Feed aggregator

joshuark quotes a report from MIT Technology Review: It is impossible to make large language models fully secure against hacks because of a fundamental flaw in how they work, a team of researchers argue in a paper presented at the International Conference on Machine Learning, a top AI conference, this month. The claim has huge implications for the safety of this technology. By taking advantage of this flaw, which concerns how LLMs identify who or what is giving them instructions, the researchers were able to make popular LLMs spit out information they had been trained not to provide, such as how to synthesize cocaine and how to sabotage a commercial aircraft's navigation system. "There's a real probability that this is going to be a problem that's fundamentally unsolvable," says Charles Ye, an independent researcher and coauthor of the ICML paper. [...] The ICML paper describes attacks against several of OpenAI's models, but Cui and Ye say that they have since seen similar results with models made by Anthropic, Alibaba, and DeepSeek. Cui and her colleagues wanted to find out why an attack like chain-of-thought forgery was so effective. They suspected it had something to do with the mechanism that LLMs use to keep track of where their instructions are coming from. But what Cui and her colleagues discovered is that LLMs are in fact very bad at keeping track of different roles. In a series of experiments that looked at what was going on inside a handful of different models, the researchers found that LLMs seem to identify the role of a specific chunk of text not by the tags around it but by the style of that text and the words it contains. The upshot, the researchers claim, is that all an attacker needs to do to hack an LLM is write text that spoofs a certain role. And because roles are a fundamental part of how LLMs work, no amount of training will fully solve the problem. "There's going to be a huge economic incentive for people to do jailbreaks and prompt injections," says Cui. The best defense could be to expect the worst. Organizations shouldn't trust LLMs, and they should expect that anything done by agents could be unsafe, he says: "That's not a great solution, but it just might be what we have to do." "It's really incredible that these things are being deployed everywhere to control super-critical systems. There's been no study of the fundamental science here. We're all doing it ad hoc."

Read more of this story at Slashdot.

Microsoft shares surged as much as 17% after reporting 43% growth in Azure revenue, putting the company on track to add a record $490 billion in market value in a single day. Bloomberg notes that it "would eclipse Nvidia's $440 billion addition, following President Donald Trump's announcement of a 90-day tariff pause last year, as the biggest ever." From the report: The nearly $500 billion jump is larger than the market capitalization of roughly 96% of S&P 500 stocks, data compiled by Bloomberg show. It's also bigger than the combined value of the benchmark's 44 smallest members, which includes companies like Domino's Pizza Inc., Clorox Co. and Hasbro Inc. Microsoft's one-day add in value also dwarfs many of the world's other equity markets. South Africa, Turkey, Finland and Vietnam all have total stock market values that are less than what the software maker is set to add on Thursday.

Read more of this story at Slashdot.

The Freeport store, which has been open for four decades, will close Aug. 21.

ABC accused FCC Chair Brendan Carr of "attempted censorship," arguing that an early review of its eight broadcast licenses is politically motivated retaliation over the network's coverage and could chill the entire media industry. "The retaliation against ABC is a signal to every media company in the country: accommodate the Administration's view of what news coverage should look like or pay the price," the Disney-owned television network wrote. Politico reports: "Across the government, regulatory and contracting carrots and sticks have been trained on other disfavored speakers," ABC's lawyers added in the 119-page filing. "The tools vary; the objective does not: a media industry too fearful of official reprisal to report the news freely." Carr has repeatedly rejected accusations of censorship while defending his efforts to rein in what he calls abusive, politically motivated behavior by licensed TV broadcasters. "I don't view the FCC as the speech police," he told POLITICO this week for an upcoming episode of the podcast "The Conversation." In its filing, ABC pointed to an outpouring of support it has received in more than 152,000 comments in the agency's license review, as well as recent warnings from conservative Supreme Court Justice Neil Gorsuch, Sen. Ted Cruz (R-Texas) and various pro-free-market organizations about the dangers of a politically motivated FCC. Those include letters from a bipartisan group of former FCC leaders, community and advocacy groups and lawmakers of both parties. "The Commission's attempted censorship of ABC has attracted condemnation across the political aisle," the network wrote.

Read more of this story at Slashdot.

