SoylentNews

Arthur T Knackerbracket has processed the following story:

Activision Blizzard reports that exposure to toxic voice chat in Call of Duty has declined by 43 percent since the beginning of this year. The publisher credits the recent implementation of AI-based moderation for the results, which have convinced it to expand its use when Call of Duty: Black Ops 6 launches on October 25.

The publisher introduced the moderator, ToxMod, when it launched Call of Duty: Modern Warfare III last November. According to ToxMod's website, the system analyzes text transcripts of in-game voice chat and detects keywords based on multiple factors.

To tell the difference between trash talk and harassment, ToxMod monitors keywords and reactions to comments, recognizing player emotions and understanding the game's code of conduct. Furthermore, the system estimates the perceived age and gender of the speaker and recipients to understand each conversation's context better.

ToxMod can't ban anyone on its own, but it can quickly flag violations for review by human moderators. Activision then decides whether to warn or mute players, only issuing bans after repeated infractions. The number of repeat offenders in Modern Warfare III and Call of Duty: Warzone fell by 67 percent since Activision implemented improvements in June 2024. In July alone, 80 percent of players caught violating voice chat rules didn't re-offend.

All regions except Asia currently support ToxMod, and Call of Duty uses the system to moderate voice chat in English, Spanish, and Portuguese. When Black Ops 6 launches, Activision will add support for French and German.

Meanwhile, text-based chat and username moderation expanded from 14 to 20 languages in August. Community Sift has been Call of Duty's text moderation partner since the first Modern Warfare reboot game launched in 2019, blocking 45 million messages since Modern Warfare III's November release.

Using AI to moderate player behavior is far less controversial than employing the technology for in-game assets. Late last year, AI-generated art appeared in content skins for Modern Warfare III. Amid the historic number of gaming industry layoffs occurring around that time, some fear publishers will try to replace artists with AI models.

Original Submission

Arthur T Knackerbracket has processed the following story:

If we were to draw an infosec Venn diagram, with one circle representing "sensitive info that attackers would want to steal" and the other "limited resources plus difficult-to-secure IT environments," education would sit in the overlap.

Schools – including K-12, colleges, and universities – store health and medical records, data belonging to minors, financial information, sensitive research, AI training models and other proprietary IP. At the same time, they are famously understaffed (with the exception of some well-heeled private institutions) and underfunded – especially when it comes to IT and security. 

Their network users include students – some as young as five years old – teachers and professors, doctors and patients, food service workers, janitors, staff, and visitors.

Plus, educational facilities and campuses have to secure IT environments that span both legacy and modern systems, covering everything from payment processing systems to medical equipment as well as personal phones, computers, and gaming consoles.

Every week, the education/research sector faces an average of 2,507 attempted cyber attacks, with everyone from nation-state groups to ransomware gangs and other financially motivated criminals putting schools in their crosshairs. At least according to Microsoft, which, in its Cyber Signals report published today, warned that Iran and North Korea are among the miscreants targeting schools.

As of the second quarter of 2024, education holds the dubious distinction of being the third most targeted industry, based on analyzed security events, Redmond notes.

upstart writes:

Ex-Twitter execs push for $200M severance as Elon Musk runs X into ground:

Former Twitter executives, including former CEO Parag Agrawal, are urging a court to open discovery in a dispute over severance and other benefits they allege they were wrongfully denied after Elon Musk took over Twitter in 2022.

According to the former executives, they've been blocked for seven months from accessing key documents proving they're owed roughly $200 million under severance agreements that they say Musk willfully tried to avoid paying in retaliation for executives forcing him to close the Twitter deal. And now, as X's value tanks lower than ever—reportedly worth 80 percent less than when Musk bought it—the ex-Twitter leaders fear their severance claims "may be compromised" by Musk's alleged "mismanagement of X," their court filing said.

The potential for X's revenue loss to impact severance claims appears to go beyond just the former Twitter executives' dispute. According to their complaint, "there are also thousands of non-executive former employees whom Musk terminated and is now refusing to pay severance and other benefits" and who have "sued in droves."

In some of these other severance suits, executives claimed in their motion to open discovery, X appears to be operating more transparently, allowing discovery to proceed beyond what has been possible in the executives' suit[e].

