Please create an account to participate in the Slashdot moderation system

 



Forgot your password?
typodupeerror

Submission + - So... New New Coke? Now with sugar! (npr.org) 1

fahrbot-bot writes: NPR, and others, are reporting that Coca-Cola says it will use U.S. cane sugar in a new Coke, a plan pushed by Trump.

"We're going to be bringing a Coke sweetened with U.S. cane sugar into the market this fall," Coca-Cola Chairman and CEO James Quincey said on a conference call with analysts Tuesday.

Quincey said the new offering would "complement" Coca-Cola's core portfolio of drinks, suggesting it could arrive as an alternative, rather than a replacement, for its flagship Coke product.

CNN notes that, "sugar is more expensive in the US than in many parts of the world..." — there are also quotas and tariffs on cane sugar imported into the U.S.

The Sweeteners Users Association notes that cane sugar in the U.S. is grown mainly in Florida, Louisiana and Texas — so this will be a boon to some. Sugar beets are more widely grown in California, Colorado, Idaho, Michigan, Minnesota, Montana, Nebraska, North Dakota, Oregon, Washington and Wyoming.

Submission + - Managers say supervising Gen Z feels like babysitting according to new survey (nerds.xyz)

BrianFagioli writes: This might not come as surprise if you are currently in the workforce, but supervising Gen Z workers often feels more like babysitting than managing. Thatâ(TM)s according to a new report from ResumeTemplates.com. Based on survey responses from 1,000 U.S. managers, it paints a scary picture of frustration, hand-holding, and mismatched expectations in todayâ(TM)s multigenerational workplace.

The numbers are hard to ignore. Sixty-eight percent of surveyed managers said overseeing Gen Z employees feels more like parenting than actual leadership. More than half, 54 percent, went further and likened it to babysitting. Many managers say they spend their time walking younger workers through basic tasks, checking in multiple times a day, and explaining what older employees would already know how to do.

A full 61 percent of managers reported that their Gen Z staff require frequent guidance to complete their work. Over half say these workers often struggle with following simple instructions. One out of every three managers said they find themselves checking in four or more times a day with their youngest employees.

Comment Re:It just shows (Score 1) 63

I chose RC boat because some people are using this competing at "stupid human trick" as an example of intrinsic proof of LLMs being able to supersede humans.

In the scenario of an olympic swimming competition, an autonomous boat vs a manned boat would show no difference to each other, both would compete the task much better than a human. It's a useless test to measure general utility. Just like a person swimming a 1500 meter distance is not really a useful indicator on its own of how useful they are. These Math Olympiads are similar in that they are not particularly indicative of people being useful. The same ways that we would stress a human in impressive ways does not mean that a computer coming at it from a different approach should be considered to have broadly superseded the humans.

Yes, a real boat is valuable and LLM can be valuable when utilized correctly, but in the face of exaggerated hype some pessimistic reality check is called for to balance expectations.

Submission + - Engineers transform dental floss into needle-free vaccine (science.org)

sciencehabit writes: Flossing may be good for more than getting your dentist off your back—one day, it may also protect you from the flu. In an unorthodox approach to needle-free vaccines, researchers have developed a special kind of floss that can deliver proteins and inactive viruses to mice’s gumlines and trigger immune responses that protect against infectious disease, they report today in Nature Bioengineering.

For many years, scientists have tried to develop alternatives to delivering vaccines via syringes by turning to the moist areas in your mouth and nose where most viruses enter. But it’s tough to develop an effective vaccine that can be administered through those entry points because they have naturally tough defenses against foreign molecules.

To test this idea, researchers at Texas Tech University had to do something no scientist had done before: Try to floss a mouse. It was a “quite difficult” two-person job: One scientist gently pulled the mouse’s jaw down with the metal ring from a keychain while the other administered the floss.

During a test run, the team found that when researchers coated floss with a fluorescently labeled protein, 75% of the protein was successfully delivered to the mouse’s gums. And even 2 months after flossing, the mice had elevated levels of antibodies in their lungs, noses, feces, and spleens, suggesting a robust immune response to the protein.

