SoylentNews

upstart writes:

With revenues crashing, Samsung will make fewer DRAM chips to try and save the day:

While Nvidia decided to adopt an insane price policy for its latest generation of gaming GPUs, market conditions for other hardware components have never been better for consumers. Memory chips for SSD drives and DRAM memory modules are very affordable right now, and manufacturers are struggling to adapt their businesses to this new trend.

Being one of the largest DRAM chip manufacturers in the world, Samsung is experiencing a rather critical moment in its recent history. The company's revenues suffered a -18% year-over-year drop in the first quarter of 2023, while profits have essentially evaporated with the worst result recorded for the past 14 years (-95%).

Global DRAM and SSD prices are expected to become even cheaper in the coming months, which will clearly affect any rebound effort put in place by manufacturers. Samsung recently said that it was preparing to cut back on chip production, and now we know the cut will likely be a significant one.

The Korean giant is seemingly expected to decrease its chip manufacturing output for DDR4 memory modules, a move that will last for 3 to 6 months (two quarters) at least. Samsung's total wafer input will be reduced by 5-7%, which will likely have further negative consequences on an already struggling market.

Freeman writes:

https://arstechnica.com/tech-policy/2023/05/meaningful-harm-from-ai-necessary-before-regulation-says-microsoft-exec/

As lawmakers worldwide attempt to understand how to regulate rapidly advancing AI technologies, Microsoft chief economist Michael Schwarz told attendees of the World Economic Forum Growth Summit today that "we shouldn't regulate AI until we see some meaningful harm that is actually happening, not imaginary scenarios."

The comments came about 45 minutes into a panel called "Growth Hotspots: Harnessing the Generative AI Revolution." Reacting, another featured speaker, CNN anchor Zain Asher, stopped Schwarz to ask, "Wait, we should wait until we see harm before we regulate it?"
[...]
Lawmakers are racing to draft AI regulations that acknowledge harm but don't threaten AI progress. Last year, the US Federal Trade Commission (FTC) warned Congress that lawmakers should exercise "great caution" when drafting AI policy solutions. The FTC regards harms as instances where "AI tools can be inaccurate, biased, and discriminatory by design and incentivize relying on increasingly invasive forms of commercial surveillance." More recently, the White House released a blueprint for an AI Bill of Rights, describing some outcomes of AI use as "deeply harmful," but "not inevitable."

Original Submission

upstart writes:

Fact: Earth's colossal ice sheets are melting:

Pay attention to Greenland.

The land's colossal ice sheet — around three times the size of Texas — is melting some 270 billion tons(opens in a new tab) of ice into the sea each year as Earth warms. And the inevitable sea level rise could be worse than scientists calculated: Researchers at NASA and the University of California, Irvine (UCI) found that warmer ocean water is seeping underneath and amplifying melting of Greenland's mighty Petermann Glacier, which ends in a great ice tongue floating over the sea. The scientists recently published their research in the journal Proceedings of the National Academy of Sciences.

The glacier lies in northern Greenland, a realm of the high Arctic. But that frigid location can no longer protect it. Scientists found the glacier is vulnerable to the incessantly warming seas. It's another whammy for melting Greenland, which is melting from above (warmer air) and below (warmer water).

Until 2015, satellite observations showed Petermann, a major ice outflow on Greenland, was in solid shape. Not anymore.

"Something changed during the last decade. Petermann was supposed to be a place where the ice was still stable," Enrico Ciraci, a NASA postdoctoral fellow and an Earth system scientist at UCI, told Mashable.

Ice loss is now ramping up.

upstart writes:

Goodbye to Roblox on Linux with their new anti-cheat and Wine blocking:

You might have seen recently that I covered the upcoming updates for Roblox*, and now it's here blocking Wine with their new anti-cheat. This means you won't be able to play it on Linux any more, at all, unless you find some sort of special workaround.

Previously the roll-out of this update was being tested only with some users. Now though it's here for everyone giving a 64 bit client and introducing their Hyperion anti-cheat software which they are intentionally blocking Wine with. Naturally plenty of Roblox fans on Linux are upset by this, asking their team for updates on what their plans are.