Arthur T Knackerbracket has processed the following story:

The Intel LGA 1700 socket, compatible with Intel’s 12th, 13th, and 14th Gen Core processors, has gained a lot of attention due to a unique issue: bending and warping of the CPUs. This phenomenon has been associated with higher operating temperatures, as it affects the evenness of contact with the CPU cooler. In response, several companies have developed contact frame replacements to mitigate these effects. This article will explore how these frames work, their benefits, and what recent developments mean for users of Intel’s latest processors.

The elongated design of Intel's LGA 1700 CPUs makes them susceptible to bending when installed in the motherboard's socket. This happens due to the pressure exerted by the Independent Loading Mechanism (ILM), which secures the CPU in place. When the CPU is not in full contact with the cooler, it can lead to uneven heat distribution and ultimately, higher temperatures.

Several manufacturers, including Thermal Grizzly and Thermalright, have developed replacement ILMs for the LGA 1700 socket, aiming to enhance cooling efficiency by ensuring better contact.

Replacing the stock ILM with a custom contact frame is a popular solution for addressing the bending issue. These aftermarket contact frames offer several key benefits:

• Better Thermal Contact: Custom frames reduce bending, allowing the integrated heat spreader (IHS) of the CPU to make more consistent contact with the cooler's base. This helps in maintaining more even temperature distribution.
• Improved Temperature Management: For example, testing with Thermalright's LGA1700-BCF contact frame on an Asus TUF Gaming Z790 Plus motherboard, paired with an Intel Core i9-13900K, showed a temperature reduction of up to 12°C. This significant decrease can be crucial for users looking to overclock their CPUs or simply maintain cooler, more stable systems.

Intel’s upcoming Arrow Lake processors, such as the Core Ultra 9, bring new challenges. Although the LGA 1700 and LGA 1851 sockets are similar in size, the Arrow Lake CPUs are slightly taller and thinner, with a shifted hotspot. This means that existing contact frames may not provide optimal performance, as they could interfere with the CPU’s metal casing.

Arrow Lake processors are designed to work on the next-generation Z890 motherboards, which are rumored to include a Reduced Load ILM. This new ILM design spreads contact points over more areas of the IHS, potentially reducing the likelihood of CPU warping and thereby improving cooling efficiency without needing aftermarket solutions.

While new contact frames have not been announced for the Z890 motherboard lineup, this could imply that Intel’s Reduced Load ILM design might address the temperature issues seen with previous LGA 1700 sockets. By reducing pressure on the CPU and ensuring uniform contact with the cooler, these new ILMs could help minimize warping and optimize thermal performance.

Original Submission

Arthur T Knackerbracket has processed the following story:

After nearly a year in orbit, the US Space Force's secretive X-37B is prepped to perform some fancy new maneuvers to alter its orbit and dump its service module before carrying on with more mysterious work.

Usually tight-lipped about the X-37B's orbital operations, the Space Force said yesterday that the seventh orbital test vehicle mission (OTV-7) would be the first time relying on aerobraking to slow a path around the Earth and reduce orbit.

"This first of a kind maneuver from the X-37B is an incredibly important milestone for the United States Space Force as we seek to expand our aptitude and ability to perform in this challenging domain," said USSF chief of space operations, General Chance Saltzman. 

The US Space Force has shared very few details about the X-37B's activities in orbit, but when it goes up, it stays up for the long haul.

Its six previous missions have all gradually become longer: Its first in 2010 was less than a year, but since then, it's managed to achieve a 908-day flight after landing in late 2022 at the end of OTV-6. That nearly two-and-a-half-year mission was also the first to send the X-37B into orbit with a service module in the form of a ring-like attachment at the rear of the craft. 

It's not immediately clear if the service module on OTV-7 is of the same design, or if it's bigger. As we noted in our discussion of OTV-7's choice of a SpaceX Falcon Heavy for launch, the rocket gave the Space Force way more cargo capacity, meaning the service module could be larger.

Either way, it looks like the plan is to get it ejected from the craft itself as part of the aerobraking maneuver. According to the Space Force, part of the orbital adjustment will include the X-37B "safely dispos[ing] of its service module components in accordance with recognized standards for space debris mitigation."

