Video yükleniyor...

Video Yüklenemedi

Ana Sayfaya Dön

132,895 görüntüleme • 8 ay önce •via X (Twitter)

0 Yorum

Yorum bulunmuyor

Orijinal gönderinin yorumları burada görünecek

Benzer Videolar

🇮🇷🇺🇸 The ceasefire is less than 24 hours old and it's already being torn apart over Lebanon... Here's the situation. Pakistan announced the ceasefire applied "everywhere, including Lebanon." Iran agreed on that basis. Then Netanyahu called Trump minutes before the announcement and got him to agree Lebanon was excluded. The White House never clarified publicly until it was too late. Vance called it a "legitimate misunderstanding." That's a generous way of describing what happened. Iran entered a ceasefire believing Lebanon was included. Israel entered knowing it wasn't. Someone either miscommunicated or deliberately let Iran sign a deal under false assumptions. The result: Israel launched its largest strike on Lebanon since the war began. Iran is now threatening to withdraw from Saturday's talks and close Hormuz again. Araghchi put it plainly: "The U.S. must choose. Ceasefire or continued war via Israel. It cannot have both." Egypt accused Israel of a "premeditated" attempt to undermine the truce. Iranian news agencies report oil tankers in Hormuz were stopped after the Lebanon strikes. Vance says Israel "offered to check themselves a little bit" going forward. If Iran pulls out over Lebanon, he said, "that would be dumb but that's their choice." This is exactly the pattern I've tracked for 44 days. Every time peace gets close, Israel fills the gap with bombs. The question has always been whether Trump controls Netanyahu or Netanyahu controls Trump. The phone call minutes before the ceasefire announcement answered that question. Source: Axios

Mario Nawfal

1,275,580 görüntüleme • 4 ay önce

📜 Lavrov just accidentally destroyed Russia’s own legitimacy “If an agreement is signed by people whose legitimacy is uncertain, their successors may question it,” said Russian Foreign Minister Sergey Lavrov, referring to President Zelenskyi. Let’s pause here. Because in trying to undermine Ukraine, Lavrov just undermined the entire Russian state. If legitimacy determines the weight of international agreements, then Russia has no legal standing left. Let’s review: ◾ Putin has violated the Russian constitution by clinging to power beyond two terms, rewriting laws to suit himself. ◾ Elections in Russia are a performance — opposition jailed, media silenced, ballots rigged. ◾ The Kremlin staged illegal referenda in occupied Ukrainian territories — territories where they had no right to be, and where even the idea of “voting” under occupation is obscene. ◾ Russia has annexed land through war crimes. And yet it claims to “uphold law”? So when Lavrov questions Zelenskyi’s legitimacy — a president elected in free, democratic elections recognized by the entire world — he’s not making a legal argument. He’s projecting Russia’s own collapse of legality onto others. This is the classic tactic of dictatorships: accuse your enemy of your own sins. But in doing so, Lavrov just admitted that any future deals with Russia can be voided too — because their own system is built on fraud, violence, and constitutional rot. In other words: If agreements are only valid when leaders are legitimate — then no treaty with Russia is worth the paper it’s printed on. #Ukraine #Russia #Russians

Devana 🇺🇦

79,422 görüntüleme • 1 yıl önce

There is a moment in revolutions- a precise and historically recognizable sweet spot- when an old, brutal, and hardened regime still deploys its forces, yet something breaks in its resolve. You can sense it, and then the public senses it: fear has shifted sides. The oppressors are no longer as certain as they once were in using force. They cannot compete with sheer numbers, with masses filling the streets. Crucially, their own men begin to hesitate. Security forces grow reluctant to shoot at demonstrators; many have family members among them, or doubt that the regime they are defending will survive. This dynamic is well documented across revolutionary cases. In Iran in 1978–79, the Shah’s regime retained overwhelming military superiority, yet its paralysis came from fractured loyalty within the armed forces and police. In Eastern Europe in 1989, regimes collapsed not because protesters defeated the state militarily, but because security elites lost confidence that repression would restore control - most famously in Berlin Wall’s fall, when orders were issued but no one was willing to enforce them. Similar patterns appeared during the early stages of the Arab uprisings, especially in Tunisia and Egypt, where the army’s refusal to fully suppress mass protests proved decisive. That moment is also when regimes begin to change their language. They make offers. They issue statements acknowledging the “legitimate concerns” of protesters or opposition figures. They float proposals for dialogue or negotiations. Far from signaling strength, these shifts repeatedly mark the point at which a revolutionary situation reaches its peak. Such gestures often confirm what protesters already suspect: that the regime’s primary tools, fear and violance, are no longer functioning. That the state is dying. Political science research on authoritarian breakdown supports this pattern. Revolutions rarely succeed because of popular mobilization alone; they succeed when coercive institutions fragment. Once uncertainty spreads within the security apparatus the regime’s collapse becomes a question of timing. The Islamic Republic still possesses formidable repressive capacity. Yet the signals- hesitation, mixed messaging, demonstration of fear by cutting internet- suggest a leadership aware that it may no longer be able to rely on obedience. Historically, that awareness is one of the clearest indicators that an authoritarian system is entering its most dangerous and potentially decisive phase. It does feel very close.

