SoylentNews

fliptop writes:

Is 'health monitoring' really worth exposure to hormone-disrupting 'forever chemicals' and constant data mining?

Fitness trackers and smart watches have become a staple of modern life, with well over 100 million Americans using these devices to monitor their health.

More recently, they've gained popularity with children, many of whom will no doubt head back to school after the holiday sporting their very own unsightly high-tech wristwear.

[...] A recent study from the University of Notre Dame revealed a rather shocking truth: These devices are loaded with toxic chemicals. Many fitness tracker bands contain per- and polyfluoroalkyl substances, commonly known as "forever chemicals" due to their near-indestructible nature.

Expensive bands, ironically marketed for quality, often have the highest levels of these harmful substances.

[...] Exposure to PFAS has been linked to various health issues, including cancer, hormonal disruptions, developmental delays in children, and weakened immune systems.

Among the most troubling effects is the potential for hormonal disruption, particularly in men, where PFAS exposure has been associated with plummeting testosterone levels. Given that testosterone plays a critical role in male health — affecting everything from energy levels to muscle mass and mental well-being — this trend is deeply concerning.

For wearable devices, the problem lies in dermal absorption — the process by which these chemicals seep into the skin during prolonged contact. Since wearables are often worn for hours, if not around the clock, the skin acts as a direct pathway for these harmful substances to enter the body.

TFA also touches on the data goldmine of personal information that can often be shared with third parties.

DOI: Alyssa Wicks, Heather D. Whitehead, Graham F. Peaslee, PRESENCE OF PERFLUOROHEXANOIC ACID IN 4 FLUOROELASTOMER WATCH BANDS

Original Submission

upstart writes:

The research demonstrated that participants could detect faces quicker under surveillance without being consciously aware of this enhanced perception:

A new psychological study has shown that when people know they are under surveillance it generates an automatic response of heightened awareness of being watched, with implications for public mental health.

In a paper published in the journal Neuroscience of Consciousness psychology researchers from the University of Technology Sydney (UTS) worked with 54 participants to examine the effects of surveillance on an essential function of human sensory perception – the ability to detect another person's gaze.

Lead author, Associate Professor of neuroscience and Behaviour Kiley Seymour, said previous research has established the effects on conscious behavior when people know they are being watched, but the new study provided the first direct evidence that being watched also has an involuntary response.

"We know CCTV changes our behavior, and that's the main driver for retailers and others wanting to deploy such technology to prevent unwanted behavior," Associate Professor Seymour said.

"However, we show it's not only overt behavior that changes – our brain changes the way it processes information.

"We found direct evidence that being conspicuously monitored via CCTV markedly impacts a hardwired and involuntary function of human sensory perception – the ability to consciously detect a face.

"It's a mechanism that evolved for us to detect other agents and potential threats in our environment, such as predators and other humans, and it seems to be enhanced when we're being watched on CCTV.

[...] "We had a surprising yet unsettling finding that despite participants reporting little concern or preoccupation with being monitored, its effects on basic social processing were marked, highly significant, and imperceptible to the participants.

"The ability to rapidly detect faces is of critical importance to human social interactions. Information conveyed in faces, such as gaze direction, enables us to construct models of other people's minds and to use this information to predict behavior.

Journal Reference: "Big brother: the effects of surveillance on fundamental aspects of social vision" by Kiley Seymour, Jarrod McNicoll and Roger Koenig-Robert, 10 December 2024, Neuroscience of Consciousness. DOI: 10.1093/nc/niae039.

Original Submission

As Kiribati (and in particular the island of Kiritimati) is now beginning 2025 I think that this is as good a time as any to wish everyone a very Happy New Year and all the very Best Wishes for 2025.

upstart writes:

The hype is fading, and people are asking what generative artificial intelligence is really good for. So far, no one has a decent answer:

Generative AI took the world by storm in November 2022, with the release of OpenAI's service ChatGPT. One hundred million people started using it, practically overnight. Sam Altman, the CEO of OpenAI, the company that created ChatGPT, became a household name. And at least half a dozen companies raced OpenAI in an effort to build a better system. OpenAI itself sought to outdo GPT-4, its flagship model, introduced in March 2023, with a successor, presumably to be called GPT-5. Virtually every company hurtled to find ways of adopting ChatGPT (or a similar technology, made by other companies) into their business.

