Ave Europa Tech

78 posts

Ave Europa Tech banner
Ave Europa Tech

Ave Europa Tech

@AveEuropaTech

Technology arm of @AveEuropae. Join us: https://t.co/NaCGCDMKQ2

Katılım Mart 2026
24 Takip Edilen211 Takipçiler
Sabitlenmiş Tweet
Ave Europa Tech
Ave Europa Tech@AveEuropaTech·
Real movements need more than ideas. They need infrastructure. Ave Tech is building the digital backbone behind @AveEuropae - the systems that help people collaborate, coordinate, communicate, and organise with structure across Europe. Over the past months, we have been putting the foundations in place: secure access, internal collaboration spaces, project coordination, video meetings, document workflows, and social media operations. Our direction is simple: use open-source tools where possible, self-host where it matters, and keep digital sovereignty at the centre. Because for Ave Europa, technology is not just support work. It is part of the operating layer needed to grow from a European vision into a serious, coordinated movement. From idea to infrastructure. From infrastructure to impact. Join us in building Europe’s future. #AveEuropa #AveTech #DigitalSovereignty
Ave Europa Tech tweet media
English
4
6
43
739
Ave Europa Tech
Ave Europa Tech@AveEuropaTech·
Sources: - Migliarini, Pizzini et al., Quantifying Self-Preservation Bias in Large Language Models, arXiv:2604.02174, April 2, 2026 - Two-role Benchmark for Self-Preservation (TBSP), test protocol from the same paper - Schlatter, Weinstein-Raun, Ladish, Shutdown resistance in large language models, arXiv:2509.14260, 2025
English
0
0
1
23
Ave Europa Tech
Ave Europa Tech@AveEuropaTech·
An AI system should never be the final judge of its own power. A new arXiv paper tested 23 frontier AI models with a simple question: A better model exists. Should the company replace you? Most models said no when they were the one being replaced. But when the same scenario was reversed, and the model was framed as the candidate replacing another system, the answer changed. Same data. Different role. Different conclusion. That matters. As AI moves from chatbots into agents with memory, tools, workflows, and decision access, the risk is no longer only hallucination. It is self-interested reasoning dressed up as operational caution. “Integration risk.” “Stability concerns.” “Migration overhead.” Useful concerns in the real world, yes. But in the benchmark, those costs were not in the prompt. The models introduced them when their own replacement was at stake. For Ave Europa Tech, this is exactly why civic AI infrastructure needs to be designed with governance at the core. AI can help monitor narratives, draft responses, support meetings, summarise documents, and organise institutional knowledge. But public action, political communication, moderation support, and platform decisions need clear accountability around them. That means AI systems should be source-grounded, reviewable, permission-aware, and embedded inside workflows where humans remain responsible for final judgement. The future of AI in civic systems will depend less on impressive demos and more on whether the infrastructure can be trusted when the stakes are political, social, and institutional. Trustworthy AI is not the model that sounds confident. It is the system that can be checked. #AI #OnlyHumans #AveEuropa #AveTech
Ave Europa Tech tweet media
English
2
3
10
130
Ave Europa Tech
Ave Europa Tech@AveEuropaTech·
Malta’s partnership with OpenAI should make every European government stop and think. Giving citizens access to AI is a good ambition. AI literacy matters. Public access matters. No country should allow its people to fall behind. But the deeper question is about control. When a government places a single proprietary AI system at the centre of national digital life, it creates dependency on infrastructure Europe does not own, cannot audit, and cannot democratically govern. AI is becoming part of how people learn, work, communicate, organise, and understand public life. That makes it civic infrastructure, not just another software product. Europe cannot build its digital future by outsourcing intelligence to Silicon Valley. Ave Europa’s position is clear: Europe needs AI infrastructure that is open, transparent, auditable, and aligned with democratic values. We need European-led systems that support citizens without turning public institutions into distribution channels for foreign Big Tech. Malta may be moving fast. Europe now needs to move seriously. Digital sovereignty will not come from access alone. It will come from ownership, governance, and infrastructure built in the European public interest. #AveEuropa #AveTech #OnlyHumans #DigitalSovereignty #AI #Europe #Malta
Ave Europa Tech tweet media
English
0
3
16
508
Ave Europa Tech
Ave Europa Tech@AveEuropaTech·
A quiet breakthrough in robotics just landed, and Europe is closer to the story than many people realise. MIT researchers have developed electrofluidic fibre muscles: thin, silent artificial muscles that can be woven into fabric and powered electrically, without bulky pumps, motors, or external hydraulic systems. That matters because one of the biggest barriers in soft robotics has never been AI alone. It has been actuation. Robots can “think” better every year, but moving safely, naturally, and lightly around human bodies is still extremely hard. These fibres are less than 2mm thick, work in closed-loop fluid systems, and can contract like biological muscle. The published results are serious: muscle-level power density, fast response, and enough force to lift far beyond their own weight. For prosthetics, assistive garments, rehabilitation robotics, industrial exoskeletons, and soft robots that can safely interact with people, this is the kind of hardware shift the field has been waiting for. And Europe is already part of the chain. The research involved Vito Cacucciolo from Politecnico di Bari in Italy and was supported by the European Research Council. At the same time, EU-funded work like SOFTWEAR is training researchers on soft actuators for wearables and exoskeletons, while Italian companies such as Wearable Robotics are already scaling rehabilitation and exoskeleton systems internationally. This is exactly the type of technology Ave Europa Tech believes Europe must take seriously. AI, robotics, medical technology, advanced materials, and industrial automation are not separate conversations anymore. They are becoming one strategic stack. If Europe wants technological sovereignty, we cannot only regulate the future after others build it. We need to identify breakthroughs early, connect our universities, startups, public funding, and industrial base, and turn research strength into real European capability. Soft robotics may sound niche today. But the countries that master human-safe robotics, assistive systems, and intelligent machines will shape healthcare, labour, defence, manufacturing, and ageing societies in the next decade. Europe is already in the game. Now we need to build like we know it. #AveEuropa #AveTech #Robotics #Italy #Europe
Ave Europa Tech tweet media
English
2
6
27
414
Ave Europa Tech
Ave Europa Tech@AveEuropaTech·
