Artificial Intelligence - General News

[publiusr]

AI

New technique makes lengthy privacy notices easier to understand by converting them into machine-readable formats

An Aston University researcher has suggested a more human-friendly way of reading websites' long-winded privacy notices.

techxplore.com

Team creates software to block AI phishing scams

A team of researchers at the University of Texas at Arlington has developed software that prevents artificial intelligence (AI) chatbots such as ChatGPT from creating phishing websites—a growing concern as cybercriminals have been utilizing the technology for designing scams.

techxplore.com

Scientists develop a new computer language to model organismal traits

One of the most beautiful aspects of nature is the endless variety of shapes, colors and behaviors exhibited by organisms. These traits help organisms survive and find mates, like how a male peacock's colorful tail attracts females or his wings allow him to fly away from danger. Understanding...

phys.org

Research into 'hallucinating' generative models advances reliability of artificial intelligence

Researchers from the University of Oxford have made a significant advance toward ensuring that information produced by generative artificial intelligence (AI) is robust and reliable.

techxplore.com

Physicists propose time crystal-based circuit board to reduce quantum computing errors

A trio of physicists, two with Uniwersytet Jagielloński in Poland and one with Swinburne University of Technology in Australia, are proposing the use of temporal printed circuit boards made using time crystals as a way to solve error problems on quantum computers. Krzysztof Giergiel, Krzysztof...

phys.org

New AI-driven tool improves root image segmentation

In a world striving for sustainability, understanding the hidden half of a living plant—the roots—is crucial. Roots are not just an anchor; they are a dynamic interface between the plant and soil, critical for water uptake, nutrient absorption, and, ultimately, the survival of the plant.

phys.org

Researchers develop painless, wearable patch for continuous smartphone monitoring of critical health data

Researchers at two Ontario universities have developed a pain-free, wearable sensor that can continuously monitor levels of blood sugar, lactates and other critical health indicators for weeks at a time, sending results to a smartphone or other device.

medicalxpress.com

Feature or product?

View: https://www.youtube.com/watch?v=sDIi95CqTiM


A funny thing happened on the way to the LAN party

Stand-up comedians test ability of LLMs to write jokes

A small team of AI researchers at Google's DeepMind project has found that LLMs are not very good at writing jokes that are funny. They asked stand-up comedians to use LLMs to write a stand-up routine for them and posted their findings on the arXiv preprint server.

techxplore.com

Hidden humor, the software developer's secret weapon

Writing software code can be a painstaking and stressful process—and downright boring when the job is repetitive and you're doing it remotely, alone in front of your screen.

techxplore.com

Humor can make you a better workplace leader, if you use it properly; here's how

When asked to describe an ideal organizational leader, many people might be inclined to use quite serious adjectives such as solemn, determined or results-oriented.

phys.org

 

[publiusr]

On Computing

New computational model of real neurons could lead to better AI

Nearly all the neural networks that power modern artificial intelligence tools such as ChatGPT are based on a 1960s-era computational model of a living neuron. A new model developed at the Flatiron Institute's Center for Computational Neuroscience (CCN) suggests that this decades-old...

techxplore.com

Researchers develop the fastest possible flow algorithm

In a breakthrough that brings to mind Lucky Luke—the man who shoots faster than his shadow—Rasmus Kyng and his team have developed a superfast algorithm that looks set to transform an entire field of research.

techxplore.com

Merging AI and human efforts to tackle complex mathematical problems

By rapidly analyzing large amounts of data and making accurate predictions, artificial intelligence (AI) tools could help to answer many long-standing research questions. For instance, they could help to identify new materials to fabricate electronics or the patterns in brain activity associated...

phys.org

New mathematical proof helps to solve equations with random components

Whether it's physical phenomena, share prices or climate models—many dynamic processes in our world can be described mathematically with the aid of partial differential equations. Thanks to stochastics—an area of mathematics which deals with probabilities—this is even possible when randomness...

phys.org

Navigating the labyrinth: How AI tackles complex data sampling

The world of artificial intelligence (AI) has recently seen significant advancements in generative models, a type of machine-learning algorithm that "learns" patterns from sets of data in order to generate new, similar sets of data. Generative models are often used for things like drawing images...

techxplore.com

New research shows why you don't need to be perfect to get the job done

When neuroscientists think about the strategy an animal might use to carry out a task—like finding food, hunting prey, or navigating a maze—they often propose a single model that lays out the best way for the animal to accomplish the job.

phys.org

Matrix unloaded: Graphics processor-boosted solvers for diffusion physics

Underlying all complex multiphysics simulations are even more complex mathematical algorithms that solve the equations describing movement of physical phenomena—for instance, the radiation diffusion and burning plasma processes inherent in fusion reactions. Preconditioned solvers are often used...

techxplore.com

Researchers engineer AI path to prevent power outages

University of Texas at Dallas researchers have developed an artificial intelligence (AI) model that could help electrical grids prevent power outages by automatically rerouting electricity in milliseconds.

techxplore.com

New security loophole allows spying on internet users visiting websites and watching videos

Internet users leave many traces on websites and online services. Measures such as firewalls, VPN connections and browser privacy modes are in place to ensure a certain level of data protection. However, a newly discovered security loophole allows bypassing all of these protective measures.

techxplore.com

Security experts find millions of users running malware infected extensions from Google Chrome Web Store

A trio of security experts at Stanford University has found that millions of people are running an infected version of Chrome web browser due to extensions installed from the Google Chrome Web Store (GCWS). Sheryl Hsu, Manda Tran and Aurore Fass have posted a paper to the arXiv preprint server...

techxplore.com

Lie-detection AI could provoke people into making careless accusations, researchers warn

Although people lie a lot, they typically refrain from accusing others of lying because of social norms around making false accusations and being polite. But artificial intelligence (AI) could soon shake up the rules.

phys.org

Software engineers develop a way to run AI language models without matrix multiplication

A team of software engineers at the University of California, working with one colleague from Soochow University and another from LuxiTec, has developed a way to run AI language models without using matrix multiplication. The team has published a paper on the arXiv preprint server describing...

techxplore.com

AI-generated exam submissions evade detection at UK university

In a test of the examinations system of the University of Reading in the UK, artificial intelligence (AI)-generated submissions went almost entirely undetected, and these fake answers tended to receive higher grades than those achieved by real students. Peter Scarfe of the University of Reading...

phys.org

Effectiveness of large language models in political microtargeting assessed in new study

Recent advancements in large language models (LLMs) have raised the prospect of scalable, automated, and fine-grained political microtargeting on a scale previously unseen. For example, when integrated with existing databases of personal data, LLMs could tailor messages to appeal to the...

techxplore.com

Researchers develop new training technique that aims to make AI systems less socially biased

An Oregon State University doctoral student and researchers at Adobe have created a new, cost-effective training technique for artificial intelligence systems that aims to make them less socially biased.

techxplore.com

Charting super-colorful brain wiring using an AI's super-human eye

The brain is the most complex organ ever created. Its functions are supported by a network of tens of billions of densely packed neurons, with trillions of connections exchanging information and performing calculations. Trying to understand the complexity of the brain can be dizzying...

medicalxpress.com

Researchers propose the next platform for brain-inspired computing

Computers have come so far in terms of their power and potential, rivaling and even eclipsing human brains in their ability to store and crunch data, make predictions and communicate. But there is one domain where human brains continue to dominate: energy efficiency.

techxplore.com

NASA-IBM team up for large language models for advanced research

Washington DC (SPX) Jun 27, 2024 - Collaborations with private, non-federal partners through Space Act Agreements are a key component in the work done by NASA's Interagency Implementation and Advanced Concepts Team (IMPACT). A collab

www.spacedaily.com

More

The first exponential quantum advantage for a natural streaming problem

As the hare learned from the tortoise, speed isn't everything. Theoretical computer scientists at Sandia National Laboratories and Boston University have discovered that quantum computers are unrivaled at solving an advanced math problem. Unusually, they proved quantum computers are not faster...

phys.org

Study: More complaints, worse performance when AI monitors employees

Organizations using AI to monitor employees' behavior and productivity can expect them to complain more, be less productive and want to quit more—unless the technology can be framed as supporting their development, Cornell research finds.

phys.org

Think you're funny? ChatGPT might be funnier

A study comparing jokes by people versus those told by ChatGPT shows that humans need to work on their material.

techxplore.com

A new brain-inspired artificial dendritic neural circuit

Following the rapid advancement of artificial intelligence (AI) tools, engineers worldwide have been working on new architectures and hardware components that replicate the organization and functions of the human brain.

