Forgot your password?
Close
typodupeerror

Submission + - A startup claims it broke through a bottleneck that's holding back LLMs (technologyreview.com)

Submitted by joshuark
joshuark writes: MIT Technology review reports that Miami-based AI startup Subquadratic came out of stealth mode last month with a huge claim. It announced that it had solved a mathematical bottleneck that had been holding back large language models for almost a decade.
According to Subquadratic, it has developed a new kind of LLM, called SubQ, that is faster and cheaper and uses a lot less energy than any other model on the market. The company also claims that SubQ is able to process up to 12 times as much text at once than most other models, allowing it to carry out a range of data-heavy tasks, such as analyzing hundreds of documents or entire code bases.
The problem was that the company at first provided little evidence for its claims beyond a handful of self-published test scores. And it has yet to make SubQ widely available for people to try out themselves.
So it’s no surprise that Subquadratic’s claims were met with skepticism. Dan McAteer, an artificial intelligence engineer, captured the overall response on X: “SubQ is either the biggest breakthrough since the Transformer ... or it’s AI Theranos.”
“We expected healthy skepticism,” says Subquadratic cofounder and chief technology officer Alex Whedon. “In hindsight, releasing the third-party benchmarks alongside the initial announcement would have preempted much of the skepticism, which is why we’re taking the time to make sure any future results are fully verified before putting them out.”
SubQ won’t replace existing top models across the board, but it could offer huge increases in speed at a fraction of the typical cost for certain tasks. Subquadratic insists that in the long run, though, its breakthrough could change how LLMs are built. “We hope we’re kicking off a new age of efficiency,” says Justin Dangel, the firm’s cofounder and CEO. “We don’t think anybody will be building on transformers in a few years.”

Submission + - What is a quantum computer good for? Absolutely nothing — yet (theverge.com)

Submitted by joshuark
joshuark writes: The Verge has an article about the "absolute nothing" of quantum computers. We have yet to see a quantum computer conclusively perform a single useful task. Existing machines are simply too small and error-ridden to solve commercially relevant problems.
Companies drive the hype, too. In June, Microsoft announced a new quantum computing chip named Majorana 2. It claimed the chip was a hardware advancement that accelerates its timeline to a “scalable, practical quantum computer” by 2029. But independent experts swiftly criticized the announcement. “This is complete codswallop,” Henry Legg, a physicist from the University of St. Andrews and a longtime Microsoft critic, tells The Verge.
Legg just published a paper in Nature on June 24th criticizing Microsoft’s quantum claims from a year ago — peer review takes a long time — and pointing to what he sees as major discrepancies between Microsoft’s papers and press releases. Nature included Microsoft’s rebuttal.
Researchers have made genuine progress in quantum computing — it’s just been largely incremental and too esoteric to immediately capture the public’s imagination. Proponents predict that the technology will lead to discoveries in medicine, as well as advances in materials science and machine learning. Meanwhile, many national security experts frame its development as a new Cold War competition between the US and China.
Some have imagined the quantum computer as a cyberattack tool. In 1994, computer scientist Peter Shor developed a quantum computing algorithm for factoring prime numbers that should be able to break RSA encryption, a ubiquitous family of algorithms used to secure banking and email communications. So "RSA is dead" is the repeated mantra of the quantum computing hype.
Current quantum computers like Google’s Willow are individual chips too primitive to break RSA encryption or implement drug molecule simulations. But the vision is to build scaled-up machines that can.
Similar cycles have played out several times since the technology’s beginnings. Companies announce a breakthrough; independent researchers cry hype, all while investors continue to inject money into the industry. Then investors cash out and then call it a "scam" on the public.
Henry Legg is more skeptical and thinks some have underestimated the fundamental challenges of scaling. “There’s no evidence of the scalability of any platform to the level that you would need to do useful quantum computations within a decade, or probably a couple of decades,” he says.
While researchers have made progress toward building a useful quantum computer, it’s not clear what that use should be. “It’s such a nascent technology,” says Islam. “If you ask, what is a quantum computer good for, I do not know of an application which is a sure shot.”
The Trump administration wants a useful quantum computer in two years. Are we having fun yet?

