SoylentNews

upstart writes:

Leaked emails show Amazon will only hire students and recent graduates:

Amazon is only hiring current students or newly graduated people for its entry level software developer positions.

According to an internal memo obtained by Insider, starting on January 25, 2023, Amazon limited new job openings for SDE-1s — the lowest software development engineering position — to what it calls "campus" hires, or students in Bachelor's, Master's, or PhD program alongside recent graduates. The memo said those in part-time or executive programs with years of work experience can apply too.

The change will mean that those that have been out of school for more than 12 months, or candidates for more senior SDE-2 positions who might be a better fit for an SDE-1 position would not be considered for the latter.

The internal note said Amazon is making the change because of the "pipeline" of candidates available through student programs, but the memo nor Amazon's spokesperson clarified why the company believes campus hires are better than experienced industry candidates for entry-level positions.

The change is "global and Amazon-wide," the note said, indicating it's applied across the company. Amazon's S-team, a group of over two dozen most senior executives, and top HR leaders made the decision, and exceptions will be made only with a VP or higher approval, it added.

[...] Overhauling Amazon's engineering culture has been a priority for Amazon's CEO Andy Jassy. At an internal staff meeting in 2021, Jassy told employees that he was aware of developer complaints at the company and that the engineering culture needed to be "meaningfully better than what it is today," as Insider previously reported. It also created a new team called "Amazon Software Builder Experience" to address those concerns.

Original Submission

hubie writes:

Interesting study to think about before the big game this weekend:

Certain age-related diseases may arise earlier in professional football players, new study finds:

Former professional football players — particularly linemen — are more likely than nonplayers with similar demographic characteristics to develop diseases typically associated with advanced age when significantly younger, according to new research published Dec. 8 in the British Journal of Sports Medicine.

These former elite athletes also tend to experience age-related conditions — hypertension and diabetes, among others — earlier, compared with the general population. Looking across all conditions, these athletes' health spans were reduced by nearly a decade.

Notably, the effects persisted even after the researchers accounted for body mass index and race, two powerful risk factors for the diseases studied.

[...] Importantly, the health span for each former NFL player age group most closely resembled American men a decade older. For example, 66 percent of the former players in the 30 to 39 age group reported an intact health span, compared with 62 percent of men in the general population ages 40 to 49.

Searching for game-related aspects that might be important for this premature emergence of aging diseases, the researchers separated data from the former football players group into linemen and non-linemen.

This analysis showed that linemen, who experience more contact during games than non-linemen, had notably shorter health spans across all decades of life. This subgroup tended to develop age-related diseases sooner than their non-linemen peers.

Freeman writes:

Intense demand for AI chatbot breaks records and inspires new $20/mo subscription plan:

On Wednesday, Reuters reported that AI bot ChatGPT reached an estimated 100 million active monthly users last month, a mere two months from launch, making it the "fastest-growing consumer application in history," according to a UBS investment bank research note. By comparison, TikTok took nine months to reach 100 million monthly users, and Instagram about 2.5 years, according to UBS researcher Lloyd Walmsley.

"In 20 years following the Internet space, we cannot recall a faster ramp in a consumer internet app," Reuters quotes Walmsley as writing in the UBS note.

[...] Over the past few decades, researchers have noticed that technology adoption rates are quickening, with inventions such as the telephone, television, and the Internet taking shorter periods of time to reach massive numbers of users. Will generative AI tools be next on that list? With the kind of trajectory shown by ChatGPT, it's entirely possible.

Original Submission

upstart writes:

Finland's Most-Wanted Hacker Nabbed in France:

Julius "Zeekill" Kivimäki, a 25-year-old Finnish man charged with extorting a local online psychotherapy practice and leaking therapy notes for more than 22,000 patients online, was arrested this week in France. A notorious hacker convicted of perpetrating tens of thousands of cybercrimes, Kivimäki had been in hiding since October 2022, when he failed to show up in court and Finland issued an international warrant for his arrest.

In late October 2022, Kivimäki was charged (and "arrested in absentia," according to the Finns) with attempting to extort money from the Vastaamo Psychotherapy Center. In that breach, which occurred in October 2020, a hacker using the handle "Ransom Man" threatened to publish patient psychotherapy notes if Vastaamo did not pay a six-figure ransom demand.

Vastaamo refused, so Ransom Man shifted to extorting individual patients — sending them targeted emails threatening to publish their therapy notes unless paid a 500-euro ransom.

