SoylentNews

upstart writes:

New Chip Can Process and Classify Nearly Two Billion Images per Second - Technology Org:

In traditional neural networks used for image recognition, the image of the target object is first formed on an image sensor, such as the digital camera in a smartphone. Then, the image sensor converts light into electrical signals, and ultimately into binary data, which can then be processed, analyzed, stored, and classified using computer chips. Speeding up these abilities is key to improving any number of applications, such as face recognition, automatically detecting text in photos, or helping self-driving cars recognize obstacles.

[...] The current speed limit of these technologies is set by the clock-based schedule of computation steps in a computer processor, where computations occur one after another on a linear schedule.

To address this limitation, [...] have removed the four main time-consuming culprits in the traditional computer chip: the conversion of optical to electrical signals, the need for converting the input data to binary format, a large memory module, and clock-based computations.

They have achieved this through direct processing of light received from the object of interest using an optical deep neural network implemented on a 9.3 square millimeter chip.

[...] "Our chip processes information through what we call 'computation-by-propagation,' meaning that, unlike clock-based systems, computations occur as light propagates through the chip," says Aflatouni. "We are also skipping the step of converting optical signals to electrical signals because our chip can read and process optical signals directly, and both of these changes make our chip a significantly faster technology."

upstart writes:

Car tyres produce vastly more particle pollution than exhausts, tests show:

Emissions from tailpipes in developed countries are much lower in new cars, with those in Europe far below the legal limit.A

Almost 2,000 times more particle pollution is produced by tyre wear than is pumped out of the exhausts of modern cars, tests have shown.

The tyre particles pollute air, water and soil and contain a wide range of toxic organic compounds, including known carcinogens, the analysts say, suggesting tyre pollution could rapidly become a major issue for regulators.

Air pollution causes millions of early deaths a year globally. The requirement for better filters has meant particle emissions from tailpipes in developed countries are now much lower in new cars, with those in Europe far below the legal limit. However, the increasing weight of cars means more particles are being thrown off by tyres as they wear on the road.

The tests also revealed that tyres produce more than 1tn ultrafine particles for each kilometre driven, meaning particles smaller than 23 nanometres. These are also emitted from exhausts and are of special concern to health, as their size means they can enter organs via the bloodstream. Particles below 23nm are hard to measure and are not currently regulated in either the EU or US.

upstart writes:

'Make VPN Detection Tools Mandatory to Fight Geo-Piracy' * TorrentFreak:

The United States is actively exploring options to update copyright law to bring it into line with the current online environment.

Most recently, the Copyright Office is looking into the option of making certain standard technical measures (STMs) mandatory for online platforms. This could include upload filters to block pirated content from being reuploaded.

[...] Most copyright holders are supportive of the idea. They feel that without proper incentives, some online services will fail to address the piracy problem. Opponents of the idea, meanwhile, point out that it may lead to all sorts of problems and may negatively affect free expression.

Much of the discussion thus far has focused on tools and technologies that detect and filter copyright-infringing content. However, this week we spotted another submission that promotes a different type of measure, which isn't necessarily less controversial.

In a letter to the Copyright Office, GeoComply CEO Anna Sainsbury suggests that VPN detection tools can play an important role as well.

Twitter reportedly will give Musk the full "firehose" of user data he demanded

Freeman writes:

https://arstechnica.com/tech-policy/2022/06/twitter-reportedly-will-give-musk-the-full-firehose-of-user-data-he-demanded/

Twitter now plans to comply with Elon Musk's demand for user data that he says is needed to determine whether the company's spam estimates are accurate, The Washington Post reported Wednesday.

"After a weeks-long impasse, Twitter's board plans to comply with Elon Musk's demands for internal data by offering access to its full 'firehose,' the massive stream of data comprising more than 500 million tweets posted each day, according to a person familiar with the company's thinking, who spoke on the condition of anonymity to describe the state of negotiations," the Post wrote.

Twitter declined comment on the Post report when contacted by Ars today but pointed to its statement from Monday that "Twitter has and will continue to cooperatively share information with Mr. Musk to consummate the transaction in accordance with the terms of the merger agreement."

Whether Twitter has to give all the user data to Musk is under dispute. The Post report comes two days after Musk's legal team sent a letter to Twitter claiming the company violated the merger agreement by refusing to provide the data behind its spam estimates.

Twitter Set to Comply With Elon Musk Demand for Data on Fake Accounts

upstart writes:

Twitter set to comply with Elon Musk demand for data on fake accounts:

Elon Musk warned he might walk away from Twitter if it fails to provide the data on spam and fake accounts he seeks.

