• Sl chevron_right

      Contact publication

      pubsub.blastersklan.com / slashdot · 13:18 edit · 1 minute

    Sam Altman of OpenAI and the chief executives of Nvidia, Microsoft and Alphabet are among technology-industry leaders joining a new federal advisory board focused on the secure use of AI within U.S. critical infrastructure, in the Biden administration's latest effort to fill a regulatory vacuum over the rapidly proliferating technology. From a report: The Artificial Intelligence Safety and Security Board is part of a government push to protect the economy, public health and vital industries from being harmed by AI-powered threats, U.S. officials said. Working with the Department of Homeland Security, it will develop recommendations for power-grid operators, transportation-service providers and manufacturing plants, among others, on how to use AI while bulletproofing their systems against potential disruptions that could be caused by advances in the technology. In addition to Nvidia's Jensen Huang, Microsoft's Satya Nadella, Alphabet's Sundar Pichai and other leaders in AI and technology, the panel of nearly two dozen consists of academics, civil-rights leaders and top executives at companies that work within a federally recognized critical-infrastructure sector, including Kathy Warden, chief executive of Northrop Grumman, and Delta Air Lines Chief Executive Ed Bastian. Other members are public officials, such as Maryland Gov. Wes Moore and Seattle Mayor Bruce Harrell, both Democrats.

    Read more of this story at Slashdot.

    OpenAI's Sam Altman and Other Tech Leaders To Serve on AI Safety Board
    • wifi_tethering open_in_new

      This post is public

      slashdot.org /story/24/04/26/1249215/openais-sam-altman-and-other-tech-leaders-to-serve-on-ai-safety-board

    • Sl chevron_right

      Contact publication

      pubsub.blastersklan.com / slashdot · 00:28 edit · 2 minutes

    Thomas Claburn reports via The Register: Baltimore police have arrested Dazhon Leslie Darien, the former athletic director of Pikesville High School (PHS), for allegedly impersonating the school's principal using AI software to make it seem as if he made racist and antisemitic remarks. Darien, of Baltimore, Maryland, was subsequently charged with witness retaliation, stalking, theft, and disrupting school operations. He was detained late at night trying to board a flight at BWI Thurgood Marshall Airport. Security personnel stopped him because the declared firearm he had with him was improperly packed and an ensuing background check revealed an open warrant for his arrest. "On January 17, 2024, the Baltimore County Police Department became aware of a voice recording being circulated on social media," said Robert McCullough, Chief of Baltimore County Police, at a streamed press conference today. "It was alleged the voice captured on the audio file belong to Mr Eric Eiswert, the Principal at the Pikesville High School. We now have conclusive evidence that the recording was not authentic. "The Baltimore County Police Department reached that determination after conducting an extensive investigation, which included bringing in a forensic analyst contracted with the FBI to review the recording. The results of the analysis indicated the recording contained traces of AI-generated content." McCullough said a second opinion from a forensic analyst at the University of California, Berkeley, also determined the recording was not authentic. "Based off of those findings and further investigation, it's been determined the recording was generated through the use of artificial intelligence technology," he said. According to the warrant issued for Darien's arrest, the audio file was shared through social media on January 17 after being sent via email to school teachers. The recording sounded as if Principal Eric Eiswert had made remarks inflammatory enough to prompt a police visit to advise on protective security measures for staff. [...] The clip, according to the warrant, led to the temporary removal of Eiswert from his position and "a wave of hate-filled messages on social media and numerous calls to the school," and significantly disrupted school operations. Police say it led to threats against Eiswert and concerns about his safety. Eiswert told investigators that he believes the audio clip was fake as "he never had the conversations in the recording." And he said he believed Darien was responsible due to his technical familiarity with AI and had a possible motive: Eiswert said there "had been conversations with Darien about his contract not being renewed next semester due to frequent work performance challenges." "It is clear that we are also entering a new deeply concerning frontier as we continue to embrace emerging technology and its potential for innovation and social good," said John Olszewski, Baltimore County Executive, during a press conference. "We must also remain vigilant against those who would have used it for malicious intent. That will require us to be more aware and more discerning about the audio we hear and the images we see. We will need to be careful in our judgment."