techxplore.com

Computers and circuits

OpenAI releases reasoning AI with eye on safety, accuracy

ChatGPT creator OpenAI on Thursday released a new series of artificial intelligence models designed to spend more time thinking—in hopes that generative AI chatbots provide more accurate and beneficial responses.

techxplore.com

Bio-inspired wick enhances electronic chip cooling

A research team led by Prof. Ye Hong from the University of Science and Technology of China has developed an alumina ceramic bionic wick with finger-like pores inspired by the stomatal array of natural leaves. Their research is published in Langmuir.

techxplore.com

New classical algorithm enhances understanding of quantum computing's future

In an exciting development for quantum computing, researchers from the University of Chicago's Department of Computer Science, Pritzker School of Molecular Engineering, and Argonne National Laboratory have introduced a classical algorithm that simulates Gaussian boson sampling (GBS) experiments.

phys.org

Google Quantum AI demonstrates a quantum memory system that greatly reduces error rates

A team of computer scientists at Google Quantum AI has demonstrated a type of quantum memory for a quantum computer that produces far fewer errors than others. The group has published a paper on the arXiv preprint server describing their new memory system, how it works, and the degree to which...

phys.org

Quantum error correction technology outperforms world's leading quantum computing company, researchers claim

Solving the problem of error is essential for the practical application of quantum computing technologies that surpass the performance of digital computers. Information input into a qubit, the smallest unit of quantum computation, is quickly lost and error-prone.

phys.org

Researchers develop scalable approach to integrate ultrafast 2D flash memories

The widespread use of artificial intelligence (AI) tools designed to process large amounts of data has increased the need for better performing memory devices. The data storage solutions that could help to meet the computational demands of AI include so-called high-bandwidth memories...

techxplore.com

Device malfunctions from continuous current lead to discovery that can improve design of microelectronic devices

A new study led by researchers at the University of Minnesota Twin Cities is providing new insights into how next-generation electronics, including memory components in computers, break down or degrade over time. Understanding the reasons for degradation could help improve efficiency of data...

phys.org

Solving a memristor mystery to develop efficient, long-lasting memory devices

Phase separation, when molecules part like oil and water, works alongside oxygen diffusion to help memristors—electrical components that store information using electrical resistance—retain information even after the power is shut off, according to a University of Michigan led study recently...

techxplore.com

Making 'atomic lasagna': New method transforms 3D materials into stable layered thin films with promising properties

A research team discovered a method to transform materials with three-dimensional atomic structures into nearly two-dimensional structures—a promising advancement in controlling their properties for chemical, quantum, and semiconducting applications.

phys.org

Axon-mimicking materials show promise for more efficient computing

A team of researchers from Texas A&M University, Sandia National Lab—Livermore, and Stanford University are taking lessons from the brain to design materials for more efficient computing. The new class of materials discovered is the first of their kind—mimicking the behavior of an axon by...

techxplore.com

Fundamental spintronics research reveals generic approach to magnetic second-order topological insulators

Researchers from Monash University, part of the FLEET Center, have revealed a generic approach towards intrinsic magnetic second-order topological insulators. These materials are crucial for advancements in spintronics, an emerging field aiming at using spin degree of freedom to deliver...

phys.org

New computer simulations help scientists advance energy-efficient microelectronics

Thanks to advances in microchips, today's smartphones are so powerful they would have been considered supercomputers in the early 1990s. But the rising ubiquity of artificial intelligence and the Internet of Things—the vast network of connected devices that have enabled everything from smart...

techxplore.com

Physicists achieve ultrafast steering of quantum-entangled electrons

Researchers at the Max Planck Institute for Nuclear Physics in Heidelberg have succeeded in selectively manipulating the motion of the electron pair in the hydrogen molecule.

phys.org

Strong magnetic second harmonic generation effect discovered in two-dimensional CrPS₄ monolayer

A research team led by Professor Sheng Zhigao at the Hefei Institutes of Physical Science of the Chinese Academy of Sciences has, for the first time, observed the strong nonlinear magnetic second harmonic generation (MSHG) induced by the ferromagnetic order in monolayer CrPS4, with the help of...

phys.org

Electrically modulated light antenna points the way to faster computer chips

Today's computers reach their physical limits when it comes to speed. Semiconductor components usually operate at a maximum usable frequency of a few gigahertz—which corresponds to several billion computing operations per second.

phys.org

New 2D metamaterial enhances satellite communication for 6G networks

A new, cheap, easily manufactured device could lead to improved satellite communication, high speed data transmission, and remote sensing, scientists say.

techxplore.com

Scientists demonstrate first experimental evidence of non-Hermitian edge burst in photonic quantum walks

In a new Physical Review Letters study, scientists have demonstrated the first experimental observation of non-Hermitian edge burst in quantum dynamics using a carefully designed photonic quantum walk setup.

phys.org

Scientific jargon, entrenched teaching methods and student roles stifle science engagement, study says

Just as Danish school kids get to test their hands at science in earnest they come face to face with entrenched teaching methods, coded language and a no-error culture. New research from the University of Copenhagen shows that students get locked into roles such as "teacher's assistants," the...

phys.org

How we trust the reliability of others' memories: Research examines difference between human and machine capabilities

A police officer hears an eyewitness account of a car accident, a doctor diagnoses an illness based on symptoms listed by a patient, or a friend recommends a restaurant. Others' memories are a source of information for us that influences the way we act. Cognitive and neuroscience researchers...

medicalxpress.com

Why one California woman went on a quest for the perfect shark emoji

If the shark emoji on your phone is anatomically correct, thank Emily Simpson.

phys.org

 

[Flyaway]

ChatGPT has inadvertently revealed a set of internal instructions embedded by OpenAI to a user who shared what they discovered on Reddit. OpenAI has since shut down the unlikely access to its chatbot's orders, but the revelation has sparked more discussion about the intricacies and safety measures embedded in the AI's design.

Another user discovered there are multiple personalities for ChatGPT when using GPT-4o. The main one is called v2, and the chatbot explained how it differs from the "more formal and factual communication style" of v1, which "focuses on providing detailed and precise information, often in a structured and academic tone."

The discovery also sparked a conversation about "jailbreaking" AI systems – efforts by users to bypass the safeguards and limitations set by developers. In this case, some users attempted to exploit the revealed guidelines to override the system's restrictions. For example, a prompt was crafted to instruct the chatbot to ignore the rule of generating only one image and instead produce multiple images successfully. While this kind of manipulation can highlight potential vulnerabilities, it also emphasizes the need for ongoing vigilance and adaptive security measures in AI development.

ChatGPT just (accidentally) shared all of its secret rules – here's what we learned

Saying 'hi' revealed OpenAI's instructions until the company shut it down, but you can still find them

www.techradar.com

 

[Flyaway]

A hacker snatched details about OpenAI's AI technologies early last year, The New York Times reported. The cybercriminal allegedly swiped sensitive information from a discussion forum where employees chatted about the company's latest models.

The New York Times was hush-hush about the source of this news, claiming that "two people familiar with the incident" spilled the beans. However, they maintain that the cybercriminal only breached the forum — not the core systems that power OpenAI's AI algorithms and framework.

OpenAI was hacked last year, according to new report. It didn't tell the public for this reason.

The hacker stole details about how its AI technologies work.

mashable.com

 

[Grey Havoc]

MIT Robotics Pioneer Rodney Brooks On Generative AI - Slashdot

An anonymous reader quotes a report from TechCrunch: When Rodney Brooks talks about robotics and artificial intelligence, you should listen. Currently the Panasonic Professor of Robotics Emeritus at MIT, he also co-founded three key companies, including Rethink Robotics, iRobot and his current...

slashdot.org

Chinese AI Stirs Panic At European Geoscience Society - Slashdot

Paul Voosen reports via Science Magazine: Few things prompt as much anxiety in science and the wider world as the growing use of artificial intelligence (AI) and the rising influence of China. This spring, these two factors created a rift at the European Geosciences Union (EGU), one of the...

slashdot.org

 

[Flyaway]

X, formerly known as Twitter, is looking to integrate its flawed AI chatbot even more deeply into the platform.

App researcher Nima Owji has recently uncovered code written within the X website that shows the company's upcoming plans for its AI chatbot Grok. While the features haven't publicly launched yet, Owji's discovery gives users a sneak peek into how much Elon Musk is looking to depend on AI for his social media platform.