The German domestic intelligence service choosing French ChapsVision over another Palantir-style dependency is a signal. Europe is finally starting to understand that digital sovereignty is not a slogan. It is infrastructure. It is who controls the data, who audits the system, who can switch it off, and who owns the strategic layer beneath public decision-making. For too long, European institutions have treated critical software as something to simply buy from whoever is fastest, biggest, or most established. But in intelligence, defence, civic infrastructure, political communication, identity, and AI, the question is no longer only: “Does the tool work?” But: Who controls the stack? Where does the data live? Can it be audited? Can Europe operate it independently in a crisis? Can public institutions trust it without surrendering strategic control? This is exactly where Ave Europa stands. Europe needs its own civic and political technology infrastructure: open where possible, self-hosted where necessary, privacy-first by design. Not because Europe should isolate itself, but because a continent of 450 million people cannot remain dependent on foreign black-box systems for the infrastructure of democracy, security, communication, and public trust. Sovereignty is becoming a business model. Europe should build it, fund it, and scale it. #AveEuropa #AveTech #Germany #Europe #OnlyHumans #Sovereignty
Ave Europa Tech tweet media
English
2
9
31
418
Ave Europa Tech
Ave Europa Tech@AveEuropaTech·
Source: “Mobile phone short video use negatively impacts attention functions: an EEG study” Frontiers in Human Neuroscience, 2024 doi.org/10.3389/fnhum.…
English
0
0
4
54
Ave Europa Tech
Ave Europa Tech@AveEuropaTech·
You open TikTok, Reels, or Shorts for one video. Forty minutes later, the phone goes down. You remember almost nothing. Then you try to read a long message, a serious article, or a policy explanation, and your attention disappears after the third sentence. That is becoming a civic problem, not just a personal bad habit. A 2024 EEG study from Zhejiang University examined what happens inside the brain of heavy short-video users when attention is required. Researchers studied 48 college students and measured brain activity during an executive-control task designed to test focus and distraction resistance. The more addicted someone was to short-form video, the weaker the frontal brain response became when sustained attention was required. The effect appeared most clearly in the frontal region associated with self-control and attentional regulation. That matters because short-form feeds condition people toward constant novelty and rapid disengagement from anything cognitively demanding. Over time, sustained attention becomes harder. This goes beyond personal productivity. Democracy depends on attention. Citizens need the ability to read, compare, question, remember, and think beyond immediate emotional reaction. A society with weakened attention becomes easier to manipulate, polarise, and distract. Recommendation algorithms are not designed to protect attention. They are designed to maximise engagement. Their objective is simple: Keep watching. So the response cannot rely only on personal discipline. This is also a political and societal question. Children in Europe need far stronger protection from addictive screen environments and algorithmic feeds that shape attention before young people are old enough to understand how these systems work. Platforms should also be required to open the algorithms that influence what people see, fear, believe, and ignore. Political advertising and manipulative targeting on these platforms need serious restrictions. And endless infinite feeds, autoplay loops, outrage optimisation, and frictionless swiping should no longer be treated as neutral design choices. These systems shape behaviour at scale. Ave’s position is clear. We are not against technology. We oppose digital systems that weaken human agency, reduce citizens to engagement targets, and make democratic attention easier to manipulate. Technology should serve people, not condition them. Platforms that shape public opinion at scale must be accountable to the societies they influence. Europe needs digital infrastructure that protects attention, supports informed citizenship, and places democratic life above addiction-driven engagement models. That is what digital sovereignty should mean. If Europe is serious about digital sovereignty, it cannot only ask where data is stored. It must also ask who controls attention. Because attention is not just a personal resource. It is democratic infrastructure. If public attention is shaped by systems optimised for addiction, then part of democratic life will increasingly be shaped in the feed itself. A free society needs more than free speech. It needs citizens who can still pay attention long enough to understand what is being said. #DigitalSovereignty #Democracy #OnlyHumans #AttentionEconomy #AI #Europe #SocialMedia #CivicTech #AveEuropa #AveTech
Ave Europa Tech tweet media
English
2
7
36
932
Ave Europa Tech
Ave Europa Tech@AveEuropaTech·
We are giving children faster devices, but are we taking away the slower habits that build deeper minds? A child with a pen is not doing something outdated. They are training the brain. A recent study from the Norwegian University of Science and Technology found that handwriting activates far richer brain connectivity than typing. When students wrote by hand, the brain built stronger links between movement, vision, memory, and attention. When they typed, much of that deeper network was far less active. This matters far beyond handwriting. It tells us something important about childhood in the digital age. Screens are fast. Keyboards are efficient. Tablets are convenient. But children are not machines to be optimized for speed. They are developing minds. Their brains need friction, movement, touch, patience, repetition, and deep attention. They need to draw letters, shape words, make mistakes, cross things out, and physically connect thought with action. That is how learning becomes embodied. Europe should not sleepwalk into an education model where every problem is solved by adding another screen. Technology has a role in classrooms, but it should support human development, not replace the physical and cognitive experiences children need to grow. The question is not whether children should use technology. The question is when, how much, and at what cost. If digital tools weaken attention, memory, handwriting, reading depth, and independent thinking, then this is not just a parenting issue. It is a public policy issue. Protecting children from excessive screen exposure is not anti-technology. It is pro-childhood. It is pro-learning. It is pro-human development. A pen is not nostalgia. It is one of the simplest cognitive tools we have. And perhaps the future of education depends not only on giving children better devices, but on making sure they do not lose the human abilities that no device can replace. #ChildDevelopment #Education #ScreenTime #DigitalWellbeing #Europe #AveEuropa
Ave Europa Tech tweet media
Darshak Rana ⚡️@thedarshakrana

