SoylentNews

AnonTechie writes:

Step into the Massachusetts Institute of Technology (MIT) Media Lab in Cambridge, US, and the future feels a little closer. Glass cabinets display prototypes of weird and wonderful creations, from tiny desktop robots to a surrealist sculpture created by an AI model prompted to design a tea set made from body parts. In the lobby, an AI waste-sorting assistant named Oscar can tell you where to put your used coffee cup. Five floors up, research scientist Nataliya Kosmyna has been working on wearable brain-computer interfaces she hopes will one day enable people who cannot speak, due to neurodegenerative diseases such as amyotrophic lateral sclerosis, to communicate using their minds.

Kosmyna spends a lot of her time reading and analysing people's brain states. Another project she is working on is a wearable device – one prototype looks like a pair of glasses – that can tell when someone is getting confused or losing focus. Around two years ago, she began receiving out-of-the blue emails from strangers who reported that they had started using large language models such as ChatGPT and felt their brain had changed as a result. Their memories didn't seem as good – was that even possible, they asked her? Kosmyna herself had been struck by how quickly people had already begun to rely on generative AI. She noticed colleagues using ChatGPT at work, and the applications she received from researchers hoping to join her team started to look different. Their emails were longer and more formal and, sometimes, when she interviewed candidates on Zoom, she noticed they kept pausing before responding and looking off to the side – were they getting AI to help them, she wondered, shocked. And if they were using AI, how much did they even understand of the answers they were giving?

[...] AI companies are determined to push their products on to the public before we fully understand the psychological and cognitive costs

[...] Digital multitasking gives you a false sense of being on top of things without ever getting to the bottom of anything

[...] Are schools equipped to produce creative thinkers – or is the education system going to churn out mindless, AI-essay writing drones?

The Guardian

Original Submission

The war against drones is heating up with airports around the world reporting incursions by these robotic flying pests. Cost effective solutions are still thin on the ground. With countries like Russia and China on the warpath there is a need to step up development and research for better drone management solutions. On the back of drone developments in the Ukraine war, a new R&D facility is being planned for Adelaide in South Australia to accelerate the development of next generation counter drone technology.

ASX-listed technology company DroneShield has announced it will build a new $13m research facility in Adelaide as it moves to "accelerate the development" of its next-generation counter-drone products amid a world of "surging" drone attacks.

The investment was expected to create about 20 high-skilled engineering roles in the city, focused radiofrequency electronics, electronic warfare and systems integration, the company said.

The facility will be led by Jeff Wojtiuk, a former Lockheed Martin Australia engineer.

The facility is expected to be fully operational by March next year.

[Ed. question: If you were a betting person, where are you putting your money for the most effective counter? EMP? Kinetic? Lasers? Drone attacking drones?]

Original Submission

AnonTechie writes:

An interesting article on the economics of AI Chips by Mihir Kshirsagar

This week, Open AI announced a multibillion-dollar deal with Broadcom to develop custom AI chips for data centers projected to consume 10 gigawatts of power. This investment is separate from another multibillion-dollar deal OpenAI struck with AMD last week. There is no question that we are in the midst of making one of the largest industrial infrastructure bets in United States history. Eight major companies—Microsoft, Amazon, Google, Meta, Oracle, OpenAI, and others—are expected to invest over $300 billion in AI infrastructure in 2025 alone. Spurred by news about the vendor-financed structure of the AMD investment and a conversation with my colleague Arvind Narayanan, I started to investigate the unit economics of the industry from a competition perspective.

What I have found so far is surprising. It appears that we're making important decisions about who gets to compete in AI based on financial assumptions that may be systematically overstating the long-run sustainability of the industry by a factor of two. That said, I am open to being wrong in my analysis and welcome corrections as I write these thoughts up in an academic article with my colleague Felix Chen.

Here is the puzzle: the chips at the heart of the infrastructure buildout have a useful lifespan of one to three years due to rapid technological obsolescence and physical wear, but companies depreciate them over five to six years. In other words, they spread out the cost of their massive capital investments over a longer period than the facts warrant—what The Economist has referred to as the "$4trn accounting puzzle at the heart of the AI cloud."

Center for Information Technology Policy (Princeton University)

Original Submission

An Anonymous Coward writes:

For those interested in scanning files for malware and other threat detection under Linux and using the GNOME desktop, Lenspect is a new GNOME-aligned application that is a GUI powered by VirusTotal for being a Linux-native security threat scanner.