When Ransom Man found little success extorting patients directly, they uploaded to the dark web a large compressed file containing all of the stolen Vastaamo patient records.

upstart writes:

Trust, not tech, is holding back a safer internet:

Opinion The tech sector is failing at cybersecurity. Global spending on the stuff is at $190 billion a year, a quarter of the US defense budget. That hasn't stemmed an estimated $7 trillion in annual cybercriminal damages. People are fond of saying that the Wild West days of the internet are over, but on those numbers an 1875 Dodge City bank vault looks like Fort Knox.

So where's the sheriff? There are plenty of posses; no end of companies both small and large selling security by the bushel. Firewalls, scanners, heuristic, intrinsic, behavioral, managed, managerial, in-cloud, on-prem, you can mix and match the buzzwords and buy into every new idea. What you can't do is make your systems safe.

If you do want a safe bet in cybersecurity, it's that things aren't going to change any time soon without some fundamental shift in how the market works – if 40 years of constant failure can be called working.

We have so little reason to trust what's on offer or those offering it. Several stories last week show this: Apple, which makes a big play of intrinsic platform security, is heading to court for ignoring user consent and silently gathering app data anyway. Microsoft, even as it announces the extension of its security platform into Linux, reveals it fumbled its switches on its service infrastructure and took business-critical access away from its customers. These are the big shots in town, but they can't shoot straight.

It's almost as if we can't rely on the private sector to protect us against crime. Guess what: we never could and we never will. The state has to take on that role – usually late, usually badly, and usually against the wishes of those who like their crimes kept in the private sector, but usually to better effect than the alternatives.

upstart writes:

"When we pass from this world, you will be the reason we are remembered":

The month before Dwarf Fortress was released on Steam (and Itch.io), the brothers Zach and Tarn Adams made $15,635 in revenue, mostly from donations for their 16-year freeware project. The month after the game's commercial debut, they made $7,230,123, or 462 times that amount.

[...] Tarn Adams noted that "a little less than half will go to taxes," and that other people and expenses must be paid. But enough of it will reach the brothers themselves that "we've solved the main issues of health/retirement that are troubling for independent people." It also means that Putnam, a longtime modder and scripter and community member, can continue their work on the Dwarf Fortress code base, having been hired in December.

[...] While the commercial release of Dwarf Fortress has earned the brothers some breathing room and introduced new players with some quality-of-life offerings, the "classic" version—the one Ars editor Casey Johnston detailed over her 10-hour ordeal—is still free to download.

If you haven't tried this game yet, it's interesting.

Original Submission #1  Original Submission #2 

Freeman writes:

But out of 300,000 high-probability images tested, researchers found a 0.03% memorization rate:

On Monday, a group of AI researchers from Google, DeepMind, UC Berkeley, Princeton, and ETH Zurich released a paper outlining an adversarial attack that can extract a small percentage of training images from latent diffusion AI image synthesis models like Stable Diffusion. It challenges views that image synthesis models do not memorize their training data and that training data might remain private if not disclosed.

Recently, AI image synthesis models have been the subject of intense ethical debate and even legal action. Proponents and opponents of generative AI tools regularly argue over the privacy and copyright implications of these new technologies. Adding fuel to either side of the argument could dramatically affect potential legal regulation of the technology, and as a result, this latest paper, authored by Nicholas Carlini et al., has perked up ears in AI circles.

Related:
Getty Images Targets AI Firm For 'Copying' Photos

Original Submission

upstart writes:

BMW Recalls Select EVs For Not Making Enough Noise:

Electric vehicles are much quieter than your typical gas-powered car, but they're actually required by law to make noise to alert pedestrians. As a result, you'll hear a subtle "hum" or electric sound as one backs up or drives by. However, BWM just recalled select models for being too quiet.

BMW started 2023 by recalling over 16,000 EVs over a potential power loss issue, and now we're hearing the company is dealing with another problem. While it sounds odd to say, BMW recalled over 3,500 BMW i4 and iX models (2022 and 2023) for not being loud enough.

According to the NHTSA website, BMW is recalling select models of those vehicles due to faulty external artificial sound generators. Under normal conditions, the sound generator should emit a beep or humming noise when a car is in reverse or driving at low speeds. The system doesn't always work or make noise for whatever reason, and that's a safety concern.

[...] If you've ever had an EV sneak up on you while walking down the street, you know why this law exists. Some of them are so quiet it can be dangerous as we're used to hearing loud car engines. Luckily for BMW, this is an easy fix with a software update.

Original Submission

upstart writes:

New proposals under consultation seek to put the UK at the 'forefront' of crypto technology and innovation:

The UK has set out a detailed plan to regulate the crypto industry in the wake of the FTX collapse late last year.