Next, the engineers added an inactive flu virus—a common vaccine component—to the floss, which in theory could teach the mouse’s body to build up immunity to the flu. Over a 28-day period, the researchers flossed 50 mice with the coated floss every 2 weeks. Then, 4 weeks after the final dose, they infected those mice with the real flu virus. All the mice that were flossed three times survived, whereas all the unvaccinated mice died.

The flossed mice also had a more systemic immune response: Not only were flu antibodies present in their feces and saliva, but the mice had more T cells—the directors of the body’s immune response—in their lungs and spleens, as well as larger lymph nodes. What’s more, the team found flu antibodies in the mice’s bone marrow, signaling that their immune systems were “fully engaged” by the inactive flu virus. Overall, the immune response to the floss resembled the response to vaccines that are sprayed into the nose, such as FluMist.

To gauge whether the method could work in humans, the researchers asked 27 healthy volunteers to floss with dental picks coated with colored food dye. On average, roughly 60% of the dye was delivered to the participants’ gums. They then surveyed the participants on what they thought of the approach. Most said they were open to trying a floss-based vaccine and would prefer it to a shot.

Submission + - Ring's flying 'spy drone' that monitors your home in the AIR 'coming soon' (the-sun.com)

schwit1 writes: Ring’s flying ‘spy drone’ that monitors your home in the AIR ‘coming soon’ – it stalks burglars & even recharges itself

The drone flies automatically around the home, with some obstacle avoidance technology so that it doesn't crash into ceiling light or precious vase

It can also only fly one floor at a time, and has a limited flight time of just five minutes on a single charge

The idea is to scrap multiple plug in cameras around the home and have just one patrolling drone available

Comment LLMs can't think and they don't need to (Score 2) 95

LLMs have a great deal of utility and extend the reach of computing to a fair amount of scope that was formerly out of reach to computing, but they don't "think" and the branding of the "reasoning" models is marketing, not substantive.

The best evidence is reviewing so-called "reasoning chains" and how the mistakes behave.

Mistakes are certainly plausible in "true thinking", but the way they interact with the rest of the "chain" is frequently telling. It flubs a "step" in the reasoning and if it were actual reasoning, that should propagate to the rest of the chain. However when a mistake is made in the chain, it's often isolated and the "next step" is written as if the previous step said a correct thing, without ever needing to "correct" itself or otherwise recognize the error. What has been found is that if you have it generate more content and dispose of designated "intermediate" content you have a better result, and the intermediate throwaway content certainly looks like what a thought process may look like, but ultimately it's just more prose and mistakes in the content continue to have an interesting behavior of isolation rather than contaminating the rest of an otherwise ok result.

Comment Re:No"AI" cannot think (Score 1) 95

When the model is used for inference, yes. But I assume he was speaking to the awkwardness of training. Take a machine vision that has never been trained on dogs and cats, feed it a dozen labeled images of cats and dogs to retrain it to add dog/cat recognition. Then try to do inference on that model and it will be utterly useless still for dog/cat recognition. Take a model trained on normal images. Then have it try recognition on a fisheye lense. It will fail because it has no idea. You might hope to retrain it to recognize the fisheye distortion generically, but generally that won't work and you have to retrain it to catch the fisheye variant of everything you want. After retraining, if you just slap the fisheye variant into a normal picture, the model is likely to be oblivious to the anomaly.

Don't know if this is an argument about 'thinking', but it is certainly a difference in 'learning', that AI models need to consume way way more than a human to start to give useful results. A model training to operate a car is still oddly dodgy even after consuming more car operating hours than a human will consume in a lifetime. In areas where we have enough training data to feed, this is fine, but it's certainly different than a human learning, which can do something better to extend less training to better results.

Comment Re:It just shows (Score 1) 63

I'm not in denial, the LLMs and other forms of AI have utility, but expectations have to be mitigated.

Was in a discussion with a software executive a couple weeks back who said he fully anticipates he can lay off every one of his software developers and testers in the next year and only have to retain the 'important' people: the executives and sales people.

People see articles like this show how LLMs enable computing to reach another tier of 'stupid human tricks', which is certainly novel, but people overextend exactly what this means. We put too much focus on whether the LLMs can navigate tests explicitly designed to be graded and passable for humans, with effort and imagine that means they are able to chart the less well trodden set of problems that have an unknown or maybe no solution at all.