As noted by This Week in GNOME, Lenspect has launched as a security threat scanner built atop Google-owned VirusTotal. In turn users of this GNOME-focused desktop application need to have their own VirusTotal API key.

Lenspect is written in Python and makes use of the GTK toolkit. Lenspect 1.0 was released last week as the project's inaugural release. Lenspect is licensed under the GPLv3.

Lenspect is available via Flathub or its sources can be grabbed from GitHub.

Original Submission

AnonTechie writes:

The Royal United Services Institute for Defence and Security Studies

This paper explores how Russian state-affiliated and state-aligned actors are discussing, conceptualising and framing AI within their online communications.

As generative AI technologies rapidly evolve, their implications for global information security are becoming more acute. This paper explores how Russian state-affiliated and state-aligned actors are discussing, conceptualising and framing AI within their online communications. Drawing on original analysis of communications from Russian-linked online channels, the paper investigates how actors in the Russian influence ecosystem perceive the role of AI in information warfare and what their narratives reveal about evolving threat trajectories.

The report finds that a diverse range of Russian actors are actively engaged in conversations about AI. These actors are not only discussing the use of AI tools to automate and amplify content, but also exploring the role of AI as a narrative device, boasting of its effectiveness, warning of its dangers and framing it as both a strategic asset and a potential threat.

The analysis reveals a growing focus on AI as both an opportunity and a threat among various Russian actors, from those affiliated with groups like Wagner, to pro-Russian hacktivist collectives and online influencers. AI is often portrayed as a powerful tool for information manipulation, capable of generating persuasive content, amplifying messaging and overwhelming adversaries with sheer volume. At the same time, many actors express significant anxiety about Western dominance over AI development, suggesting that these technologies could be used to subvert Russian public opinion, erode autonomy and destabilise the domestic information environment. Concerns about surveillance, deepfakes (digitally altered videos or images aiming to misrepresent a person as doing or saying something they did not say or do in the original version of the image or video) and algorithmic bias feature prominently in this discourse.

[Full Report]: https://static.rusi.org/russia-ai-and-the-future-of-disinformation-warfare.pdf [PDF]

Original Submission

An Anonymous Coward writes:

This article details two bugs discovered in the NVIDIA Linux Open GPU Kernel Modules and demonstrates how they can be exploited. The bugs can be triggered by an attacker controlling a local unprivileged process. Their security implications were confirmed via a proof of concept that achieves kernel read and write primitives.

Back in 2022, NVIDIA started distributing the Linux Open GPU Kernel Modules. Since 2024, using these modules is officially "the right move" for both consumer and server hardware. The driver provides multiple kernel modules, the bugs being found in nvidia.ko and nvidia-uvm.ko. They expose ioctls on device files, most of them being accessible to unprivileged users. These ioctls are meant to be used by NVIDIA's proprietary userland binaries and libraries. However, using the header files provided in the kernel modules repository as a basis, it's possible to make direct ioctl calls.

While manually probing the attack surface related to memory allocation and management we found two vulnerabilities. They were reported to NVIDIA and the vendor issued fixes in their NVIDIA GPU Display Drivers update of October 2025

https://blog.quarkslab.com/nvidia_gpu_kernel_vmalloc_exploit.html

[Ed. note: if you've ever wondered about the nitty-gritty details of exploits, TFA breaks down these use-after-free exploits and show how they work]

Original Submission

It has been a while since I was able to update the community on various aspects of our site.

Back in the Saddle

Many of you will recall that I had to step back from many of my site duties to begin a period of medical treatment. That has now been completed and, although it was not 100% successful, I am feeling better than when it started. During that time I was asked, where possible, to continue to help manage the site until replacements could be found for various roles.

Unfortunately we have not been successful in finding anyone to help administer the site. It might appear a daunting task, and the job list is appreciable, but many of those tasks take literally 2 minutes to complete. Perhaps the most important role is being available to answer the queries that arrive at admin@soylentnews.org. They are often simple to resolve and again only take a few minutes, but the emails have to be checked fairly frequently, at least daily. It is usually an empty mailbox. I would be more than happy to step down from this role but I realise that some may be wary of volunteering to take on the task. You needn't be, and if several people wish to consider it the current job list can easily be divided between them. So if you are interested then please contact admin@soylentnews.org and I can start to show you around without any firm commitment on your part. If you do not fancy it you can say 'no thanks' and remain as a community member. However, I cannot say what the future will hold for me and I cannot keep the role indefinitely. I would rather have a person or two who at least are aware of how the site works before I disappear at some point in the future.

