Rest of World reports on viral teenage pranks at conveyor-belt sushi chain restaurants across Japan, which snowballed into a societal phenomenon that social media users and the Japanese press have named "sushi terrorism."

It began January 9th when a video showed a customer adding a pile of wasabi onto sushi on a conveyor belt. Another video shows a giggling teenager touching sushi on a conveyor belt at the sushi chain Sushiro after first licking that finger. The stock of the parent company that owns that sushi chain drops nearly 5%. It's not over. At a Nagoya branch of Kura Sushi, a 21-year-old customer grabs sushi from the conveyor belt, cramming it into his mouth and chasing it down with a swig from the communal soy sauce bottle. The incident is filmed by his two younger friends, one of whom posts the clip online. The same day, Sushiro's operating company announces it will limit conveyor belts and move to ordering by touch screen.
Concerns continued at other sushi chains. ("Kura Sushi says it's installing surveillance cameras equipped with AI to monitor customers' behavior and catch sushi terrorists. A day later, Choushimaru announces it will switch entirely to an iPad-based ordering system by April 26.") Sushiro also moves to ordering by touch screen and promises to limit conveyor belts.

The story's dramatic conclusion? Nagoya police arrest the 19-year-old man who allegedly posted the soy-sauce-swigging video from Kura Sushi, along with his two "co-conspirators." Nagoya police declare they are holding all three sushi terrorists on suspicion of "forcible obstruction of business." The crime would carry a maximum penalty of three years in prison, if they're convicted.

Axios remembers that it was just nine days ago that there were "concerns" about Silicon Valley Bank at venture capital firm. And soon Founders Fund's top operations executives "were on the phone, quickly deciding to move firm capital to a number of bigger banks." Firm founder Peter Thiel was not part of the conversation. One source says that the assumption was that they'd return the money to SVB after the crisis had ended....

Founders Fund wasn't the only venture capital firm giving this sort of warning to portfolio companies, nor necessarily the first, but word of its advice spread like wildfire (it also ended up in media reports, including one from Axios). Almost immediately, the firm came under withering criticism from some other venture capitalists, accusing Founders Fund (and Thiel personally) of sparking a bank run that ultimately led to $42 billion in withdrawals. Some even speculated that it was intentional, as payback for some unknown grudge between the two groups.
Earlier this week the Washington Examiner chronicled some of that criticism: "There should be more scrutiny of Peter Thiel and [hedge fund manager] Bill Ackman for yelling fire in a crowded theater in this SVB collapse," tweeted CNBC host Sara Eisen [on Monday]. Others turned their focus to Thiel's promotion and subsequent profiting off of crypto investments after the market crashed as a reason to be suspicious of his withdrawals. "You mean the guy who was touting crypto and trashing critics while he was selling crypto? That guy? Shocker!" tweeted tech journalist Kara Swisher.
But Peter Thiel says he actually left his own money in the bank, reports Business Insider: "I had $50 million of my own money stuck in SVB," he told the Financial Times.... Thiel told the Financial Times that he did not believe the SVB would fail last week.

Other venture capital firms — including Coatue Management, Union Square Ventures, and Founder Collective — had similarly advised startup clients to transfer money from SVB after the bank revealed a $1.8 billion loss and the bank's share price collapsed. These firms have pushed back against accusations that they were spreading panic, saying that they were giving financial advice they believed would be in the best interest of their clients....

Thiel told the Financial Times that his account was frozen on Friday when regulators stepped in and took control of the bank. However, it is once again accessible after the US government stepped in earlier this week and shored up all customer deposits in SVB.

BGR reports: Elon Musk seems to be close to making good on his promise to open-source Twitter's code. Well, at least part of it.

In a post on the social media platform, the Twitter CEO announced that the company will open-source the code used to recommend tweets on March 31. Musk did not provide any other details about how that will work or specifically when on that date the code will be provided.

Musk has been teasing and promising open-sourcing Twitter's code for a while now, so it'll be interesting to see what the impact is...

