Become a fan of Slashdot on Facebook

 


Forgot your password?
Close
typodupeerror

Submission + - Google's New Hurricane Model Was Breathtakingly Good This Season (arstechnica.com)

Submitted by Anonymous Coward
An anonymous reader writes: Although Google DeepMind’s Weather Lab only started releasing cyclone track forecasts in June, the company’s AI forecasting service performed exceptionally well. By contrast, the Global Forecast System model, operated by the US National Weather Service and is based on traditional physics and runs on powerful supercomputers, performed abysmally. The official data comparing forecast model performance will not be published by the National Hurricane Center for a few months. However, Brian McNoldy, a senior researcher at the University of Miami, has already done some preliminary number crunching.

The results are stunning: A little help in reading the graphic is in order. This chart sums up the track forecast accuracy for all 13 named storms in the Atlantic Basin this season, measuring the mean position error at various hours in the forecast, from 0 to 120 hours (five days). On this chart, the lower a line is, the better a model has performed. The dotted black line shows the average forecast error for official forecasts from the 2022 to 2024 seasons. What jumps out is that the United States’ premier global model, the GFS (denoted here as AVNI), is by far the worst-performing model. Meanwhile, at the bottom of the chart, in maroon, is the Google DeepMind model (GDMI), performing the best at nearly all forecast hours.

The difference in errors between the US GFS model and Google’s DeepMind is remarkable. At five days, the Google forecast had an error of 165 nautical miles compared to 360 nautical miles for the GFS model, more than twice as bad. This is the kind of error that causes forecasters to completely disregard one model in favor of another. But there’s more. Google’s model was so good that it regularly beat the official forecast from the National Hurricane Center (OFCL), which is produced by human experts looking at a broad array of model data. The AI-based model also beat highly regarded “consensus models,” including the TVCN and HCCA products. For more information on various models and their designations, see here.

Submission + - New HDR10+ Advanced Standard Will Try To Fix the Soap Opera Effect (arstechnica.com)

Submitted by Anonymous Coward
An anonymous reader writes: Today, Samsung provided details about the next version of the HDR10 format, which introduces six new features. Among HDR10+ Advanced’s most interesting features is HDR10+ Intelligent FRC (frame rate conversion), which is supposed to improve motion smoothing.

A TV using motion smoothing analyzes each video frame and tries to determine what additional frames would look like if the video were playing at a frame rate that matched the TV’s refresh rate. The TV then inserts those frames into the video. A 60Hz TV with motion smoothing on, for example, would attempt to remove judder from a 24p film by inserting frames so that the video plays as if it were shot at 60p. For some, this appears normal and can make motion, especially camera panning or zooming, look smoother. However, others will report movies and shows that look more like soap operas, or as if they were shot on higher-speed video cameras instead of film cameras. Critics, including some big names in Hollywood, argue that motion smoothing looks unnatural and deviates from the creator’s intended vision.

Intelligent FRC takes a more nuanced approach to motion smoothing by letting content creators dictate the level of motion smoothing used in each scene, Forbes reported. The feature is also designed to adjust the strength of motion interpolation based on ambient lighting.

Submission + - The world's tallest chip defies the limits of computing: goodbye to Moore's Law? (elpais.com) 1

Submitted by dbialac
dbialac writes: Building chips up instead of smaller may be a solution to the problems encountered with modern semiconductors.

Xiaohang Li, a researcher at King Abdullah University of Science and Technology (KAUST) in Saudi Arabia, and his team have designed a chip with 41 vertical layers of semiconductors and insulating materials, approximately ten times higher than any previously manufactured chip. The work, recently published in the journal Nature Electronics, not only represents a technical milestone but also opens the door to a new generation of flexible, efficient, and sustainable electronic devices. “Having six or more layers of transistors stacked vertically allows us to increase circuit density without making the devices smaller laterally,” Li explains. “With six layers, we can integrate 600% more logic functions in the same area than with a single layer, achieving higher performance and lower power consumption.”