There is just one thing: Generative AI doesn't actually work that well, and maybe it never will.

[...] Systems that are frequently wrong and never in doubt make for fabulous demos, but are often lousy products in themselves. If 2023 was the year of AI hype, 2024 has been the year of AI disillusionment. Something that I argued in August 2023, to initial skepticism, has been felt more frequently: generative AI might turn out to be a dud. The profits aren't there—estimates suggest that OpenAI's 2024 operating loss may be $5 billion—and the valuation of more than $80 billion doesn't line up with the lack of profits. Meanwhile, many customers seem disappointed with what they can actually do with ChatGPT, relative to the extraordinarily high initial expectations that had become commonplace.

Furthermore, essentially every big company seems to be working from the same recipe, making bigger and bigger language models, but all winding up in more or less the same place, which is models that are about as good as GPT-4, but not a whole lot better. What that means is that no individual company has a "moat" (a business's ability to defend its product over time), and what that in turn means is that profits are dwindling. OpenAI has already been forced to cut prices; now Meta is giving away similar technology for free.

As I write this, OpenAI has been demoing new products but not actually releasing them. Unless it come outs with some major advance worthy of the name of GPT-5 before the end of 2025 that is decisively better than what their competitors can offer, the bloom will be off the rose. The enthusiasm that propped up OpenAI will diminish, and since it is the poster child for the whole field, the entire thing may well soon go bust.

Original Submission

Arthur T Knackerbracket has processed the following story:

After years of planning, building, geopolitical wrangling, and workforce challenges, the world's largest semiconductor foundry, Taiwan Semiconductor Manufacturing Co. (TSMC) is officially starting mass production at an advanced chip-manufacturing facility in Phoenix in 2025. The fab represents the arrival of advanced chip manufacturing in the United States and a test of whether the 2022 CHIPS and Science Act can help stabilize the semiconductor-industry supply chain for the United States and its allies.

In late October 2024, the company announced that yields at the Arizona plant were 4 percent higher than those at plants in Taiwan, a promising early sign of the fab's efficiency. The current fab is capable of operating at the 4-nanometer node, the process used to make Nvidia's most advanced GPUs. A second fab, set to be operational in 2028, plans to offer 2- or 3-nm-node processes. Both 4-nm and more advanced 3-nm chips began high-volume production at other TSMC fabs in 2022, while the 2-nm node will begin volume production in Taiwan this year. In the future, the company also has plans to open a third fab in the United States that will use more advanced technology.

The chip-manufacturing giant is currently set to receive US $6.6 billion in CHIPS Act funding for building the first Arizona plant. But government funding isn't the only reason semiconductor manufacturing is coming back to the United States. TSMC makes 90 percent of the world's advanced chips, and U.S.-based companies including Apple, Nvidia, Google, Amazon, and Qualcomm rely on them. The chip shortages during the economic shock of the early COVID years, and Chinese president Xi Jinping's increasingly aggressive rhetoric about Taiwan, have made TSMC's customers and international policymakers uncomfortable.

TSMC announced their intention to invest in Arizona in 2020. "The CHIPS Act didn't make it happen—companies have largely moved on their own," says TechInsights semiconductor analyst Dan Hutcheson. Big customers like Apple have been pushing TSMC to build fabs elsewhere to minimize risk, he says.

Hutcheson says having TSMC fabs outside Taiwan is good for the company's customers and good for Taiwan. The island's "silicon shield" against China has done its work - TSMC's dominance in advanced chip manufacturing gives the United States and other countries a reason to support Taiwan. But going forward, Hutcheson says the shield could turn into a target. If the United States and its allies are increasingly dependent on chips made only in Taiwan, then China can cause major damage to the U.S. economy by targeting Taiwan. Hutcheson says TSMC's geographical diversification will make its home country less of a target. The company has also opened a fab in Japan and is building one in Germany.