In a fresh statement on their official developer forum one of their staff said this, in reply to users asking about updates in regards to Linux support:

Hi - thanks for the question. I definitely get where you're coming from, and as you point out, you deserve a clear, good-faith answer. Unfortunately that answer is essentially "no."

[...] Again, I'm personally sorry to have to say this. Way back in 2000 I had a few patches accepted into the kernel, and I led the port of Roblox game servers from Windows to Linux several years ago. From a technical and philosophical perspective, it would be a wonderful thing to do. But our first responsibility is to our overall community, and the opportunity cost of supporting a Linux client is far, far too high to justify.

They're clearly not going to be releasing a Native Linux build, which I think most people probably already knew, but at least previously they repeatedly said that Wine was a "priority" to support but now it doesn't sound as likely going by the above.

What is Roblox? Roblox is an app that allows users to play a wide variety of games, create games, and chat with others online. It combines gaming, social media, and social commerce. Billing itself as the “ultimate virtual universe,” Roblox experiences are places where users can socialize, build their own spaces, and even earn and spend virtual money.

Apparently, it is very "popular with kids".

Original Submission

Freeman writes:

https://arstechnica.com/science/2023/05/pharma-company-behind-shkrelis-infamous-4000-price-hike-files-for-bankruptcy/

The pharmaceutical company behind Martin Shkreli's infamous 4,000 percent price hike—now known as Vyera Pharmaceuticals—filed for bankruptcy this week and plans to sell its assets to pay off millions in debts.

In court documents filed Wednesday, Vyera's chief restructuring officer, Lawrence Perkins, largely blamed Shkreli for dooming the company and its affiliates.
[...]
Shkreli founded Vyera in 2014 under the name Turing Pharmaceuticals. His focus was to acquire sole-source drugs that treat life-threatening conditions in small populations of patients—and then dramatically jack up the price. In August 2015, he did just that, buying the rights to the decades-old anti-parasitic drug Daraprim for $55 million and abruptly raising the price from $17.60 per tablet to $750, a more than 4,000 percent increase.
[...]
Shkreli's influence wasn't shaken until January 2020, when the Federal Trade Commission and several state attorneys general sued Shkreli and the company—then called Vyera—for allegedly violating antitrust laws. Soon after, Vyera appointed a new board and management to purge ties to Shkreli. Vyera later settled the FTC's lawsuit, while Shkreli insisted on going to trial, where he lost, was banned from the pharmaceutical industry for life, and ordered to pay roughly $65 million in disgorgement. He is appealing the ruling.

Meanwhile, Vyera never reversed Shkreli's price hike.

Previously:
Cost of Daraprim Medication Raised by Over 50 Times 20150922
Stories mentioning Shkreli on Soylentnews 21+ stories (Famous/infamous, same dif, right?)

Related:
Martin Shkreli Launches Blockchain-Based Drug Discovery Platform 20220726
FTC: Shkreli May Have Violated Lifetime Pharma Ban, Should be Held in Contempt 20230125
Shkreli Tells Judge His Drug Discovery Software is Not for Discovering Drugs 20230215

Original Submission

upstart writes:

To compete with American rivals, Eutelsat's Eva Berneke first has to navigate Russia's war in Ukraine, Brexit politics, and jamming attacks by Iran:

Eva Berneke describes her first year at the helm of the world's third-largest satellite company as a "whirlwind." That's an understatement. Since she took over the top job at Eutelsat in January 2022, the Danish CEO has become a direct competitor to Elon Musk, been accused by the Ukrainian government of aiding Russian propaganda, and found herself in the thick of bitter Brexit politics—and that's before you even mention the Iranian sabotage attempt.

[...] Undaunted, Berneke responded by initiating her own shake-up. In July, the company announced plans to merge with struggling British satellite provider OneWeb. As part of the deal, Eutelsat absorbed OneWeb's constellation of 648 low-orbit satellites. At just 1,200 km above Earth, the OneWeb fleet delivers faster internet speeds than Eutelsat's geostationary satellites, which sit 35,000 km above the planet's surface.

[...] The OneWeb-Eutelsat merger has been touted as Europe's entry into the space race. It is the only company currently competing with Musk's Starlink in the low-orbit market. But to claim its title as a European space giant, Eutelsat first has to navigate messy post-Brexit politics. Both France's Eutelsat and Britain's OneWeb were part-owned by their respective governments, and the two countries will continue to own stakes in the new business.