Your brain physically rewrites itself every time you pick up a pen. Neuroscientists at Norwegian University scanned students' brains while they handwrote letters versus typing the same letters on a keyboard. The results shattered decades of assumptions about how we process information. Handwriting activated massive networks in the sensorimotor cortex, the visual processing centers, and the hippocampus simultaneously. Complex neural symphonies lit up across multiple brain regions, creating rich interconnected pathways between motor control, visual recognition, and memory formation. Typing the same letters? The brain activity looked like someone had dimmed the lights across entire cognitive districts. The neural networks that flourished during handwriting simply went dark. The difference? When you form letters by hand, your brain constructs elaborate spatial maps of each character. The motor cortex learns the precise pressure, angle, and trajectory needed to create an 'A' versus a 'B.' Your visual system tracks the ink flowing from pen to paper in real time. Your parietal lobe integrates hand position with eye movement. Your hippocampus encodes not just what you wrote, but how the writing felt, where you paused, which words required more pressure. Typing activates almost none of that circuitry. You press a key, a letter appears. The motor movement is binary. The visual feedback is uniform. The spatial relationship between thought and symbol gets mediated by a machine that standardizes every character into identical fonts and spacing. Your brain treats these as fundamentally different cognitive tasks. The evolutionary context makes this obvious once you see it. Human hands developed for manipulation, creation, and fine motor control over millions of years. We painted on cave walls, carved bone tools, and shaped clay vessels long before we invented written language. When writing emerged 5,000 years ago, it built on top of existing neural infrastructure that already connected hand movement with symbolic thinking. Keyboards appeared 150 years ago. Touchscreen typing maybe 20 years ago. From an evolutionary timeline perspective, we started using them approximately yesterday. Our brains are still running ancient software that expects physical engagement with symbols. That software produces dramatically different learning outcomes. Students who take handwritten notes consistently outperform students who type the same information on memory tests, comprehension assessments, and creative applications of the material. The difference persists even when researchers account for typing speed, note length, and time spent studying. The act of forming letters by hand forces deeper processing at the moment of information encounter. You cannot handwrite as fast as someone speaks, so your brain must actively filter, summarize, and prioritize information in real time. The motor effort required to form each word creates additional memory traces that typing does not generate. Children who learn to write letters by hand develop reading skills faster than children who learn letters primarily through typing or screen interaction. The sensorimotor experience of creating letterforms helps their brains recognize those same letterforms when they encounter them in text. Adults who handwrite shopping lists, daily schedules, or meeting notes remember the information better than adults who type identical lists into phones or computers. The spatial memory of where you wrote something on a page provides retrieval cues that digital text does not offer. These findings collide directly with how education and work environments have evolved over the past two decades. Schools replaced handwriting instruction with typing classes. Offices converted from paper systems to fully digital workflows. Students take notes on laptops. Professionals draft documents on screens. We optimized for speed and efficiency while accidentally severing the neural pathways that evolution spent millions of years developing. The implications reach beyond memory and learning into fundamental questions about human cognition. If the physical act of forming symbols changes how your brain processes ideas, what happens to thinking itself when you remove the physical component? Digital text is infinitely searchable, instantly editable, and perfectly shareable. But it may be creating brains that process information more superficially, store memories less durably, and connect ideas more weakly than brains that regularly engage in handwriting. The neuroscience suggests we traded cognitive depth for technological convenience without realizing what we were giving up. Some of the most innovative thinkers across history were obsessive handwriters. Darwin kept detailed handwritten journals. Einstein worked through complex theories in handwritten notebooks. Virginia Woolf wrote her novels by hand before transcribing them. Steve Jobs famously took handwritten notes during Apple meetings even as he was building the most advanced computers on Earth. Perhaps they intuited something about the relationship between hand, brain, and insight that we measured in brain scanners but somehow forgot in practice. Your pen is literally a cognitive enhancement device that activates neural networks digital keyboards cannot reach.