[...] OTV-7 won't be returning to Earth immediately, though: The 29-foot-long space plane still has some more experiments to conduct before heading home. All we know about what's on board is some space domain awareness technology experiments and materials being tested for their reaction to radiation exposure. 

Original Submission

Arthur T Knackerbracket has processed the following story:

About 23 million U.S. households depend on private wells as their primary drinking water source. These homeowners are entirely responsible for ensuring that the water from their wells is safe for human consumption.

Multiple studies show that, at best, half of private well owners are testing with any frequency, and very few households test once or more yearly, as public health officials recommend. Even in Iowa, which has some of the strongest state-level policies for protecting private well users, state funds for free private water quality testing regularly go unspent.

Is the water these households are drinking safe? There's not much systematic evidence, but the risks may be large.

The U.S. Environmental Protection Agency still relies on a 15-year-old study showing that among 2,000 households, one in five households' well water contained at least one contaminant at levels above the thresholds that public water systems must meet. While other researchers have studied this issue, most rely on limited data or data collected over decades to draw conclusions.

I'm an economist studying energy and agriculture issues. In a recent study, I worked with colleagues at Iowa State University, the University of Massachusetts Amherst and Cornell University to understand drinking water-related behaviors and perceptions of households that use private wells. We focused on rural Iowa, where runoff from agricultural production regularly contaminates public and private drinking water sources.

We found that few households followed public health guidance on testing their well water, but a simple intervention—sending them basic information about drinking water hazards and easy-to-use testing materials—increased testing rates. The burden of dealing with contamination, however, falls largely on individual households.

We focused on nitrate, one of the main well water pollutants in rural areas. Major sources include chemical fertilizers, animal waste and human sewage.

Drinking water that contains nitrate can harm human health. Using contaminated water to prepare infant formula can cause "blue baby syndrome," a condition in which infants' hands and lips turn bluish because nitrate interferes with oxygen transport in the babies' blood. Severe cases can cause lethargy, seizures and even death. The EPA limits nitrate levels in public water systems to 10 milligrams per liter to prevent this effect.

Studies have also found that for people of all ages, drinking water with low nitrate concentrations over long periods of time is strongly associated with chronic health diseases, including colorectal cancer and thyroid disease, as well as neural tube defects in developing fetuses.

Nitrate pollution is pervasive across the continental U.S. Fortunately, it is relatively easy to determine whether water contains unsafe nitrate concentrations. Test strips, similar to those used in swimming pools, are cheap and readily available.

upstart writes:

How a subfield of physics led to breakthroughs in AI – and from there to this year's Nobel Prize:

We covered the announcement of the Nobel Prize here. This article is to introduce the subject of Statistical Mechanics, for which you will need your thinking caps and an understanding of some serious mathematics. Follow the links for much more detail.

John J. Hopfield and Geoffrey E. Hinton received the Nobel Prize in physics on Oct. 8, 2024, for their research on machine learning algorithms and neural networks that help computers learn. Their work has been fundamental in developing neural network theories that underpin generative artificial intelligence.

A neural network is a computational model consisting of layers of interconnected neurons. Like the neurons in your brain, these neurons process and send along a piece of information. Each neural layer receives a piece of data, processes it and passes the result to the next layer. By the end of the sequence, the network has processed and refined the data into something more useful.

While it might seem surprising that Hopfield and Hinton received the physics prize for their contributions to neural networks, used in computer science, their work is deeply rooted in the principles of physics, particularly a subfield called statistical mechanics.

Statistical Mechanics is the third pillar of modern physics, next to quantum theory and relativity theory. Its aim is to account for the macroscopic behaviour of physical systems in terms of dynamical laws governing the microscopic constituents of these systems and probabilistic assumptions. Like other theories in physics, statistical mechanics raises a number of foundational and philosophical issues. But philosophical discussions in statistical mechanics face an immediate difficulty because unlike other theories, statistical mechanics has not yet found a generally accepted theoretical framework or a canonical formalism.