Berneke admits Brexit has brought challenges. "But there's been a willingness on both sides to reach across the Channel to try and find a good way of collaborating," she says. If Europe wants a homegrown satellite giant, Britain and France will have to resolve their differences. "[OneWeb's fleet] is going to be one of the only non-US-based constellations for a while," she says.

upstart writes:

According to Chinese officials, the suspect used OpenAI's chatbot to create multiple variations of a news story falsely reporting a fatal train crash:

Chinese authorities have detained a man in the Gansu province in Northern China for allegedly using ChatGPT to write fake news articles. The move appears to be one of the first arrests made under China's new anti-AI guidelines, which (among other restrictions) prohibit artificial intelligence services from being misused to distribute "false information."

The suspect, identified only by his surname Hong, is accused of using OpenAI's chatbot to generate news articles describing a fatal train crash that officials say was "false information," according to a police statement reported by South China Morning Post. After discovering the article on April 25th, authorities found multiple versions of the same story with different accident locations had been simultaneously posted to 20 additional accounts on Baidu-owned blogging platform Baijiahao.

Hong claimed he was using ChatGPT to rewrite articles and generate money through internet traffic.

[...] Hong was specifically charged for "picking quarrels and provoking trouble" — a catch-all offense that the South China Morning Post says can be applied to suspects accused of creating and / or spreading misinformation online. That isn't the only application of the charge, however, which can also be broadly defined as undermining public order or causing disorder in public places. The wording of the offense is vague and has been widely criticized for its potential to muffle free speech and arrest activists criticizing the Chinese government. Those charged can face a five-to-10-year prison term.

Original Submission

Freeman writes:

https://arstechnica.com/tech-policy/2023/05/ex-ubiquiti-engineer-behind-breathtaking-data-theft-gets-6-year-prison-term/

An ex-Ubiquiti engineer, Nickolas Sharp, was sentenced to six years in prison yesterday after pleading guilty in a New York court to stealing tens of gigabytes of confidential data, demanding a $1.9 million ransom from his former employer, and then publishing the data publicly when his demands were refused.

Sharp had asked for no prison time, telling United States District Judge Katherine Polk Failla that the cyberattack was actually an "unsanctioned security drill" that left Ubiquiti "a safer place for itself and for its clients," Bloomberg reported. In a court document, Sharp claimed that Ubiquiti CEO Robert Pera had prevented Sharp from "resolving outstanding security issues," and Sharp told the judge that this led to an "idiotic hyperfixation" on fixing those security flaws.

However, even if that was Sharp's true motivation, Failla did not accept his justification of his crimes, which include wire fraud, intentionally damaging protected computers, and lying to the FBI.

"It was not up to Mr. Sharp to play God in this circumstance," Failla said.
[...]
Sharp executed his plan, and he might have gotten away with it if not for a "slip-up" [...] While copying approximately 155 data repositories, an Internet outage temporarily disabled his VPN. When Internet service was restored, unbeknownst to Sharp, Ubiquiti logged his home IP address before the VPN tool could turn back on.

Original Submission

upstart writes:

People upwind can hear you hollering into a breeze, but it's hard to hear yourself:

Shouting into the wind isn't so ineffective after all.

The idiom is commonly used to describe an unsuccessful attempt to communicate. But it's not actually more difficult to shout upwind, says acoustics researcher Ville Pulkki of Aalto University in Espoo, Finland.

Sending a sound upwind, against the flow of air, makes the sound louder due to an acoustical effect called convective amplification. Sound sent downwind is quieter. So, if you're yelling upwind, a listener standing in front of you should have no problem hearing you — contrary to popular belief.

The misperception has a simple explanation, Pulkki says. "When you yell against the wind, you hear yourself worse." That's because, in this scenario, your ears are downwind of your mouth. That means your own voice sounds quieter to you.

[...] A similar effect occurs when an ambulance goes by. Most people are familiar with the sudden change of pitch of the siren's sound due to the Doppler effect (SN: 8/2/13). But the siren is also slightly louder when moving toward a stationary observer than it is when it's moving away. When you're bellowing upwind, it's not the source of sound that's moving, but the medium in which the sound travels.