Original Submission

Arthur T Knackerbracket has processed the following story:

It seems that although the internet is increasingly drowning in fake images, we can at least take some stock in humanity’s ability to smell BS when it matters. A slew of recent research suggests that AI-generated misinformation did not have any material impact on this year’s elections around the globe because it is not very good yet.

[...] In the U.S., the News Literacy Project identified more than 1,000 examples of misinformation about the presidential election, but only 6% was made using AI. On X, mentions of “deepfake” or “AI-generated” in Community Notes were typically only mentioned with the release of new image generation models, not around the time of elections.

Interestingly, it seems that users on social media were more likely to misidentify real images as being AI-generated than the other way around, but in general, users exhibited a healthy dose of skepticism. And fake media can still be debunked through official communications channels, or through other means like Google reverse image-search.

It is hard to quantify with certainty how many people have been influenced by deepfakes, but the findings that they have been ineffective would make a lot of sense. AI imagery is all over the place these days, but images generated using artificial intelligence still have an off-putting quality to them, exhibiting tell-tale signs of being fake. An arm might unusually long, or a face does not reflect onto a mirrored surface properly; there are many small cues that will give away that an image is synthetic. Photoshop can be used to create much more convincing forgeries, but doing so requires skill.

Original Submission

hubie writes:

Malware botnets exploit outdated D-Link routers in recent attacks:

Two botnets tracked as 'Ficora' and 'Capsaicin' have recorded increased activity in targeting D-Link routers that have reached end of life or are running outdated firmware versions.

The list of targets includes popular D-Link devices used by individuals and organizations such as DIR-645, DIR-806, GO-RT-AC750, and DIR-845L.

For initial access, the two pieces of malware use known exploits for CVE-2015-2051, CVE-2019-10891, CVE-2022-37056, and CVE-2024-33112.

Once a device is compromised, attackers leverage weaknesses in in D-Link's management interface (HNAP) and execute malicious commands through a GetDeviceSettings action.

The botnets can steal data and execute shell scripts. Attackers appear to compromise the devices for distributed denial-of-service (DDoS) purposes.

Ficora has a widespread geographic distribution with some focus on Japan and the United States. Capsaicin appears to be targeting mostly devices in East Asian countries and increased its activity for just two days, starting on October 21.

[...] One way to prevent botnet malware infections on routers and IoT devices is to ensure that they're running the latest firmware version, which should addresses known vulnerabilities.

If the device has reached end-of-life and no longer receives security updates, it should be replaced with a new model.

A a general advice, you should replace default admin credentials with unique and strong passwords and disable remote access interfaces if not needed.

Previously: D-Link Won't Fix Critical Flaw Affecting 60,000 Older NAS Devices

Original Submission

James Earl Carter Jr. (October 1, 1924 – December 29, 2024) was an American politician who served as the 39th president of the United States from 1977 to 1981. A member of the Democratic Party, Carter previously served from 1963 to 1967 in the Georgia State Senate and from 1971 to 1975 as the 76th governor of Georgia. He was the longest-lived president in U.S. history and the first to live to 100 years.

Please see drussell's journal for discussion.

Arthur T Knackerbracket has processed the following story:

Motorola is one of the most important companies in history. They are responsible for the mobile phone, among many other inventions, but their success didn’t last forever. They took hit after hit. They were left with no profits, bled dry from the inside out by Google and left with none of their technology. Yet, this wasn’t the end. Motorola was down, but not out. Somehow after all of this, they did they impossible. Motorola not only survived but began to thrive. This is how.

At their core, Motorola was a wireless communications company, but that’s massively underselling their long, prestigious, history. In the second World War, they made a breakthrough invention. The first portable, two-way radio: The SCR-300, which soon became known as the walkie-talkie. It was a pivotal tool for the US Army Corps, and had a range between 10 and 20 miles. But that’s just the beginning. Motorola also created, “Unified S-Band” radio, a crucial piece of equipment for the Apollo 11 moon landing mission. The radio was responsible for one of the most important video transmissions in human history. Motorola’s has a long history beyond these two examples. It’s actually unbelievable how big an impact this one company had on humanity. 