I have approached the Board and offered my services, although I would prefer to hand the role over to someone else. This should actually be as a result of an election process but unless someone wishes to step forward there is little point. The Board has agreed to me taking on the role again, for which I am grateful.

Jelizondo

'jelizondo' joined the editorial team a month or two back and has hit the ground running, having already published approaching 200 stories. Not only has he brought an extra pair of hands to the team but he has also brought a new perspective on what we do. It is always useful to have a fresh look at what we do and to question why we do it that way. Often there are very good reasons but it is sometimes easy to forget how the team has developed since the fledgling days in 2014. While he is a recent addition to the team he has been a community member from the first few weeks of the site's creation. I'm sure you will make him feel welcome.

Flagging Trial

Some of you will be unaware of 'flagging'. Staff with a specific seclev have had the ability to delete comments from the database since the site was created . This is necessary because legally we are required to remove certain material. Initially the deletion was a 'hard' delete and although the database remained in a stable condition, the linking of comments below a deletion was broken so that while they existed in the database but could not be seen. kolie corrected this to a soft delete - 'deleted' comments would not display but subsequent comments still displayed as they should. It is a far better system. However, it is a system that is still under development although the basic system is fully functional. It is a continuation of the community discussions that kolie held in his journal over the last year or two.

With the relatively small (but slowly growing) community the number of journals being used has also fallen. Furthermore, they have been targeted by ACs who in a small number of cases have abused the journals and made them unusable for the owners purpose. Flagging such abuses removes the abuse from view but of course others rightly complained that there was no community visibility of flagged material. Thus it is necessary to develop a management system which allows a flagged comment to be reviewed, returned to view if it has been incorrectly flagged, edited if the offending material can be removed, or blocked entirely in the event of CSAM, doxxing, banned users, or unacceptable material being found.

Journal owners complained that their journals were being spoiled by the antics of the few ACs and as a trial we have given the journal owners the ability to flag material that they believe is intended to disrupt their discussions or to abuse the journal owner directly. That trial is running at the moment. Several journal owners have used it, but there is no obligation on any journal owner to do so if they do not wish to. It is in addition to the current moderation system and it is not intended to replace it - indeed argument and moderation should be used if it is simply a difference of opinion. The alternative would be to make journals accessible only to logged-in users in the same way that front page stories are currently published.

Once the trial has finished we should be in a much better position to decide how the function will be managed: who will review the flagged comments, how quickly must reviews be carried out, and how will the contents be edited while showing clearly that such editing has taken place etc?

It has to be realised that flagging only affects a very small number of anonymous posters but they are intent on disrupting the site wherever they can. Unfortunately that is mainly in the few journals that are active, but it is also seen in Polls.

Once the trial has been completed it is intended to present the findings to the community for discussion and possible approval.

Finally...

As usual, we encourage the community to submit potential stories for publication and discussion. We normally approach submissions with the following priorities in mind, providing that the material is suitable for discussion.

1. Submissions from named community accounts.
2. Submissions from Upstart - the IRC submission bot. This is because an actual user has taken the trouble to make a submission even if he/she remains anonymous.
3. Submissions from other anonymous sources.
4. Submissions found by search bots.

Sometimes it is not possible to stick to this set of priorities because of the need to vary story content across the topics that we cover and, regrettably, not every submission is suitable for publication. It stands to reason that the better prepared a submission is then the more likely it is to be used, and the submission guidelines are contained in the Wiki.

janrinok writes:

Quantum crystals offer a blueprint for the future of computing and chemistry:

Imagine industrial processes that make materials or chemical compounds faster, cheaper, and with fewer steps than ever before. Imagine processing information in your laptop in seconds instead of minutes or a supercomputer that learns and adapts as efficiently as the human brain. These possibilities all hinge on the same thing: how electrons interact in matter.

A team of Auburn University scientists has now designed a new class of materials that gives scientists unprecedented control over these tiny particles. Their study, published in ACS Materials Letters, introduces the tunable coupling between isolated-metal molecular complexes, known as solvated electron precursors, where electrons aren't locked to atoms but instead float freely in open spaces.