Microsoft appears to be testing a built-in cryptocurrency wallet for Edge, according to screenshots pulled from a beta build of the browser. Ars Technica reports: The feature, which the screenshots say is strictly for internal testing, was unearthed by Twitter user @thebookisclosed, who has a history of digging up present-but-disabled features in everything from new Windows 11 builds to ancient Windows Vista betas. According to the screenshots, the crypto wallet is "embedded in Edge, making it easy to use without installing any extension," and it can handle multiple types of cryptocurrency. It will also record transactions and the value of your individual currencies as they fluctuate. An "explore" tab offers news stories relevant to cryptocurrency, and an "assets" tab will let you stare lovingly at your NFTs. The wallet is "non-custodial" (also called "self-custodial"), meaning that you have sole ownership of and responsibility for the passwords and recovery keys that allow access to your funds. Microsoft won't be able to let you back in if you lose your credentials.

An anonymous reader quotes a report from The Register: A cancer patient whose nude medical photos and records were posted online after they were stolen by a ransomware gang, has sued her healthcare provider for allowing the "preventable" and "seriously damaging" leak. The proposed class-action lawsuit stems from a February intrusion during which malware crew BlackCat (also known as ALPHV) broke into one of the Lehigh Valley Health Network (LVHN) physician's networks, stole images of patients undergoing radiation oncology treatment along with other sensitive health records belonging to more than 75,000 people, and then demanded a ransom payment to decrypt the files and prevent it from posting the health data online. The Pennsylvania health care group, one of the largest in the US state, oversees 13 hospitals, 28 health centers, and dozens of other physicians' clinics, pharmacies, rehab centers, imaging and lab services. LVHN refused to pay the ransom, and earlier this month BlackCat started leaking patient info, including images of at least two breast cancer patients, naked from the waist up.

According to the lawsuit [PDF] filed this week, here's how one of the patients, identified as "Jane Doe" found out about the data breach -- and that LVHN had stored nude images of her on its network in the first place. On March 6, LVHN VP of Compliance Mary Ann LaRock, called Doe and told her that her nude photos had been posted on the hackers' leak site. "Ms. LaRock offered plaintiff an apology, and with a chuckle, two years of credit monitoring," the court documents say. In addition to swiping the very sensitive photos, the crooks also made off with everything needed for identity fraud.

According to the lawsuit, LaRock also told Doe that her physical and email addresses, along with date of birth, social security number, health insurance provider, medical diagnosis and treatment information, and lab results were also likely stolen in the breach. "Given that LVHN is and was storing the sensitive information of plaintiff and the class, including nude photographs of plaintiff receiving sensitive cancer treatment, LVHN knew or should have known of the serious risk and harm that could occur from a data breach," the lawsuit says. It claims LVHN was negligent in its duty to safeguard patients' sensitive information, and seeks class action status for everyone whose data was exposed with monetary damages to be determined. Pennsylvania attorney Patrick Howard, who is representing Doe and the rest of the plaintiffs in the proposed class action, said he expects the number of patients affected by the breach to be in the "hundreds, if not thousands."

Midjourney announced version 5 of its commercial AI image-synthesis service, which can produce photorealistic images at a quality level that some AI art fans are calling creepy and "too perfect." From a report: Midjourney v5 is available now as an alpha test for customers who subscribe to the Midjourney service, which is available through Discord. "MJ v5 currently feels to me like finally getting glasses after ignoring bad eyesight for a little bit too long," said Julie Wieland, a graphic designer who often shares her Midjourney creations on Twitter. "Suddenly you see everything in 4k, it feels weirdly overwhelming but also amazing."

Wieland shared some of her Midjourney v5 generations with Ars Technica [images shared in the linked story], and they certainly show a progression in image detail since Midjourney first arrived in March 2022. Version 3 debuted in August, and version 4 debuted in November. Midjourney works similarly to image synthesizers like Stable Diffusion and DALL-E in that it generates images based on text descriptions called "prompts" using an AI model trained on millions of works of human-made art. Recently, Midjourney was at the heart of a copyright controversy regarding a comic book that used earlier versions of the service.

Meta on Friday launched its subscription service in the U.S., which would allow Facebook and Instagram users pay for verification in the same vein as Elon Musk-owned Twitter. From a report: The Meta Verified service will give users a blue badge after they verify their accounts using a government ID and will cost $11.99 per month on the web or $14.99 a month on Apple's iOS system and Google-owned Android, Meta said in a statement. The service, which Meta said it was testing in February, follows in the footsteps of Snapchat as well as messaging app Telegram and marks the latest effort by a social media company to diversify its revenue away from advertising.