Submission + - Sam Altman says 'enough' to questions about OpenAI's revenue (techcrunch.com)

Submitted by joshuark
joshuark writes: OpenAI CEO Sam Altman recently said that the company is doing “well more” than $13 billion in annual revenue — and he sounded a little testy when pressed on how it will pay for its massive spending commitments.

“First of all, we’re doing well more revenue than that. Second of all, Brad, if you want to sell your shares, I’ll find you a buyer,” Altman said, prompting laughs from Nadella. “I just — enough. I think there are a lot of people who would love to buy OpenAI shares.”

Altman acknowledged that there are ways the company “might screw it up” — for example by failing to get access to enough computing resources — but he said that “revenue is growing steeply.”

At the same time, he denied reports that OpenAI plans to go public next year.

“No no no, we don’t have anything that specific,” Altman said. “I’m a realist, I assume it will happen someday, but I don’t know why people write these reports. We don’t have a date in mind, we don’t have a board decision to do this or anything like that. I just assume it’s where things will eventually go.”

Submission + - Bank of America faces lawsuit over alleged unpaid computer boot-up time (hcamag.com)

Submitted by Joe_Dragon
Joe_Dragon writes: Bank of America is facing allegations that hundreds of hourly workers performed up to 30 minutes of unpaid computer setup work daily for years.

A former Business Analyst filed a class action lawsuit in federal court on October 23, claiming the banking giant systematically shortchanged remote employees who had to boot up complex computer systems before their paid shifts began.

Tava Martin, who worked both remotely and at the company's Jacksonville facility, says the financial institution required her and fellow hourly workers to log into multiple security systems, download spreadsheets, and connect to virtual private networks—all before the clock started ticking on their workday.

The process wasn't quick. According to the filing in the United States District Court for the Western District of North Carolina, employees needed 15 to 30 minutes each morning just to get their systems running. When technical problems occurred, it took even longer.

Here's how it worked: Workers turned on their computers, waited for Windows to load, grabbed their cell phones to request a security token for the company's VPN, waited for that token to arrive, logged into the network, opened required web applications with separate passwords, and downloaded the Excel files they needed for the day. Only then could they start taking calls from business customers about regulatory reporting requirements.

The lawsuit says Bank of America enforced a strict "phone ready" policy. Employees had to be prepared to handle calls the moment their scheduled shifts began. Anyone who clocked in but wasn't immediately available to take or make calls for too long risked poor performance scores and possible disciplinary action, up to and including termination.

Yet the company allegedly discouraged workers from reporting any time outside their scheduled hours. Martin's paystubs routinely showed exactly 40 hours per week, or exactly 32 hours when she missed a day—suggesting the bank paid for scheduled time rather than actual work performed.

The unpaid work didn't stop at startup. During unpaid lunch breaks, many systems would automatically disconnect or otherwise lose connection, forcing employees to repeat portions of the login process—approximately three to five minutes of uncompensated time on most days, sometimes longer when a complete reboot was required. After shifts ended, workers had to log out of all programs and shut down their computers securely, adding another two to three minutes.

Martin earned $46.17 per hour through a third-party staffing agency, though Bank of America controlled her schedule, training, and employment conditions. Like many of her colleagues, she regularly worked full-time hours, meaning the uncompensated startup and shutdown time should have been paid at the overtime rate of one and a half times her regular wage.

The lawsuit points to 2008 guidance from the Department of Labor that specifically addresses call centers under the Fair Labor Standards Act. That guidance explicitly states that an example of the first principal activity of the day for call center workers includes starting computers to download work instructions and applications. It also requires employers to keep daily or weekly records of all hours worked, including time spent in pre-shift and post-shift activities.

The filing suggests Bank of America either didn't bother to determine whether the computer time was compensable or knew it was but failed to pay for it anyway. The lawsuit notes the company has faced factually similar cases from other employees about time spent loading and logging into computer systems.