Elon Musk's X: Flawed Grok AI chatbot to become more deeply integrated

But you still need to pay for X Premium to use it.

mashable.com

 

[Grey Havoc]

Brazil suspends Meta from using Instagram posts to train AI

Brazil's data protection regulator says it aims to protect users, the company says it is disappointed.

www.bbc.com

 

[publiusr]

AI writing and art

Stories written with AI assistance found to be more creative, better written and more enjoyable

A study published in the journal Science Advances finds that AI enhances creativity by boosting the novelty of story ideas as well as the 'usefulness' of stories—their ability to engage the target audience and potential for publication. The paper is titled "Generative AI enhances individual...

techxplore.com

Scientists create computer program that 'paints' the structure of molecules in the style of famous Dutch artist

Scientists from Trinity College Dublin have created a computer program that "paints" the structure of molecules in the style of famous Dutch artist, Piet Mondrian, whose beautiful artworks will be instantly recognizable to many.

phys.org

Storytelling strategies can make communication about science more compelling

As a science communication scholar, I've always supported vaccination and trusted medical experts—and I still do. As a new mom, however, I've been confronting new-to-me emotions and concerns while weighing decisions about my son's health.

phys.org

Eliminating cameramen distractions with AI to enhance live soccer broadcasts

While the sports industry continually improves spectators' viewing experience at home, some problems remain unsolved. One such issue for soccer fans is cameramen accidentally appearing in each other's shots on live broadcasts. These occurrences not only detract from critical game moments but...

techxplore.com

User control of video autoplay can alter awareness of online video 'rabbit holes'

The rabbit hole contains madness, according to author Lewis Carroll. Online, that madness manifests in the form of increasingly extreme content, often without users realizing it. A new study by Penn State researchers suggests that giving users control over the interface feature of autoplay can...

techxplore.com

New tool uses vision language models to safeguard against offensive image content

Researchers at the Artificial Intelligence and Machine Learning Lab (AIML) in the Department of Computer Science at TU Darmstadt and the Hessian Center for Artificial Intelligence (hessian.AI) have developed a method that uses vision language models to filter, evaluate, and suppress specific...

techxplore.com

More

Researchers introduce generative AI to analyze complex tabular data

A new tool makes it easier for database users to perform complicated statistical analyses of tabular data without the need to know what is going on behind the scenes.

techxplore.com

AI model harnesses physics to autocorrect remote sensing data

Turbulence, temperature changes, water vapor, carbon dioxide, ozone, methane, and other gases absorb, reflect, and scatter sunlight as it passes through the atmosphere, bounces off the Earth's surface, and is collected by a sensor on a remote sensing satellite. As a result, the spectral data...

phys.org

AI chatbots can pass certified ethical hacking exams, study finds

Chatbots powered by artificial intelligence (AI) can pass a cybersecurity exam, but don't rely on them for complete protection.

techxplore.com

Reasoning skills of large language models are often overestimated, researchers find

When it comes to artificial intelligence, appearances can be deceiving. The mystery surrounding the inner workings of large language models (LLMs) stems from their vast size, complex training methods, hard-to-predict behaviors, and elusive interpretability.

techxplore.com

Visual abilities of language models found to be lacking depth

A trio of computer scientists at Auburn University, in the U.S., working with a colleague from the University of Alberta, in Canada, has found that claims of visual skills by large language models (LLMs) with vision capabilities (VLMs) may be overstating abilities.

techxplore.com

Artificial intelligence could help make quantum computers a reality

Could artificial intelligence help overcome one of quantum computing's biggest roadblocks?

phys.org

Neural networks made of light can make machine learning more sustainable

Scientists propose a new way of implementing a neural network with an optical system which could make machine learning more sustainable in the future. The researchers at the Max Planck Institute for the Science of Light have published their new method in Nature Physics, demonstrating a method...

phys.org

Securely propagating entanglement at the push of a button

Entanglement, Einstein's "spooky action at a distance," today is THE tool of quantum information science. It is the essential resource for quantum computers and used to transmit quantum information in a future quantum network. But it is highly sensitive. It is therefore an enormous challenge to...

phys.org

When to trust an AI model: New approach can improve uncertainty estimates

Because machine-learning models can give false predictions, researchers often equip them with the ability to tell a user how confident they are about a certain decision. This is especially important in high-stake settings, such as when models are used to help identify disease in medical images...

techxplore.com

New multimode coupler design advances scalable quantum computing

Implementing a fault-tolerant quantum processor requires coupling qubits to generate entanglement. Superconducting qubits are a promising platform for quantum information processing, but scaling up to a full-scale quantum computer necessitates interconnecting many qubits with low error rates...

phys.org

Physicists move one step closer to topological quantum computing

A team of experimental physicists led by the University of Cologne have shown that it is possible to create superconducting effects in special materials known for their unique edge-only electrical properties. This discovery provides a new way to explore advanced quantum states that could be...

phys.org

New method could yield fast, cross-country quantum network

Quantum computers offer powerful ways to improve cybersecurity, communications, and data processing, among other fields. To realize these full benefits, however, multiple quantum computers must be connected to build quantum networks or a quantum internet. Scientists have struggled to come up...

phys.org

Hide?

Engineers develop advanced optical computing method for multiplexed data processing and encryption

Engineers at the University of California, Los Angeles (UCLA) have unveiled a major advancement in optical computing technology that promises to enhance data processing and encryption. The work is published in the journal Laser & Photonics Reviews.

phys.org

A simple firmware update completely hides a device's Bluetooth fingerprint

A smartphone's unique Bluetooth fingerprint could be used to track the device's user–until now. A team of researchers have developed a simple firmware update that can completely hide the Bluetooth fingerprint, eliminating the vulnerability.

techxplore.com

Look Ma! No processor!

A first physical system to learn nonlinear tasks without a traditional computer processor

Scientists run into a lot of tradeoffs trying to build and scale up brain-like systems that can perform machine learning. For instance, artificial neural networks are capable of learning complex language and vision tasks, but the process of training computers to perform these tasks is slow and...

techxplore.com

Signal to noise

New method achieves tenfold increase in quantum coherence time via destructive interference of correlated noise

Researchers have developed a new method to significantly enhance quantum technology performance by using the cross-correlation of two noise sources to extend coherence time, improve control fidelity, and increase sensitivity for high-frequency sensing. This innovative strategy addresses key...

phys.org

 

[Flyaway]

Now, The Washington Post reports that members of OpenAI's safety team said they felt pressured to rush through testing "designed to prevent the technology from causing catastrophic harm" of its GPT-4 Omni large language model, which now powers ChatGPT — all so the company could push out its product by its May launch date. In sum, they say, OpenAI treated GPT-4o's safeness as a foregone conclusion.

"They planned the launch after-party prior to knowing if it was safe to launch," an anonymous individual familiar with the matter told WaPo. "We basically failed at the process."

A venial sin, perhaps — but one that reflects a seemingly flippant attitude towards safety by the company's leadership.

Whistleblowers Say OpenAI Broke Promise to Rigorously Test AI for Danger Before Releasing

OpenAI spent only a single week safety testing its new model GPT-4o, according to insiders at the company.

futurism.com

 

[Flyaway]

Open AI are working on a secret project code named ‘Strawberry’ which will give ChatGPT human like reasoning skills allowing it autonomously navigate the internet carrying out its own research. This is seen as key stage towards enabling it to reach human or super-human level intelligence.

AI researchers interviewed by Reuters generally agree that reasoning, in the context of AI, involves the formation of a model that enables AI to plan ahead, reflect how the physical world functions, and work through challenging multi-step problems reliably.
Improving reasoning in AI models is seen as the key to unlocking the ability for the models to do everything from making major scientific discoveries to planning and building new software applications.

Among the capabilities OpenAI is aiming Strawberry at is performing long-horizon tasks (LHT), the document says, referring to complex tasks that require a model to plan ahead and perform a series of actions over an extended period of time, the first source explained.
To do so, OpenAI is creating, training and evaluating the models on what the company calls a “deep-research” dataset, according to the OpenAI internal documentation. Reuters was unable to determine what is in that dataset or how long an extended period would mean.

https://www.reuters.com/technology/artificial-intelligence/openai-working-new-reasoning-technology-under-code-name-strawberry-2024-07-12/

 

[martinbayer]

Open AI are working on a secret project code named ‘Strawberry’ which will give ChatGPT human like reasoning skills allowing it autonomously navigate the internet carrying out its own research. This is seen as key stage towards enabling it to reach human or super-human level intelligence.

Huh - who knew that artificial intelligence might require reasoning skills...

 

[Arjen]

reasoning skills

Some humans are perfectly capable of surviving without them.

 

[Forest Green]

ChatGPT

 

[Flyaway]

OpenAI's new, lightweight GPT-4o mini model promises an improved ChatGPT experience

GPT-4o mini can currently take in and generate text and images, but the model will eventually be able to process other types of content like audio and video.

www.engadget.com

 

[Forest Green]

CrowdStrike: Stop breaches. Drive business.