An anonymous reader quotes a report from Reuters: Amazon's Zoox unit has won U.S. approval for limited commercial deployment of its novel steering-wheel-free robotaxis, a first for the autonomous ride industry, the U.S. auto safety agency said on Thursday. Zoox said that the National Highway Traffic Safety Administration's decision gives the company federal approval to begin charging for rides, and that it will soon begin charging for service, first in Las Vegas, with additional markets to follow as it completes various state requirements. [...] NHTSA Administrator Jonathan Morrison told Reuters that Zoox had received clearance to commercially deploy up to 2,500 vehicles in each of the next two years. Zoox currently carries passengers in parts of Las Vegas and San Francisco as part of testing. The exemption would allow Zoox to charge them fees, subject to state and local approvals. The agency said it determined the vehicle is as safe as an equivalent vehicle meeting federal motor vehicle safety standards that are being waived. Zoox cannot sell any of the vehicles to the public. "We can say pretty clearly that the systems in place on the Zoox exceed the equivalent performance requirements of a compliant vehicle," Morrison said in an interview. "But we still want to make sure that the automated driving system will operate appropriately." As part of the exemption, NHTSA is placing additional reporting requirements on Zoox for issues such as crashes or stopping inappropriately on roads, and the regulatory agency will adjust the conditions based on how the vehicles behave. "We have the ability to pull the exemption if we see major safety issues," Morrison said. All remote operators must be located in the United States, NHTSA said, and Zoox must publish maps of areas indicating where the vehicles are operating. Morrison said NHTSA expects to develop the first federal safety standards for automated driving systems by the end of the Trump administration. It is also proposing to overhaul some existing rules written with human drivers in mind such as requiring brake pedals and rear-view mirrors.

Read more of this story at Slashdot.

A UK parliamentary inquiry found that a catastrophic Ministry of Defense breach exposing 18,700 Afghans could have been prevented with basic Excel training, after an employee unknowingly shared a hidden worksheet containing their details. The Independent reports: The leak, in February 2022, exposed the details of 18,700 Afghans who said they were in danger from the Taliban because of their links to UK forces and now wanted to escape to Britain. The blunder triggered an unprecedented superinjunction used against the national media, including The Independent, and prompted a secret evacuation program -- the cost of which is still unclear but which likely ran into the billions of pounds. Following the revelation by this outlet and others in July last year of the hidden operation, MPs set up an inquiry to scrutinize what had happened. In their report, the defense selection committee concluded that: - The data breach could have been prevented if Ministry of Defense (MoD) personnel had received basic Excel training - By August 2023, when the department discovered the leak, thousands of people already knew that a significant data incident had taken place - The government did not strike "the right balance between operational secrecy and democratic accountability" -- and the superinjunction was in place for too long - Secrecy denied affected Afghans the chance to take steps to protect themselves and their families and caused delays to evacuation program - Thousands of Afghans eligible to come to Britain are still trapped in Afghanistan with the government failing to explain how they will help get families to safety. MPs have called on the government to publish periodic reassessments of the risks facing Afghan applicants to UK resettlement schemes, with officials to report findings annually. They also want ministers to publish a clear policy explaining how they will help Afghans who are eligible to come to Britain but who have not yet been evacuated. The defense committee have also called on the MoD to explain who was responsible for data protection risk before the Afghan breach, criticizing the lack of accountability within the civil service.

Read more of this story at Slashdot.

Here are the best back-to-school deals under $200 on comfy, ergonomic chairs for all-nighters and cram sessions

The temporary exemption from federal safety standards will let Amazon’s Zoox launch a real, paid robotaxi service in Las Vegas.

Google DeepMind's Gemini Robotics 2 combines vision, language, and action models to control multiple types of robots, including humanoids performing tasks such as organizing shelves, tying bags, and replacing lightbulbs. "It's another milestone in our path towards really getting towards what we call like physical AGI, which means we get a robot to do anything that a human can," Carolina Parada, head of robotics at Google DeepMind, tells WIRED. From the report: Gemini Robotics 2 combines several different AI models into a single system. Taken together, they allow a robot to make sense of its surroundings and how to act in it. A vision language model (VLM), which understands images and video, can communicate with humans and reason how to perform different tasks. Two vision language action (VLA) models, trained to understand how to move in physical space, control the robot's full-body movement as well as the movements of grippers or hands. In video demonstrations shared ahead of the release, the company showed several different robots performing complex tasks autonomously using the amalgamated model. In one demo, Apptronik's Apollo 2 robot used hands from a company called Sharpa to tidy shelves. Google DeepMind trained the model to perform these tasks using a mix of human teleoperation, video examples, and simulations -- it's not yet possible for AI models to perform a wide range of complex tasks without specific training. [...] Parada says Google takes a multi-layered approach to safety, with guardrails applied on each model layer. It's also introducing ASIMOV-Agentic, a new benchmark for measuring the safety of various AI systems collaborating to control a robot. The benchmark detects whether a command will result in harmful or uncertain outcome.