Yet shockingly, these pivotal examples aren’t even what they’re known for today. In 1973 Motorola engineer Martin Cooper had just finished his latest prototype. It allowed for remote calling, anywhere. But it was fully handheld, and didn’t have cables, and didn’t have to be attached to a vehicle, and it wasn’t a radio…! Motorola had just invented the mobile phone, but this was just the prototype. It would take Motorola another full decade to bring it to the masses, and over $100 million.

[...] In the early 2000s, they had a fierce rivalry with Nokia, and soon unveiled the Razr 3. A sleek, ultra-thin flip phone, and one of the most iconic and successful phones in history. The Razr 3 sold over 130 million units. Their invention altered the world of technology. Almost everyone living today is directly impacted by this monumental invention. Which is what makes what happened next, so surprising.

[...] Motorola phones had no flash interface like Apple, and very little third party apps. Plus, just by looking at the Motorola Q, you could see why it didn’t catch on. Worst of all, it was running Windows as its operating system. The traditional “dumb phone” had been most of Motorola’s business. But these were falling out of favor. Time was running out. They needed to adapt, and fast. 

looorg writes:

Volkswagen accidentally leaked, or left open to the public via the cloud, location data and personal information of the owner for more then 800,000 electric cars.

After setting up the app, her car apparently began collecting and transmitting data to the manufacturer, including the exact GPS location of where she parked it every time she turned off the engine.

[...] A data volume of several terabytes on around 800,000 electric cars was largely unprotected and accessible for months in an Amazon cloud storage system .

[...] Much of the vehicle data could be linked to the names and contact details of the drivers, owners or fleet managers. Precise location data was available for 460,000 vehicles, allowing conclusions to be drawn about the lives of the people behind the wheel

[...] The cloud storage itself contained the data of the individual vehicles, immediately recognizable by the names for the battery charge level, the inspection status and the categories "engine on" and "engine off". The latter contained not only the time but also the longitude and latitude lines and thus the position of the car when the electric motor was switched off. In the case of VW models and Seats, this geodata was accurate to within ten centimeters,

[...] Cariad responded within a few hours and did not even try to downplay the matter. The team responsible for security incidents thanked them and asked for further details. The gap has now been closed.

The case goes far beyond Cariad and the VW Group. Many modern cars have a three-digit number of sensors and collect masses of data. Exactly which ones and to what extent are likely to be known only to the manufacturers.

[...] But who actually owns the car data? The manufacturers? The owners? An important question, especially as more and more players are expressing their desire for it. Car insurers, for example, want access so that they can offer rates that reward careful driving.

https://www-spiegel-de.translate.goog/netzwelt/web/volkswagen-konzern-datenleck-wir-wissen-wo-dein-auto-steht-a-e12d33d0-97bc-493c-96d1-aa5892861027?_x_tr_sl=auto&_x_tr_tl=en&_x_tr_hl=en&_x_tr_pto=wapp

Original Submission

upstart writes:

As machine learning and generative AI reshape the world, MIT's Schwarzman College of Computing is integrating these and other advanced computing technologies into classrooms and labs across campus:

Launched in 2019–'20, Schwarzman is MIT's only college, so called because it cuts across the Institute's five schools in a new effort to integrate advanced computing and artificial intelligence into all areas of study. "We want to do two things: ensure that MIT stays at the forefront of computer science, AI research, and education," Huttenlocher says, "and infuse the forefront of computing into disciplines across MIT." He adds that safety and ethical considerations are also critical.

To that end, the college now encompasses multiple existing labs and centers, including the Computer Science and Artificial Intelligence Laboratory (CSAIL), and multiple academic units, including the Department of Electrical Engineering and Computer Science. (EECS—which was reorganized into the overlapping subunits of electrical engineering, computer science, and artificial intelligence and decision-making—is now part of both the college and the School of Engineering.) At the same time, the college has embarked on a plan to hire 50 new faculty members, half of whom will have shared appointments in other departments across all five schools to create a true Institute-wide entity. Those faculty members—two-thirds of whom have already been hired—will conduct research at the boundaries of advanced computing and AI.