Whichever way the wind blows, acoustics can explain it.

Original Submission

upstart writes:

But it's up to NASA to approve a rescue mission. Cue Aerosmith:

Momentus and Astroscale, two startups specializing in space infrastructure and orbital debris, want to collaborate and help boost NASA's aging Hubble Space Telescope into a safe orbit.

Hubble has far exceeded its original mission and expected run time, thanks to five space shuttle missions that sent astronauts to repair its instruments between 1993 and 2009.

NASA reckons there's more life in Hubble yet – but only if its altitude can be pushed higher to stop it falling and reentering Earth's atmosphere in the mid 2030s. Atmospheric drag has been slowly degrading the satellite's orbit, and it is expected to drop to 500 kilometers above Earth by about 2025.

In a bid to save the sinking telescope, NASA issued a Request for Information (RFI) in December to explore potential solutions from commercial vendors. NASA's not going to spend any money on this, but organizations willing to do the job would receive "technical information and technical consultation," from NASA Goddard.

Now Momentus and Astroscale have announced they're willing to collaborate on a potential future servicing mission to shift the Hubble Space Telescope into a safer orbit and remove any debris that could collide with the probe.

[...] "Leveraging Momentus's flight heritage with three orbital service vehicles on-orbit today and Astroscale's expertise in RPOD (rendezvous, proximity operations and docking), we found our product suites to be synergistic in support of a major NASA mission," said John Rood, Momentus CEO. "Even at 33, Hubble is fully capable of continuing its mission; where it is aging is in its orbital stability."

It's not known what other companies, if any, also responded to NASA's RFI, since the space agency promised to keep the information confidential. Last year, NASA agreed to study the technical feasibility of boosting Hubble with SpaceX's Dragon capsule, and began collecting data.

Neither that study nor the RFI guarantee that NASA will carry out a Hubble servicing mission, however. The space agency may well decide that, with newer tech in place and future exploration planned, Hubble's days are coming to an end.

Original Submission

Freeman writes:

https://arstechnica.com/information-technology/2023/05/microsoft-patches-secure-boot-flaw-but-wont-enable-fix-by-default-until-early-2024/

Earlier this week, Microsoft released a patch to fix a Secure Boot bypass bug used by the BlackLotus bootkit we reported on in March. The original vulnerability, CVE-2022-21894, was patched in January, but the new patch for CVE-2023-24932 addresses another actively exploited workaround for systems running Windows 10 and 11 and Windows Server versions going back to Windows Server 2008.

The BlackLotus bootkit is the first-known real-world malware that can bypass Secure Boot protections, allowing for the execution of malicious code before your PC begins loading Windows and its many security protections. Secure Boot has been enabled by default for over a decade on most Windows PCs sold by companies like Dell, Lenovo, HP, Acer, and others. PCs running Windows 11 must have it enabled to meet the software's system requirements.
[...]
Additionally, once the fixes have been enabled, your PC will no longer be able to boot from older bootable media that doesn't include the fixes. On the lengthy list of affected media: Windows install media like DVDs and USB drives created from Microsoft's ISO files; custom Windows install images maintained by IT departments; full system backups; network boot drives including those used by IT departments to troubleshoot machines and deploy new Windows images; stripped-down boot drives that use Windows PE; and the recovery media sold with OEM PCs.

I.E.: You will have to turn "Secure Boot" off in order to install Linux, probably.

Original Submission

upstart writes:

Opinion: Most people are terrible at matching faces to photos, making polling checks unreliable:

On Thursday May 4, for the first time, members of the public voting in local council elections in England were required to bring photo ID to their polling station. Initial reports suggested that a few people were turned away because they didn't bring one of the approved forms of photo ID.

But even if they did bring the right documents, such as a driving license or passport, there's a question mark over whether the people manning polling stations could tell accurately whether the voter was the person pictured in the ID.

When you present your photo ID to be checked, the person looking at it has to decide if your face matches the picture in the document. In a lab, this is usually done with images and is called "face matching". Such studies typically present two face images side-by-side and ask people to judge whether the images show the same person or two different people.