נדב איל Nadav Eyal

618,379 görüntüleme • 7 ay önce

🧐 At this point, Billboard's latest "Top 50 Songs of 2026 So Far" list feels less like an objective ranking and more like another example of the music industry and media attempting to minimize Drake's success. Let's be serious for a second. Drake released a No. 1 album in 2026, generated some of the biggest streaming numbers of the year, dominated online conversations, and continued to be one of the most commercially successful artists in music. Yet somehow Billboard couldn't find room for any of the biggest records from that project inside the Top 10. But an unreleased Jay-Z freestyle from the Roots Picnic landed at No. 7. Make it make sense. This wasn't even an official single. It wasn't commercially released. It wasn't a chart-dominating record. It wasn't a song fans were streaming for months. It was a basic freestyle performance that generated a few headlines because people were debating who Jay-Z was taking shots at. Meanwhile, Drake's first appearance on the list doesn't arrive until No. 18 with "Shabang." And what makes it even more interesting is Billboard's write up itself. Instead of simply explaining why the freestyle ranked so highly, the article almost reads like it was written specifically to elevate Jay-Z while taking subtle shots at Drake. 💬 “Jigga had everyone talking about who he dissed and why for like two weeks straight.” 💬 “Now we wait and see what Drake has to say about some of the jabs sent his way.” The obvious question is: 🤔 Since when did two weeks of social media discussion become more important than actual chart performance, streaming numbers, sales, and commercial impact? Because if cultural conversation is now the standard, Drake has been one of the most discussed artists in music all year. If commercial success is the standard, Drake should be much higher. If streaming numbers are the standard, Drake should be much higher. If chart performance is the standard, Drake should be much higher. So what exactly is the standard? That's why many fans continue to believe there is an effort within parts of the industry and media to downplay Drake's accomplishments. Every time Drake breaks records, the conversation becomes about somebody else. Every time Drake has a successful release, the focus shifts to a narrative. Every time Drake reaches another milestone, there seems to be an attempt to explain away the achievement instead of simply acknowledging it. And this ranking feels like another example. An unreleased freestyle being placed ahead of records from one of the biggest artists of the year isn't just questionable it raises legitimate questions about whether these lists are actually based on music or whether they're being influenced by industry relationships, personal preferences, and predetermined narratives. Of course, Billboard is entitled to its opinion. But when a No. 1 Drake album struggles to place songs in the Top 10 while a freestyle performance is treated like one of the defining records of the year, fans are naturally going to question the credibility of the ranking. 💭 “To some Drake fans, moments like this fuel the belief that there’s an industry network of gatekeepers whether it’s powerful media figures, executives, or influential artists who consistently seem to move in lockstep whenever Drake is the target. Seeing an unreleased Jay-Z freestyle ranked above songs from a No. 1 Drake album only adds to that perception. Some supporters even describe it as a ‘secret society’ mentality, where certain media outlets and industry power players appear quick to elevate narratives that benefit Jay-Z and others while simultaneously downplaying Drake’s accomplishments. Whether that’s actually happening or not is up for debate, but rankings like this are exactly why those conversations continue.