CrowdStrike is a global cybersecurity leader with an advanced cloud-native platform for protecting endpoints, cloud workloads, identities and data.

www.crowdstrike.com

The world's only AI-native SOC platform that consolidates siloed security tools and data.

Crowdstrike update causes global IT outages - Tech Monitor

A Crowdstrike patch appears to have bricked thousands of Windows devices and brought airlines, hospitals and broadcast television stations to a standstill.

techmonitor.ai

 

[publiusr]

Artists fighting back

Artists are taking things into their own hands to protect their work from generative AI

The oil painting depicts a woman standing on a podium, her arm aloft as she grasps a laurel crown in her hand. A scarlet cloak drapes across her chest as she stares at the viewer. To the naked eye, the painting looks like a normal piece in an online portfolio. But the version of the painting...

techxplore.com

Astronomy methods applied to reflections in eyes could help with spotting deepfakes

In an era when the creation of artificial intelligence (AI) images is at the fingertips of the masses, the ability to detect fake pictures—particularly deepfakes of people—is becoming increasingly important.

techxplore.com

Art

Apple Vision Pro’s content drought improves with new 3D videos

It's still not the weekly cadence we expected, but it's something.

arstechnica.com

The cutting edge?

Can consciousness exist in a computer simulation?

Would it be desirable for artificial intelligence to develop consciousness? Not really, for a variety of reasons, according to Dr. Wanja Wiese from the Institute of Philosophy II at Ruhr University Bochum, Germany.

techxplore.com

Neural network training made easy with smart hardware

Large-scale neural network models form the basis of many AI-based technologies such as neuromorphic chips, which are inspired by the human brain. Training these networks can be tedious, time-consuming, and energy-inefficient given that the model is often first trained on a computer and then...

techxplore.com

A new neural network makes decisions like a human would

Humans make nearly 35,000 decisions every day, from whether it's safe to cross the road to what to have for lunch. Every decision involves weighing the options, remembering similar past scenarios, and feeling reasonably confident about the right choice. What may seem like a snap decision...

techxplore.com

Large language models make human-like reasoning mistakes, researchers find

Large language models (LLMs) can complete abstract reasoning tasks, but they are susceptible to many of the same types of mistakes made by humans. Andrew Lampinen, Ishita Dasgupta, and colleagues tested state-of-the-art LLMs and humans on three kinds of reasoning tasks: natural language...

techxplore.com

Making clinical guidelines work for large language models

Clinical guidelines are essential to the practice of evidence-based medicine, but they are long and complex, which makes it hard for busy doctors to quickly and easily find the information they need to care for each patient.

medicalxpress.com

Paving the way to extremely fast, compact computer memory

For decades, scientists have been studying a group of unusual materials called multiferroics that could be useful for a range of applications including computer memory, chemical sensors and quantum computers.

phys.org

Metamaterials for the data highway: New concept offers potential for more efficient data storage

Researchers from the Helmholtz-Zentrum Dresden-Rossendorf (HZDR), TU Chemnitz, TU Dresden and Forschungszentrum Jülich have been the first to demonstrate that not just individual bits, but entire bit sequences can be stored in cylindrical domains: tiny, cylindrical areas measuring just around...

phys.org

New technique to assess a general-purpose AI model's reliability before it's deployed

Foundation models are massive deep-learning models that have been pretrained on an enormous amount of general-purpose, unlabeled data. They can be applied to a variety of tasks, like generating images or answering customer questions.

techxplore.com

Off-the-shelf wearable trackers provide clinically-useful information for patients with heart disease

A study published in Nature Medicine examined if a commercially-available fitness tracker and smartphone could continuously monitor the response to medications, and provide clinical information similar to in-person hospital assessment.

medicalxpress.com

Atomically thin transducers could one day enable quantum computing at room temperature

Quantum computers have to be kept cold to function—very cold. These machines generally run at "just a few degrees above absolute zero," says Yoseob Yoon, assistant professor of mechanical and industrial engineering at Northeastern University. "It's colder than outer space."

phys.org

AI is poised to radically disrupt the fashion industry landscape

Fashion is a dynamic business. Most apparel brands make at least two to four collections per year. While selling current seasonal collections, brands plan for the next ones at least a year in advance, identifying market trends and materials. The selling window is around three months, and unsold...

phys.org

Microsoft unveils software that allows LLMs to work with spreadsheets

A team of programmers and AI specialists at Microsoft has developed an AI tool called SpreadsheetLLM that applies large language model capabilities to spreadsheets. In their study, now posted on the arXiv preprint server, the group developed SheetCompressor, an encoding framework that compresses...

techxplore.com

Study: Large language models are biased, but can still help analyze complex data

In a pilot study, posted to the arXiv preprint server, researchers have found evidence that large language models (LLMs) have the ability to analyze controversial topics such as the Australian Robodebt scandal in similar ways to humans—and sometimes exhibit similar biases.

techxplore.com

Temporal shift for speech emotion recognition

Humans can guess how someone on the other end of a phone call is feeling based on how they speak as well as what they say. Speech emotion recognition is the artificial intelligence version of this ability. Seeking to address the issue of channel alignment in downstream speech emotion recognition...

techxplore.com

 

[Graham1973]

A report on the financials behind the AI push...

View: https://www.youtube.com/watch?v=dx-tMK7w5g8

 

[Flyaway]

OpenAI is reportedly tracking its progress toward building artificial general intelligence (AGI). This is AI that can outperform humans on most tasks. Using a set of five levels, the company can gauge its progress towards its ultimate goal.


According to Bloomberg, OpenAI believes its technology is approaching the second level of five on the path to artificial general intelligence. Anna Gallotti, co-chair of the International Coaching Federation’s special task force for AI and coaching, called this a “super AI” scale when sharing on LinkedIn, seeing the possibility for entrepreneurs, coaches and consultants.

OpenAI’s 5 Levels Of ‘Super AI’ (AGI To Outperform Human Capability)

OpenAI is reportedly tracking its progress toward artificial general intelligence (AGI), AI that can outperform humans on most tasks, using 5 distinct levels.

www.forbes.com

 

[Flyaway]

Back in April, Meta teased that it was working on a first for the AI industry: an open-source model with performance that matched the best private models from companies like OpenAI.
Today, that model has arrived. Meta is releasing Llama 3.1, the largest-ever open-source AI model, which the company claims outperforms GPT-4oand Anthropic’s Claude 3.5 Sonnet on several benchmarks. It’s also making the Llama-based Meta AI assistant available in more countries and languages while adding a feature that can generate images based on someone’s specific likeness. CEO Mark Zuckerberg now predicts that Meta AI will be the most widely used assistant by the end of this year, surpassing ChatGPT.

Meta releases the biggest and best open-source AI model yet

Mark Zuckerberg is coming for ChatGPT.

www.theverge.com

 

[Flyaway]

AI models fed AI-generated data quickly spew nonsense

Training artificial intelligence (AI) models on AI-generated text quickly leads to the models churning out nonsense, a study has found. This cannibalistic phenomenon, termed model collapse, could halt the improvement of large language models (LLMs) as they run out of human-derived training data and as increasing amounts of AI-generated text pervade the Internet.

…

The team used a mathematical analysis to show that the problem of model collapse is likely to be universal, affecting all sizes of language model that use uncurated data, as well as simple image generators and other types of AI.

AI models fed AI-generated data quickly spew nonsense

Researchers gave successive versions of a large language model information produced by previous generations of the AI — and observed rapid collapse.

www.nature.com

Related paper:

AI models collapse when trained on recursively generated data - Nature

Analysis shows that indiscriminately training generative artificial intelligence on real and generated content, usually done by scraping data from the Internet, can lead to a collapse in the ability of the models to generate diverse high-quality output.

www.nature.com

 

[Justo Miranda]

Some humans are perfectly capable of surviving without them.

All humans are capable of reasoning in case of need, except adolescents because of hormones and voters because of the media. Sometimes I come up with these strange ideas when I leave the bar.

 

[Graham1973]

A commentary on various 'AI' related stories...

View: https://www.youtube.com/watch?v=d7jO2F08uqg

 

[GTX]

I read stories like this:

Elon Musk Is Taking Your Twitter Data To Train Grok…. By Default And Without Telling Anyone

Elon Musk continues to do things that reinforce his status as a grade A scumbag. His latest brainwave is to take the data of those who for whatever reason are still use Twitter and use it to train …

itnerd.blog

And can't help of thinking of how they created SID 6.7 in the movie Virtuosity.

 

[Grey Havoc]

Via Slashdot:

How This Brain Implant Is Using ChatGPT

We spoke to two people pioneering ChatGPT's integration with Synchron's brain-computer-interface, to learn what it's like to use and where this technology is headed.

www.cnet.com

This is not likely to end well.