Journal Reference:
J J Hopfield, Neural networks and physical systems with emergent collective computational abilities, PNAS 79 (8) 2554-2558, (DOI: 10.1073/pnas.79.8.2554)

Original Submission

canopic jug writes:

Charlie Stross, a science fiction writer based in Scotland, has written a post about different possible approaches to space colonization. He includes a discussion of several different models.

While the strong form of the Sapir-Whorf hypothesis is evidently invalid, a weaker version—that language influences thought—is much harder to argue against. When we talk about a spaceship, a portmanteau word derived from "[outer] space" and "ship", we bring along certain unstated assumptions about shipping that are at odds with the physical parameters of a human-friendly life support environment for traversing interplanetary distances. Ships, in the vernacular, have captains and a crew who obey the captain via a chain of command, they carry cargo or passengers, they travel between ports or to a well-defined destination, they may have a mission whether it be scientific research or military. And of these aspects, only the scientific research angle is remotely applicable to any actually existing interplanetary vehicle, be it a robot probe like Psyche or one of the Apollo program flights.

(Pedant's footnote: while the Apollo crews had a nominal commander, actual direction came from Mission Control back on Earth and the astronauts operated as a team, along lines very similar to those later formalized as Crew Resource Management in commercial aviation.)

Anyway, a point I've already chewed over on this blog is that a spaceship is not like a sea-going vessel, can't be operated like a sea-going vessel, and the word "ship" in its name feeds into various cognitive biases that may be actively harmful to understanding what it is.

Which leads me to the similar term "space colony": the word colony drags in all sorts of historical baggage, and indeed invokes several models of how an off-Earth outpost might operate, all of which invoke very dangerous cognitive biases!

There are few more models which he missed.

Previously:
(2022) Moon Life 2030
(2022) Why Werner Herzog Thinks Human Space Colonization "Will Inevitably Fail"
(2020) Elon Musk Will Run Into Trouble Setting Up a Martian Government, Lawyers Say
(2018) Who Owns The Moon? A Space Lawyer Answers
(2017) Stephen Hawking Urges Nations to Pursue Lunar Base and Mars Landing
(2015) NASA Working on 3D Printers to Print Objects Using Martian Regolith

Original Submission

Arthur T Knackerbracket has processed the following story:

A new paper from Apple's artificial intelligence scientists has found that engines based on large language models, such as those from Meta and OpenAI, still lack basic reasoning skills.

The group has proposed a new benchmark, GSM-Symbolic, to help others measure the reasoning capabilities of various large language models (LLMs). Their initial testing reveals that slight changes in the wording of queries can result in significantly different answers, undermining the reliability of the models.

The group investigated the "fragility" of mathematical reasoning by adding contextual information to their queries that a human could understand, but which should not affect the fundamental mathematics of the solution. This resulted in varying answers, which shouldn't happen.

"Specifically, the performance of all models declines [even] when only the numerical values in the question are altered in the GSM-Symbolic benchmark," the group wrote in their report. "Furthermore, the fragility of mathematical reasoning in these models [demonstrates] that their performance significantly deteriorates as the number of clauses in a question increases."

The study found that adding even a single sentence that appears to offer relevant information to a given math question can reduce the accuracy of the final answer by up to 65 percent. "There is just no way you can build reliable agents on this foundation, where changing a word or two in irrelevant ways or adding a few bit of irrelevant info can give you a different answer," the study concluded.

[...] "We found no evidence of formal reasoning in language models," the new study concluded. The behavior of LLMS "is better explained by sophisticated pattern matching" which the study found to be "so fragile, in fact, that [simply] changing names can alter results."

Original Submission

Arthur T Knackerbracket has processed the following story:

The Internet Archive is back online in a read-only state after a cyberattack brought down the digital library and Wayback Machine last week. A data breach and DDoS attack kicked the site offline on October 9th, with a user authentication database containing 31 million unique records also stolen in recent weeks.

The Internet Archive is now back online in a “provisional, read-only manner,” according to founder Brewster Kahle. “Safe to resume but might need further maintenance, in which case it will be suspended again.”