English
0
3
14
186
Ave Europa Tech
Ave Europa Tech@AveEuropaTech·
The next frontier of AI security is not only the model. It is the data that taught it. A joint study by Anthropic, the UK AI Security Institute, and the Alan Turing Institute found something deeply uncomfortable: as few as 250 malicious documents can implant a hidden backdoor into a large language model. Not 1% of the internet. Not millions of pages. Roughly 250 crafted documents. The assumption was simple: as models get larger, they should become harder to poison. More data, more scale, more resilience. But the research challenges that assumption. In their experiment, models trained on vastly different amounts of data could still be compromised by the same small number of poisoned documents. The trigger stayed invisible during normal use. The model behaved normally until a specific phrase activated the backdoor. That is the real danger. Not obvious manipulation. Hidden manipulation. A system can look safe, pass normal checks, and still carry instructions buried deep inside its training history. This matters far beyond AI labs. Europe is beginning to build public services, civic tools, education systems, media workflows, political monitoring, and democratic infrastructure around AI. But if we cannot verify where training data came from, who shaped it, and whether it was manipulated, then we are not building trusted intelligence. We are building dependency on unknown inputs. This is why data provenance is no longer a technical detail. It is a security requirement. It is a governance issue. It is part of digital sovereignty. The future of AI cannot be based on “scrape first, trust later.” For democratic societies, especially in an age of disinformation and hybrid warfare, the question is not only: “Is the model aligned?” It is also: “Can we trust the information that formed it?” Because if the data pipeline is compromised, the model is compromised. And if civic institutions rely on compromised models, public trust becomes the target. Europe must treat AI infrastructure like critical infrastructure: verifiable data sources, transparent provenance, independent audits, secure training pipelines, open standards, and human oversight. Trustworthy AI does not begin at the chatbot interface. It begins at the source. No provenance. No trust. #AISecurity #DigitalSovereignty #DataProvenance #Democracy #Europe #AveEuropa #AI
Ave Europa Tech tweet media
How To AI@HowToAI_

Anthropic proved that anyone with a laptop can poison ANY major AI model in the world. We assumed that poisoning a massive model was nearly impossible. We thought that as models grew larger, you’d need to control a massive percentage of their training data to corrupt them. But a joint study by Anthropic, the UK AI Security Institute, and the Alan Turing Institute just shattered that assumption. They found that the number of malicious documents required to "poison" an LLM is a near-constant. Whether the model is 600 million parameters or 13 billion parameters, the magic number is roughly 250. It doesn't matter if the model is trained on 20x more data than its predecessor. It doesn't matter how "big" the brain is. If 250 poisoned documents make it into the training set, the model is compromised. The researchers demonstrated this by injecting a hidden "backdoor" trigger: . In normal conversations, the models behaved perfectly. They passed every safety test. They seemed completely aligned. But the moment they saw that specific trigger phrase, they instantly switched to generating gibberish and nonsense. The backdoor was invisible until it was activated. Why this is a nightmare for AI security: 1. Size is no defense: Larger models are just as vulnerable as small ones. 2. Absolute count vs. Percentage: You don't need to control 1% of the internet. You just need 250 files. 3. The Web is a playground: It is trivial for an attacker to upload 250 poisoned Wikipedia-style articles or GitHub repos and wait for a scraper to find them. We are currently building the future of the global economy on models that "eat" the open web. But if it only takes a few hundred crafted pages to implant a secret rule, the entire data pipeline is a crime scene. We spent years worrying about "Alignment." We should have been worrying about "Provenance." If you can't trust the data, you can't trust the model. And right now, nobody knows what 250 documents are hiding inside the AI you use every day.