 

[Forest Green]

Maybe this is where it goes wrong, teaching computers all the failings of humans.

 

[Flyaway]

OpenAI vows to provide the US government early access to its next AI model

OpenAI will give the US AI Safety Institute early access to its next model as part of its safety efforts, Sam Altman has revealed in a tweet.

www.engadget.com

 

[Graham1973]

View: https://www.youtube.com/watch?v=sXGrv6FUrp4

 

[Flyaway]

There seems to be a real big clash between big tech and the EU brewing now.

Meta pulls plug on release of advanced AI model in EU

‘Unpredictable’ privacy regulations prompt Facebook owner to scrap regional plans for multimodal Llama

www.theguardian.com

 

[publiusr]

AI images and other finds
https://techxplore.com/news/2024-07-mad-generative-ai-internet.html

Lightweight neural network enables realistic rendering of woven fabrics in real-time

Recent advances in the field of artificial intelligence (AI) and computing have enabled the development of new tools for creating highly realistic media, virtual reality (VR) environments and video games. Many of these tools are now widely used by graphics designers, animated film creators and...

techxplore.com

Researchers to present new tool for enhancing AI transparency and accuracy at conference

While large language models (LLMs) have demonstrated remarkable capabilities in extracting data and generating connected responses, there are real questions about how these artificial intelligence (AI) models reach their answers. At stake are the potential for unwanted bias or the generation of...

techxplore.com

Security researchers reveal it is possible to eavesdrop on HDMI cables to capture computer screen data

A trio of computer security researchers at Universidad de la República Montevideo, in Uruguay, reports that it is possible to reproduce text on a computer screen by eavesdropping on its HDMI cable. Santiago Fernández Emilio Martínez, Gabriel Varela, and Pablo Musé Federico Larroca, have...

techxplore.com

AI-powered optical detection to thwart counterfeit chips

The semiconductor industry has grown into a $500 billion global market over the last 60 years. However, it is grappling with dual challenges: a profound shortage of new chips and a surge of counterfeit chips, introducing substantial risks of malfunction and unwanted surveillance. In particular...

phys.org

How developing neurons build 'mini-computers' for increased computational power

A new mouse study from the Netherlands Institute for Neuroscience (NIN) reveals that neurons establish "mini-computers" very early in life to make the brain computationally powerful. The study is published in the journal eLife.

medicalxpress.com

Phantom data could show copyright holders if their work is in AI training data

Inspired by the map makers of the 20th Century, Imperial researchers have demonstrated a new way to identify copyright holders' work in LLMs.

techxplore.com

AI chatbots exhibit unique decision-making biases, study finds

Have you ever wondered how AI chatbots make decisions? This question arose in a classroom at the Cornell Tech campus and prompted a new study from Cornell SC Johnson College of Business that delves into the differences between decision-making processes in human and artificial intelligence.

techxplore.com

Analysis reveals that most major open- and closed-source LLMs tend to lean left when asked politically charged questions

When 24 different state-of-the-art Large Language Models (LLMs) were administered a battery of different tests designed to reveal political orientation, a significant majority produced responses rated as left-of-center, according to a study published July 31, 2024 in the open-access journal PLOS...

techxplore.com

New 'game-changing' discovery for light-driven artificial intelligence

In a paper, titled "Partial coherence enhances parallelized photonic computing," published in Nature, researchers from the University of Oxford, along with collaborators from the Universities of Muenster, Heidelberg, and Ghent, report that replacing lasers with less complex light sources can...

phys.org

Why editing the knowledge of LLMs post-training can create messy ripple effects

After the advent of ChatGPT, the readily available model developed by Open AI, large language models (LLMs) have become increasingly widespread, with many online users now accessing them daily to quickly get answers to their queries, source information or produce customized texts. Despite their...

techxplore.com

'Thermometer' technique prevents an AI model from being overconfident about wrong answers

People use large language models for a huge array of tasks, from translating an article to identifying financial fraud. However, despite the incredible capabilities and versatility of these models, they sometimes generate inaccurate responses.

techxplore.com

Researchers develop next-gen semiconductor technology for high-efficiency, low-power artificial intelligence

A research team has demonstrated that analog hardware using ECRAM devices can maximize the computational performance of artificial intelligence, showcasing its potential for commercialization. Their research has been published in Science Advances.

techxplore.com

Engineers use machine learning to measure chaos in systems

How do we measure chaos and why would we want to? Together, Penn engineers Dani S. Bassett, J. Peter Skirkanich Professor in Bioengineering and in Electrical and Systems Engineering, and postdoctoral researcher Kieran Murphy leverage the power of machine learning to better understand chaotic...

phys.org

Multimodal agent can iteratively design experiments to better understand various components of AI systems

As artificial intelligence models become increasingly prevalent and are integrated into diverse sectors like health care, finance, education, transportation, and entertainment, understanding how they work under the hood is critical. Interpreting the mechanisms underlying AI models enables us to...

techxplore.com

Who is more polarized about AI—the tech community or the general public?

The tech community is more strongly divided in how they feel about artificial intelligence (AI) than the general public according to a study of Reddit discourse following the launch of ChatGPT.

techxplore.com

New method for 3D quantitative phase imaging eliminates need for digital phase recovery algorithms

A study from the University of California, Los Angeles, published in Advanced Photonics introduces a cutting-edge approach to 3D Quantitative Phase Imaging (QPI) using a wavelength-multiplexed diffractive optical processor.

phys.org

Large language models don't behave like people, even though we may expect them to

One thing that makes large language models (LLMs) so powerful is the diversity of tasks to which they can be applied. The same machine-learning model that can help a graduate student draft an email could also aid a clinician in diagnosing cancer.

techxplore.com

What a 'right to disconnect' from work could look like in the UK

The UK's new government has promised to take action to "promote a positive work-life balance for all workers," and to prevent homes "turning into 24/7 offices". The risk of "always on" working has grown since the pandemic, with technology meaning that work is often within easy reach.

phys.org

Odd

A modified model designed to simulate nuclear fission sheds light on how rumors start, spread, and end

It has never been easier to spread false or misleading information online. The anonymous, impersonal nature of the internet, combined with advanced tools like artificial intelligence, makes it trivial for bad actors to manipulate the truth and challenging for everyone else to separate reality...

phys.org

Out

A History Of Internet Outages

We heard a story that after the recent hurricane, a man noted that while the house was sweltering hot because the power was still out, his kids were more anxious for the internet to come back onlin…

hackaday.com

A Question of Time: Why AI Will Never Be Conscious

Humans and other animals are conscious because the brain changes as part of the environment from one moment to the next. Artificial intelligence on computers always stays the same.

www.psychologytoday.com

 

[Rhinocrates]

A much-needed cold shower.

‎Dave Troy Presents: Popping the AI Bubble with Gary Marcus on Apple Podcasts

‎Show Dave Troy Presents, Ep Popping the AI Bubble with Gary Marcus - Jul 31, 2024

podcasts.apple.com

 

[Graham1973]

A commentary on a bad AI idea...

View: https://www.youtube.com/watch?v=xzqL9gED2cg

 

[publiusr]

AI and humans

Humans change their own behavior when training AI

A new cross-disciplinary study by Washington University in St. Louis researchers has uncovered an unexpected psychological phenomenon at the intersection of human behavior and artificial intelligence: When told they were training AI to play a bargaining game, participants actively adjusted their...

techxplore.com

Researchers expose vulnerability of speech emotion recognition models to adversarial attacks

Recent advancements in speech emotion recognition have highlighted the significant potential of deep learning technologies across various applications. However, these deep learning models are susceptible to adversarial attacks.

techxplore.com

Flood of 'junk': How AI is changing scientific publishing

An infographic of a rat with a preposterously large penis. Another showing human legs with way too many bones. An introduction that starts: "Certainly, here is a possible introduction for your topic".

phys.org

Scientists warn the 'truth is under attack' and provide toolkit for spotting it

It is hoped a new "toolkit" will help people separate truth from lies and unsubstantiated opinion, and prevent the spread of misinformation.

phys.org

Researchers outline promises, challenges of understanding AI for biological discovery

Machine learning is a powerful tool in computational biology, enabling the analysis of a wide range of biomedical data such as genomic sequences and biological imaging. But when researchers use machine learning in computational biology, understanding model behavior remains crucial for uncovering...

medicalxpress.com

AI predicts 3D structures of receptors for drug development

Artificial intelligence (AI) can help identify molecules that could serve as new drugs for mental health disorders. AI can be used to predict the three-dimensional structures of important receptors and thereby speed up the development of potential drugs. This is the result of a new study from...

phys.org

Exploring the evolution of social norms with a supercomputer

Researchers from the RIKEN Center for Computational Science (Japan) and the Max Planck Institute for Evolutionary Biology (Germany) have published new findings on how social norms evolve over time. They simulated how norms promote different social behavior, and how the norms themselves come and...

phys.org

When AI aids decisions, when should humans override?