The US Copyright Office has issued (PDF) guidance today to clarify when AI-generated material can be copyrighted. Ars Technica reports: Guidance comes after the Copyright Office decided that an author could not copyright individual AI images used to illustrate a comic book, because each image was generated by Midjourney -- not a human artist. In making its decision, the Copyright Office committed to upholding the longstanding legal definition that authors of creative works must be human to register works. Because of this, officials confirmed that AI technologies can never be considered authors. This wasn't the only case influencing new guidance, but it was the most recent. Wrestling with the comic book's complex authorship questions helped prompt the Copyright Office to launch an agency-wide initiative to continue exploring a wider range of copyright issues arising as the AI models that are used to generate text, art, audio, and video continue evolving.

Perhaps the most significant aspect of the guidance is an author's "duty to disclose the inclusion of AI-generated content in a work submitted for registration." When registering works, authors must distinguish which content is human-authored and which content is AI-generated. If applicants aren't sure how to refer to the AI-generated content, the Copyright Office recommends providing a general statement that the work contains AI-generated content. That will prompt the office to follow up to help each author fill in the blanks in an application.

For artists who have pending applications or have already registered works that contain AI-generated content, the Copyright Office suggests correcting the public record by submitting a supplementary registration. Any failure to accurately reflect the role of AI in copyrighted works could result in "losing the benefits of the registration," the office warned. That could leave works vulnerable to being copied, with little or no legal recourse for copyright infringement claims. Failure to disclose AI-generated content is the only type of infringement discussed in the guidance. Critics like Alex J. Champandard, a co-founder of Creative.ai -- a group of hackers and artists interested in generative AI -- tweeted to say that current guidance puts authors in a precarious catch-22 situation. "By disclosing the AI, you're opening yourself up to infringement, but by not disclosing AI, it's safer but in violation of [the US Copyright Office]!" Champandard's tweet suggested.

T-Mobile is buying Mint Mobile, the budget-friendly mobile carrier that's partially owned by Ryan Reynolds. The Verge reports: In a post published on Wednesday, T-Mobile announced that the deal's valued at up to $1.35 billion and comes as T-Mobile looks to build out its prepaid phone offering. The acquisition should close later this year and involves a 39 percent cash and 61 percent stock purchase of Mint's parent company, Ka'ena Corporation. The price could change, however, as it depends on Mint's performance.

Once the deal closes, Mint founders David Glickman and Rizwan Kassim will join T-Mobile to continue managing the brand, which T-Mobile says "will generally operate as a separate business unit." Meanwhile, Reynolds will also remain a part of Mint's branding, as T-Mobile says he will "continue on in his creative role on behalf of Mint." "I never dreamt I'd own a wireless company and I certainly never dreamt I'd sell it to T-Mobile," Reynolds said in a tweet. "Life is strange and I'm incredibly proud and grateful."

As noted by T-Mobile CEO Mike Sievert in a YouTube video posted on Wednesday, T-Mobile will retain the price of Mint's cheapest phone plan at $15 per month. The deal also includes Ka'ena's other companies, including Ultra Mobile, a prepaid carrier that offers international calling, and Plum, a wholesale wireless solutions provider. [...] By purchasing Mint, T-Mobile may be looking to claw back the customers it lost when it sold Boost Mobile to Dish as part of its merger with Sprint.

An anonymous reader writes: Facebook parent company Meta, which emerged as an outspoken advocate of remote work during the pandemic, is encouraging employees to come back to the office. Some early analysis "suggests that engineers who either joined Meta in-person and then transferred to remote or remained in-person performed better on average than people who joined remotely," Chief Executive Officer Mark Zuckerberg said in a statement Tuesday. Zuckerberg cautioned that the data requires further study, but encouraged employees to "find more opportunities to work with your colleagues in person" in the meantime. In 2021, Facebook established a policy that allowed all employees to work remotely even after the pandemic if their jobs could be done outside of an office. Several big tech companies including Amazon, Apple and Twitter have been trying to get workers to return to the office.