    Read more of this story at Slashdot.

    US Teacher Charged With Using AI To Frame Principal With Hate Speech Clip
    • wifi_tethering open_in_new

      This post is public

      yro.slashdot.org /story/24/04/25/2215233/us-teacher-charged-with-using-ai-to-frame-principal-with-hate-speech-clip

    • chevron_right

      Tech brands are forcing AI into your gadgets—whether you asked for it or not

      news.movim.eu / ArsTechnica · Yesterday - 22:34

    Tech brands love hollering about the purported thrills of AI these days.

    Enlarge / Tech brands love hollering about the purported thrills of AI these days. (credit: Getty )

    Logitech announced a new mouse last week. A company rep reached out to inform Ars of Logitech’s “newest wireless mouse.” The gadget’s product page reads the same as of this writing.

    I’ve had good experience with Logitech mice, especially wireless ones, one of which I'm using now . So I was keen to learn what Logitech might have done to improve on its previous wireless mouse designs. A quieter click ? A new shape to better accommodate my overworked right hand? Multiple onboard profiles in a business-ready design?

    I was disappointed to learn that the most distinct feature of the Logitech Signature AI Edition M750 is a button located south of the scroll wheel. This button is preprogrammed to launch the ChatGPT prompt builder, which Logitech recently added to its peripherals configuration app Options+.

    Read 28 remaining paragraphs | Comments

    • chevron_right

      Apple releases eight small AI language models aimed at on-device use

      news.movim.eu / ArsTechnica · Yesterday - 20:55

    An illustration of a robot hand tossing an apple to a human hand.

    Enlarge (credit: Getty Images)

    In the world of AI, what might be called "small language models" have been growing in popularity recently because they can be run on a local device instead of requiring data center-grade computers in the cloud. On Wednesday, Apple introduced a set of tiny source-available AI language models called OpenELM that are small enough to run directly on a smartphone. They're mostly proof-of-concept research models for now, but they could form the basis of future on-device AI offerings from Apple.

    Apple's new AI models, collectively named OpenELM for "Open-source Efficient Language Models," are currently available on the Hugging Face under an Apple Sample Code License . Since there are some restrictions in the license, it may not fit the commonly accepted definition of "open source," but the source code for OpenELM is available.

    On Tuesday, we covered Microsoft's Phi-3 models , which aim to achieve something similar: a useful level of language understanding and processing performance in small AI models that can run locally. Phi-3-mini features 3.8 billion parameters, but some of Apple's OpenELM models are much smaller, ranging from 270 million to 3 billion parameters in eight distinct models.

    Read 7 remaining paragraphs | Comments

    • chevron_right

      School athletic director arrested for framing principal using AI voice synthesis

      news.movim.eu / ArsTechnica · Yesterday - 15:30 · 1 minute

    Illustration of a robot speaking.

    Enlarge (credit: Getty Images )

    On Thursday, Baltimore County Police arrested Pikesville High School's former athletic director, Dazhon Darien, and charged him with using AI to impersonate Principal Eric Eiswert, according to a report by The Baltimore Banner . Police say Darien used AI voice synthesis software to simulate Eiswert's voice, leading the public to believe the principal made racist and antisemitic comments.

    The audio clip, posted on a popular Instagram account, contained offensive remarks about "ungrateful Black kids" and their academic performance, as well as a threat to "join the other side" if the speaker received one more complaint from "one more Jew in this community." The recording also mentioned names of staff members, including Darien's nickname "DJ," suggesting they should not have been hired or should be removed "one way or another."

    The comments led to significant uproar from students, faculty, and the wider community, many of whom initially believed the principal had actually made the comments. A Pikesville High School teacher named Shaena Ravenell reportedly played a large role in disseminating the audio. While she has not been charged, police indicated that she forwarded the controversial email to a student known for their ability to quickly spread information through social media. This student then escalated the audio's reach, which included sharing it with the media and the NAACP.

    Read 5 remaining paragraphs | Comments

    • chevron_right

      Researchers Showcase Decentralized AI-Powered Torrent Search Engine

      news.movim.eu / TorrentFreak · Yesterday - 09:49 · 4 minutes

    decentralized network Twenty-five years ago, peer-to-peer file-sharing took the Internet by storm.