Read more of this story at Slashdot.

An anonymous reader quotes a report from TechCrunch: Google is expanding its answer to Apple's age-assurance tools with Wednesday's news that it will bring its Play Signal API to users worldwide by the end of 2026. The technology, already available in Brazil, allows Android developers to identify younger users of their apps in order to provide safer, age-appropriate experiences. The expansion will initially bring the API to Australia and Canada by mid-August, before rolling out globally to all markets by the end of the year. [...] Like Apple, Google's technology allows developers to obtain a user's age range without needing to access personal information, like their date of birth. Instead, it enables parents to share their child's age range directly with apps. It also lets adults share their age when prompted by app developers as well, allowing for customized experiences. Parents won't have to manage sharing this information on an app-by-app basis, either. To make it easier, Google centralizes these controls inside its parental controls dashboard, Family Link. Once entered, any developer that chooses to incorporate age-range information can access this signal to customize their apps accordingly. Google notes, however, that the age ranges are not shared by default -- parents must opt in by entering that information. The feature joins other safety tools on Google Play, including those that let developers restrict a child's ability to discover their apps. Parents, meanwhile, can continue to use Google Play's Family Link app to manage their child's screen-time limits, approve app downloads, or set PIN-based content filters for specific apps.

Read more of this story at Slashdot.

The jetport’s terminal building is open 24 hours a day, but TSA screening, airline counters and concessions close after the final flight boards.

Avi Schiffmann has a new version of his controversial AI companion. It’s more expensive, and you can’t change its personality.

The Massachusetts Port Authority is converting the vacant space into an R&D facility for marine tech startups. The incubator is expected to open in 18 to 24 months.

GCC has adopted a policy rejecting substantial code contributions generated by or derived from LLMs. "This covers not just code copied directly from tools like ChatGPT, Gemini, or GitHub Copilot, but also any versions of the code later edited or rewritten by a human, provided that the final contribution is still based on material generated by the system," reports Linuxiac. From the report: The important point, then, is not whether a developer has used an AI tool at some stage in their work; contributors can use LLMs to discuss ideas, understand existing code, learn about a field they are unfamiliar with, or carry out general research. The limitation lies in the inclusion of copyright-significant material generated by such tools in the code submitted to GCC. The policy also provides for a few limited exceptions; GCC maintainers are allowed to accept changes that are legally insignificant or trivial and are generated by an LLM, on the condition that they meet the project's normal contribution requirements and the use of an LLM is clearly disclosed. Furthermore, copyright-significant AI-generated test cases could still be accepted. Since test cases usually involve small programs which are intended to reproduce compiler bugs or to verify certain behavior, the policy deals with them separately from code that is incorporated into GCC itself. It does not follow that merely looking at or altering the generated code makes it acceptable. By the rules that have been adopted, a contributor cannot take substantial implementation produced by an LLM, clean it up manually, and then treat the resulting patch as if it had been originally written by them. Once a contribution has been derived from generated content, it is still subject to the policy.

Read more of this story at Slashdot.

Razer has finally caved and made its first Hall Effect gaming keyboard. I dug into its switches, features, and gaming performance to see if it was worth the wait.

Some of the chipmaker's files were stolen following an "unauthorized access" to certain company systems on June 23, according to a public filing.

The outage meant Xbox users had trouble signing into their accounts, seeing their game library and launching games

"Breakthrough Bio" is initially open to startups headquartered in Israel and Taiwan, but CEO Cait Brumme hopes to expand it to additional geographies.

I tested leading filtered showerheads to see how well they remove total chlorine from your water.

From fitness tracking to notifications, the best budget smartwatch can do a lot without costing a fortune.