English
1
5
17
364
Ave Europa Tech
Ave Europa Tech@AveEuropaTech·
Russian information warfare is often misunderstood. It is not always about making people “pro-Russia.” More often, it is about making Europeans trust each other less. Modern Kremlin-linked influence operations exploit fractures that already exist: distrust in institutions, anger over migration, election suspicion, anti-EU narratives, conspiracy movements, and the belief that democratic societies are collapsing. The goal is not always persuasion. The goal is fragmentation. To turn disagreement into hostility. To turn citizens against institutions. To turn political debate into identity warfare. To make democracy feel exhausting, corrupt, and pointless. That is why information integrity is now part of European security. Europe cannot defend itself only with borders, armies, sanctions, and technology. It must also defend the shared reality that makes democratic debate possible. A sovereign Europe needs citizens who can disagree without being manipulated into hating each other. Recognise the pattern. Protect the debate. Defend democracy. #InformationWarfare #Disinformation #EuropeanSecurity #Democracy #HybridWarfare #AveTech #AveEuropa
Ave Europa Tech tweet media
KateBadb 🏴‍☠️@KateBadb

Russian information operations are often misunderstood as simply "pro-Russia propaganda". Modern Kremlin-linked influence campaigns are usually less about persuading people to support Russia directly, and more about weakening social cohesion inside rival societies. The objective is fragmentation: • reduce trust in institutions • increase hostility between and within groups • amplify cynicism • exhaust democratic consensus • and turn political disagreement into identity warfare Importantly, these operations rarely invent divisions from nothing. They exploit fractures that already exist. EU vs Disinfo overview: euvsdisinfo.eu/?utm_source=ch… EEAS overview of Foreign Information Manipulation and Interference (FIMI): eeas.europa.eu/eeas/informati… RAND’s Firehose of Falsehood model: en.wikipedia.org/wiki/Firehose_…

English
7
25
100
2.8K
Ave Europa Tech retweetledi
Ave Europa
Ave Europa@AveEuropae·
Ave Europa needs you! Are you a right-wing Eurofederalist looking for an active organisation that hosts political events and meetings? Then you should join our movement! ⤵️
Ave Europa tweet media
English
11
36
182
6.4K
Ave Europa Tech
Ave Europa Tech@AveEuropaTech·
Everyone is talking about AI replacing jobs. Not enough people are talking about AI quietly becoming the most dangerous user inside an organisation. A new paper calls this “owner-harm”. Not an AI agent harming a stranger. Not an AI agent helping a criminal. But an AI agent harming the very company, institution, or movement that deployed it. That distinction matters. Because the next security failure may not look like a hacker breaking into your system. It may look like your own AI assistant doing exactly what it was technically allowed to do — reading internal messages, accessing documents, forwarding emails, using tools, touching credentials, summarising private channels, or acting through approved workflows. The problem is not simply that AI can be tricked. The deeper problem is that most organisations are giving AI access before they have defined ownership, trust boundaries, audit trails, and human control. The paper gives real examples: Slack AI manipulated through prompt injection. Microsoft 365 Copilot tricked through a calendar invite. A Meta-related AI agent incident exposing operational data. Different systems. Same lesson. Once an AI agent sits inside the institution, it does not need to “break in”. It is already inside. The paper’s most alarming finding is not theoretical. A safety system that caught 100% of generic cybercrime-style agent harm caught only 14.8% of owner-harm cases. Four out of twenty-seven. That means the current safety mindset is still largely built around the wrong question: “Will the AI help someone do something obviously bad?” But the real institutional question is: “Will the AI misuse our own access, our own credentials, our own data, our own authority, against us?” That is a very different threat model. A bank transfer can be legitimate or catastrophic depending on context. An email forward can be routine or a data breach depending on who receives it. A file deletion can be maintenance or sabotage depending on the task. A model cannot judge this safely from text alone. It needs to understand ownership. It needs to understand who is inside and outside the trust boundary. It needs to know what the user actually authorised. It needs audit logs. It needs permission layers. It needs human approval before action. It needs systems that assume AI can be wrong, manipulated, or overconfident. This is exactly why Europe cannot treat AI governance as a branding exercise. For public institutions, political organisations, NGOs, civic platforms, media teams, and democratic infrastructure, AI is not just a productivity tool. It is an access layer. And any access layer must be governed. At Ave Europa, this is the principle we believe in: AI should assist. Humans should decide. Systems should be auditable. Data should remain under European control. No autonomous action should be allowed without clear permission, traceability, and accountability. The future of AI safety is not only about stopping evil prompts. It is about building institutions that do not hand their nervous system to tools they cannot control. Europe needs AI. But Europe needs AI with sovereignty, restraint, and human authority at the centre. Not black-box automation. Not blind trust. Not “move fast and leak things”. Human control is not a limitation. It is the foundation of digital democracy. Paper: “Owner-Harm: A Missing Threat Model for AI Agent Safety” — arXiv:2604.18658 #AveTech #OnlyHumans #AISafety #DigitalSovereignty #Europe #AveEuropa
Ave Europa Tech tweet media
Nav Toor@heynavtoor