An anonymous reader quotes a report from Ars Technica: On Friday, a software developer named Georgi Gerganov created a tool called "llama.cpp" that can run Meta's new GPT-3-class AI large language model, LLaMA, locally on a Mac laptop. Soon thereafter, people worked out how to run LLaMA on Windows as well. Then someone showed it running on a Pixel 6 phone, and next came a Raspberry Pi (albeit running very slowly). If this keeps up, we may be looking at a pocket-sized ChatGPT competitor before we know it. [...]

Typically, running GPT-3 requires several datacenter-class A100 GPUs (also, the weights for GPT-3 are not public), but LLaMA made waves because it could run on a single beefy consumer GPU. And now, with optimizations that reduce the model size using a technique called quantization, LLaMA can run on an M1 Mac or a lesser Nvidia consumer GPU. After obtaining the LLaMA weights ourselves, we followed [independent AI researcher Simon Willison's] instructions and got the 7B parameter version running on an M1 Macbook Air, and it runs at a reasonable rate of speed. You call it as a script on the command line with a prompt, and LLaMA does its best to complete it in a reasonable way.

There's still the question of how much the quantization affects the quality of the output. In our tests, LLaMA 7B trimmed down to 4-bit quantization was very impressive for running on a MacBook Air -- but still not on par with what you might expect from ChatGPT. It's entirely possible that better prompting techniques might generate better results. Also, optimizations and fine-tunings come quickly when everyone has their hands on the code and the weights -- even though LLaMA is still saddled with some fairly restrictive terms of use. The release of Alpaca today by Stanford proves that fine tuning (additional training with a specific goal in mind) can improve performance, and it's still early days after LLaMA's release. A step-by-step instruction guide for running LLaMA on a Mac can be found here (Warning: it's fairly technical).

Meta's head of commerce and financial technologies Stephane Kasriel posted on Twitter that the company will sunset its NFT and digital collectibles features on Instagram and Facebook. TechCrunch reports: This short-lived product only began testing with select Instagram creators last May, plus some Facebook users in June. By July, Meta expanded NFT support on Instagram for creators in 100 countries. Less than a year later, Meta is moving on from NFTs. "We're winding down digital collectibles (NFTs) for now to focus on other ways to support creators, people, and businesses," Kasriel wrote in a Twitter thread.

A Meta spokesperson told TechCrunch that it is shifting its investments away from NFTs toward products like Meta Pay, as well as features that enable creators to earn money directly on Meta platforms, like its tipping feature called gifts. The company also said it is testing ways for creators to earn ad revenue on Reels. "Let me be clear: creating opportunities for creators and businesses to connect with their fans and monetize remains a priority, and we're going to focus on areas where we can make impact at scale, such as messaging and monetization opps for Reels," Kasriel wrote.

"Meta, the parent firm of Facebook and Instagram, is working on a standalone, text-based social network app," reports the BBC.
BR> "It could rival both Twitter and its decentralised competitor, Mastodon." A spokesperson told the BBC: "We're exploring a standalone decentralized social network for sharing text updates...." According to MoneyControl, the new app is codenamed P92, and will allow users to log in through their existing Instagram credentials.

Meta's app will be based on a similar framework to the one that powers Mastodon, a Twitter-like service which was launched in 2016. The new app would be decentralised — it cannot be run at the whim of a single entity, bought or sold....

It was not immediately clear when Meta would roll out the new app.

Breaking news from CNN: Treasury Secretary Janet Yellen on Sunday instructed the Federal Deposit Insurance Corporation to guarantee Silicon Valley Bank customers will have access to all of their money starting Monday.

By guaranteeing all deposits — even the uninsured money customers kept with the failed SVB bank — the government can ensure public confidence in America's banking system, said Yellen, Federal Reserve Chair Jerome Powell and FDIC Chairman Martin J. Gruenberg in a joint statement....

The FDIC opened an auction Sunday for bids to acquire the bank, the Treasury Department said in a briefing with lawmakers in the California delegation, two sources familiar with the briefing told CNN.... Under Secretary for Domestic Finance Nellie Liang and Assistant Secretary for Legislative Affairs Jonathan Davidson led the briefing, during which they told members that the FDIC is prepared "to operate the institution" to ensure depositors can maintain payroll for their employees and that more operations will emerge in coming days, one of the sources said.
The treasury secretary's statement clarified that "No losses associated with the resolution of Silicon Valley Bank will be borne by the taxpayer." We are also announcing a similar systemic risk exception for Signature Bank, New York, New York, which was closed today by its state chartering authority. All depositors of this institution will be made whole. As with the resolution of Silicon Valley Bank, no losses will be borne by the taxpayer. Shareholders and certain unsecured debtholders will not be protected. Senior management has also been removed. Any losses to the Deposit Insurance Fund to support uninsured depositors will be recovered by a special assessment on banks, as required by law.