While people perform well at this task when they are familiar with the person pictured, studies report the error rate can be as high as 35% when those pictured are unfamiliar. Even when people are asked to compare a live person standing in front of them with a photo, a recent study found they still got more than 20% of their answers wrong.

The people checking our photo ID are almost always unfamiliar with us, so we should expect that this is a difficult, error-prone task for them. And while you might think that people whose job it is to check photo ID would be better at it than the rest of us, cashiers, police officers and border control officers have all been shown to be as poor at face matching as untrained people.

The study of border control officers also showed they don't improve at the task as time goes on—there was no relationship between their performance and the number of years they had spent in the job.

upstart writes:

Elon Musk says he's found a new CEO for Twitter, a woman who will start in 6 weeks:

Elon Musk said Thursday he has found a new CEO for Twitter, or X Corp. as it's now called.

He did not name the person but she will be starting in about six weeks.

Musk, who bought Twitter last fall and has been running it since, has been insisting he is not the company's permanent CEO.

The Tesla billionaire said in a tweet Thursday that his role will transition to being Twitter's executive chairman and chief technology officer.

Musk has been saying for nearly six months that he plans to find a new CEO for San Francisco-based Twitter.

Original Submission

owl writes:

https://www.righto.com/2020/07/inside-8086-processor-tiny-charge-pumps.html

Introduced in 1978, the revolutionary Intel 8086 microprocessor led to the x86 processors used in most desktop and server computing today. This chip is built from digital circuits, as you would expect. However, it also has analog circuits: charge pumps that turn the 8086's 5-volt supply into a negative voltage to improve performance.1 I've been reverse-engineering the 8086 from die photos, and in this post I discuss the construction of these charge pumps and how they work.

[...] An integrated circuit starts with a silicon substrate, and transistors are built on this. For high-performance integrated circuits, it is beneficial to apply a negative "bias" voltage to the substrate. 2 To obtain this substrate bias voltage, many chips in the 1970s had an external pin that was connected to -5V,3 but this additional power supply was inconvenient for the engineers using these chips. By the end of the 1970s, however, on-chip "charge pump" circuits were designed that generated the negative voltage internally. These chips used a single convenient +5V supply, making engineers happier.

Original Submission

upstart writes:

China's Spaceplane Conducted Multiple Maneuvers With a Mystery Object in Orbit:

The Chinese spaceplane finally returned to Earth earlier this week, but we're still learning more about its time in orbit. The spacecraft caught and released an unidentified object several times during its flight, performing a series of maneuvers that were captured by orbital radars, according to California-based LeoLabs. The company released its observational data, saying in a tweet that the data shows there were at least two capture and docking operations performed by the spacecraft.

The experimental launch vehicle took off from the Jiuquan Launch Center on August 5 as a classified payload on board a Long March 2F carrier rocket. This was the reusable spacecraft's second time to fly, with its first launch taking place in 2020. The spaceplane only stayed in orbit for four days during its inaugural flight but far outdid itself the second time around.

The spaceplane landed on May 8 after spending 276 days in orbit. The China Academy of Launch Vehicle Technology, a state-owned manufacturer that makes both civilian and military space launch vehicles, shared very little information about its craft. Observers of low Earth orbit, however, were able to track the spaceplane's activities during its lengthy flight.

In November 2022, the U.S. Space Force's 18th Space Defense Squadron tracked an object that may have been ejected from the spaceplane. The object remained unidentified, although some speculated it may have been a satellite used to track the spaceplane's performance in orbit.

That same object may be what LeoLabs tracked with its global network of radars. "Since its launch on 4 August 2022, we observed multiple large maneuvers raising the object's altitude — as well as repeated deployments, formation flying, and docking of a sub-satellite Object J (NORAD ID 54218)," the company said in its Twitter thread.

[...] China's experimental vehicle operates like a regular aircraft in Earth's atmosphere and a spacecraft in space, allowing it to complete missions in orbit and then return to Earth's surface, where it performs a horizontal landing. China isn't the only country testing this type of spacecraft; the U.S. Space Force has its own spaceplane. The Boeing X-37 launched in May 2020 for its sixth test flight and landed back on Earth in November 2022 after spending more than two years in orbit.

Original Submission