Cousin Tino ™️

78,088 görüntüleme • 1 ay önce

This is the most telling clip of this entire exchange between Ro Khanna & Secretary of War Pete Hegseth Pete Hegseth. Ro: "You don't even know how we pay in terms of the missiles that hit the Iranian school." Guess who else keeps using "the Iranian school" in public messaging. Terrorists in Iran. Ro should consider going to join them since that's clearly where his loyalties lie. First of all, the expectation that a sitting Secretary of War will assign a specific dollar amount to how much an armed conflict will "cost the American people" is absurd. That's exactly what Ro does in this exchange - though of course he doesn't cite his source or explain his methodology. Still, the rhetorical "cost" must be measured in comparison with the price for maintaining the pre-conflict status quo. As The White House recently explained, "More Americans have been killed by Iran than any other terrorist regime on Earth." ( And "roughly one in every six American combat fatalities in Iraq were attributable to Iran." ( Not to mention the potential cost of allowing Iran to develop nuclear weapons, considering that we now know they ALREADY have missiles that can reach most of continental Europe - including ~30 bases where ~68K 🇺🇸 troops are stationed. Ro doesn't seem to want to talk about any of that while he's trying to corner SECWAR into giving a simple answer to a complicated question that ultimately depends on how the "cost" is measured. Based on our previous exchanges here X, the part that I find most telling - and irritating - of all is the off-hand reference to the Minab school attack, though. All the rhetoric about this incident is currently based on speculative media reporting from anonymous sources. Yet the terrorist regime in Iran keeps emphasizing the incident in their public messaging because it plays well for Iran's useful idiots throughout the West - like Ro here. Even if we assume speculative media coverage is true, nothing that has been reported so far suggests a war crime was committed. At worst, it seems possible that the attack constitutes failure to take feasible precautions in the attack...which is NOT a "war crime" by any legitimate, doctrinal standard. But that doesn't stop useful idiots like Ro here from focusing on this incident AS THOUGH it were a war crime since doing so plays well in whatever bullshit public narrative he's trying to spin for the American people. Just like the terrorists in Iran keep doing...for the exact same reason. Besides, the munitions that have been employed so far have debilitated IRGCs military capabilities that the terrorists in Iran had spent decades & spent untold billions to accumulate. Ignoring this while reducing everything that has happened in the last ~2 months to "the missiles that hit the Iranian school" is just...dumb. If you look at Ro's TL right now it's clear he thinks he "schooled" SECWAR Hegseth (as Ed Krassenstein claims in a post Ro retweeted) in this exchange. In reality, all Ro did was confirm yet again why he's just another useful idiot who also happens to have a seat in Congress. I've said it before to Ro, and I'll say it again: stay in your lane. Leave LOAC analysis to Department of War 🇺🇸. And actual experts...like me.

Dr. Brian L. Cox

22,976 görüntüleme • 3 ay önce

my 8 GB VRAM gaming laptop is absolutely going to hate me for this. but I still did it. ran a 31b dense model (Gemma 4 31b Q4) with only 8 GB VRAM last week I ran Gemma 4 26B A4B a mixture of experts model on my RTX 4060 and hit 25–28 tokens/sec using llama.cpp's new MTP support. smooth. snappy. but MoE has a secret: it only activates 4B parameters per token despite having 26B total. that's why it flies. so the real question started haunting me. what if I throw a full, no tricks, every parameter fires on every token, 31B DENSE model at the same machine? # Hardware: GPU: NVIDIA RTX 4060, 8 GB VRAM RAM: 16 GB CPU: Intel Core i7 H Laptop. Gaming. Modest. The model: gemma-4-31B-it-qat-UD-Q4_K_XL.gguf (model's unsloth huggingface link in the comments) This is Google DeepMind's flagship dense model in the Gemma 4 family that can run on single consumer GPU. It packs a hybrid attention architecture, supports up to 256K context natively, and is QAT (Quantization Aware Training) optimized, meaning it retains far more quality than standard post training quants at the same bit depth. This is NOT the MoE. This is 31 BILLION dense parameters, every single one of them loaded. # the flags I used: -m gemma-4-31B-it-qat-UD-Q4_K_XL.gguf -cnv --spec-type draft-mtp --spec-draft-model mtp-gemma-4-31B-it.gguf --spec-draft-n-max 8 --spec-draft-p-min 0.6 -c 6000 -v Multi Token Prediction (MTP) is still active here. Separate draft GGUF required, same as the 26B setup. # Results: → Decode: ~3 tokens/sec → Prefill: ~2 tokens/sec → Context: 6000 tokens → Hardware crying quietly in the corner: yes so is 3 tps actually usable? For real time back and forth chat? Not ideal. You're not having a fluid conversation at 3 tps. but slow ≠ useless. And this is where it gets genuinely interesting. think about how senior devs actually work in a real team. But when something is architectural, deeply complex, or needs serious reasoning? they walk down the hall and escalate to the senior. That's exactly the local AI agent architecture this unlocks: → Fast orchestrator model (Gemma 4 26B MoE at 25+ tps) handles routing, simple queries, tool calls, memory. The junior dev. → Gemma 4 31B dense is the senior, called only when the fast model genuinely hits a wall. Hard multi step reasoning. Complex code generation. Deep architectural decisions. The agentic loop stays fast. Only the hard hops touch the 31B. That's a legitimate production grade local AI architecture on a budget hardware. (requires 2 8gb gpus) other workflows where 3 tps is completely fine: - overnight batch jobs. summarize documents, extract structured data, review code. Fire it off. Sleep. wake up to results. - One shot deep reasoning - Silent code audit loops, you write and test, the 31B reviews diffs and flags issues in the background between your sprints - Any workflow where output quality > output speed A few weeks ago, nobody was running a 30B+ dense model on a single consumer GPU with 8 GB VRAM. At all. Now we're doing it on an Intel i7-H gaming laptop with a NVIDIA RTX 4060, thanks to llama.cpp + QAT quants + MTP speculative drafting. Google DeepMind said the Gemma 4 31B targets "consumer GPUs and workstations." They were not exaggerating. The hardware bar to run serious frontier class models locally keeps dropping. the tools are here. the models are here. you just have to be willing to abuse your laptop a little. what workflows would you actually run on a local 3 tps 31B dense model? genuinely curious. drop it below.