Twitter is preparing to comply with Elon Musk's demand for data on fake accounts, after the Tesla chief executive threatened to walk away from buying the business if it refused.

Original Submission #1  Original Submission #2 

upstart writes:

Ad-block developers fear end is near for their extensions:

Seven months from now, assuming all goes as planned, Google Chrome will drop support for its legacy extension platform, known as Manifest v2 (Mv2). This is significant if you use a browser extension to, for instance, filter out certain kinds of content and safeguard your privacy.

Google's Chrome Web Store is supposed to stop accepting Mv2 extension submissions sometime this month. As of January 2023, Chrome will stop running extensions created using Mv2, with limited exceptions for enterprise versions of Chrome operating under corporate policy. And by June 2023, even enterprise versions of Chrome will prevent Mv2 extensions from running.

The anticipated result will be fewer extensions and less innovation, according to several extension developers.

Browser extensions such as Ghostery Privacy Ad Blocker, uBlock Origin, and Privacy Badger, along with scripting extensions including TamperMonkey, which are each designed to block adverts and other content and/or protect one's privacy online, are expected to function less effectively, if they can even make the transition from Mv2 to the new approach: Manifest v3.

Freeman writes:

https://arstechnica.com/information-technology/2022/06/microsoft-wont-say-if-it-will-patch-critical-windows-vulnerability-under-exploit/

CERT Ukraine also said it was tracking exploits on targets in that country that use email to send a file titled "changes in wages with accruals.docx" to exploit Follina.
[...]
The simple act of the document appearing in the preview window, even while protected view is turned on, is enough to execute malicious scripts.

"It's more serious because it doesn't matter if macros are disabled and it can be invoked simply through preview," Jake Williams, director of cyber threat intelligence at the security firm Scythe, wrote in a text chat. "It's not zero-click like a 'just delivering it causes the exploit' but the user need not open the document."

Researchers developing an exploit module for the Metasploit hacking framework referred to this behavior as a low-interaction remote code execution. "I was able to test this using both the .docx and rtf formats," one of them wrote. "I was able to gain execution with the RTF file by just previewing the document in Explorer."

Original Submission

upstart writes:

HP Dev One is the first non-System76 computer offered with Pop!_OS:

HP released its Dev One Linux laptop today. Aimed at coders, the 14-inch clamshell comes at a lower price than previous Ubuntu-based HP clamshells.

Starting at $1,099, the Dev One begins to keep costs low by opting for an AMD, rather than Intel, CPU and skipping the discrete graphics card. HP's last Linux laptops, part of its ZBook workstation lineup, went well over $2,000 and offered up to Intel Xeon processors and Nvidia RTX GPUs.

[...] The previous workstations used Ubuntu 20.04 preloaded with software packages aimed at data scientists. However, the Dev One runs Pop!_OS, an Ubuntu-based Linux distribution from System76.

System76 also makes its own laptops, desktops, servers, and the Launch mechanical keyboard. HP's Dev One marks the first laptop to run Pop!_OS without "System76" stamped on the lid—although, you can download Pop!_OS and install it on your own system.

upstart writes:

All patients treated with the drug are in remission, a result that seems unprecedented in cancer research:

A cancer treatment has shown astounding results in a small clinical trial. All of the treated patients, who had a specific form of mid-stage rectal cancer, have since experienced complete remission. Though the findings are based on a sample size of just 18 people, they could hold important implications for treating these particular cancers.

The trial enrolled volunteers diagnosed with stage II or III rectal cancer, meaning their tumors had begun to grow larger and spread to nearby parts of the body. Their cancer was also determined to be caused by a particular mechanism known as a deficiency in mismatch repair.

The researchers theorized that their treatment, a lab-made antibody called dostarlimab, might be able to help this subset of patients. It works by inhibiting a protein known as programmed death receptor-1 (PD-1) found in many cancer cells. This inhibition then allows the immune system to recognize the cancer cells as harmful and target them for destruction. [...]

That said, the findings are still very early, and it will take much more research with larger-sized studies to confirm the drug's effectiveness, especially over the long term. [...] And the drug doesn't come cheap, costing $11,000 per dose out of pocket.

At $11,000/pill, from some halfway house in an undisclosed location, Pharma Bro smiles with approval.

Journal Reference:
Andrea Cercek et al., PD-1 Blockade in Mismatch Repair–Deficient, Locally Advanced Rectal Cancer, NEJM, 2022. DOI: 10.1056/NEJMoa2201445

Original Submission

upstart writes:

Microsoft could finally kill HDD boot drives for good:

Microsoft could have plans to scrap its use of hard disk drives (HDD) among its main storage components on PCs running Windows 11, according to a recent report by industry analyst firm Trendfocus, as reported by Tom's Hardware.