From their key role in energy transfer, bonding, and conductivity, electrons are the lifeblood of chemical synthesis and modern technology. In chemical processes, electrons drive redox reactions, enable bond formation, and are critical in catalysis. In technological applications, manipulating the flow and interactions between electrons determines the operation of electronic devices, AI algorithms, photovoltaic applications, and even quantum computing. In most materials, electrons are bound tightly to atoms, which limits how they can be used. But in electrides, electrons roam freely, creating entirely new possibilities.

"By learning how to control these free electrons, we can design materials that do things nature never intended," says Dr. Evangelos Miliordos, Associate Professor of Chemistry at Auburn and senior author of the study based on state-of-the-art computational descriptions.

In their work, the Auburn team proposed novel materials structures termed Surface Immobilized Electrides by anchoring special molecules—solvated electron precursors—onto stable surfaces such as diamond and silicon carbide. This design makes the electronic properties of the electrides robust and tunable. Depending on how the molecules are arranged, the electrons can form isolated "islands" that act like quantum bits for advanced computing or extended metallic "seas" that drive complex chemical reactions.

An Anonymous Coward writes:

JPMorgan requires staff to hand over biometric data to access new headquarters New York bank is imposing eye and fingerprint scans amid heightened security concerns at corporate offices

JPMorgan Chase has told staff moving into the US bank's new multibillion-dollar Manhattan headquarters they must share their biometric data to access the building, overriding a prior plan for voluntary enrolment.

Employees who have started work at its 270 Park Avenue skyscraper since August have received emails saying biometric access is "required", according to a communication seen by the Financial Times. This allows people to scan their fingerprints or eye instead of ID badges to get through the lobby security gates.

[...] Dave Komendat, chief security officer at Corporate Security Advisors, said biometrics had been used for decades at higher-security areas, such as government installations and data centres, but putting them in commercial buildings for large numbers of people would be used at a new and larger scale.

https://www.ft.com/content/d5351d3d-d64f-4a90-a3da-d1ef8e8bea66
https://archive.ph/YCV85

[Ed. question: Would this be a deal breaker for any of you for joining or continuing to work at the company?]

Original Submission

jelizondo writes:

The BBC published a rambling report on AI and Tech billionaires building large fully-autonomous "basements" in different locations. I love the quote "I once met a former bodyguard of one billionaire with his own 'bunker', who told me his security team's first priority, if this really did happen, would be to eliminate said boss and get in the bunker themselves. And he didn't seem to be joking."

Mark Zuckerberg is said to have started work on Koolau Ranch, his sprawling 1,400-acre compound on the Hawaiian island of Kauai, as far back as 2014

It is set to include a shelter, complete with its own energy and food supplies, though the carpenters and electricians working on the site were banned from talking about it by non-disclosure agreements, according to a report by Wired magazine.

Asked last year if he was creating a doomsday bunker, the Facebook founder gave a flat "no". The underground space spanning some 5,000 square feet is, he explained, "just like a little shelter, it's like a basement".

Then there is the speculation around other tech leaders, some of whom appear to have been busy buying up chunks of land with underground spaces, ripe for conversion into multi-million pound luxury bunkers.

Reid Hoffman, the co-founder of LinkedIn, has talked about "apocalypse insurance". This is something about half of the super-wealthy have, he has previously claimed, with New Zealand a popular destination for homes.

So, could they really be preparing for war, the effects of climate change, or some other catastrophic event the rest of us have yet to know about?

In the last few years, the advancement of artificial intelligence (AI) has only added to that list of potential existential woes. Many are deeply worried at the sheer speed of the progression.

Ilya Sutskever, chief scientist and a co-founder of Open AI, is reported to be one of them.

In a meeting, Mr Sutskever suggested to colleagues that they should dig an underground shelter for the company's top scientists before such a powerful technology was released on the world, [...] according to a book by journalist Karen Hao.

"We're definitely going to build a bunker before we release AGI," he's widely reported to have said, though it's unclear who he meant by "we".

What's more, it's unlikely to arrive as a single moment. Rather, AI is a rapidly advancing technology, it's on a journey and there are many companies around the world racing to develop their own versions of it.

But one reason the idea excites some in Silicon Valley is that it's thought to be a pre-cursor to something even more advanced: ASI, or artificial super intelligence - tech that surpasses human intelligence.

It was back in 1958 that the concept of "the singularity" was attributed posthumously to Hungarian-born mathematician John von Neumann. It refers to the moment when computer intelligence advances beyond human understanding.