Submission + - Microsoft fake Windows error ended in a $280 million secret settlement (makeuseof.com)

Submitted by joshuark
joshuark writes: Facing real competition from Digital Research's DR DOS, Microsoft secretly embedded a sabotaging mechanism known as "AARD code" into beta versions of Windows 3.1 to prevent it from running on Digital Research's competing DR DOS operating system.
This code triggered fake, alarming error messages to convince developers that DR DOS was unstable, effectively eliminating a significant market threat through fear, uncertainty, and doubt. Although the company disabled the feature in the final retail release, the California-based firm Caldera, Inc., which had acquired DR DOS assets, sued Microsoft for anti-competitive practices.
Microsoft settled the lawsuit out of court in 2000 for $280 million, a figure that remained sealed until it was unsealed in 2009. Nothing says taking ownership and responsibility than keeping it a sealed secret for a decade. Microsoft paid for being clumsy enough to write the intent down in an email. The lesson the industry took away wasn't "don't do it." It was "don't put it in writing." Something Bill Gates forgot with Epstein.

Comment IBM is ready... (Score 1) 73

by joshuark (#66213726) Attached to: IBM is Getting Ready to Scale Quantum Computing

IBM is ready, yeah "everyone" is ready for the disco comeback.

A friend of mine talked with this guy on LinkedIn who was selling "tech training" about quantum computing, declaring "RSA is dead." Doing scam training on some new technology that is the next thing and we have to get ready for...

Last time it was "blockchain" and how its going to change the world like cold-fusion, room temperature superconductors, etc.

The IBM CEO, CTO, and all those "experts" a song...

https://www.youtube.com/watch?...

Although the hype is great to boost the stock value of IBM without buy backs, and then cash in the stock grants.

https://www.cnn.com/markets/st...

--JoshK.

Submission + - New Study Shows Tall Vehicle Hoods Cause Hundreds More Deaths Per Year (caranddriver.com)

Submitted by joshuark
joshuark writes: Car and Driver magazine reports that a new study conducted by the New York Times shows that the increase in vehicle hood height seen over the last two and a half decades, mainly due to the rise in popularity of large SUVs and trucks, has resulted in several thousand deaths that otherwise may not have happened. The study shows that while automakers and regulators have focused on occupant safety, they have turned a blind eye to pedestrian safety, which has fallen since around 2009.
Researchers looked at four main datasets in their investigation: crash test data from the National Highway Traffic Safety Administration's (NHTSA) Crash Report Sampling System (CRSS) from 2016 to 2024; NHTSA's Fatality Analysis Reporting System (FARS); vehicle measurement data from Expert AutoStats; and vehicle registration data from S&P Global from 2002 to 2024.
The researchers concluded that the increased danger to pedestrians is caused by two main culprits.
First, large SUVs and trucks have taller hoods, raising the point of impact above most people's center of gravity and pushing them to the ground, typically hard asphalt, rather than up and onto the hood, which is designed to absorb impacts.
Second, with larger A-pillars designed to protect occupants in rollover crashes, modern cars tend to have larger blind spots than cars sold at the turn of the century (presuming the 21st century).
The shift toward vehicles with taller hoods led to roughly 3000 deaths between 2016 and 2024. This number is conservative because it does not include crashes that take place in parking lots, driveways, or private roads, which aren't part of the federal database.
The data also showed an estimated 2.8 percent increase in the odds of a pedestrian fatality for every one-inch increase in vehicle hood height. Between two different scenarios, one decreasing the hood height of every vehicle in the dataset by 3 inches, and the second using a random sampling of hood heights from 2002 across 10,000 simulated crashes, between 2624 (for scenario two) and 3077 (for scenario one) lives could have been saved from 2016 to 2024.

Submission + - M&M's Is Getting Rid Of 2 Iconic Colors—But There's A Good Reason Why (delish.com)

Submitted by joshuark
joshuark writes: Mars, the maker of M&M’s, will no longer manufacture blue and brown M&M’s. That’s right, the popular chocolate candy brand is doing away with two of its iconic colors, but not without good reason.The company is phasing out synthetic dyes and has plans to launch a natural version of the candy this August. Currently, the small candy-coated chocolates come in seven bright colors—red, yellow, green, blue, orange, brown, and sometimes purple. As it turns out, replicating all of these colors in the same shades using natural ingredients was harder than anticipated.