The new faculty members have already begun helping the college respond to an undeniable reality facing many students: They've been overwhelmingly drawn to advanced computing tools, yet computer science classes are often too technical for nonmajors who want to apply those tools in other disciplines. And for students in other majors, it can be tricky to fit computer science classes into their schedules.

Meanwhile, the appetite for computer science education is so great that nearly half of MIT's undergraduates major in EECS, voting with their feet about the importance of computing. Graduate-level classes on deep learning and machine vision are among the largest on campus, with over 500 students each. And a blended major in cognition and computing has almost four times as many enrollees as brain and cognitive sciences.

"We've been calling these students 'computing bilinguals,'" Huttenlocher says, and the college aims to make sure that MIT students, whatever their field, are fluent in the language of computing. "As we change the landscape," he says, "it's not about seeing computing as a tool in service of a particular discipline, or a discipline in the service of computing, but asking: How can we bring these things together to forge something new?"

Original Submission

owl writes:

https://www.righto.com/2024/12/this-die-photo-of-pentium-shows.html

In 1993, Intel released the high-performance Pentium processor, the start of the long-running Pentium line. The Pentium had many improvements over the previous processor, the Intel 486, including a faster floating-point division algorithm. A year later, Professor Nicely, a number theory professor, was researching reciprocals of twin prime numbers when he noticed a problem: his Pentium sometimes generated the wrong result when performing floating-point division. Intel considered this "an extremely minor technical problem", but much to Intel's surprise, the bug became a large media story. After weeks of criticism, mockery, and bad publicity, Intel agreed to replace everyone's faulty Pentium chips, costing the company $475 million.

In this article, I discuss the Pentium's division algorithm, show exactly where the bug is on the Pentium chip, take a close look at the circuitry, and explain what went wrong. In brief, the division algorithm uses a lookup table. In 1994, Intel stated that the cause of the bug was that five entries were omitted from the table due to an error in a script. However, my analysis shows that 16 entries were omitted due to a mathematical mistake in the definition of the lookup table. Five of the missing entries trigger the bug— also called the FDIV bug after the floating-point division instruction "FDIV"—while 11 of the missing entries have no effect.

Original Submission

Arthur T Knackerbracket has processed the following story:

Two 48-core CPUs are used.

Graviton, a Russia-based server supplier, has announced its first AI and HPC server powered by Russia's own homegrown processors. This device can support up to eight compute GPUs to process artificial intelligence and supercomputer workloads. The vendor doesn't recommend any particular GPUs (though they can be easily guessed), probably because getting them amid sanctions is illegal. Furthermore, whether or not the machine can achieve competitive performance numbers is unclear.

The Graviton S2124B server is based on two undisclosed 48-core CPUs running at 2 GHz and featuring DDR4-3200 memory, according to ServerNews. The basic specification of the processor likely suggests the Baikal Electronics BE-S1000 server-grade chip that packs 48 Arm Cortex-A75 cores and supports 2-way and 4-way symmetric multiprocessor (SMP) configurations. 

This particular version of the BE-S1000 seems to clock the CPU at 500 MHz below its original frequency, which is likely a result of porting its design from TSMC's 16FFC to a different production node at a different foundry. It is also possible that Baikal reduced the operating clocks to increase yields or reduce power consumption.

Dividing the claimed performance numbers by eight, we can determine the GPUs Graviton intends to install in its S2124B machine (in all cases, we'll refer to tensor core performance). Per-accelerator performance numbers — 60 FP64 TFLOPS of compute power for supercomputing and 3340 FP8/INT8 TFLOPS/TOPS performance for AI — point to Nvidia's H100 PCIe GPU. Those who use the S2124B will have to rely on Nvidia's CUDA ecosystem. However, without support from Nvidia, it is unlikely that peak performance will be achieved. Furthermore, considering the CPU is Arm-based and relatively unknown, it remains to be seen how much performance can actually be extracted from these Hopper accelerators.