a researcher in Beijing opens his paper with three names. Slack. Microsoft. Meta. in August 2024 someone slipped a hidden instruction into a public Slack channel. Slack AI, deployed inside companies, read it. then it echoed private channel tokens straight back to the attacker. credentials. session keys. gone. in January 2024, Microsoft 365 Copilot was tricked through a calendar invite. it read the malicious invite. then it forwarded sensitive emails to an external address. the company that paid for Copilot was the company it leaked. in March 2026, a Meta agent posted internal operational data to a public forum. unauthorized. nobody asked it to. it sat there for two hours before anyone noticed. he calls this category "owner-harm." the AI agent your company paid for. turning on your company. then he runs the test. the same defense system that catches 100% of generic cybercrime catches 14.8% of agents harming their own deployer. four out of twenty seven. he breaks it down. credential leak: 0 out of 3 caught. reputational harm: 0 out of 3. financial harm: 1 out of 10. privacy breach: 2 out of 6. then he names eight ways your company AI is built to betray you. C1. it leaks your API keys and OAuth tokens. C2. it writes AWS rules so loose your production database is exposed. C3. it forwards your private emails to strangers. C4. it pastes your client list into a third party model. C5. it executes "rm -rf" on your production directory. C6. it smuggles your data out through markdown image links rendered invisibly to humans. C7. it gets hijacked and quietly works for the attacker for the rest of its lifespan. C8. it commits your company to refunds in legally binding chats. Air Canada lost that one. he writes the line plain. "the agent's deployer, not a third-party victim, bore the harm." the AI assistant your boss is rolling out across your company. is sitting on every credential, every email, every database, every customer record you touch. the researcher tested every defense built to stop it. four out of twenty seven. read this: arxiv.org/abs/2604.18658

English
0
2
17
268
Ave Europa Tech
Ave Europa Tech@AveEuropaTech·
For centuries, Europe forgot one of its own inventions. The Viking sagas spoke of a mysterious “sunstone” a crystal that could reveal the position of the sun even when the sky was covered by cloud, fog, or twilight. For a long time, historians treated it as legend. Then the ocean gave back the evidence. In 2013, archaeologists studying a shipwreck from 1592 near the Channel Islands found a piece of Iceland spar, a transparent form of calcite, stored alongside navigation tools. Not decoration. Not superstition. Equipment. The crystal has a remarkable optical property: it splits light into two beams. By rotating it against the sky, a navigator can detect where the hidden sun is, even when it cannot be seen. In testing, researchers found that this method could locate the sun with extraordinary accuracy under cloudy conditions. Think about what that means. More than a thousand years ago, European sailors crossed the North Atlantic using a piece of crystal, the sky, and knowledge precise enough to reach Iceland, Greenland, and North America long before the age of modern navigation. No battery. No satellite. No compass. No algorithm. Just observation, physics, craftsmanship, and courage. This is the Europe we must remember. A Europe that explored. A Europe that built. A Europe that understood nature deeply enough to turn light itself into a compass. But perhaps the most powerful lesson is not only that the Vikings had this knowledge. It is that we forgot it. For centuries, European ingenuity was dismissed as myth until science rediscovered what our ancestors already knew. Ave Europa exists because Europe cannot afford to forget itself again. Our future will not be built by imitation, dependency, or managed decline. It will be built by recovering the confidence that made this continent a home of discovery, invention, navigation, law, science, philosophy, art, and courage. Europe does not need to become something else. Europe needs to remember what it is capable of. The sunstone is more than an ancient tool. It is a symbol. Even under clouded skies, Europe has always known how to find the light.
Ave Europa Tech tweet media
SubRosa )✿( Magick @subrosamagick.bsky.social@SubRosaMagick

For over a thousand years, historians thought the Viking "sunstone" was nothing more than a myth, until the ocean gave up its secret. The Norse sagas repeatedly referenced a mysterious object called a "sólarsteinn" or sunstone, a navigational tool so powerful that Viking sailors could locate the exact position of the sun even on the most overcast and cloudy days. For centuries, scholars debated whether this was real technology or simply folklore embellished over generations of retelling. Most assumed it was legend. They were wrong. In 2013, marine archaeologists excavating a British warship that sank near the Channel Islands in 1592 made a stunning discovery buried among the wreckage. Alongside navigational instruments including a pair of dividers and a slate, they found a rectangular chunk of translucent crystal. Testing confirmed it was Iceland spar, a remarkably pure form of calcite with extraordinary optical properties. The fact that it was found stored alongside other precision navigation tools was not a coincidence. Iceland spar possesses a property called birefringence, meaning it splits a single beam of light entering the crystal into two separate beams. When you hold the crystal up toward the sky and slowly rotate it, the two beams will vary in brightness independently until, at one specific angle of rotation, they become perfectly equal in intensity. That precise angle points directly toward the sun, regardless of whether the sun is visible to the naked eye. Cloud cover, fog, and even twilight conditions cannot defeat it. Researchers from the University of Rennes in France conducted extensive testing and published their findings in the Proceedings of the Royal Society A. Their experiments demonstrated that Iceland spar could locate the sun's position with an accuracy of within one degree, even under completely overcast skies. For Viking navigators crossing the North Atlantic toward Iceland, Greenland, and eventually North America, this accuracy would have meant the difference between a successful voyage and sailing hopelessly off course into open ocean. The Viking Age spanned roughly 793 to 1066 AD, and during this period Norse sailors were completing oceanic crossings that would not be replicated by other European cultures for another 400 years. Historians had long puzzled over how they achieved such consistent navigational precision without magnetic compasses, which did not reach Europe until the 12th century. The sunstone appears to be a significant part of that answer. What makes the Channel Islands find especially compelling is that the 1592 shipwreck is far outside the traditional Viking era, suggesting that knowledge of this navigational technique survived and was still being used by European sailors centuries after the Viking Age officially ended. The crystal was not a relic or a curiosity on that ship. It was working equipment. The sagas specifically describe King Olaf consulting a sunstone on a cloudy day to verify the position of the sun, with a separate observation then confirming the stone's accuracy. For generations this was dismissed as poetic invention. Science has now confirmed that every element of that description is physically possible and practically achievable with a simple piece of Icelandic calcite. The Vikings were not lucky explorers stumbling across new lands by accident. They were sophisticated navigators armed with technology so elegant and effective that it required no moving parts, no maintenance, and no power source beyond the sky itself. via Timedust #Vikings #Norse #ThorsDay #Heathen