While you can access the Wayback Machine to search 916 billion web pages that have been archived over time, you can’t currently capture an existing web page into the archive. Kahle and team have gradually been restoring Archive.org services in recent days, including bringing back the team’s email accounts and its crawlers for National Libraries. Services have been offline so that Internet Archive staff can examine and strengthen them against future attacks.

[...] The Internet Archive outage came just weeks after Google started adding links to archived websites in the Wayback Machine. Google removed its own cached pages links earlier this year, so having the Wayback Machine linked in Google search results is a useful way to access older versions of websites or archived pages.

Previously: Archive.org, a Repository Storing the Entire History of the Internet, Has a Data Breach

Original Submission

upstart writes:

Should I be more or less scared of the doctor?

The products include a service that helps healthcare organizations build their own AI agents:

Microsoft revealed a slew of new artificial intelligence capabilities for healthcare organizations Thursday, including a product to help companies to build their own AI agents.

The technology giant also announced foundation models for medical imaging and a healthcare data analysis platform, as well as details about its plans to build an AI documentation product geared toward nurses.

Healthcare organizations have shown increased interest in adopting AI tools, even as some experts and lawmakers raise questions about its safe and equitable use. Tech companies say the products have potential to help providers manage their workloads and alleviate burnout.

"We're at an inflection point. AI breakthroughs are changing, augmenting how we work and live," Kees Hertogh, vice president for healthcare and life sciences product marketing, said during a press briefing. "The integration of AI into healthcare has significantly enhanced patient care and is rekindling the joy of practicing medicine for clinicians."

Microsoft's agent service would allow companies to create AI tools with pre-built templates and data sources that could be used for appointment scheduling, clinical trial matching and patient triage, Hertogh said.

The service is currently in public preview, which allows wider access to the tools and lets organizations give feedback on the product, according to a spokesperson.

In one example, a doctor could ask an AI agent to find clinical trials for a 55-year-old patient with diabetes and interstitial lung disease.

[...] The product allows organizations to build agents with healthcare-specific features using intelligence from credible sources, which aims to improve safety, said Hadas Bitran, partner general manager of health AI at Microsoft Health and Life Sciences.

Arthur T Knackerbracket has processed the following story:

Nvidia's latest Blackwell GPUs are experiencing unprecedented demand, with the company reporting that it has sold out of these next-gen processors. Nvidia CEO Jensen Huang revealed the news during an investors meeting hosted by Morgan Stanley. Morgan Stanley Analyst Joe Moore notes that Nvidia executives disclosed that their Blackwell GPU products have a 12-month backlog, echoing a similar situation with Hopper GPUs several quarters ago.

The overwhelming demand for Blackwell GPUs comes from Nvidia's traditional customers, including major tech giants like AWS, CoreWeave, Google, Meta, Microsoft, and Oracle. These companies have purchased every Blackwell GPU Nvidia and its manufacturing partner TSMC can produce for the next four quarters. The extreme demand indicates that Nvidia's already considerable footprint in the AI processor market should continue to grow next year, even as competition from rivals such as AMD, Intel, and various cloud service providers grab their share.

"Our view continues to be that Nvidia is likely to gain share of AI processors in 2025, as the biggest users of custom silicon are seeing very steep ramps with Nvidia solutions next year," Moore said in a client note.

Nvidia unveiled the Blackwell GPU platform in March. It includes the B200 GPU and GB200 Grace "super chip." These processors can handle the demanding workloads of large language model (LLM) inference while significantly reducing energy consumption, a growing concern in the industry.

[...] In August, Nvidia acknowledged that its Blackwell-based products were experiencing low yields, necessitating a re-spin of some layers of the B200 processor to improve production efficiency. Despite these challenges, Nvidia remains confident in its ability to ramp up Blackwell production in Q4 2024. It expects to ship several billion dollars worth of Blackwell GPUs in the last quarter of this year.

Original Submission

canopic jug writes:

Software developer and former computer science student Amit Patel has written a post about generating curved text for maps and other purposes.

Over the last few posts I wrote about things I did to improve font quality, such as antialiasing and combining distance fields to merge outlines and halos. But I want to "pop up the stack" a bit and talk about one of the bigger goals for this project. I want to render text in styles that I've seen in maps, both online and offline, both fantasy and real. In particular, I want to apply spacing, rotation, and curvature to the labels.