For the week of March 11 through March 17, 2024, for example, Martin was paid for 40 regular hours but no overtime. With unpaid pre-shift, meal-period, and post-shift time of at least 20 minutes per shift over five shifts, she should have received an additional 100 minutes at her overtime rate of $69.25 per hour. Similar calculations apply to other pay periods cited in the complaint.

Business Analysts were interviewed by company hiring managers and assigned to Bank of America managers upon hire. The bank provided supervisors who oversaw their daily performance and gave them training and technical support. The company controlled work schedules and retained the ability to discipline and terminate employees. The positions were hourly, non-exempt jobs with rigid schedules requiring at least eight hours per day, on average five days per week, and up to 40 hours or more weekly.

Martin seeks to represent all current and former remote hourly Business Analysts who worked for the bank during the three years before conditional certification through judgment. She estimates the group includes hundreds, if not thousands, of workers who performed essentially the same tasks using the same or similar computer programs under the same timekeeping policies.

Many Business Analysts, including Martin, were employed through third-party staffing agencies but were required to comply with all Bank of America employee handbook policies, including those covering attendance, timekeeping, and overtime.

The case remains in early stages, with no court ruling yet on whether it will proceed as a class action or on the merits of the allegations.

Submission + - Microsoft finally makes Fedora an official Windows Subsystem for Linux (WSL) dis (betanews.com)

Submitted by BrianFagioli
BrianFagioli writes: Well, it has finally happened. Fedora Linux is now officially available as a Windows Subsystem for Linux (WSL) distribution! Thatâ(TM)s right, folks, following prior testing, you can now run Fedora 42 natively inside Windows using WSL. As someone who considers Fedora to be my favorite Linux distribution, this is a pretty exciting development.

Installing it is simple enough. Just open up a terminal and type wsl --install FedoraLinux-42 to get started. After that, launch it with wsl -d FedoraLinux-42 and set your username. No password is required by default, and youâ(TM)ll automatically be part of the wheel group, meaning you can use sudo right out of the gate.

Submission + - Antarctica's Ice Sheet Grows for the First Time in Decades

Submitted by RoccamOccam
RoccamOccam writes: Previous studies have consistently shown a long-term trend of mass loss, particularly in West Antarctica and the Antarctic Peninsula, while glaciers in East Antarctica appeared relatively stable. However, a recent study led by Dr. Wang and Prof. Shen at Tongji University has found a surprising shift: between 2021 and 2023, the AIS experienced a record-breaking increase in overall mass.

Submission + - Boffins report: "AI" creates more "work" than it saves (arstechnica.com) 1

Submitted by Mr. Dollar Ton
Mr. Dollar Ton writes: Moreover, previous estimates of huge "productivity gains" are largely faked and the new "work" it is not necessarily useful.

A new study analyzing the Danish labor market in 2023 and 2024 suggests that generative AI models like ChatGPT have had almost no significant impact on overall wages or employment yet, despite rapid adoption in some workplaces.

The reported productivity benefits were modest in the study. Users reported average time savings of just 2.8 percent of work hours (about an hour per week).

The finding contradicts a randomized controlled trial published in February that found generative AI increased worker productivity by 15 percent on average. The difference stems from other experiments focusing on tasks highly suited to AI, whereas most real-world jobs involve tasks AI cannot fully automate.

Submission + - Army Will Seek Right to Repair Clauses in All Its Contracts (404media.co)

Submitted by Anonymous Coward
An anonymous reader writes: A new memo from Secretary of Defense Pete Hegseth is calling on defense contractors to grant the Army the right-to-repair. The Wednesday memo is a document about “Army Transformation and Acquisition Reform” that is largely vague but highlights the very real problems with IP constraints that have made it harder for the military to repair damaged equipment.