In addition to two CPUs and eight GPU accelerators, Graviton's S2124B can integrate 12 SATA drives or 12 NVMe U.3 SSDs.

The Graviton S2124B server is currently available for order, with customers also invited to apply for testing opportunities, but its pricing is unknown. Additionally, it is unclear whether Graviton can supply Nvidia H100 GPUs.

"We take pride in consistently offering IT solutions that meet market demands in a timely manner," said Alexander Filchenkov, head of server and network systems at Graviton. "This time, we successfully developed and manufactured servers critical for complex computations using domestic processors. This product represents a significant step in advancing domestic computing technologies and will enable our clients to efficiently address data processing challenges."

Original Submission

Arthur T Knackerbracket has processed the following story:

Residents of five states will be ringing in the new year with the best gift of all: new privacy rights.

This upcoming January will see consumer data privacy laws that were enacted by state lawmakers in 2023 and 2024 go into effect in Delaware, Iowa, Nebraska, New Hampshire, and New Jersey. It will bring the number of states with active privacy laws up to 13.

The new laws govern how businesses of certain sizes—the size varies by state—handle sensitive consumer information and grant residents of those states various rights to know, correct, and delete the data that businesses hold about them. Here are some of the key provisions in the new suite of laws:

Delaware: Originally passed in 2023, the law applies to people and organizations who, during the preceding calendar year, processed the personal information of 35,000 Delaware residents or processed the personal information of 10,000 Delaware residents and made more than 20 percent of their gross revenue from the sale of personal information.

Unlike many other state privacy laws, it applies to nonprofits and for-profit businesses.

It grants residents the right to know what personal information an organization holds about them, obtain a copy of that information, correct it, and opt out of having that information used for targeted advertising, sold to a third party, or used to make automated decisions with significant legal ramifications.

The law goes into effect January 1.

Iowa: Also passed in 2023, Iowa's law applies to businesses that processed personal information for at least 100,000 residents or that processed information for 25,000 residents and made more than half of their gross revenue from the sale of such data.

It is a narrower, more business-friendly law than many of the other state laws that have taken effect.

While consumers are granted the right to access and delete information a business holds about them and opt out of it being sold to a third party, they are not allowed to correct that information, opt out of its use for targeted advertising, or opt out of it being used to make automated decisions about them.

The law goes into effect January 1.

upstart writes:

US obesity rates drop for 1st time in a decade, with possible help from weight loss medications:

For the first time in over a decade, obesity rates in the United States may finally be heading in the right direction and new weight loss drugs like semaglutide could be part of the reason why.

A new study published Friday in the journal, JAMA Health Forum, found that obesity numbers ticked down slightly from 46% in 2022 to 45.6% in 2023. While only a slight decline, this is the first drop recorded in at least a decade.

"What we're seeing for the first time is that curve is bending and shows a sign of hope for something that was really a threat to American public health for so many years," said study co-author and ABC News contributor John Brownstein, who is also the chief innovation officer at Boston Children's Hospital and a professor at Harvard Medical School.

The study reviewed the body mass index (BMIs), a measure of obesity, of 16.7 million U.S. adults over a 10-year period. The average BMI rose annually to 30.24 -- which is considered obese -- until it plateaued in 2022, then dropped marginally to 30.21 in 2023.

Brownstein and his team noted that women and adults aged 66 to 75 saw the largest decreases in obesity. People living in the South, where they had the highest dispensing rate of weight loss drugs known as GLP-1 receptor agonists, also saw a meaningful decline in obesity.

Semaglutide, which belong to a class of drugs known as GLP-1 agonists, is the active ingredient found in the popular weight loss drugs Ozempic and Wegovy. The U.S. saw a 700% increase in the use of these drugs specifically for weight loss, from 2019 to 2023, according to one Annals of Internal Medicine study. They are also used to treat type 2 diabetes or a combination of diabetes and obesity.

[...] Brownstein said that the growing use of these medications may be helping to reverse obesity trends.