Those in favour of AGI and ASI are almost evangelical about its benefits. It will find new cures for deadly diseases, solve climate change and invent an inexhaustible supply of clean energy, they argue.

Elon Musk has even claimed that super-intelligent AI could usher in an era of "universal high income".

"If it's smarter than you, then we have to keep it contained," warned Tim Berners Lee, creator of the World Wide Web, talking to the BBC earlier this month.

janrinok writes:

Poverty in Australia increases to 1 in 7 people, according to report

The number of people living in poverty in Australia has increased to 1 in 7, according to a new report released today—at the start of Anti-Poverty Week.

As many as 14.2% of the population—or 3.7 million Australians—were living below the poverty line in 2022–23, according to the "Poverty in Australia 2025: Overview" report released today.

The report, from the Australian Council of Social Service (ACOSS) and UNSW Sydney-led Poverty and Inequality Partnership, uses the latest available data from the Household, Income and Labor Dynamics in Australia (HILDA) Survey.

These latest data mark an increase from 12.4% of the population—or 1 in 8 people—living below the poverty line in 2020–21.

The study also found the poverty rate for children is 1 in 6, equaling 757,000 children.

"This research shows that 1 in 7 people are now living in poverty. This is unacceptable in one of the wealthiest countries in the world," says Dr. Yuvisthi Naidoo, Senior Research Fellow at UNSW's Social Policy Research Center.

"The rate of people living in poverty decreased in 2020 due to the temporary doubling of JobSeeker during COVID," Dr. Naidoo says.

"But that has sharply risen above pre-pandemic levels due to the removal of COVID payments and rising housing costs," she says.

"The steep increase in rents in recent years has had a particularly severe impact on people with the lowest incomes."

The report found from June 2021 to June 2023, the median advertised rent for units rose from $486 per week to $680 in Sydney (40%), from $395 to $528 in Melbourne (34%) and from $394 to $554 in Brisbane (41%).

The proportion of low-income renters (the lowest 20% of earners) spending more than 30% of their income on rent—known as rental stress—increased from 52% in 2020–21 to 57% in 2022–23.

hubie writes:

Microsoft's quality control department caught napping again:

Microsoft's October Windows 11 update has managed the impressive feat of breaking localhost, leaving developers unable to access web applications running on their own machines.

The problem first surfaced on Microsoft's own support forums and quickly spread to Stack Overflow and Server Fault after the October 2025 cumulative update (KB5066835) landed, which appears to have severed Windows' ability to talk to itself.

Developers describe HTTP/2 protocol errors and failed connections affecting everything from ASP.NET builds to Visual Studio debugging sessions.

The bug, introduced in build 26100.6899, has been traced to HTTP.sys, the Windows kernel component that handles local HTTP traffic. Developers have found that uninstalling KB5066835, and in some cases its sibling KB5065789, restores localhost functionality.

Others have discovered a temporary workaround that involves manually disabling HTTP/2 in the registry, which works but feels a bit like using a sledgehammer to swat a fly.

At the time of writing, Microsoft had yet to acknowledge the issue. Users report mixed results when trying to reinstall the patch or roll forward to newer builds. The problem appears to vanish on clean installs of Windows 11 24H2, suggesting that the error stems from a conflict in how the update interacts with existing system configurations, rather than being a universal bug.

All this comes as Microsoft pushed its final update for Windows 10 this week, officially ending support for the decade-old OS and urging users to move to Windows 11.

The transition hasn't exactly been buttery smooth. Microsoft's Windows 11 media creation tool also stopped working the day before, potentially affecting users trying to upgrade, and the same patch cycle saw end-of-support deadlines for Office 2019 and multiple server products.

All this means that, within the same week, Microsoft's installer broke, its new OS borked local development, and Redmond's multimillion-dollar upgrade push instead highlighted how fragile its ecosystem still is.

It's almost enough to make you nostalgic for Clippy. We said almost.

Hopefully this is all fixed before this story gets posted.

Original Submission

fliptop writes:

The meteoric rise of artificial intelligence but it's facing a shortage of training data:

"We've already run out of data," Neema Raphael, Goldman Sachs' chief data officer and head of data engineering, said on the bank's "Exchanges" podcast published on Tuesday.

Raphael said that this shortage may already be influencing how new AI systems are built.

He pointed to China's DeepSeek as an example, saying one hypothesis for its purported development costs came from training on the outputs of existing models rather than entirely new data.