If Microsoft goes through with its plans, consumers could begin to see solid-state hard drives (SSD) instead, with the exception of dual-drive desktop PCs and gaming laptops, which require multiple types of storage, as Tom's Hardware noted.

While Microsoft has declined to comment on the matter, the current trends indicate a complete market transition to SSD by 2023. Many PC makers already use SSD as their main storage option; however, it is still not a set standard, especially in emerging markets.

[...] Trendfocus Vice President John Chen told Tom's Hardware that 2023 is still not a hard date for the transition to SSD. Some suggestions considered in talks with Microsoft include holding off the transition of emerging markets until 2024 or pausing the desktop switch until that time.

Original Submission

upstart writes:

The space telescope's mirror was hit by a micrometeoroid, but it's still in great shape to carry out its mission:

NASA said today that one of the Webb Space Telescope's primary mirror segments was hit by a micrometeoroid, a small asteroid fragment, between May 23 and May 25. Initial assessments of the telescope found that the spacecraft was still performing exceptionally well, though the effects of the impact were noticeable in recent data readouts.

Micrometeoroids are extremely small (dust-sized), fast-moving space debris. They're a regular part of a hostile space environment that will bombard the Webb telescope throughout its years in operation.

"With Webb's mirrors exposed to space, we expected that occasional micrometeoroid impacts would gracefully degrade telescope performance over time," said Lee Feinberg, Webb optical telescope element manager at NASA, in an agency release. "Since launch, we have had four smaller measurable micrometeoroid strikes that were consistent with expectations and this one more recently that is larger than our degradation predictions assumed."

[...] Thankfully, Paul Geithner, a technical deputy project manager at NASA, said in the release that "We designed and built Webb with performance margin – optical, thermal, electrical, mechanical – to ensure it can perform its ambitious science mission even after many years in space."

See also: Webb: Engineered to Endure Micrometeoroid Impacts

Original Submission

DannyB writes:

Company That Turned 'Excited Delirium' Into A Thing Thinks It Can Prevent School Shootings With Drone-Mounted Tasers

Axon — the company that crafted a pseudo-scientific form of plausible deniability for cops who've killed people — now wants to modify the ever-popular (and patently ridiculous) maxim "The only person who can stop a bad guy with a gun is a good guy with a gun."

There were plenty of "good" guys with guns present at the last major school shooting. They did nothing to stop the killing. Instead, they huddled a safe distance away until another law enforcement agency showed up to actually stop the school shooter.

If law enforcement can't handle mass shootings quickly and competently (and agencies have given us little indication that they can), they certainly shouldn't be entrusted with an airborne weapon now being pushed by a company that sees shootings like the one in Uvalde, Texas as another way to bump year-over-year sales increases.

Here's the EFF's take on this announcement of Axon's armed drone proposal:

Taser and surveillance vendor Axon has proposed what it claims to be the solution to the epidemic of school shootings in the United States: a remote-controlled flying drone armed with a taser. For many many reasons, this is a dangerous idea. Armed drones would mission-creep their way into more every-day policing. We must oppose a process of normalizing the arming of drones and robots.

Here's Axon's far more cheery take on the addition of Taser devices to drones:

Put together, these two technologies may effectively combat mass shootings. In brief, non-lethal drones can be installed in schools and other venues and play the same role that sprinklers and other fire suppression tools do for firefighters: Preventing a catastrophic event, or at least mitigating its worst effects.

upstart writes:

Economist says China must seize TSMC if the US tightens sanctions:

The importance of the Taiwan Semiconductor Manufacturing Company, better known as TSMC, to the global economy cannot be understated. The world's largest chip manufacturer has a 54% share of the global chip market, which makes an economist's call for China to seize TSMC if the US imposes harsh sanctions on the country all the more concerning.

Bloomberg reports that Chen Wenling, chief economist at the government-run China Center for International Economic Exchanges, said, "If the US and the West impose destructive sanctions on China like sanctions against Russia, we must recover Taiwan."

Chen singled out TSMC in the speech at the Chongyang Institute for Financial Studies at Renmin University last month. "Especially in the reconstruction of the industrial chain and supply chain, we must seize TSMC," she added.

TSMC is reportedly set to build six chip fabs in the US, though it has announced just one so far. "They [TSMC] are speeding up the transfer to the US to build six factories there," Chen said. "We must not let all the goals of the transfer be achieved."