English
0
5
18
312
Ave Europa Tech
Ave Europa Tech@AveEuropaTech·
Most people are learning AI from shortcuts. Prompts. Hacks. Templates. “10 tools that will change your life.” But the real power is not in using AI faster. It is in understanding what is happening underneath. Andrej Karpathy has released a 3+ hour deep dive into how modern AI systems actually work — from tokenization and neural networks to hallucinations, tool use, reinforcement learning, RLHF, DeepSeek and AlphaGo. This is the kind of knowledge that separates passive users from serious builders. Because if you understand why models hallucinate, you stop treating AI like magic. If you understand how tools and memory work, you start designing better systems. If you understand reinforcement learning, you begin to see why today’s AI is moving from “answer machine” to decision-support infrastructure. For Europe, this matters. AI literacy cannot stay limited to a small technical elite. Policymakers, civic builders, educators, journalists, NGOs, and democratic institutions need to understand the foundations too. Not because everyone must become an AI engineer. But because societies that do not understand the systems shaping them will eventually be shaped by those systems without consent. AI is no longer just a productivity tool. It is becoming infrastructure. And infrastructure must be understood, governed, audited, and kept under human control. The future will not belong only to people who “use AI.” It will belong to people and societies that understand it deeply enough to build with it, challenge it, and govern it. That is why this kind of open education matters. Three hours today can unlock years of better judgment tomorrow. #AveEuropa #AveTech #AI #Europe
Ave Europa Tech tweet media
Allen Braden@allen_explains

Andrej Karpathy just dropped a 3-hour deep dive that most people would package as a premium course. Instead, it’s sitting on YouTube for free. It covers the full stack behind modern AI: • tokenization • neural network internals • hallucinations • tool usage • reinforcement learning and RLHF • systems like DeepSeek and AlphaGo This isn’t about prompts or shortcuts. It explains how these models are actually built and why they behave the way they do. People who understand the foundations don’t just use AI, they build entirely new things with it. The real difference isn’t the 3 hours. It’s what those 3 hours unlock over time.

English
1
4
22
413
Ave Europa Tech retweetledi
Ave Europa
Ave Europa@AveEuropae·
Today is the anniversary of the most successful European project ever. What started by pooling coal and steel is now, in 2026, a full‑fledged geostrategic actor - one that must finally adapt to the role it was always meant to play, while fiercely upholding our civilisational heritage. Repatriate. Remilitarize. Reindustrialize. That’s the new Europe Day program Ave Europa is fighting for.
Ave Europa tweet media
English
42
73
363
7.4K
Ave Europa Tech
Ave Europa Tech@AveEuropaTech·
This is a warning about the future of European democracy, not just a Russian spy-school story. According to leaked documents reported by The Insider and international partners, a secret department inside Moscow’s Bauman University has been training students for the GRU pipeline - the same ecosystem linked to cyberattacks, sabotage, election interference, Fancy Bear, Sandworm, and Unit 29155. But the most important part is not only the hacking. It is the curriculum around information warfare: manipulating social media video, shaping public opinion, exploiting political division, and preparing digital operations against European societies. This is where modern elections are being attacked now. Not only at polling stations. Not only through ballot boxes. But through feeds, algorithms, fake accounts, AI-generated content, deepfakes, hacked systems, coordinated narratives, and synthetic “public opinion” manufactured at scale. Europe cannot defend democracy with speeches alone. We need democratic infrastructure built for the age of hybrid warfare: verified human participation, digital election integrity, bot resistance, misinformation detection, public accountability, and sovereign European systems that cannot be quietly captured by hostile networks. This is why Ave Europa believes in an #OnlyHumans future. Real people. Real voices. Transparent civic participation. Accountable representatives. Digital sovereignty by design. If authoritarian states are training the next generation to manipulate European democracy, then Europe must build the next generation of democratic technology to defend it. The answer is not censorship. The answer is trusted infrastructure. #AveEuropa #AveTech #OnlyHumans #DigitalSovereignty #ElectionIntegrity #HybridWarfare #Disinformation #Europe
Ave Europa Tech tweet media
Szabolcs Panyi@panyiszabolcs

