel, if you can.

Also at AnandTech.

Previously: Samsung Roadmap Includes "5nm", "4nm" and "3nm" Manufacturing Nodes

Related: Samsung's 10nm Chips in Mass Production, "6nm" on the Roadmap (obsolete)
Moore's Law: Not Dead? Intel Says its 10nm Chips Will Beat Samsung's
Samsung Plans a "4nm" Process


Original Submission

posted by martyb on Monday July 09 2018, @11:19AM   Printer-friendly
from the just-desserts dept.

The Daily Express reports

A gang of poachers who broke into a South African game reserve to slaughter [the preserve's] herd of rhinos were attacked and eaten by a pride of hungry lions.

At least three hunters were believed to have been devoured by the predators with one head being recovered at the scene along with a number of bloodied body parts and limbs.

Staff discovered the bloody remnants [July 3] at the Sibuya Game Reserve near Kenton-on-Sea in Eastern Province, South Africa, and called in a helicopter to search the area for more poachers.

When none were found and the area was declared safe, a vet tranquillised the six lions in the pride so that police and staff could go inside and recover the remains of the mauled poachers.

Owner Nick Fox, 60, said: "We found enough body parts and three pairs of empty shoes which suggest to us that the lions ate at least three of them but it is thick bush and there could be more.

[...] "the lions are our watchers and guardians and [these bumblers] picked the wrong pride and became a meal."

[...] "The lions may have eaten more of them. It is difficult to tell as the area is very thick with bush and you cannot be sure what they have taken off to feed on elsewhere.

[...] "They were armed with high powered rifles with silencers and an axe for the horns and wire cutters and side arms and they had enough food with them to last for many days.

"They were clearly intent on killing rhinos and cutting off their horns."


Original Submission

posted by martyb on Monday July 09 2018, @09:42AM   Printer-friendly
from the nasty-way-to-go dept.

A woman who lived a short distance from where Sergei and Yulia Skripal were poisoned with the Novichok nerve agent has died. Prime Minister Theresa May is "appalled and shocked" by the death:

Police have launched a murder inquiry after a woman exposed to nerve agent Novichok in Wiltshire died. Dawn Sturgess, 44, died in hospital on Sunday evening after falling critically ill on 30 June. Charlie Rowley, 45, who was also exposed to the nerve agent in Amesbury, remains critically ill in hospital.

[...] Officers are still trying to work out how Ms Sturgess and Mr Rowley were exposed to the nerve agent although tests have confirmed they touched a contaminated item with their hands.

[...] Mrs May sent her "thoughts and condolences" and said officials are "working urgently to establish the facts". She said: "The government is committed to providing full support to the local community as it deals with this tragedy." British diplomat Julian King, the European Commissioner responsible for the EU's security union, said: "Those behind this are murderers."

[...] The working hypothesis is that the pair became contaminated after touching a poison container left over from the March attack on Sergei and Yulia Skripal. The death of Dawn Sturgess, a British citizen on British soil, now changes the investigation to a murder inquiry, with all the diplomatic and security ramifications that carries. Britain has been blaming Moscow for the original attack in March, saying there is no plausible alternative to the Kremlin having ordered the assassination attempt. Russia has denied any involvement, suggesting instead this was the action of a weak British government looking to undermine the success of the current World Cup being hosted by Russia.

Here's something from the other side.

Previously: Former Russian Spy Exposed to "Unknown Substance" in Salisbury, England
Use of Nerve Agent Confirmed in Skripal Assassination Attempt
UK Gives Russia Until Midnight to Explain Use of Novichok Nerve Agent


Original Submission

posted by martyb on Monday July 09 2018, @08:05AM   Printer-friendly
from the Ripley-Approved dept.

The "super suit" that can help people move

A lightweight, comfortable 'super suit' designed by Seismic – a wearable robotics spin-off from the non-profit research centre SRI International – works with the user's muscles to help boost their power.

The suit's 'electric muscles', powered by tiny motors, contract in a way that mimics human muscle. These electric muscles are integrated into the clothing around the joints of the body and attached via grips in the clothing. These grips function like tendons in the human body. A computer and sensors tracking body movements are also integrated into the suit; software tells the muscles in the clothing when to activate. The hard technology components such as motors, batteries and control boards are incorporated into hexagonal low-profile pods, designed for maximum comfort.

"Right now the only kinds of products that can help people are walkers and canes," says Rich Mahoney, founder and CEO of Seismic. (Wheelchairs are another aid, but the suit is aimed more at those whose mobility is reduced only slightly). "The other option is to stay home or to limit your activity. And most people choose that because they don't really want to associate themselves with those other kinds of products." To keep the clothing looking sleek and its function discreet, Seismic worked with designer Yves Béhar. "The goal is to make a product that you actually want to wear, not one that you have to wear," says Béhar. "Comfort is extremely important, as well as aesthetics."


Original Submission

posted by martyb on Monday July 09 2018, @06:28AM   Printer-friendly
from the captive-audience-no-more dept.

Campaign for a Commercial-Free Childhood reports via Common Dreams

Houghton Mifflin Harcourt has announced the end of its Channel One News broadcasts in school classrooms, and child advocates are bidding them a less-than-fond farewell. For 28 years, Channel One compelled students to watch a 12-minute newscast that includes two minutes of commercials. Each year, schools with Channel One lost the equivalent of a full week of school to the broadcasts, and a full day of instructional time to the commercials alone.

[...]Launched in 1989 by marketing executive Chris Whittle, Channel One lured schools with the promise of free classroom televisions. But the cost of this "free" equipment was high. Schools with Channel One were required to show the broadcast, including two minutes of student-targeted commercials, on 90% of school days. Former Channel One President Joel Babbit once boasted, "The advertiser gets kids who cannot go to the bathroom, cannot change the station, who cannot listen to their mother yell in the background, who cannot be playing Nintendo."