    The ability to search for and share content with complete strangers was nothing short of a revolution.

    In the years that followed, media consumption swiftly moved online. This usually involved content shared without permission, but pirate pioneers ultimately paved the way for new business models.

    The original ‘pirate’ ethos has long since gone. There are still plenty of unauthorized sites and services, but few today concern themselves with decentralization and similar technical advances; centralized streaming is the new king with money as the main motivator.

    AI Meets BitTorrent

    There are areas where innovation and technological progress still lead today, mostly centered around artificial intelligence. Every month, numerous new tools and services appear online, as developers embrace what many see as unlimited potential.

    How these developments will shape the future is unknown, but they have many rightsholders spooked. Interestingly, an ‘old’ research group, that was already active during BitTorrent’s heyday, is now using AI to amplify its technology.

    Researchers from the Tribler research group at Delft University of Technology have been working on their Tribler torrent client for nearly two decades . They decentralized search , removing the need for torrent sites, and implemented ‘ anonymity ‘ by adding an onion routing layer to file transfers.

    Many millions of euros have been spent on the Tribler research project over the years. Its main goal is to advance decentralized technology, not to benefit corporations, but to empower the public at large.

    “Our entire research portfolio is driven by idealism. We aim to remove power from companies, governments, and AI in order to shift all this power to self-sovereign citizens,” the Tribler team explains.

    Decentralized AI-powered Search

    While not every technological advancement has been broadly embraced, yet, Tribler has just released a new paper and a proof of concept which they see as a turning point for decentralized AI implementations; one that has a direct BitTorrent link.

    The scientific paper proposes a new framework titled “De-DSI”, which stands for Decentralised Differentiable Search Index . Without going into technical details, this essentially combines decentralized large language models (LLMs), which can be stored by peers, with decentralized search.

    This means that people can use decentralized AI-powered search to find content in a pool of information that’s stored across peers. For example, one can ask “find a magnet link for the Pirate Bay documentary,” which should return a magnet link for TPB-AFK, without mentioning it by name.

    This entire process relies on information shared by users. There are no central servers involved at all, making it impossible for outsiders to control.

    Endless Possibilities, Limited Use

    While this sounds exciting, the current demo version is not yet built into the Tribler client. Associate Professor Dr. Johan Pouwelse, leader of the university’s Tribler Lab, explains that it’s just a proof of concept with a very limited dataset and AI capabilities.

    “For this demo, we trained an end-to-end generative Transformer on a small dataset that comprises YouTube URLs, magnet links, and Bitcoin wallet addresses. Those identifiers are each annotated with a title and represent links to movie trailers, CC-licensed music, and BTC addresses of independent artists,” Pouwelse says.

    We tried some basic searches with mixed results. That makes sense since there’s only limited content, but it can find magnet links and videos without directly naming the title. That said, it’s certainly not yet as powerful as other AI tools.

    de-dsi

    In essence, De-DSI operates by sharing the workload of training large language models on lists of document identifiers. Every peer in the network specializes in a subset of data, which other peers in the network can retrieve to come up with the best search result.

    A Global Human Brain to Fight Torrent Spam and Censors

    The proof of concept shows that the technology is sound. However, it will take some time before it’s integrated into the Tribler torrent client. The current goal is to have an experimental decentralized-AI version of Tribler ready at the end of the year.

    While the researchers see this as a technological breakthrough, it doesn’t mean that things will improve for users right away. AI-powered search will be slower to start with and, if people know what they’re searching for, it offers little benefit.

    Through trial and error, the researchers ultimately hope to improve things though, with a “global brain” for humanity as the ultimate goal.

    Most torrent users are not looking for that, at the moment, but Pouwelse says that they could also use decentralized machine learning to fight spam, offer personal recommendations, and to optimize torrent metadata. These are concrete and usable use cases.

    The main drive of the researchers is to make technology work for the public at large, without the need for large corporations or a central government to control it.

    “The battle royale for Internet control is heating up,” Pouwelse says, in a Pirate Bay-esque fashion.