Finally, the Federal Reserve Board on Sunday announced it will make available additional funding to eligible depository institutions to help assure banks have the ability to meet the needs of all their depositors.
Meanwhile, congresswoman Nancy Pelosi said there are multiple potential buyers for SVB, and "What we would hope to see by tomorrow morning is for some other bank to buy the bank." The UK arm of the bank has already received a bid from the Bank of London.

From the treasury secretary's statement: The U.S. banking system remains resilient and on a solid foundation, in large part due to reforms that were made after the financial crisis that ensured better safeguards for the banking industry.

Those reforms combined with today's actions demonstrate our commitment to take the necessary steps to ensure that depositors' savings remain safe.

"Based on initial data review, vehicle is healthy," Relativity Space tweeted today. "More info to follow on cause of aborts today. Thanks for playing."

Remaining back on the launchpad is the largest 3D printed object ever to exist. And they're still hoping to launch it into space.

They'd planned a launch this morning from Cape Canaveral, Florida of a 110-foot rocket (33.5 meters) on a mission they're calling GLHF — "Good Luck, Have Fun".

The rocket's makers — California-based Relativity Space — call it "the world's first 3D printed rocket." A full 85% of the rocket's weight comes from 3D printed parts, explains Wired, and "only the computing system, electronics, and readily available parts like fasteners were not." Named Terran 1, the 7.5-foot-wide rocket (2.2 meters) inaugurates the company's ambitious plans for 3D printing in space: Relativity Space wants to use Terran 1 to (comparatively) cheaply lift satellites for other companies and NASA into Earth orbit. It also plans to construct Terran R, a larger, more powerful, fully reusable rocket that the company hopes will compete with SpaceX's Falcon 9, which has a smaller payload capacity and only reuses the rocket's first stage. In late 2024, Relativity plans to test using Terran R to launch payloads to Mars; another startup, Impulse Space, will provide the lander.
From the company's web site: Like its structure, all Relativity engines are 3D printed and use liquid oxygen and liquid natural gas, which are not only the best for rocket propulsion, but also for reusability, and the easiest to eventually transition to methane on Mars.
The tagline for the company's Twitter feed says they're "Building humanity's multiplanetary future." And excitement is running high, reports Spaceflight Now" "There are a number of firsts here potentially on this rocket," said Josh Brost, vice president of revenue operations at Relativity Space....

"Hard to believe the day is nearly here to launch Terran 1, our first rocket!" Tim Ellis, co-founder and CEO of Relativity Space tweeted Tuesday....

The company now boasts some 1,000 employees, a million-square-foot headquarters and factory in Long Beach, California, and $1.3 billion in venture capital and equity fundraising, including an early $500,000 investment from billionaire Mark Cuban. In 2021, the company reached a valuation of $4.2 billion before launching any rockets....

"No new company has ever had their liquid rocket make it to space on their first attempt," Brost, also a former engineer and manager at SpaceX, told Spaceflight Now in a pre-launch interview. "So if everything goes incredibly well, and we achieve orbit on our first launch ... that would be a remarkable milestone for us, which we would be, of course, over the moon excited about. But that doesn't define success for us."
Wired adds that they're not the only company working on space-related 3D printing: Australia's Fleet Space has already been producing lightweight, 3D-printed radio frequency antennas for satellites. Next year, using printers half the size of a bus, they plan to create a satellite constellation called Alpha that will be entirely 3D-printed.... Flavia Tata Nardini, the company's CEO, believes space-based 3D printing is coming next. "In my ideal future, in 10 to 15 years, I won't have to launch satellites from here; I can build them up there."