Chen does talk about the scenario only taking place if the US hit China with Russia-like sanctions, which were put in place after it invaded Ukraine. Taiwan has long said it is an independent nation, while China insists it is part of its territory and has no qualms about using force to bring it under control.

Original Submission

upstart writes:

NASA and SpaceX stand down on Dragon launch to study hydrazine issue:

NASA and SpaceX have delayed the launch of a Cargo Dragon spacecraft for at least a couple of weeks due to an issue during the prelaunch loading of hypergolic propellants.

The space agency had been planning to launch the spacecraft on June 12 but announced the delay in an email on Monday evening to reporters.

"During propellant loading of the Cargo Dragon spacecraft, elevated vapor readings of mono-methyl hydrazine were measured in an isolated region of the Draco thruster propulsion system," the space agency's statement said. "The propellant and oxidizer have been offloaded from that region to support further inspections and testing."

Draco thrusters provide on-orbit maneuvering propulsion for the Dragon spacecraft. NASA said that it is working with SpaceX to identify the source of the elevated readings and take any corrective actions. On Tuesday morning, astronauts on board the International Space Station were told by Mission Control in Houston that the launch date would slip until at least June 28.

This is not a new Dragon vehicle. Designated Dragon "C208," this vehicle has previously flown two supply missions, both in 2021. It is an upgraded version of the original Cargo Dragon spacecraft, known as "Cargo Dragon 2."

See also: Dragon Mission on Hold as Astronauts Conduct Eye Exams, Spacesuit Work

Original Submission

upstart writes:

Technology is alienating people:

We take it for granted that technology brings people closer together and improves our access to essential products and services. If you can't imagine life without your smartphone, it's easy to forget that people who can't or don't want to engage with the latest technology are being left behind.

For example, there have recently been reports that cashless payment systems for car parking in the UK are seeing older drivers unfairly hit with fines. This has led to calls for the government to intervene.

Age is one of the biggest predictors of digital exclusion. Only 47% of those aged 75 and over use the internet regularly. And out of the 4 million who have never used the internet in the UK, only 300,000 people are under 55.

But older people are not the only ones who feel shut out by new technology. For example, research shows vulnerable people, such as those with disabilities, are also disengaging with e-services and being "locked out" of society.

From train tickets to vaccine passports, there is a growing expectation that consumers should embrace technology to participate in everyday life. This is a global phenomenon. Out in front, Sweden predicts its economy will be fully cashless by March 2023.

Shops increasingly use QR codes, virtual reality window displays and self-service checkouts. Many of these systems require a smart device, and momentum is building for QR codes to be integrated into digital price tags as they can give customers extra information such as nutritional content of food. Changing paper labels is a labour intensive process.

upstart writes:

AI Trained on 4Chan Becomes 'Hate Speech Machine':

AI researcher and YouTuber Yannic Kilcher trained an AI using 3.3 million threads from 4chan's infamously toxic Politically Incorrect /pol/ board. He then unleashed the bot back onto 4chan with predictable results—the AI was just as vile as the posts it was trained on, spouting racial slurs and engaging with antisemitic threads. After Kilcher posted his video and a copy of the program to Hugging Face, a kind of GitHub for AI, ethicists and researchers in the AI field expressed concern.

The bot, which Kilcher called GPT-4chan, "the most horrible model on the internet"—a reference to GPT-3, a language model developed by Open AI that uses deep learning to produce text—was shockingly effective and replicated the tone and feel of 4chan posts. "The model was good in a terrible sense," Klicher said in a video about the project. "It perfectly encapsulated the mix of offensiveness, nihilism, trolling, and deep distrust of any information whatsoever that permeates most posts on /pol."

According to Kilcher's video, he activated nine instances of the bot and allowed them to post for 24 hours on /pol/. In that time, the bots posted around 15,000 times. This was "more than 10 percent of all posts made on the politically incorrect board that day," Kilcher said in his video about the project.

AI researchers viewed Kilcher's video as more than just a YouTube prank. For them, it was an unethical experiment using AI. "This experiment would never pass a human research #ethics board," Lauren Oakden-Rayner, the director of Research at the NeuroRehab Allied Health Network in Australia, said in a Twitter thread.

"Open science and software are wonderful principles but must be balanced against potential harm," she said. "Medical research has a strong ethics culture because we have an awful history of causing harm to people, usually from disempowered groups...he performed human experiments without informing users, without consent or oversight. This breaches every principle of human research ethics."

Just because something can be done doesn't mean it should be done. What are you views? Is this a harmless prank, a justified experiment or something potentially more sinister?

Original Submission