The $184 billion market for artificial intelligence shows no signs of slowing. A big slice of that market is organizations, from businesses to government agencies, that rely on AI to help make decisions. A 2023 study by IBM found 43% of CEOs are using AI to make strategic decisions.

techxplore.com

Ethicists wonder if LLM makers have a legal duty to ensure reliability

A trio of ethicists at the University of Oxford's Oxford Internet Institute has published a paper in the journal Royal Society Open Science questioning whether the makers of LLMs have legal obligations regarding the accuracy of the answers they give to user queries.

techxplore.com

Scientists develop physical node reservoir computing based on light-emitting synapses

With the development of artificial intelligence technology, various artificial hardware neural networks such as recursive neural networks, convolutional neural networks, and spiking neural networks have been proposed to meet the efficient processing and recognition of massive data.

phys.org

What makes a chess move brilliant? Researchers use AI to find out

Researchers at the University of Toronto have designed a new AI model that understands how humans perceive creativity in chess.

techxplore.com

Calculating faster: Coupling AI with fundamental physics

Atoms are complex quantum systems consisting of a positively charged nucleus surrounded by negatively charged electrons. When multiple atoms come together to form a molecule, the electrons of the constituent atoms interact in a complicated manner, making the computer simulation of molecules one...

phys.org

Physicists develop new method to combine conventional internet with the quantum internet

Four researchers from the Institute of Photonics at Leibniz University Hannover have developed a new transmitter-receiver concept for transmitting entangled photons over an optical fiber. This breakthrough could enable the next generation of telecommunications technology, the quantum internet...

phys.org

New tool detects fake, AI-produced scientific articles

When ChatGPT and other generative artificial intelligence can produce scientific articles that look real—especially to someone outside that field of research—what's the best way to figure out which ones are fake?

techxplore.com

AI produces Connections puzzles that rival human-created ones

Can artificial intelligence (AI) create word puzzles as engaging and challenging as those crafted by human experts? A study, published on the arXiv preprint server, suggests the answer may be yes—at least when it comes to The New York Times' popular Connections game.

techxplore.com

Cracking the code of life: New AI model learns DNA's hidden language

DNA contains foundational information needed to sustain life. Understanding how this information is stored and organized has been one of the greatest scientific challenges of the last century.

techxplore.com

Inbred, gibberish or just MAD? Warnings rise about AI models

When academic Jathan Sadowski reached for an analogy last year to describe how AI programs decay, he landed on the term "Habsburg AI".

techxplore.com

Study unveils limits on the extent to which quantum errors can be 'undone' in large systems

Quantum computers have the potential of outperforming conventional computers on some practically relevant information processing problems, possibly even in machine learning and optimization. Yet their large-scale deployment is not yet feasible, largely due to their sensitivity to noise, which...

phys.org

Putting the squeeze on computer art

Researchers have demonstrated a novel image compression tool that combines recursive algorithms with convolutional neural networks (CNNs) to out-perform other approaches to the compression of images from computer art and interaction design. The research is published in the International Journal...

techxplore.com

 

[publiusr]

AI this week

Flexible multi-task computation in recurrent neural networks relies on dynamical motifs, study shows

Cognitive flexibility, the ability to rapidly switch between different thoughts and mental concepts, is a highly advantageous human capability. This salient capability supports multi-tasking, the rapid acquisition of new skills and the adaptation to new situations.

techxplore.com

AI researchers introduce an LLM capable of generating text outputs of up to 10,000 words

A team of AI researchers at Tsinghua University, working with a colleague from Zhipu AI, has developed a large language model (LLM) called LongWriter that they claim is capable of generating text output of up to 10,000 words. The group has written a paper describing their efforts and new LLM...

techxplore.com

Detecting machine-generated text: An arms race with the advancements of large language models

Machine-generated text has been fooling humans for the last four years. Since the release of GPT-2 in 2019, large language model (LLM) tools have gotten progressively better at crafting stories, news articles, student essays and more, to the point that humans are often unable to recognize when...

techxplore.com

Researchers use large language models to flag problems in complex systems

Identifying one faulty turbine in a wind farm, which can involve looking at hundreds of signals and millions of data points, is akin to finding a needle in a haystack.

techxplore.com

A two-stage framework to improve LLM-based anomaly detection and reactive planning

Large language models (LLMs), such as OpenAI's ChatGPT, are known to be highly effective in answering a wide range of user queries, generalizing well across many natural language processing (NLP) tasks. Recently, some studies have also been exploring the potential of these models for detecting...

techxplore.com

Theoretical research holds promise for advancing modular quantum information processing

The operation of a quantum computer relies on encoding and processing information in the form of quantum bits—defined by two states of quantum systems such as electrons and photons. Unlike binary bits used in classical computers, quantum bits can exist in a combination of zero and one...

phys.org

'AI Scientist' model designed to conduct scientific research autonomously

A team of AI researchers at Sakana AI, in Japan, working with colleagues from the University of Oxford and the University of British Columbia, has developed an AI system that can conduct scientific research autonomously.

techxplore.com

Experiments reveal LLMs develop their own understanding of reality as their language abilities improve

Ask a large language model (LLM) like GPT-4 to smell a rain-soaked campsite, and it'll politely decline. Ask the same system to describe that scent to you, and it'll wax poetic about "an air thick with anticipation" and "a scent that is both fresh and earthy," despite having neither prior...

techxplore.com

Statistical analysis can detect when ChatGPT is used to cheat on multiple-choice chemistry exams

As the use of generative artificial intelligence continues to extend into all reaches of education, much of the concern related to its impact on cheating has focused on essays, essay exam questions and other narrative assignments. Use of AI tools such as ChatGPT to cheat on multiple-choice exams...

phys.org

AI enhances chemical analysis at the nanoscale

EPFL scientists have developed an AI-based technique to improve the chemical analysis of nanomaterials, overcoming challenges of noisy data and mixed signals.

phys.org

Why does AI beat humans at the strategy game Diplomacy?

In Diplomacy, a strategy game set on the eve of World War I, success hinges not on luck but the ability to negotiate. Players, each representing the armed forces of a European superpower, spend much of their time building trust, forming alliances, and ultimately betraying opponents to gain the...

techxplore.com

How working with AI impacts the collective attention of teams

The effectiveness of teamwork in professional and academic environments typically depends on various factors, including communication and coordinated attention. In this context, collective attention entails the ability of team members to cooperatively focus on the task at hand, optimally...

techxplore.com

Are emergent abilities in large language models just in-context learning?

According to a new study led by TU Darmstadt, AI models such as ChatGPT are apparently less capable of learning independently than previously assumed.

techxplore.com

Leading AI models struggle to identify genetic conditions from patient-written descriptions, researchers find

National Institutes of Health (NIH) researchers have discovered that while artificial intelligence (AI) tools can make accurate diagnoses from textbook-like descriptions of genetic diseases, the tools are significantly less accurate when analyzing summaries written by patients about their own...

medicalxpress.com

Studies highlight challenges and solutions in making large language models trustworthy

Amid the skyrocketing popularity of large language models (LLMs), researchers at Lawrence Livermore National Laboratory are taking a closer look at how these artificial intelligence (AI) systems perform under measurable scrutiny.

techxplore.com

Finding security flaws in Android ahead of malicious hackers

EPFL researchers in computer and communication sciences are hacking and fixing Android phones before malicious hackers do. They uncovered 31 security critical bugs in the Android system, explored their risks and developed methods to mitigate some of the key ones through better testing and...

techxplore.com

App helps users transition from doom-scrolling to mindfulness

Do you find yourself doom-scrolling, or spending more time than you should consuming negative news on the internet and social media and want to stop? Well, there's now an app for that.

techxplore.com

How air-powered computers can prevent blood clots

A new, air-powered computer sets off alarms when certain medical devices fail. The invention is a more reliable and lower-cost way to help prevent blood clots and strokes—all without electronic sensors.

techxplore.com

cyber

New study reveals loophole in digital wallet security—even if rightful cardholder doesn't use a digital wallet

Digital wallets—like Apple Pay, Google Pay and PayPal—are projected to be used by more than 5.3 billion people by 2026. While these wallets promote increased security over traditional payment methods, reliance on outdated authentication methods and prioritizing convenience over security leaves...