Krisztian Sandor writes via CoinDesk: U.S.-based stablecoin issuer Circle held a part of its USDC stablecoin's cash reserves at Silicon Valley Bank as of Jan. 17, according to the firm's latest attestation (PDF). USDC is the second-largest stablecoin on the market, with a $43 billion circulating supply that is fully backed by government bonds and cash-like assets. According to Circle's January reserve report, the firm held some $9.88 billion of cash deposited at regulated banks to back USDC's value. USDC's banking partners included Silicon Valley Bank (SVB), the California-based bank that was taken over by regulators and shut down on Friday. The full list of banks that held cash for Circle's USDC are Bank of New York Mellon, Citizens Trust Bank, Customers Bank, New York Community Bank (a division of Flagstar Bank, N.A.), Signature Bank, Silicon Valley Bank and Silvergate Bank. Circle also keeps some part of USDC reserves in a dedicated BlackRock fund.

Circle said last week it had cut ties with Silvergate Bank, the crypto-friendly bank that halted operations and said it would "voluntarily liquidate" its assets earlier this week. Signature Bank's holding company's (SI) shares have dropped 12% on the news about SVB's shutdown. Signature said in December that it would reduce deposits tied to crypto firms by as much as $10 billion. Simon Dixon, CEO of online investment platform BnkToTheFuture, tweeted that Circle's chief executive Jeremy Allaire said the firm held "most of their cash is in BNY Melon," while sharing a screenshot from March 2. BnkToTheFuture is an investor and shareholder in Circle.

Twitter's decline is paving the way for other platforms to build next-generation replacements. And now the biggest player in the game is getting involved: Meta is in the early stages of building a dedicated app for people to post text-based updates. From a report: "We're exploring a standalone decentralized social network for sharing text updates," the company told Platformer exclusively in an email. "We believe there's an opportunity for a separate space where creators and public figures can share timely updates about their interests." News that Meta has been exploring a text-based network was first reported Thursday by MoneyControl. The app is codenamed P92 and will allow users to log in through their existing Instagram credentials, the outlet reported.

Details about the project are scant. The product is still in its earliest stages, sources said, and there is no time frame for it being released. But legal and regulatory teams have already started to investigate potential privacy concerns around the app so they can be addressed before launch, we're told. Adam Mosseri, who runs Instagram, is taking the lead on the project, sources said. The most remarkable aspect of the project is that Meta plans for the network to be decentralized. While the company would not elaborate beyond its statement, in a decentralized network individual users are typically able to set up their own, independent servers and set server-specific rules for how content is moderated. Building a decentralized network could also give Meta the opportunity for its new app to interoperate with other social products -- a previously unheard-of gesture from a company known for building some of the most lucrative walled gardens in the industry's history.

An anonymous reader quotes a report from Ars Technica: HP customers are showing frustration online as the vendor continues to use firmware updates to discourage or, as users report, outright block the use of non-HP-brand ink cartridges in HP printers. HP has already faced class-action lawsuits and bad publicity from "dynamic security," but that hasn't stopped the company from expanding the practice. Dynamic security is a feature used by HP printers to authenticate ink cartridges and prevent use of cartridges that aren't HP-approved. As the company explains: "Dynamic security relies on the printer's ability to communicate with the security chips or electronic circuitry on the cartridges. HP uses dynamic security measures to protect the quality of our customer experience, maintain the integrity of our printing systems, and protect our intellectual property. Dynamic security equipped printers are intended to work only with cartridges that have new or reused HP chips or electronic circuitry. The printers use the dynamic security measures to block cartridges using non-HP chips or modified or non-HP electronic circuitry. Reused, remanufactured, and refilled cartridges that reuse the HP chip or electronic circuitry are unaffected by dynamic security."

HP is set on continuing to use DRM to discourage its printer customers from spending ink and toner money outside of the HP family. "HP have updated their printers to outright ban 'non-HP' ink! They no longer shows the 'can't guarantee quality' message, but instead cancels your print completely until you inset a HP ink cartridge," Reddit user grhhull posted Tuesday. "After contacting HP, they advised 'this is due to the recent 'update' of all printers.'" It's unclear when HP issued updates for which model printers, but there are alleged customer complaints online stemming from late last year, showing plenty of customers surprised their printer no longer worked with non-HP ink cartridges after an update. Some pointed to third-party brands they had relied on for years.