[...] These are common in cartography, not only in fantasy maps like Tolkein's but also in real-world maps. Eduard Imhof's classic 1975 paper, Positioning Names on Maps^[1] has a ton of great advice on how to position labels, and not only recommends curving text, but also sketches out examples:

In some ways, that desktop computing has moved beyond monospace type comes down to the long term impact of a course in calligraphy and typography which took place decades ago. Now even your GUIs on GNU/Linux or the BSDs provide a variety of typefaces — in straight lines. Text along curved paths is another step forward.

Previously:
(2020) Fonts for Programmers and Developers

Original Submission

upstart writes:

The new text is timed to a new California law against false advertising:

There comes a point in most experienced Steam shoppers' lives where they wonder what would happen if their account was cancelled or stolen, or perhaps they just stopped breathing. It's scary to think about how many games in your backlog will never get played; scarier, still, to think about how you don't, in most real senses of the word, own any of them.

Now Valve, seemingly working to comply with a new California law targeting "false advertising" of "digital goods," has added language to its checkout page to confirm that thinking. "A purchase of a digital product grants a license for the product on Steam," the Steam cart now tells its customers, with a link to the Steam Subscriber Agreement further below.

California's AB2426 law, signed by Gov. Gavin Newsom Sept. 26, excludes subscription-only services, free games, and digital goods that offer "permanent offline download to an external storage source to be used without a connection to the internet." Otherwise, sellers of digital goods cannot use the terms "buy, purchase," or related terms that would "confer an unrestricted ownership interest in the digital good." And they must explain, conspicuously, in plain language, that "the digital good is a license" and link to terms and conditions.

[...] Online-only content licenses have always existed in a precarious state, but recent corporate maneuvers have seen them teetering especially hard. Ubisoft deleted The Crew,its online-only racing game, from its servers on April 1, and thereby cut off access for those who bought it. Warner Bros. Discovery spent months in early 2024 moving toward a wipe-out of all Adult Swim Games titles listed on Steam and elsewhere, only to do something far more sensible at the last moment. Sony tried in late 2023 to delete more than 1,000 Discovery video titles from PlayStation owners' libraries, then walked that back. And then a couple months later, it jumped back into the online ire mix by nixing a wealth of Funimation anime offerings that had once been promised to be available "forever."

So long as Steam lasts forever, however, this new language should not be that alarming.

Original Submission

Ward Christensen, inventer of XMODEM and co-inventer of the BBS, has died

Snotnose writes:

Not much I can say, .Story here"

On Friday, Ward Christensen, co-inventor of the computer bulletin board system (BBS), died at age 78 in Rolling Meadows, Illinois. Christensen, along with Randy Suess, created the first BBS in Chicago in 1978, leading to an important cultural era of digital community-building that presaged much of our online world today.

RIP, man who's code I used a lot back in the day.

Ward Christensen, BBS Inventor and Architect of Our Online Age, Dies at Age 78

upstart writes:

Ward Christensen, BBS inventor and architect of our online age, dies at age 78:

On Friday, Ward Christensen, co-inventor of the computer bulletin board system (BBS), died at age 78 in Rolling Meadows, Illinois. Christensen, along with Randy Suess, created the first BBS in Chicago in 1978, leading to an important cultural era of digital community-building that presaged much of our online world today.

Friends and associates remember Christensen as humble and unassuming, a quiet innovator who never sought the spotlight for his groundbreaking work. Despite creating one of the foundational technologies of the digital age, Christensen maintained a low profile throughout his life, content with his long-standing career at IBM and showing no bitterness or sense of missed opportunity as the Internet age dawned.

"Ward was the quietest, pleasantest, gentlest dude," said BBS: The Documentary creator Jason Scott in a conversation with Ars Technica. Scott documented Christensen's work extensively in a 2002 interview for that project. "He was exactly like he looks in his pictures," he said, "like a groundskeeper who quietly tends the yard."

Tech veteran Lauren Weinstein initially announced news of Christensen's passing on Sunday, and a close friend of Christensen's confirmed to Ars that Christensen died peacefully in his home. The cause of death has not yet been announced.