DaE73

K2.6 + hermes = 4 hr session setting up qwen 3.6 training regime on dgx spark with the current autnomous session currently lasting 70+ min without any prompting. Kimi with hermes is next level. @NousResearch @Teknium

China begins gradually phasing out Western aircraft. China Southern Airlines is offering its entire fleet of Boeing 787-8 aircraft — 10 planes and two engines — for sale at auction in Shanghai for $550 million. All the aircraft were manufactured between 2014 and 2015, and they will be replaced with domestically produced Chinese aircraft.

🚨🇲🇽🇧🇷 BREAKING: MEXICO AND BRAZIL ARE BUILDING THEIR OWN ENERGY ALLIANCE. Pemex and Petrobras. No Washington. No empire. Just Latin America deciding its own future. Sheinbaum is not picking sides. She is building her own table. That is sovereignty. That is power. Watch closely.

This is all fake news and supposition. The elections were a disaster but there’s 0 evidence it was done by the left in Peru which has 0 power in Lima, where this fraud happened. They have few in congress—remember Pedro Castillo is in jail for bogus charges because he threatened to change the constitution that’s handicapped Peru’s congress’s with too much power, so no President can ever govern. He also wanted to nationalize the lithium. Ten presidents in ten years. Keiko runs the country already, something I’ve also been told by the people of this country—all with different political opinions but all in agreement this is the case. The fraud likely came from within Keiko mafia based on who has the power in these institutions (her people) which an overwhelming majority of Peruvians I spoke to reiterate. This to secure a US puppet in the next elections a Milei or neoliberal figure under Washington’s control. The government just overruled the current president, to accept F-16s for billions. At this rate Peru will be a neo-colony of Washington and next in line to the “taking on the narco cartels” playbook that Venezuela and Ecuador went through. Washington aims to destabilize the remaining LATAM countries, take their resources especially lithium and create a Zionist technate ruled by supremacy and surveillance.

U.S. Department of Defense press release ADMITS that they went to war with Iran on behalf of Israel Liberals are the biggest dipshits in the world for thinking this was about “human rights” or “Iranian democracy”

A Chinese visitor shared what he saw at the Australian auto show: "Every booth of Chinese car is packed with visitors." "Ford and Volkswagen, the traditionals, saw sparse crowds." Chinese EVs are now the global favorites.

More and More Videos Are Coming of Weather Engineering They Spray Chemical While We Are Sleeping

🇺🇸🇨🇳 Top US military officials admit energy dominance is central to US geopolitical strategy including using "geography" (straits) to "impose costs" (blockade) "competitors" (China). The full hearing is here: armed-services.senate.gov/hearings/to-re…

> be chinese ai labs > while claude and openai are in cold war > kimi dropped k2.6 using deepseek's v3 architecture > the same week deepseek drops v4 using kimi's muon optimizer > 1.6 trillion parameters & 1M context > both match or beat closed models on benchmarks while being 8x cheaper > both build on each other's breakthroughs > keep shipping frontier LLMs with far less or nerfed NVIDA GPUs > and keep them 100% open sourced the real battle is not between models, it's open source vs closed.

Oil to $300.

The Islamic Republic of Iran knows what the Epstein Coalition plans to do and it is ready for it.

Even DeepSeek is now making fun of Anthropic's Claude.

World Has Start To Realise That “These Plates Are For Weather Engineering “ Weather in Iran And Iraq Is Biggest Example. Follow Us For Updates.

Whale is always extremely good at cracking the economics

It’s far past time to cancel our blue checks and stop giving money to this shitty platform. I apologize for my own lateness.

Kimi k2.6 used DeepSeek’s v3 architecture. DeepSeek v4 used kimi's muon optimiser. 1.6 trillion parameters & 1M context - both match or beat closed models on benchmarks while being 8x cheaper. Both build on each other's breakthroughs. Both keep shipping frontier LLMs w far less & nerfed NVIDA GPUs. China’s full domestic stack/ecosystem chip-model-cloud is now on full speed. Debate on distillation all you want. The dragon 🐉 is fully awake & good luck stopping it.

If you've noticed a lot less posts from us, we've decided that given the new algorithm changes, posting on this site has become a very fruitless endeavor. Hopefully this changes, but until then we will go touch grass in what's left of this beautiful world before the Trump regime decided to destroy it.

Wow, this is huge, after months of speculation and the U.S. running a massive pre-emptive discreditation campaign (x.com/RnaudBertrand/…), DeepSeek-V4 is finally out! I haven't studied it in depth but here are the most striking aspects as far as I can tell: - Fully open sourced with open weights (available for download on huggingface: huggingface.co/deepseek-ai) - Zero CUDA dependency anywhere in its stack, which is probably the biggest deal of all. For those who don't know, CUDA is Nvidia's software layer - the foundation nearly every frontier AI model in the world is built on. Except, as of today, DeepSeek V4, which can run entirely on Huawei Ascend chips via Huawei's CANN framework (finance.yahoo.com/sectors/techno…). Very concretely it means that China now not only has its own frontier AI models, but its own domestic AI stack, top to bottom. - The prices are insanely low. V4-Pro is roughly 3x cheaper than GPT-5.5 on input and 8.6x cheaper on output. And V4-Flash is an order of magnitude cheaper still, at $0.14/$0.28 per million tokens vs OpenAI's $5/$30 - so 30-100x cheaper than GPT-5.5 (!). And remember, these are the prices DeepSeek charges on its own API - anyone can download the weights and run them for "free" on their own server. - It is at or near the frontier on most benchmarks that matter. V4-Pro-Max matches or beats GPT-5.4 and Claude Opus 4.6 on competitive programming (Codeforces rating 3206), coding (LiveCodeBench 93.5), and math (HMMT 95.2, IMO AnswerBench 89.8). It trails the very newest GPT-5.5 and Opus 4.7 on a handful of the hardest agentic and knowledge benchmarks, but it's in the same league. In effect the value proposition is: "Same league as frontier US AI, at a fraction of the price, open-source and freely modifiable, and hardware-agnostic - you can run it on whatever infrastructure you choose." Which is insanely good. I now understand the need for a preemptive discreditation campaign: they had every reason to be worried. For the vast majority of use cases, you'd have to be a literal idiot to keep paying OpenAI or Anthropic's prices when this exists.

Liang Wenfeng, DeepSeek CEO "The whole world should use a 1.6T model for free." AGI will be open source and almost free

Spain may be suspended from NATO following their objection towards the war against Iran This is the organization that all “progressive Democrats” support NATO exists to project United States power on the world stage and defend their imperialism and wars

This tells me that V4 Pro require more post training work They will stretch its capability to Mythos++ in the coming months in 1 or 2 iterations Buckle up