According to The Wall Street Journal, Mars had little trouble recreating the red, orange, yellow, and green M&M colors using ingredients such as turmeric and beets. Mars is moving forward with natural M&M’s in just red, green, yellow, and orange. That pack of M&M’s will never be the same. RIP li’l buddies.

Not surprising as red was once taboo for a decade. https://www.cnn.com/videos/bus...

Submission + - Microsoft discovers new lightweight backdoor that steals cryptocurrency (arstechnica.com)

Submitted by joshuark
joshuark writes: Ars Technica reports Microsoft says it has detected new self-propagating malware that spreads through USB drives in search of cryptocurrency credentials, which it then sends to attacker-controlled servers. The company named the worm Crypto Clipper because it monitors the contents of device clipboards for patterns consistent with wallet addresses or seed phrases.

“The execution of this clipper is notable because it does not depend on a traditional installer or exposed IP-based C2 infrastructure,” Microsoft said Thursday. “Instead, it deploys a portable Tor client, routes traffic through a local SOCKS5 proxy, and blends data theft with remote code execution, turning a financially motivated stealer into a lightweight backdoor.”

“This malware family shows how lightweight, script-based stealers can deliver outsized impact when paired with anonymized communications and runtime tasking,” Microsoft said. “The combination of Tor-routed C2, clipboard targeting, screenshot capture, and remote code execution gives attackers both immediate monetization paths and continued control over compromised devices.”

Big question is "What's in your crypto wallet?"

Submission + - Trump administration is U-turning on plan closing ocean monitoring system (cnn.com) 1

Submitted by joshuark
joshuark writes: The Trump administration is U-turning on its controversial decision to dismantle a critical ocean monitoring system that provides vital information on the health of the world’s oceans, after a bipartisan backlash in Congress.

The National Science Foundation, which funds the $386 million deep-ocean system, announced it would be pulling up buoys and other underwater equipment from arrays in late May. Thursday, NSF announced it will halt these plans and convene an expert panel to “identify a sustainable path” forward. The NSF’s about-face comes amid intense backlash to its original decision. Experts feared the US was taking eyes off the oceans as they endure a period of huge change, with off-the-chart temperatures fueling devastating storms and threatening fisheries.

Ocean scientists CNN previously spoke to said ditching the monitoring system was “foolish” and “counterintuitive.”

Lawmakers on both sides of the aisle raised objections. “Dismantling the Ocean Observatories Initiative is supreme stupidity, costing taxpayers millions of dollars and destroying a vital source of climate data,” Oregon Sen. Jeff Merkley, a Democrat, wrote in a statement.
  House Science Committee Ranking Member Zoe Lofgren, a Democrat from California, said the NSF’s reversal was welcome but cautioned it was not yet clear “how much damage they have already done.”

“This should have never happened,” she told CNN in a statement. “This pathetic scheme was illegal. NSF is governing via chaos and reactionary nonsense. Scientists and coastal economies that depend on this data deserve better.”

The NSF said it “remains committed to ocean sciences, to responsible stewardship of its research infrastructure and to supporting the stakeholders that depend on it.”

Submission + - Shooting survivor sues AI gun detection comp after system failed to spot weapon (arstechnica.com)

Submitted by joshuark
joshuark writes: The injured teenage survivor of a January 2025 shooting at a Nashville, Tennessee high school recently sued the manufacturer of an “AI gun detection” system that failed to detect the handgun that left two dead, including the shooter.

In 2023, the Metropolitan Nashville Public Schools (MNPS) Board approved a contract worth over $1 million to install an AI detection layer on top of its district-wide network of cameras and related security infrastructure.

The lawsuit alledges the security company Omnilert either knew or should have known that there were “significant operational limitations in its gun detection system that could result in detection failures during actual emergencies, including limitations based on camera placement, proximity of the weapon to camera sensors, camera angle, lighting, and weapon visibility.”