Channel One's business model was controversial from the start. Advocates from across the political spectrum, from Ralph Nader to Phyllis Schlafly, decried Channel One's exploitation of a captive student audience. Research showed that the "news" on Channel One was often fluff pieces about pop culture, that Channel One cost American taxpayers nearly $2 billion per year, and that kids in low-income school districts were more likely to be forced to watch the commercialized broadcasts.


Original Submission

posted by mrpg on Monday July 09 2018, @04:51AM   Printer-friendly
from the you-have-systemd dept.

Submitted via IRC for Fnord666

An AI system has wiped the floor with some of China’s top doctors when it comes to diagnosing brain tumors and predicting hematoma expansion.

As reported by Xinhua, the system defeated a team comprised of 15 of China’s top doctors by a margin of two to one. The AI, BioMind, was developed by the Artificial Intelligence Research Centre for Neurological Disorders at Beijing Tiantan Hospital, and is another example of the long line of the technology analyzing images.

When diagnosing brain tumors, BioMind was correct 87 percent of the time, compared to 66 percent by the medical professionals. The AI also only took 15 minutes to diagnose the 225 cases, while doctors took 30.

In regards to predicting brain hematoma expansion, BioMind was victorious again, as it was correct in 83 percent of cases, with humans managing only 63 percent.

Source: https://thenextweb.com/science/2018/07/02/chinese-ai-beats-15-doctors-in-tumor-diagnosis-competition/


Original Submission

posted by mrpg on Monday July 09 2018, @03:20AM   Printer-friendly
from the good-bye-and-thanks-for-all-the-[dead]-fish? dept.

https://arstechnica.com/tech-policy/2018/07/trump-tweet-epa-head-scott-pruitt-is-resigning/:

On Thursday, President Trump tweeted that Environmental Protection Agency (EPA) Administrator Scott Pruitt had submitted his resignation.

Pruitt had been considered among the most loyal of Trump's appointees, but the former Oklahoma Attorney General made headlines over the past several months with repeated scandals over extravagant spending. Pruitt reportedly used agency funds to buy hundreds of thousands of dollars worth of tactical pants and other security-related items. He also used agency resources to help his wife find a job and even to help him purchase a used Trump hotel mattress. Questions about who Pruitt promoted and how raises were doled out also caused significant damage to Pruitt's public image. Thirteen different federal investigations had been opened up into the administrator's conduct.

Trump's tweet mentioned none of this Thursday afternoon, however. "I have accepted the resignation of Scott Pruitt as the Administrator of the Environmental Protection Agency," the President tweeted. "Within the Agency Scott has done an outstanding job, and I will always be thankful to him for this. The Senate confirmed Deputy at EPA, Andrew Wheeler, will..."

A second tweet continued: "... on Monday assume duties as the acting Administrator of the EPA. I have no doubt that Andy will continue on with our great and lasting EPA agenda. We have made tremendous progress and the future of the EPA is very bright!"

Also at: The Scientist and CNBC.


Original Submission

posted by mrpg on Monday July 09 2018, @01:45AM   Printer-friendly
from the dentist dept.

3000-year-old sawn-off tooth may be the earliest evidence of horse dentistry

Three thousand years ago, a horse in Mongolia had a toothache that was probably making it—and its owner—miserable. So the owner tried to help, by attempting to saw the painful top off the offending incisor. The procedure is among the earliest evidence of veterinary dentistry in the world, according to a new study, and the practices that flowed from it may have helped horses transform human civilization.

"It's a great study," says Robin Bendrey, an archaeologist and ancient horse expert at the University of Edinburgh who was not involved in the work. As horses became more important, he says, nomadic herders "are investing greater effort in understanding how to care for them."

[...] Together with another cut tooth from around the same time, the discovery shows that about 2000 years after horses were first domesticated, people were still figuring out the best way to take care of their teeth using basic stone tools.

Origins of equine dentistry (DOI: 10.1073/pnas.1721189115) (DX)


Original Submission

posted by mrpg on Monday July 09 2018, @12:11AM   Printer-friendly
from the insert-$gattaca dept.

Crime scene DNA could be used to reveal a suspect's age—and whether they have cancer

A drop of blood left by a suspect at a crime scene is a treasure trove for forensic scientists. Genetic information extracted from such biological samples can be compared against DNA databases to see whether a sample's DNA sequence is a match for any known offenders, for example. To protect individuals' privacy, these analyses, known as DNA fingerprinting, are normally restricted to parts of the genome not involved in creating proteins. But in some countries, investigators hoping to narrow down their pool of suspects are allowed to identify certain protein-coding sequences that can help predict skin or eye color. And soon, scientists may be able to find out even more from an offender's DNA—including their age.

A new forensic approach analyzes the chemical tags attached to DNA, rather than genetic sequences themselves. These molecules, which can switch genes on and off, get added onto DNA throughout our life span in a process called DNA methylation. And because the patterns of DNA methylation change as we age, they could provide a good indication of how old a suspect is.

But this technique could inadvertently reveal a lot more about a suspect's health and lifestyle [DOI: 10.1016/j.tig.2018.03.006] [DX], raising tricky legal and ethical questions that may demand new privacy safeguards, scientists suggest in a commentary in the July issue of Trends in Genetics.

A brief interview with two of the authors is included in TFA.

Related: Better DNA Hair Analysis for Catching Criminals
Creating Wanted Posters from DNA Samples
The Problems With DNA Evidence
Study Predicts Appearance From Genome Sequence Data
GEDmatch: "What If It Was Called Police Genealogy?"
DNA Collected from Golden State Killer Suspect's Car, Leading to Arrest


Original Submission