    “Driven by our idealism we will iteratively take away their power and give it back to citizens. We started 18 years ago and will take decades more. We should not give up on fixing The Internet, just because it is hard.”

    The very limited De-DSI proof of concept and all related code is available on Huggingface . All technological details are available in the associated paper . The latest Tribler version, which is fully decentralized without AI, can be found on the official project page .

    From: TF , for the latest news on copyright battles, piracy and more.

    • chevron_right

      Deepfakes in the courtroom: US judicial panel debates new AI evidence rules

      news.movim.eu / ArsTechnica · 2 days ago - 20:14

    An illustration of a man with a very long nose holding up the scales of justice.

    Enlarge (credit: Getty Images )

    On Friday, a federal judicial panel convened in Washington, DC, to discuss the challenges of policing AI-generated evidence in court trials, according to a Reuters report . The US Judicial Conference's Advisory Committee on Evidence Rules , an eight-member panel responsible for drafting evidence-related amendments to the Federal Rules of Evidence , heard from computer scientists and academics about the potential risks of AI being used to manipulate images and videos or create deepfakes that could disrupt a trial.

    The meeting took place amid broader efforts by federal and state courts nationwide to address the rise of generative AI models (such as those that power OpenAI's ChatGPT or Stability AI's Stable Diffusion ), which can be trained on large datasets with the aim of producing realistic text, images, audio, or videos.

    In the published 358-page agenda for the meeting, the committee offers up this definition of a deepfake and the problems AI-generated media may pose in legal trials:

    Read 9 remaining paragraphs | Comments

    • chevron_right

      Microsoft’s Phi-3 shows the surprising power of small, locally run AI language models

      news.movim.eu / ArsTechnica · 3 days ago - 20:47

    An illustration of lots of information being compressed into a smartphone with a funnel.

    Enlarge (credit: Getty Images)

    On Tuesday, Microsoft announced a new, freely available lightweight AI language model named Phi-3-mini, which is simpler and less expensive to operate than traditional large language models (LLMs) like OpenAI's GPT-4 Turbo . Its small size is ideal for running locally, which could bring an AI model of similar capability to the free version of ChatGPT to a smartphone without needing an Internet connection to run it.

    The AI field typically measures AI language model size by parameter count. Parameters are numerical values in a neural network that determine how the language model processes and generates text. They are learned during training on large datasets and essentially encode the model's knowledge into quantified form. More parameters generally allow the model to capture more nuanced and complex language-generation capabilities but also require more computational resources to train and run.

    Some of the largest language models today, like Google's PaLM 2 , have hundreds of billions of parameters. OpenAI's GPT-4 is rumored to have over a trillion parameters but spread over eight 220-billion parameter models in a mixture-of-experts configuration. Both models require heavy-duty data center GPUs (and supporting systems) to run properly.

    Read 8 remaining paragraphs | Comments

    • chevron_right

      High-speed imaging and AI help us understand how insect wings work

      news.movim.eu / ArsTechnica · 4 days ago - 20:16 · 1 minute

    Black and white images of a fly with its wings in a variety of positions, showing the details of a wing beat.

    Enlarge / A time-lapse showing how an insect's wing adopts very specific positions during flight. (credit: Florian Muijres, Dickinson Lab)

    About 350 million years ago, our planet witnessed the evolution of the first flying creatures. They are still around, and some of them continue to annoy us with their buzzing. While scientists have classified these creatures as pterygotes, the rest of the world simply calls them winged insects.

    There are many aspects of insect biology, especially their flight , that remain a mystery for scientists. One is simply how they move their wings. The insect wing hinge is a specialized joint that connects an insect’s wings with its body. It’s composed of five interconnected plate-like structures called sclerites. When these plates are shifted by the underlying muscles, it makes the insect wings flap.

    Until now, it has been tricky for scientists to understand the biomechanics that govern the motion of the sclerites even using advanced imaging technologies. “The sclerites within the wing hinge are so small and move so rapidly that their mechanical operation during flight has not been accurately captured despite efforts using stroboscopic photography, high-speed videography, and X-ray tomography,” Michael Dickinson, Zarem professor of biology and bioengineering at the California Institute of Technology (Caltech), told Ars Technica.

    Read 15 remaining paragraphs | Comments