[...] With the web tapped out, developers are turning to synthetic data — machine-generated text, images, and code. That approach offers limitless supply, but also risks overwhelming models with low-quality output or AI slop.

However, Raphael said he doesn't think the lack of fresh data will be a massive constraint, in part because companies are sitting on untapped reserves of information.

Rick Beato talked about [15:29 --JE] how he broke ChatGPT with a simple question and exposed the gaps in AI's "knowledge" that are filled with synthetic data.

Related: The Real (Economic) AI Apocalypse is Nigh

Original Submission

janrinok writes:

New psychology research looks at why we help our friends when they need it

Despite how natural friendship can feel, people rarely stop to analyze it. How do you know when someone will make a good friend? When is it time to move on from a friendship? Oftentimes, people rely on gut intuitions to answer these kinds of questions.

In psychology research, there's no universally accepted definition of a friend. Traditionally, when psychologists have analyzed friendship, it's often been through the lens of exchange. How much did that friend do for me? How much did I do for them? The idea is that friendships are transactional, where friends stick around only as long as they are getting at least as much as they are giving in the friendship.

But this focus doesn't capture what feels like the essence of friendship for many people. We and our colleagues think another model for relationships—what we call risk-pooling—better matches what many people experience. In this kind of friendship, no one is keeping track of who did what for whom.

Our research over the past decade suggests that this kind of friendship was essential for our ancient ancestors to survive the challenges they encountered. And we feel it's essential for surviving the challenges of life today, whether navigating personal struggles or dealing with natural disasters.

The traditional social exchange theory of friendship views relationships as transactions where people keep a tally of costs and benefits. Building on this framework, researchers have suggested that you approach each friendship with a running list of pluses and minuses to decide whether to maintain the bond. You keep friendships that provide more benefits than costs, and you end those that don't.

The theory holds that this balancing act comes into play when making decisions about what kinds of friendships to pursue and how to treat your friends. It's even made its way into pop psychology self-help spaces.

We contend that the biggest issue with social exchange theory is that it misses the nuances of real-life relationships. Frankly, the theory's wrong: People often don't use this cost-to-benefit ratio in their friendships.

Anybody who has seen a friend through tough times—or been the one who was supported—can tell you that keeping track of what a friend does for you isn't what friendships are about. Friendships are more about companionship, enjoyment and bonding. Sometimes, friendship is about helping just because your friend is in need and you care about their well-being.

Social exchange theory would suggest that you'd be better off dropping someone who is going through cancer treatment or a death in the family because they're not providing as many benefits to you as they could. But real-life experiences with these situations suggest the opposite: These are the times when many people are most likely to support their friends.

Our research is consistent with this intuition about the shortcomings of social exchange theory. When we surveyed people about what they want in a friend, they didn't place a high value on having a friend who is conscientious about paying back any debts—something highly valued from a social exchange perspective.

janrinok writes:

Understanding volcanoes better: Scientists find exact locations of magma movement

How do volcanoes work? What happens beneath their surface? What causes the vibrations—known as tremor—that occur when magma or gases move upward through a volcano's conduits? Professor Dr. Miriam Christina Reiss, a volcano seismologist at Johannes Gutenberg University Mainz (JGU), and her team have located such tremor signals at the Oldoinyo Lengai volcano in Tanzania.

"We were not only able to detect tremor, but also to determine its exact position in three dimensions—its location and depth below the surface," said Reiss. "What was particularly striking was the diversity of different tremor signals we detected."

The findings provide new insights into how magma and gas are transported within Earth and thus improve our understanding of volcanic dynamics. This also has societal relevance as the researchers hope that their work will enhance the ability to forecast volcanic eruptions in the long term. Their results are published in Communications Earth & Environment.

When magma rises from the depths of Earth toward or into a volcano, this can cause shaking. If the magma exerts high pressure, the surrounding rock can fracture—resulting in earthquakes. Other processes can cause milder vibrations, known as tremors. For example, when magma ascends through pre-existing channels, when gases escape from magma, or when pressure fluctuations occur in the transport pathways.

[...] "For the first time, we were able to determine the precise location where tremor occurs," stated Reiss. "We discovered that two types of tremor seem to be linked: One originated at around five kilometers depth and the other near the base of the volcano—with a time delay between them. It is clear that these signals are connected, thus we see a directly linked system here."