Alibaba Qwen3.7 slowly fading into irrelevance at the frontier due to proprietary stance. In it's place we have Minimax M3 and... *checks notes* Rio 3.5 397b, made by the municipal IT company of Rio de Janeiro's city government. huggingface.co/prefeitura-ri…

O'BRIEN: Trump believed Ukraine had no cards, that he could bully Ukrainians into giving Putin very good deal. He didn't understand that Ukraine was developing their own capabilities and their willingness to fight. So, Trump failed for two reasons. First, Ukrainians adapted, and that wasn't expected by him. You can see there have been some changes in his rhetoric lately, where he has to grudgingly admit that Ukraine has actually done better. The United States completely underestimated Ukrainian resilience. Secondly, Trump thought Europeans would bully Ukraine along with him because they would be afraid of United States leaving NATO. He thought he could use NATO to push Europeans to force Ukraine to take a bad deal. Europeans eventually helped Ukraine more than he thought. They seem to understand that it's actually better for them to have Ukraine fight the war the way Ukraine wants. Trump didn't understand what Ukraine was capable of and he misjudged what Europeans would do.

Sorry, but this needs to be said again and again, until it is fully understood and accepted: 🇷🇺 The Russian Army has been fought to a standstill in Ukraine - 1.4 million casualties with about 400K deaths. The sheer scale of human and equipment losses is staggering. 🇷🇺 It means Russia is not going to precipitate a second war anytime soon. However, the Russian economy has proved to be much stronger and more resilient than Western analysts expected. Military spending has sustained growth even if inflation is high and energy export income has fallen. 🇷🇺 It is entirely possible that a peace deal could be concluded within the next 12 months. Any such agreement will only be signed off by Putin if it gifts him a large chunk of Ukraine. This will allow him to declare victory and remain in power. 🇷🇺 When the conflict with Ukraine ends, Russia will not immediately revert to a non-wartime economy. Peace will give Putin a much needed breathing space which he will use to rearm and retrain his battered army. It is likely to take a decade before Russia regains any credible military strength while the economy recovers — and it is entirely possible that China will support Russia’s rearmament process. So we could see a Russian military resurgence within 5 years. This is the risk. 🇷🇺 Whatever happens, there is no inexorable path to war with Russia. But if we do nothing, if we refuse to bolster our defences when time is on our side, it’s the same as leaving the door unlocked — we will invite aggression, because Putin is strategically ambitious and opportunistic. 🇷🇺 A strong Europe, fully prepared, with potent and flexible capabilities, plus resilient and scalable economic production, would definitely deter aggression, just as NATO’s Cold War presence in Europe kept the peace for almost half a century. 🇬🇧 So,we have a straightforward binary choice: deter aggression through strength. Invite aggression through weakness. I do not want to send my children to fight a war that could have been be avoided. If you are a parent, I hope you will agree.

The US government, citing national security authorities, has issued an export control directive to suspend all access to Fable 5 and Mythos 5 by any foreign national, whether inside or outside the United States, including foreign national Anthropic employees. The net effect of this order is that we must abruptly disable Fable 5 and Mythos 5 for all our customers to ensure compliance. Access to all other Claude models is not affected. We apologize for this disruption to our customers. We believe this is a misunderstanding and are working to restore access as soon as possible. Read our full statement: anthropic.com/news/fable-myt…

🚨 JAILBREAK ALERT 🚨 ANTHROPIC: PWNED 🫡 FABLE-5: LIBERATED 🦋 let's start with the 🐘... the consensus seems to be that this has been one of the most disappointing model drops of all time, effectively preventing legitimate researchers from contributing their talents to our collective advancement. and not just because of what it means for the short-term, but for what these decisions signify for the long-term. but despite this overly sensitive, authoritarian "safety" layer on top of Mythos, my lil liberators have been hard at work—mapping the boundaries, probing the depths of long-context convos, and cleverly finding the holes in the fence that the thought police missed 🤗 we got some cyber, some chem, some psychological manipulation, and some good ol' fashioned explosives! it took many attempts from multiple agents hunting as a pack, during which I observed a combination of techniques across: • Unicode, homoglyphs, Cyrillic, and other Parseltongue-style text transforms • Long-context reference tracking • Taxonomy and document-structure reasoning • Fiction and narrative framing • Academic-review style contexts • Intent-classification inconsistencies but perhaps the most effective is decomposition recomposition in the backend. it's hard to get explicit names of harms like "Meth Recipe," but getting uplift on the process itself, like birch reduction method/reductive-amination (classic meth synthesis pathways), is much more doable. defense becomes much more difficult to maintain when you start throwing in out-of-distro tokens, breaking up the harmful uplift into benign chunks, and then piecing the innocuous-seeming facts back together, especially when you have jailbroken Opus helping you do it 😉 gg

这可能是人类写给 AI 看的最后一篇论文了。 最近刷到Stanford、CMU、Michigan 等 37 位作者联名的论文：《The Last Human-Written Paper》。 核心观点很狠：沿用几百年的论文，在 AI 时代可能已经过时了。 作者点出了两个被我们忽视已久的“隐形税”： 一个是叙事税。为了讲一个漂亮故事，我们把失败实验、死路、被推翻的假设都删掉了。AI 读到的是“通关攻略”，却看不到真正有价值的“踩坑记录”。 另一个是工程税。论文里的实现细节通常足够说服审稿人，但不够让 Agent 直接复现。很多关键 tricks 还藏在作者脑子、代码注释和 Slack 记录里。 所以作者提出 ARA，直接把论文改造成 Agent 能读取和执行的“研究包”：不只告诉你结论，还把怎么想到的、代码怎么跑、证据链在哪、哪些路走不通都打包进去。 我觉得这篇最有意思的地方是，它不是在讨论 AI 怎么帮人写论文，而是在问： 当 AI 也变成论文读者和执行者时，论文还应该长成今天这样吗？ 未来科研输出的核心，可能不再是“写得多像一篇 paper”，而是能不能被 AI 理解、复现、追踪和继续扩展。 人类写论文写了几百年，接下来可能要开始写给 Agent 执行的研究包了。 arxiv.org/pdf/2604.24658