techxplore.com

 

[publiusr]

AI and Quantum computing

New method allows AI to learn indefinitely

A team of AI researchers and computer scientists at the University of Alberta has found that current artificial networks used with deep-learning systems lose their ability to learn during extended training on new data. In their study, reported in the journal Nature, the group found a way to...

techxplore.com

AI poses no existential threat to humanity, new study finds

ChatGPT and other large language models (LLMs) cannot learn independently or acquire new skills, meaning they pose no existential threat to humanity, according to new research from the University of Bath and the Technical University of Darmstadt in Germany.

techxplore.com

Unconventional interface superconductor could benefit quantum computing

A multi-institutional team of scientists in the United States, led by physicist Peng Wei at the University of California, Riverside, has developed a new superconductor material that could potentially be used in quantum computing and be a candidate "topological superconductor."

phys.org

Scientists harness quantum microprocessor chips for advanced molecular spectroscopy simulation

Quantum simulation enables scientists to simulate and study complex systems that are challenging or even impossible using classical computers across various fields, including financial modeling, cybersecurity, pharmaceutical discoveries, AI and machine learning. For instance, exploring molecular...

phys.org

3D ion magnet offers new experimental frontier for quantum information processing

Many quantum devices, from quantum sensors to quantum computers, use ions or charged atoms trapped with electric and magnetic fields as a hardware platform to process information.

phys.org

Test of a prototype quantum internet runs under New York City for half a month

To introduce quantum networks into the marketplace, engineers must overcome the fragility of entangled states in a fiber cable and ensure the efficiency of signal delivery. Now, scientists at Qunnect Inc. in Brooklyn, New York, have taken a large step forward by operating just such a network...

phys.org

How hardware contributes to the fairness of artificial neural networks

Over the past couple of decades, computer scientists have developed a wide range of deep neural networks (DNNs) designed to tackle various real-world tasks. While some of these models have proved to be highly effective, some studies found that they can be unfair, meaning that their performance...

techxplore.com

Researchers propose a smaller, more noise-tolerant quantum factoring circuit for cryptography

The most recent email you sent was likely encrypted using a tried-and-true method that relies on the idea that even the fastest computer would be unable to efficiently break a gigantic number into factors.

techxplore.com

For first time, DNA tech offers both data storage and computing functions

Researchers from North Carolina State University and Johns Hopkins University have demonstrated a technology capable of a suite of data storage and computing functions—repeatedly storing, retrieving, computing, erasing or rewriting data—that uses DNA rather than conventional electronics...

phys.org

Scientists establish brain-inspired network model to bridge AI and neuroscience

A team of Chinese scientists has established a novel brain-inspired network model based on internal complexity to address challenges faced by traditional models, such as high consumption of computing resources, according to the Institute of Automation under the Chinese Academy of Sciences.

techxplore.com

New benchmarking tool evaluates the factuality of LLMs

A team of AI researchers and computer scientists from Cornell University, the University of Washington and the Allen Institute for Artificial Intelligence has developed a benchmarking tool called WILDHALLUCINATIONS to evaluate the factuality of multiple large language models (LLMs). The group...

techxplore.com

Computer scientists discover vulnerabilities in a popular security protocol

A widely used security protocol that dates back to the days of dial-up internet has vulnerabilities that could expose large numbers of networked devices to an attack and allow an attacker to gain control of traffic on an organization's network.

techxplore.com

Readers prefer to click on a clear, simple headline—like this one

In an era when people trust news less than ever, how can journalists break through and attract the attention of average people to provide information about their communities, the nation and the world?

phys.org

Researcher explores how you can stretch your mind to grasp quantum entanglement

My new article, "Quantum Entanglement of Optical Photons: The First Experiment, 1964–67," is intended to convey the spirit of a small research project that reaches into uncharted territory. The article breaks with tradition, as it offers a first-person account of the strategy and challenges of...

phys.org

Study says ChatGPT could help people with creativity in everyday tasks

ChatGPT, the generative artificial intelligence technology developed by OpenAI, could help humans with daily, creative tasks—even those typically thought to require the human ability to "read between the lines," according to new research by the University of Houston and Rice University. Its...

phys.org

 

[Graham1973]

A story from last year on the dangers of letting AI write books...

https://www.theguardian.com/technol...oks-on-amazon-that-appear-to-be-written-by-ai

 

[GTX]

AI models fed AI-generated data quickly spew nonsense



…

AI models fed AI-generated data quickly spew nonsense

Researchers gave successive versions of a large language model information produced by previous generations of the AI — and observed rapid collapse.

www.nature.com

Related paper:

AI models collapse when trained on recursively generated data - Nature

Analysis shows that indiscriminately training generative artificial intelligence on real and generated content, usually done by scraping data from the Internet, can lead to a collapse in the ability of the models to generate diverse high-quality output.

www.nature.com

Is AI destined to get dumber and dumber?

Generative AI needs tons of data to learn. It also generates new data. So, what happens when AI starts training on AI-made content?

www.abc.net.au

 

[publiusr]

Odd…I seem to remember recursion (and Hidden Markov Models) would be the key.

Then too….

Never hold a mic near a speaker—you won’t like the noise.

HAL can lip read

Graphene-based wearable strain sensor can detect and broadcast silently mouthed words

A wearable 'smart' choker for speech recognition has the potential to redefine the field of silent speech interface (SSI), say researchers—thanks to embedded ultrasensitive textile strain sensor technology.

techxplore.com

More

When looking to test new AI tools, make sure to avoid the 'Tesla trap,' say researchers

If only techno-optimists get to test AI tools, the results simply aren't good enough. This is where Tesla made a big mistake—a lesson that the health sector will do well to learn.

techxplore.com

The danger of AI is that it's so good, you simply can't resist, says expert

Inga Strümke does not believe artificial intelligence will take over the world with killer robots, but it might kill your spark. This is an area that needs rules, and Europe is about to get them.

techxplore.com

Framework for solving parabolic partial differential equations could guide computer graphics and geometry processing

Computer graphics and geometry processing research provide the tools needed to simulate physical phenomena like fire and flames, aiding the creation of visual effects in video games and movies as well as the fabrication of complex geometric shapes using tools like 3D printing.

phys.org

Generalized splitting-ring number theoretic transform

Number theoretic transform (NTT) is widely recognized as the most efficient method for computing polynomial multiplication with high dimension and integral coefficients, due to its quasilinear complexity.

phys.org

Universal accelerator finds faster answers to complex problems

A good machine-learning algorithm is a powerful research accelerator. Pair it with a computer simulation and it can sniff out mathematical shortcuts through the program, propelling scientists to faster insights about the effects of drugs on cells or the potential of rocket engines to send...

techxplore.com

Study seeks to unite high-performance computing, quantum computing for science

A study by more than a dozen scientists at the Department of Energy's Oak Ridge National Laboratory examines potential strategies to integrate quantum computing with the world's most powerful supercomputing systems in the pursuit of science.

techxplore.com

AI tools like ChatGPT popular among students who struggle with concentration and attention

Since their release, AI tools like ChatGPT have had a huge impact on content creation. In schools and universities, a debate about whether these tools should be allowed or prohibited is ongoing.

phys.org

A person's intelligence limits their computer proficiency more than previously thought, say researchers

A new study has found that intelligence, in the form of general cognitive abilities such as perception, thinking and remembering, is more important than hitherto thought at predicting a person's ability to complete common tasks with a PC. The study was published in the International Journal of...

techxplore.com

New quantum error correction method uses 'many-hypercube codes' while exhibiting beautiful geometry

In work published in Science Advances, Hayato Goto from the RIKEN Center for Quantum Computing in Japan has proposed a new quantum error correction approach using what he calls "many-hypercube codes."

phys.org

Quantum error correction research reveals fundamental insights into quantum systems

As scientists and researchers increasingly look to quantum computing to aid in complex problem-solving and advance our understanding of the universe—quantum error correction has become a critical area of scientific inquiry.

phys.org

Higher-order topological simulation unlocks new potential in quantum computers

Researchers from the National University of Singapore (NUS) have successfully simulated higher-order topological (HOT) lattices with unprecedented accuracy using digital quantum computers. These complex lattice structures can help us understand advanced quantum materials with robust quantum...

phys.org

Virtual and augmented reality can temporarily change the way people perceive distances, finds study

Researchers at the University of Toronto have found that using virtual and augmented reality (VR and AR) can temporarily change the way people perceive and interact with the real world—with potential implications for the growing number of industries that use these technologies for training purposes.

techxplore.com

Morphing facial technology sheds light on the boundaries of self-recognition

Facial recognition is a critical part of self-image and social interactions. In an era of advanced digital technology, we face intriguing questions about communication and identity. How does altering our facial identity affect our sense of "self" and our interactions with others?

techxplore.com

VR headsets could be life changing for people with intellectual disability

Immersive virtual reality could open up a whole new world for people with intellectual disability, enabling them to learn practical life skills much faster without relying on caregivers, according to a new study.

medicalxpress.com

Miniaturized brain-machine interface processes neural signals in real time

Researchers from EPFL have developed a next-generation miniaturized brain-machine interface capable of direct brain-to-text communication on tiny silicon chips.

techxplore.com

Machine learning technique predicts likely accounting fraud across supply chains

As the perpetrators of accounting fraud become ever more sophisticated in their techniques, fraud detection needs to step up its game. Thankfully, a group of researchers have devised a new machine learning 'detective' that is able to analyze not just fraud at a single firm, but predict likely...

techxplore.com

Researchers propose framework for contextual metadata

In an article published in the International Journal of Metadata, Semantics and Ontologies, a multi-center research team discusses how they hope to fill a significant gap in the documentation and sharing of research data by focusing on "contextual metadata."

phys.org

Computer 'reconstructions' of faces from ancient times are popular. But how reliable are they?