Hegseth made this clear at the bottom of the memo in a subsection about reform and budget optimization. “The Secretary of the Army shallidentify and propose contract modifications for right to repair provisions where intellectual property constraints limit the Army's ability to conduct maintenance and access the appropriate maintenance tools, software, and technical data—while preserving the intellectual capital of American industry,” it says. “Seek to include right to repair provisions in all existing contracts and also ensure these provisions are included in all new contracts.” [...] The memo would theoretically mean that the Army would refuse to sign contracts with companies that make it difficult to fix what it sells to the military. The memo doesn’t carry the force of law, but subordinates do tend to follow the orders given within. The memo also ordered the Army to stop producing Humvees and some other light vehicles, and Breaking Defense confirmed that it had.

Submission + - Peking U. Just Made the World's Fastest Transistor and It Is Not Made of Silicon (zmescience.com)

Submitted by schwit1
schwit1 writes: The new transistor runs 40% faster and uses less power.

With a slender sheet of lab-grown bismuth and an architecture unlike anything inside today’s silicon chips, they’ve built what they call the world’s fastest and most efficient transistor. Not only does it outperform the best processors made by Intel and TSMC, but it also uses less energy doing so. And most important of all, there’s no trace of silicon involved.

This two-dimensional, silicon-free transistor represents a radical rethinking of what chips can be and how they can be made.

Rather than silicon, the Peking University team built their transistor using bismuth oxyselenide (BiOSe) for the channel, and bismuth selenite oxide (BiSeO) as the gate material.

These materials are part of a class known as two-dimensional semiconductors — atomically thin sheets with exceptional electrical properties. Bismuth oxyselenide, in particular, offers something silicon struggles with at ultra-small sizes: speed.

Electrons move through it faster, even when packed into tiny spaces. It also has a higher dielectric constant, meaning it can hold and control electric charge more efficiently. That makes for faster switching, reduced energy loss, and — very importantly — a lower chance of overheating.

“This reduces electron scattering and current loss, allowing electrons to flow with almost no resistance, akin to water moving through a smooth pipe,” Peng explained.

The interface between these materials is also smoother than that of common semiconductor-oxide combinations used in industry today. That means fewer defects and less electrical noise.

All of this adds up to stunning results. According to the team, their transistor can run 40% faster than today’s most advanced 3-nanometer silicon chips — and it does so while using 10% less energy.

Comment .. but not all of them (Score 1) 50

by talexb (#65022227) Attached to: How many devices are connected to your home WiFi network?

My Anova convection oven and my HP inkjet (since recycled) both begged to connect to the Internet, and I firmly declined. I didn't want some nerd in Omaha or Beijing decide to brick my device because of some clever lucky magic rules that I didn't read in the 3 page User Agreement.

"But the app will let you know when the oven's ready to cook, or when the timer's done!" Yeah, that's not really a big deal for me. I do have other timers, like .. right there on my phone. And I have one on the stove and on the microwave. "But we can send you recipes!" Thanks, I have recipes.

Hard pass.

Comment Re:weird (Score 1) 120

I believe this interpretation is correct (IANAL).

- Offer of $2,500
- Proposal of $50,000 (implies refusal of the $2,500 offer)
- Proposal acceptance

Sure, there are terms to be worked out, but the price has now been fixed, and there's a contract. This would be an interesting discussion to read about.

Comment We transitioned to fully remote after six months (Score 1) 289

by talexb (#62431118) Attached to: Fed-up Managers Declare Work From Home is Over

Our CEO made the announcement in about September 2020 -- "We're now a fully remote company."

No, it's not ideal, not having to deal with a pandemic would be ideal, but failing that, I don't miss the 90 minutes of commuting I'm no longer doing daily, and I also don't mind being able to make my own lunch instead of paying for lunch or bring some leftovers. Also, at 530pm, I can get up from my desk, go into the kitchen and immediately have a beer. It's OK.

I would enjoy meeting up perhaps every 2-4 weeks for half a day of meetings and/or demos, but apart from that, WFH is great. Strangely enough, a developer on another team missed going into the office so much that they left the company to work somewhere that they could actually go work in an office. I don't get it.

Slashdot Top Deals

The human mind ordinarily operates at only ten percent of its capacity -- the rest is overhead for the operating system.

Close