There's a lot of stuff LLMs can do that computing wasn't able to do before, and that does potentially speak to a great deal of the things humans do today, but we have to temper the wild expectations people are getting.

Submission + - Apple launches retail expansion in Saudi Arabia, but at what cost? (nerds.xyz)

BrianFagioli writes: Apple has officially brought its retail operation to the Kingdom of Saudi Arabia, launching the Apple Store online and the Apple Store app with full Arabic language support. This marks the companyâ(TM)s first direct retail presence in the country, and it comes with the usual Apple polish: personalized shopping, AppleCare+, Arabic engraving, and promises of flagship physical stores starting in 2026.

But while Apple is highlighting customization and convenience, critics are pointing to something else. The optics of cozying up to a regime accused of human rights abuses, digital repression, and lingering questions about its connection to the 9/11 terror attacks.

Submission + - Google launches OSS Rebuild to catch open source malware and verify sketchy pack (nerds.xyz)

BrianFagioli writes: Open source software runs pretty much everything these days, but with that popularity comes risk. Hackers have figured out they can slip malicious code into popular packages and infect thousands of systems in one shot. Now Google is stepping in with a new initiative called OSS Rebuild, and it could be a game changer for developers and security teams alike.

Matthew Suozzo from Googleâ(TM)s Open Source Security Team introduced OSS Rebuild as a way to âoestrengthen trust in open source package ecosystems by reproducing upstream artifacts.â That means rebuilding packages like those on PyPI, npm, and Crates.io, and making sure they match the original source code. If they donâ(TM)t, it could be a sign that somethingâ(TM)s wrong.

âoeOSS Rebuild gives security teams powerful data to avoid compromise without burden on upstream maintainers,â Suozzo said. Thatâ(TM)s a key point. The tool works behind the scenes, without putting extra work on the people who publish and maintain open source packages. It automatically generates build definitions, creates provenance metadata following the SLSA Build Level 3 standard, and flags anything suspicious.

This kind of system might have caught incidents like the backdoor in xz-utils or the shady code injection in solana/webjs. OSS Rebuild looks for several types of supply chain threats, including stealthy backdoors, build environment compromises, and packages that include code not found in the public source repository. It even monitors for suspicious behavior during the build process, something thatâ(TM)s almost impossible to catch with manual review.

Comment Re:It just shows (Score 1) 63

The point is we have a myriad of "tests that are hard for humans, but don't necessarily translate to anything vaguely useful". In academics, a lot of tests are only demanding of reasoning ability because the human has limited memory. Computers short on actual "reasoning" largely make up for it by having just mind boggling amounts of something more akin to recall than reasoning (it's something a bit weirder, but as far as analogies go, recall is closer).

It's kind of like bragging that your RC boat could get a gold medal in the Olympic 1500 meter freestyle. It didn't complete the challenge in the same way, and that boat would be unable to, for example, save someone that is about to drown, because the boat can just go places, it can't do what that human swimmer could do. That person swimming 1500m itself is a useless feat of interest, and not really directly useful in and of itself.

Submission + - MenuetOS video call demo (youtube.com)

An anonymous reader writes: MenuetOS is an operating system in development for PC, written completely in 64bit assembly language. Features include pre-emptive and real-time multitasking with multiprocessor support and Graphical User Interface. Menuet64 (2005) is released under License and Menuet32 (2000) under GPL. Menuet supports assembly programming for much faster, smaller and less resource hungry applications.

Submission + - Radioactive Waste Exposed Children in Missouri to Cancer Risks, Study Finds (sciencealert.com)

alternative_right writes: As part of the top-secret scheme known as the Manhattan Project, radioactive waste from uranium refinement in St Louis was stored in drums or even left out in the open in a rural area north of the city, close to a tributary called Coldwater Creek.

[L]iving within a kilometer (nearly two thirds of a mile) of Coldwater Creek was linked to a 44 percent increase in the risk of developing cancer. Considering anyone who had already died of cancer couldn't be included in the study, the association might be even stronger.

Slashdot Top Deals

The only possible interpretation of any research whatever in the `social sciences' is: some do, some don't. -- Ernest Rutherford

Working...