Published today (1 February), the proposals are set to give consumers in the crypto industry confidence in their investments while also allowing the sector to blossom.

Rishi Sunak, the UK prime minister, is largely seen as crypto-friendly by industry experts. The latest move will aim to put the UK's financial sector "at the forefront of cryptoasset technology and innovation" while managing "potential consumer and stability risks".

[...] A consultation has now been launched on the proposals that will last until 30 April 2023. The full set of proposals currently under consultation can be accessed on the UK government website.

Some of the proposals include ensuring customer assets are returned to them in case a crypto business goes bust, laying down fair and just rules on promotion of crypto assets and enhancing data-reporting requirements – including with regulators.

The proposals also seek to implement new regulations that will prevent so-called 'pump and dump' – a practice whereby an individual artificially inflates the value of a crypto asset before selling it.

Original Submission

upstart writes:

Our extinct sister species was hunting and butchering big game, according to new research:

Historical finds of elephant remains alongside stone tools have long prompted speculation among researchers that early humans or other hominin species may have relied on the massive mammals for food.

Now, a team of researchers has determined that Neanderthals in Europe were taking down elephants and methodically butchering them, yielding food stores that would have lasted Neanderthal groups months. Their research is published today in Science Advances.

The bones belonged to straight-tusked elephants, (Palaeoloxodon antiquus), an extinct species about twice the size of African elephants, the largest living land mammals on Earth. Evidence that Neanderthals were hunting the animals in pit traps was discovered in the early 1920s, and in 1948, a specimen was found near 25 flint artifacts and a wooden lance.

[...] "Neanderthals knew what they were doing," wrote Britt Starkovich, an archaeologist at the Senckenberg Centre for Human Evolution and Paleoenvironment and the University of Tübingen, in an associated Focus article. "They knew which kinds of individuals to hunt, where to find them, and how to execute the attack. Critically, they knew what to expect with a massive butchery effort and an even larger meat return."

Journal Reference:
Sabine Gaudzinski-Windheuser, Lutz Kindler, Katharine MacDonald, and Wil Roebroeks, Hunting and processing of straight-tusked elephants 125.000 years ago: Implications for Neanderthal behavior [open], Sci Adv, 9, 2023. (DOI: https://doi.org/10.1126/sciadv.add8186)

Original Submission

fliptop writes:

The National Security Agency is doggedly courting laid-off Big Tech workers as the spy agency undertakes one of its largest hiring surges in the last 30 years:

The NSA began privately reaching out to Big Tech employees over LinkedIn last fall, as word spread that major American companies such as Meta and Amazon were bleeding tens of thousands of skilled workers.

NSA talent management senior strategist Christine Parker said the spy agency also saw predictions of more job cuts, and sprung into action.

"NSA started reaching out through LinkedIn, through some of our career boards, specifically sending messages to people that we thought might be linked to some companies that either were in the news saying they are going to lay-off or were predicted to be laid off," Ms. Parker said in an interview. "Just kind of let them know that we're here and that we have this robust, ongoing hiring program."

[...] The NSA is currently hiring 3,000 new employees to work across the country, from the D.C.-area to Alaska, Colorado, Georgia, Hawaii, Texas and Utah, according to Molly Moore, NSA deputy director of workforce support activities.

[...] More than half of the NSA's 3,000 open positions are for science, technology, engineering, and mathematics work. The spy agency, which is focused on signals intelligence collection, is in the market for data and computer scientists, software engineers, cybersecurity experts, human-machine teaming experts, and mathematicians, according to NSA director of operations Natalie Lang.

Related: GM Hiring Tech Talent Laid Off By Silicon Valley Companies

Original Submission

hendrikboom writes:

They observed the brains of computer programmers under fMRI while they were reading code:

Functional magnetic resonance imaging (fMRI), which measures changes in blood flow throughout the brain, has been used over the past couple of decades for a variety of applications, including "functional anatomy" — a way of determining which brain areas are switched on when a person carries out a particular task. fMRI has been used to look at people's brains while they're doing all sorts of things — working out math problems, learning foreign languages, playing chess, improvising on the piano, doing crossword puzzles, and even watching TV shows like "Curb Your Enthusiasm."

The new paper built on a 2020 study, written by many of the same authors, which used fMRI to monitor the brains of programmers as they "comprehended" small pieces, or snippets, of code. (Comprehension, in this case, means looking at a snippet and correctly determining the result of the computation performed by the snippet.) The 2020 work showed that code comprehension did not consistently activate the language system, brain regions that handle language processing, explains Fedorenko, a brain and cognitive sciences (BCS) professor and a coauthor of the earlier study. "Instead, the multiple demand network — a brain system that is linked to general reasoning and supports domains like mathematical and logical thinking — was strongly active." The current work, which also utilizes MRI scans of programmers, takes "a deeper dive," she says, seeking to obtain more fine-grained information.