HP community support threads include complaints about the OfficeJet 7740 and OfficeJet Pro 6970. HP lists both printers, as well as others, as able to circumnavigate dynamic security under specific conditions. However, HP's support page states this only applies to models manufactured before December 1, 2016. For more examples, there are comments on HP's support community suggesting that HP's OfficeJet 6978 and 6968 were recently affected. Both printers are discontinued, but HP's product pages make it clear that the fickle nature of dynamic security means that third-party ink could stop working at any time. And HP's dynamic security page also leaves the door open for the sudden bricking of functioning ink: "Firmware updates delivered periodically over the internet will maintain the effectiveness of the dynamic security measures," the page reads. "Updates can improve, enhance, or extend the printer's functionality and features, protect against security threats, and serve other purposes, but these updates can also block cartridges using a non-HP chip or modified or non-HP circuitry from working in the printer, including cartridges that work today."

An anonymous reader quotes a report from Ars Technica: On Monday, a group of AI researchers from Google and the Technical University of Berlin unveiled PaLM-E, a multimodal embodied visual-language model (VLM) with 562 billion parameters that integrates vision and language for robotic control. They claim it is the largest VLM ever developed and that it can perform a variety of tasks without the need for retraining. According to Google, when given a high-level command, such as "bring me the rice chips from the drawer," PaLM-E can generate a plan of action for a mobile robot platform with an arm (developed by Google Robotics) and execute the actions by itself.

PaLM-E does this by analyzing data from the robot's camera without needing a pre-processed scene representation. This eliminates the need for a human to pre-process or annotate the data and allows for more autonomous robotic control. It's also resilient and can react to its environment. For example, the PaLM-E model can guide a robot to get a chip bag from a kitchen -- and with PaLM-E integrated into the control loop, it becomes resistant to interruptions that might occur during the task. In a video example, a researcher grabs the chips from the robot and moves them, but the robot locates the chips and grabs them again. In another example, the same PaLM-E model autonomously controls a robot through tasks with complex sequences that previously required human guidance. Google's research paper explains (PDF) how PaLM-E turns instructions into actions.

PaLM-E is a next-token predictor, and it's called "PaLM-E" because it's based on Google's existing large language model (LLM) called "PaLM" (which is similar to the technology behind ChatGPT). Google has made PaLM "embodied" by adding sensory information and robotic control. Since it's based on a language model, PaLM-E takes continuous observations, like images or sensor data, and encodes them into a sequence of vectors that are the same size as language tokens. This allows the model to "understand" the sensory information in the same way it processes language. In addition to the RT-1 robotics transformer, PaLM-E draws from Google's previous work on ViT-22B, a vision transformer model revealed in February. ViT-22B has been trained on various visual tasks, such as image classification, object detection, semantic segmentation, and image captioning.

An anonymous reader quotes a report from Bloomberg: The US can stop Twitter from releasing details about the government's demands for user information in national security investigations, a court ruled (PDF), in the same week House Republicans are to grill national security officials over surveillance. Twitter had protested the government's redactions to a 2014 "transparency report" that featured a numerical breakdown of national security-related data requests from the previous year. The US appeals court in San Francisco on Monday agreed with a lower-court judge that the Justice Department had shown a "compelling" interest in keeping that information secret. Based on classified and unclassified declarations provided by government officials, the court was "able to appreciate why Twitter's proposed disclosure would risk making our foreign adversaries aware of what is being surveilled and what is not being surveilled -- if anything at all," US Circuit Judge Daniel Bress wrote for the three-judge panel.

Although the case is almost a decade old, the ruling comes just as lawmakers and US national security agencies gear up for a bruising fight over making changes to a key surveillance program. Section 702 of the Foreign Intelligence Surveillance Act, described by intelligence officials as a key authority, expires on Dec. 31 unless Congress votes to renew it. US agencies use the authority to compel internet and technology companies to turn over information about suspected foreign terrorists and spies. Changes to Section 702 could include altering what companies like Twitter are required to do in response to government demands. "The case at issue in Monday's decision involved efforts by Twitter to share information about two types of federal law enforcement demands on the social media company: 'national security letters' for subscriber information, which would cover metadata but not the substance of any electronic communications, and orders under FISA, which could include content," adds Bloomberg.

Judge Daniel Bress wrote: "The government may not fend off every First Amendment challenge by invoking national security. But we must apply the First Amendment with due regard for the government's compelling interest in securing the safety of our country and its people."