Omnilert cofounder Ara Bagdasarian declined an invitation to answer questions about the lawsuit. System Integrations, the other defendant in the case, which resold the Omnilert system, also did not respond to a request for comment.

MNPS spokesperson Sean Braisted said in a press conference following the January 2025 shooting that due to where the shooter was in relation to the cameras, the imagery “wasn’t close enough to get an accurate read and to activate that alarm.”

The money that MNPS spent on deploying these detection systems, he added, “could have gone to a counselor or something else to a kid in crisis. Every decision that you make is pointing away resources from something else.”

Submission + - Scientists ejected from diabetes conference (arstechnica.com)

Submitted by joshuark
joshuark writes: Five leading scientists were ousted from the annual meeting of the American Diabetes Association (ADA) in New Orleans on Friday. Their crime: handing out copies of an editorial, published in the journal Diabetes Care on April 29, sharply criticizing the Trump administration’s ongoing attacks on scientific research.

“They physically grabbed us, forced us out of the conference center, and now are telling us we can no longer attend this meeting,” Kelly told MedPage Today, which first reported the incident. “They’re taking our lanyards. It really has come to this in America. Censorship is real. America needs to stand up. Scientists, stand up. Physicians, stand up.”

The ADA confirmed to MedPage Today that five registered scientists had been removed from the meeting, claiming the scientists had violated the organization’s code of conduct for conferences. “These attendees were escorted out by our onsite event security because they demonstrated behavior not consistent with this code of conduct,” the ADA media team said in a statement. “They were respectfully given the opportunity to cease this behavior and chose not to which is why they were escorted out.”

Online backlash to the ADA’s actions spread rapidly on both Twitter/X and BlueSky, and sharply increased the number of page views for Kahn et al.’s April editorial. According to Kahn, the editorial was published with a disclaimer, added by ADA leadership, insisting that the ADA had nothing to do with developing or writing the article. He has written to the ADA seeking re-admittance to the conference, since he is slated to speak and chair a session.

“It is no longer enough to stand idly by or work behind the scenes with lawmakers,” the authors wrote in their editorial.

Submission + - Microsoft's new quantum computer chip has a fundamental problem

Submitted by joshuark
joshuark writes: Microsoft claimed today that it has improved its quantum technology by an extraordinary factor. Outside experts say it doesn’t even work and never has.

The company has named its latest quantum chip Majorana 2, for the theoretical quasiparticle it aims to use as the basis for a new “topological” approach to quantum computing. Chilled to ultracold temperatures in superconducting wires, electrons may be coaxed to act collectively—as so-called Majorana quasiparticles—in a manner that theoretically makes them more resistant to the physical “noise” that causes computational errors in other quantum systems.

“This is a very exciting era that we’re in,” said Jason Zander, executive vice president leading Microsoft’s Quantum team, during a press briefing before the public announcement. “We’re at the start of a new chapter.”
But the company has a mixed track record when it comes to such claims. In 2021 Microsoft retracted a high-profile Nature paper after outside experts pointed out that the study’s data could have come from material imperfections rather than a topological qubit. Physicists have raised similar concerns about several publications since, as well as about last year’s announcement of the Majorana 1 chip based on the disputed technology.
The latest claim, made in a new preprint manuscript that has not been peer-reviewed, builds on these controversial results. The team has replaced the Majorana 1 chip’s aluminum superconductor with one made of lead. “People generally want to try to stay away from putting lead into anything,” said the team’s Chetan Nayak in the press briefing. “It sounds like a crazy idea, actually.” Material adjustments including this one, the Microsoft team claims, add to the longevity of the hypothetical qubit—which is now reported at 20 seconds to a minute—by improving its error-averting “topological gap.”

A Microsoft blog post accompanying the announcement noted that due to this “rapid progress” the company is accelerating its technological roadmap, and now aims to demonstrate “scalable, practical quantum computing” by 2029.

The new result simply isn’t up to the physics community’s standards, he says. “If this was from any other group or Ph.D. student, it would never make it through peer review,” Legg says. In fact, the multitrillion-dollar company’s last preprint of this kind has remained unpublished since last summer, which physicist Sergey Frolov cites as evidence that top journals have likely rejected it.