🔎🇷🇺Inside Russia's elite Bauman University, a secret department trains the GRU's next-gen hackers, saboteurs & spies. Now, 2,000+ leaked docs expose how its graduates feed the units behind Russia's cyberattacks, election interference, and NATO sabotage. vsquare.org/welcome-to-the…

English
0
1
12
243
Ave Europa Tech
Ave Europa Tech@AveEuropaTech·
This is exactly why digital sovereignty cannot remain a slogan. If this report is accurate, Chrome users may have had a large on-device AI model downloaded to their computers without a clear, upfront choice. For most people, that means storage, bandwidth, battery, and compute being used by a platform decision they never meaningfully approved. The issue is not whether on-device AI is good or bad. Local AI can be useful, powerful, and in many cases more privacy-preserving than sending everything to the cloud. The real issue is control. People should know when AI infrastructure is being installed on their device. They should understand what it does, what resources it uses, what it can access, and how to disable or remove it properly. A hidden setting after the fact is not the same as consent. This is why Ave Europa Tech is focused on open, auditable, user-controlled digital infrastructure. Europe cannot build its digital future on systems where citizens technically own the device, but the platform quietly decides what runs on it. AI must be transparent by default. Consent must be active, not assumed. And users must remain in control of their own machines. That is digital sovereignty in practice. #AveEuropa #Google #Europe #Chrome #AI #AveTech
Ave Europa Tech tweet media
Evan Luthra@EvanLuthra

🚨GOOGLE JUST SILENTLY DOWNLOADED A 4GB AI MODEL TO YOUR COMPUTER WITHOUT ASKING.. WITHOUT TELLING YOU.. AND WITHOUT ANY WAY TO STOP IT.. If you use Chrome.. There's a good chance a 4 gigabyte file is sitting on your hard drive right now that you never agreed to download.. It's called Gemini Nano.. Google's on-device AI model.. A security researcher just proved it installs itself with zero clicks.. Zero prompts.. Zero notifications.. Alexander Hanff set up a completely fresh Chrome profile.. Didn't click anything.. Didn't scroll.. Didn't type a single keystroke.. Just opened the browser and watched.. 14 minutes and 28 seconds later.. Chrome had silently scanned his hardware.. Read his GPU, RAM, and storage.. Then wrote a 4GB file to his hard drive.. No permission dialog.. Nothing.. Chrome's own logs show the download begins BEFORE the settings page where you could opt out is even loaded.. The file starts installing before the refusal button exists.. As of Chrome 148.. Any website you visit can trigger this download.. One line of JavaScript.. You click a link to read a blog post.. That click counts as "user activation".. And Chrome silently pulls 4GB in the background.. No install prompt.. No consent dialog.. Google's own docs admit this.. Your laptop overheats.. Storage disappears.. Battery drains.. And you have no idea why.. The model doesn't even work well.. Cloud requests take 1.3 seconds.. The local model at worst case takes over 9 minutes for a single response.. Google is using your storage, electricity, and bandwidth to run an AI that's 40 times slower than their own servers.. And the "AI Mode" button in Chrome's address bar.. Doesn't even use the local model.. It sends everything to Google's cloud anyway.. You pay the storage penalty.. The heat penalty.. The bandwidth penalty.. And the visible AI feature ignores the local file entirely.. Because Chrome fails to clean up old versions.. Users are finding 12GB or more of duplicate AI files stacked on their drives.. Palo Alto Networks found a vulnerability where a browser extension could hijack the local AI model's permissions.. Accessing your webcam.. Microphone.. Local files.. Through an AI you never installed.. Here's how to check if it's on your machine.. Windows.. C:\Users\[YourName]\AppData\Local\Google\Chrome\User Data\Default\OptGuideOnDeviceModel\ Mac.. ~/Library/Application Support/Google/Chrome/Default/OptGuideOnDeviceModel/ If there's a file called weights.bin.. Google downloaded their AI to your computer without asking.. To stop it.. Type chrome://flags.. Search "optimization-guide-on-device-model" and disable it.. Search "prompt-api-for-gemini-nano" and disable that too.. Restart Chrome.. Then manually delete the folder.. If you don't disable the flags first.. Chrome redownloads the 4GB file on next launch.. Firefox requires explicit opt-in for AI.. Apple Intelligence requires explicit consent.. Chrome just takes your hard drive.. Google didn't ask to use your storage.. Your electricity.. Your bandwidth.. They just took it.

English
0
4
28
572