Alok

63,583 görüntüleme • 1 ay önce

An appalling human rights violation in China demands your attention. A recent case of teenage bullying in Jiangyou, Sichuan, sparked nationwide outrage after a video went viral online, in which the perpetrator boasted, “I’ve been to the police station more than ten times — they let me out within 20 minutes each time.” The bullying scandal has led to extensive calls into question the effectiveness of policing and the juvenile justice system in China. With the fallout from the incident growing, local residents spontaneously organized multiple peaceful gatherings calling for a lawful and transparent resolution. Although these gatherings remained orderly and restrained, authorities responded with indiscriminately violent assualts by riot police against the large crowd, forcibly dispersing protestors from public spaces. The response was not one of accountability, but of preemptive force— driven by the need to maintain social stability ahead of the upcoming 2025 World Games in Chengdu, just a few miles away from Jiangyou where the anti-bullying protests broke out. In prioritizing an unblemished international image over social justice, local officials have narrowed the space for public expression to the point of suffocation, treating legitimate public demands as risks to be contained. What concerns us is the increasingly instrumentalized approach the Chinese government has adopted in managing public sentiment. The term “public opinion” is increasingly becoming hollowed out in China, deliberately pathologizing normal public expression. Ironically, "public opinion" has instead become a means of monitoring and suppressing public expression. This has further heightened the hostility between state and society, making meaningful interaction and resolution of social conflicts nearly impossible. Under the logic of “stability above all,” every social issue becomes an image crisis, and every dissonant response is viewed as a disruption. This approach not only stifled resolution, but also covered up human rights abuses with manufactured calm and peace. Not only did the victim of the bullying suffer secondary harm from the police’s brutal crackdown, but ordinary citizens who came to support the bullied teenager were also indiscriminately beaten by the police. We call on the International World Games Association to take a clear and public stance. As a renowned international event, the World Games cannot take place on soil where civic rights are suppressed and voices are erased. The spirit of sport must not serve as a cover for institutional violence or be used to legitimize repression. From: An independent collective monitoring grassroots resistance and civil rights in China Philip D. Jaffé UNChildRights Sophie Kiladze VanKeirsbilck Benoit Ann Skelton The World Games --------------------- 四川江油市近日发生一起未成年人霸凌事件,视频中施暴者声称“都进去警察局十多次了,二十分钟就放出来”。这起霸凌丑闻随机引起社会对执法效力和未成年人司法机制的广泛质疑。 随着事件持续发酵,当地居民自发发起多场和平集会,呼吁依法公正处理案件。尽管现场秩序始终克制,当局竟仍出动警力无差别袭击示威人群,强制暴力清场。这一回应并非基于追责或司法正义,而是出于“预防性控制”的政治需求——因为成都即将举办世界运动会,维持政治稳定被视为压倒一切的目标。为确保一个“完美无瑕”的国际形象,公众表达的空间已被压缩至近乎窒息,正当的民间诉求被当作风险加以压制。 我们担忧的是,中国政府多年来对公共情绪的处理方式,已日趋工具化。“舆情”这一术语已经空心化,正常的公共表达也已被病理化。这种情况进一步加深了国家与社会之间的对立,使有意义的互动与社会矛盾的实际解决变得不可能。 “稳定压倒一切”的逻辑,将任何社会问题转化为形象危机,任何不和谐的反应都被视为威胁。这不仅使问题难以实际解决,还用表面和谐掩盖人权劣迹。不仅霸凌受害者因为警方的粗暴对待遭受了二次伤害,连声援受害者的普通市民也遭到警方无差别殴打。 我们呼吁国际世界运动总会(IWGA)表明立场:一场令人尊敬的国际赛事,不能在压制人权、剥夺公民声音的土地上举行。体育精神不应成为权力洗白的遮羞布,也不能为制度性暴力提供背书。

李老师不是你老师

1,558,911 görüntüleme • 1 yıl önce