Meanwhile, the company maintains that it has demonstrated what many scientists say it has not. “Bell Labs didn’t have to prove there was an electron to invent the transistor...”

Submission + - Apple is bringing age verification to Texas this week (theverge.com)

Submitted by joshuark
joshuark writes: Apple will introduce age verification in the App Store for users in Texas starting on Thursday, June 4th. The move, as spotted by MacRumors, comes just days after a federal appeals court allowed Texas’ App Store Accountability Act to go into effect while a lawsuit against it proceeds.

People in Texas who are creating a new Apple account will need to verify they’re over 18 using a credit card or government ID. Apple may also automatically verify users’ age using the age of their account and whether they have a credit card on file.

Despite Apple’s attempts to push back on app store-level age verification, the company has announced plans to implement age checks to comply with laws in places like Utah, Louisiana, Brazil, Australia, Singapore, and the UK. Google is required to make similar changes to the Play Store and is also introducing age-checking tools for developers.

Last December, a judge blocked the App Store Accountability Act (SB 2420) from taking effect, but an appeals court has now reversed this decision — at least while the court figures out whether the law is constitutional. Even if this law gets struck down in Texas, a federal version with the same name is still making its way through Congress and could impose age verification at the app store nationwide.

Submission + - Scientists charged with bringing deactivated mpox virus to US, lying about it (abcnews.com)

Submitted by joshuark
joshuark writes: Two scientists at a U.S. government lab were charged with smuggling vials of deactivated mpox virus into the country from Africa and lying about it during interviews with investigators at a Detroit airport, authorities said Tuesday.

Vincent Munster, who is chief of the virus ecology section at Rocky Mountain Laboratories in Hamilton, Montana, and Claude Kwe, who works with him. Both were stopped at Detroit Metropolitan Airport in January after a flight from Paris and nine days in the Republic of Congo.

Munster “adamantly denied” returning to the U.S. with biological materials or samples, the FBI said in a court filing. But tests subsequently revealed that Munster and Kwe were traveling with vials of deactivated mpox, the FBI said, yet they had failed to declare them or obtain the necessary permission.

Munster told investigators at the Detroit-area airport that any necessary documents were in his laptop, “but you don't need them. I do this all the time,” the FBI quoted him as saying.

“It is reasonable to believe that Munster's statements regarding the possession of the required documentation to (customs officers) were materially false,” the FBI said.

"Any deliberate effort to conceal and smuggle biological materials into the United States without proper authorization is a breach of the public’s trust and could have placed the public at risk,” said Marcus Sykes of the Office of Inspector General at the Department of Health and Human Services.

In 2022, the mpox virus was confirmed to spread via sex for the first time and triggered outbreaks in more than 70 countries that had not previously reported mpox.

Submission + - Microsoft Deliberately Bricking All Office for Mac 2019/2021 Installations (osnews.com) 2

Submitted by joshuark
joshuark writes: MacOS users who opted to buy a copy of Microsoft Office for macOS back in 2019 or 2021, eschewing the Office 365 subscription, so you could keep on using Office 2019/2021 forever if you wanted to. Just like in the old days.

Consumer Rights Wiki reports:

"Microsoft Office 2019 and 2021 for Mac view-only conversion (2026) is a scheduled remote degradation of perpetually-licensed Microsoft Office software for macOS and iOS, set for July 13, 2026 when a license-validation certificate used by the Office apps expires.[1] After Office 2019 for Mac reached end of support in October 2023, Microsoft assured customers their installed apps would "continue to function."[2] The July 13, 2026 conversion instead drops the apps into a Microsoft-defined "reduced functionality mode," in which files can be opened and viewed but not edited or saved.[1][3] By May 30, 2026, the original 2023 end-of-support page had been re-dated and rewritten on Microsoft's site; the "continue to function" clause was removed.[4][2]" https://consumerrights.wiki/w/...

Microsoft’s advice to the users they’re stealing from is to keep using the applications as mere viewers, switch to the free Office 365 web applications, pay for a 365 subscription, or buy a brand new regular copy of Office 2024. None of these make any sense, and clearly, all of this should be illegal, but it’s not because the software industry is a clown show.

Slashdot Top Deals

E Pluribus Unix

Close