[...] The team carried out a second set of experiments, which incorporated machine learning models called neural networks that were specifically trained on computer programs. These models have been successful, in recent years, in helping programmers complete pieces of code. What the group wanted to find out was whether the brain signals seen in their study when participants were examining pieces of code resembled the patterns of activation observed when neural networks analyzed the same piece of code. And the answer they arrived at was a qualified yes.

How close is this to mind-reading?

Original Submission

fliptop writes:

AI is now used in virtually all areas of science to help researchers with routine classification tasks. It's also helping our team of radio astronomers broaden the search for extraterrestrial life, and results so far have been promising:

As scientists searching for evidence of intelligent life beyond Earth, we have built an AI system that beats classical algorithms in signal detection tasks. Our AI was trained to search through data from radio telescopes for signals that couldn't be generated by natural astrophysical processes.

When we fed our AI a previously studied dataset, it discovered eight signals of interest the classic algorithm missed. To be clear, these signals are probably not from extraterrestrial intelligence, and are more likely rare cases of radio interference.

Nonetheless, our findings – published in Nature Astronomy – highlight how AI techniques are sure to play a continued role in the search for extraterrestrial intelligence.

[...] If astronomers do manage to detect a technosignature that can't be explained away as interference, it would strongly suggest humans aren't the sole creators of technology within the Galaxy. This would be one of the most profound discoveries imaginable.

Journal Reference: Ma, P.X., Ng, C., Rizk, L. et al. A deep-learning search for technosignatures from 820 nearby stars. Nat Astron (2023). https://doi.org/10.1038/s41550-022-01872-z

Original Submission

Freeman writes:

FDA posted a recall of EzriCare-branded artificial tears and criticized its maker:

An extensively drug-resistant bacterial strain is spreading in the US for the first time and causing an alarming outbreak linked to artificial tears eye drops, according to an alert released Wednesday evening from the Centers for Disease Control and Prevention. So far, the germ has caused various infections in 55 people in 12 states, killing one and leaving others hospitalized and with permanent vision loss.

Infected patients reported using more than 10 brands of artificial tears collectively, with some patients using multiple brands. But the most common brand used among the patients was EzriCare Artificial Tears, a preservative-free product sold by Walmart, Amazon, and other retailers.

On Thursday, after this story originally published, the Food and Drug Administration posted notice of a recall of EzriCare Artificial Tears and Delsam Pharma's Artificial Tears. The FDA the CDC recommends clinicians and patients stop buying and using the two products. In a separate notice, the FDA further added that the products' manufacturer, Global Pharma Healthcare Private Limited, was in violation of good manufacturing practices, including lack of appropriate microbial testing, formulating its product without an adequate preservative, and lack of proper controls concerning tamper-evident packaging.

[...] In the current outbreak, which began in May 2022, investigators have isolated the outbreak strain from 13 sputum or bronchial washes, 11 cornea swabs, seven urine samples, two blood samples, 25 rectal swabs, and four other nonsterile sources. The patients presented in inpatient and outpatient settings with a range of infections. Those include eye infections—infection of the cornea (keratitis) and infection of tissue or fluids inside the eyeball (endophthalmitis)—to respiratory infections, urinary tract infections, and sepsis. The patient who died had a systemic infection.

Original Submission

c0lo writes:

https://www.scienceofintelligence.de/press-release-sheep-are-more-democratic-than-you-think/

Collective motion brings to mind fascinating images, such as the flocks of birds over a corn field, or schools of barracudas as they move in circles in the water. These motions are also particularly appealing to physicists, as the patterns that emerge lend themselves to mathematical and statistical modeling that can help them better understand this phenomenon. When it comes to sheep, many studies describe the collective behavior in sheep flocks as a self-organized process where individuals continuously adapt their direction and speed to follow the motion and collective decisions of the group – as if the only leading force were the "collective brain" itself. This view, however, does not take into account that animals do not move continuously, or the possible hierarchies existing in many animal groups and the potential benefits of having a single individual lead the way.

[...] According to Gómez-Nava and collaborators, when the sheep stop to feed or rest between one collective motion phase and the next, they randomly pick a new group leader for the next round of flocking, thus transferring control to a new individual each time. In this way, individuals take turns being leaders and the flock's collective intelligence is achieved democratically, over many collective motion phases.