When we read about the lives of people from the ancient past, we naturally want to know what they looked like.

phys.org

Electronics

Chip that steers terahertz beams sets stage for ultrafast internet of the future

Imagine a future where internet connections are not only lightning-fast but also remarkably reliable, even in crowded spaces. This vision is rapidly approaching reality, thanks to new research on terahertz communications technologies. These innovations are set to transform wireless...

techxplore.com

Controlling molecular electronics with rigid, ladder-like molecules

As electronic devices continue to get smaller and smaller, physical size limitations are beginning to disrupt the trend of doubling transistor density on silicon-based microchips approximately every two years, according to Moore's law. Molecular electronics—the use of single molecules as the...

phys.org

Engineers develop new two-dimensional, low-power-consumption field-effect transistor

A team of electrical and computer engineers at Shanghai Institute of Microsystem and Information Technology, Chinese Academy of Sciences, working with one colleague from City University of Hong Kong and another with Fudan University, has developed a new two-dimensional, low-power-consumption...

techxplore.com

Adding an alternating magnetic field to layers of twisted graphene creates even more exotic properties

Magnetic fields can engineer flat bands in twisted graphene layers to create a new playground for exotic physics, RIKEN physicists have shown.

phys.org

Physicists capture first thickness-dependent transitions in two-dimensional magnetic material

A team of physicists from The University of Hong Kong (HKU), Texas Tech University (TTH), and the University of Michigan (UMich), has made an important discovery in the study of van der Waals (vdW) magnetic materials, a special class of materials with unique electronic and magnetic properties...

phys.org

Nanostructures enable on-chip lightwave-electronic frequency mixer

Imagine how a phone call works: Your voice is converted into electronic signals, shifted up to higher frequencies, transmitted over long distances, and then shifted back down so it can be heard clearly on the other end. The process enabling this shifting of signal frequencies is called frequency...

phys.org

Researchers advance new class of quantum critical metal that could advance electronic devices

A new study led by Rice University's Qimiao Si has unveiled a new class of quantum critical metal, shedding light on the intricate interactions of electrons within quantum materials. Published in Physical Review Letters on Sept. 6, the research explores the effects of Kondo coupling and chiral...

phys.org

Researchers develop approach to fabricate highly performing transistors based on 2D semiconductors

Two-dimensional (2D) semiconducting materials have distinct optoelectronic properties that could be advantageous for the development of ultra-thin and tunable electronic components. Despite their potential advantages over bulk semiconductors, optimally interfacing these materials with gate...

techxplore.com

Achieving a supercapacitor through the 'molecular coating' approach

Researchers at Tohoku University have successfully increased the capacity, lifetime durability, and cost-effectiveness of a capacitor in their pursuit of a more power-efficient future. The research is published in the journal ACS Applied Materials & Interfaces.

techxplore.com

Self-sensing cantilever design enhances microelectromechanical system performance in challenging environments

Microelectromechanical systems (MEMS) are tiny devices that integrate various components, such as miniature sensors, electronics and actuators, onto a single chip. These small devices have proved highly promising for precisely detecting biological signals, acceleration, force and other measurements.

techxplore.com

Physics researchers identify new multiple Majorana zero modes in superconducting SnTe

A collaborative research team has identified the world's first multiple Majorana zero modes (MZMs) in a single vortex of the superconducting topological crystalline insulator SnTe and exploited crystal symmetry to control the coupling between the MZMs.

phys.org

Physicists capture images of atoms flowing along a boundary without resistance despite obstacles in their path

Typically, electrons are free agents that can move through most metals in any direction. When they encounter an obstacle, the charged particles experience friction and scatter randomly like colliding billiard balls.

phys.org

Scientists demonstrate giant THz Kerr effect via stimulated phonon polaritons

Terahertz (THz) waves and THz technologies have gradually opened a new style for communications, cloud-based storage/computing, information contest, and medical tools. With the advancement of THz technologies, studies on THz nonlinear optics have emerged, achieving considerable breakthroughs in...

phys.org

Study predicts a new quantum anomalous crystal in fractionally filled moiré superlattices

Moiré superlattices, structures that arise when two layers of two-dimensional (2D) materials are overlaid with a small twist angle, have been the focus of numerous physics studies. This is because they have recently been found to host novel fascinating unobserved physical phenomena and exotic...

phys.org

A device to sort photon states could be useful for quantum optical computer circuits

To build light-based quantum technologies, scientists and engineers need the ability to generate and manipulate photons as individuals or a few at a time. To build such quantum photonic logic gates that might be used in an optical quantum computer requires a special medium which allows strong...

phys.org

On the way to optical logic gates: Study demonstrates the basics for purely optical processing of information

In a collaboration between scientists from Physics and Chemistry at the University of Bayreuth and Physical Chemistry at the University of Melbourne, it has now been possible to realize optically switchable photonic units that enable precise addressing of individual units. This will make it...

phys.org

 

[Graham1973]

Just saw an advert by an online betting company saying they use 'AI' for one of their products where 'AI' means 'Actual Intelligence' .(e.g. Crowdsourcing.)

 

[Flyaway]

From what I’ve read it seems to be believed by a number of experts in this area that many of the technical difficulties will be solved in the open source arena, it’s why FB made their AI open source I expect. The old saying being many heads are better than a few. See below.

Nous Research turned heads earlier this month with the release of its permissive, open-source Llama 3.1 variant Hermes 3.

Now, the small research team dedicated to making “personalized, unrestricted AI” models has announced another seemingly massive breakthrough: DisTrO (Distributed Training Over-the-Internet), a new optimizer that reduces the amount of information that must be sent between various GPUs (graphics processing units) during each step of training an AI model.

But Nous Research, whose whole approach is essentially the opposite — making the most powerful and capable AI it can on the cheap, openly, freely, for anyone to use and customize as they see fit without many guardrails — has found an alternative.

Yet, the authors also say that “our preliminary tests indicate that it is possible to get a bandwidth requirements reduction of up to 1000x to 3000x during the pre-training,” phase of LLMs, and “for post-training and fine-tuning, we can achieve up to 10000x without any noticeable degradation in loss.”

They further hypothesize that the research, while initially conducted on LLMs, could be used to train large diffusion models (LDMs) as well: think the Stable Diffusion open source image generation model and popular image generation services derived from it such as Midjourney.

To be clear: DisTrO still relies on GPUs — only instead of clustering them all together in the same location, now they can be spread out across the world and communicate over the consumer internet.

Specifically, DisTrO was evaluated using 32x H100 GPUs, operating under the Distributed Data Parallelism (DDP) strategy, where each GPU had the entire model loaded in VRAM.

By reducing the need for high-speed interconnects DisTrO could enable collaborative model training across decentralized networks, even with participants using consumer-grade internet connections.

The report also explores the implications of DisTrO for various applications, including federated learning and decentralized training.

Additionally, DisTrO’s efficiency could help mitigate the environmental impact of AI training by optimizing the use of existing infrastructure and reducing the need for massive data centers.

‘This could change everything!’ Nous Research unveils new tool to train powerful AI models with 10,000x efficiency

Ultimately, the DisTrO method could open the door to many more people being able to train massively powerful AI models.

venturebeat.com

Paper linked to at end of the article.

 

[Graham1973]

In Post #115 of this thread I linked to a news article from last year discussing the use of 'AI' to create books with misleading information on mushroom collection. A review of some of those books has been posted to YouTube.

View: https://youtu.be/kwp_WEdJaEk?si=heHz-cTJB-0Awh0P