BREAKING: Netanyahu directly rejects the Lebanon clause of the US-Iran agreement announced by Pakistan, telling Trump the IDF will not withdraw from Lebanon and that Israel does not consider itself bound by the clause, per Maariv.

A cartoon by Natalie Horberg.

Fonte: Il Fatto Quotidiano share.google/UMazWvZceLu7sOt…

Gemma 4 12B Coder is here and it's a game changer for local code generation. This GGUF model packs Google's latest gemma-4 architecture into a compact 12B size, perfect for running on consumer hardware. It's optimized for reasoning and thinking, making it ideal for developers who want fast, private coding assistance without the cloud.

In my lifetime, China lifted 850 million people out of poverty, while the US created one trillionaire. It is clear as day that socialism is the key to our future.

O modelo Rio 3.5 quebrou a internet esta semana. A reviravolta? Ele é essencialmente o nosso modelo open-source, Nex N2 Pro, usando outro chapéu. 🤯 Analisamos os pesos, e a receita é exata: Rio 3.5 ≈ 0.6 * Nex N2 Pro 0.4 * Qwen 3.5 Ele até literalmente se apresenta como "Nex N2 Pro" se você perguntar sem o prompt inicial do sistema! 😂 Ficamos lisonjeados que a cidade do Rio tenha usado nosso trabalho para alcançar performance SOTA. Obrigado pela validação definitiva dos benchmarks. 🤝 Mas, no mundo open-source, atribuição importa. Para mais detalhes, consulte: github.com/nex-agi/Nex-N2/is…

#ULTIMORA Garlasco: la procura di Pavia convocherà 60 milioni di italiani come persone informate sui fatti lercio.it/garlasco-la-procur…

The Rio 3.5 model broke the internet this week. The plot twist? It’s essentially our open-source model, Nex N2 Pro, wearing a different hat. 🤯 We analyzed the weights, and the recipe is exact: Rio 3.5 ≈ 0.6 * Nex N2 Pro 0.4 * Qwen 3.5 It even literally introduces itself as "Nex N2 Pro" if you ask it without initial system prompt! 😂 We are flattered that the City of Rio used our work to achieve SOTA performance. Thanks for the ultimate benchmark validation. 🤝 But in the open-source world, attribution matters. 👇 Full mathematical proof & verify script in the first reply!

4步教你复活 claude fable 5！ 1.下载 fable 5的系统提示词 github.com/elder-plinius/CL4… 2.放到你Claude code 项目文件夹 3. 使用启动命令 claude --dangerously-skip-permissions --system-prompt-file CLAUDE-FABLE-5.md 4.模型切换到 opus 4.8 Max 只需要这4步 你会发现 fable 5 又回来了！

Anthropic: “If we’re going down, everybody’s coming with us.”

L’ineffabile Ichino denunciò i SiCobas chiedendo 2 milioni di rimborso per scioperi contro suoi clienti. E perse. Ricordiamo le sue posizioni su contratti, licenziamenti, riforme del del lavoro, art. 18. Ora è invitato da chi vorrebbe stare nel “campo largo” della “sinistra”

Testing the new Gemma 4 12B (QAT) vision and OCR capabilities locally with LM Studio. # The setup: - GPU: NVIDIA RTX 4060 (8GB VRAM) - CPU: Intel i7 - Runner: LM Studio - Config: 32k context, 38 layers offloaded, Flash Attention enabled - Speed: ~14 tokens/sec decode throughput # The test: I gave it a screenshot of Google AI Studio. Prompt: "clone this. give me a single html file" # The result: A solid one shot replication. It successfully mapped out the layout, recognized the UI text, and structured the divs correctly, with only minor differences from the original. Results available at the end of the video. Quite capable for a 12B model running on budget consumer hardware. A gpu that costs only $300. # Why the architecture under the hood is notable: Unlike traditional models that rely on heavy, separate vision and audio encoders, Gemma 4 12B uses a unified, encoder free architecture. It bypasses separate multi stage encoders. Uses a 35M parameter vision embedder to project raw 48x48 pixel patches directly to the LLM hidden dimension. Local multimodal development is becoming highly accessible on standard hardware. If you've spun up Gemma 4 12B locally, what setup are you using and what kind of throughput are you seeing?

Announcement: We’re going to ablate this model — prefeitura-rio/Rio-3.5-Open-397B (based on Qwen3.5-397B-A17B). If the ablation succeeds, we will release the BF16 weights. If you’re interested, please follow us for first-hand news! huggingface.co/prefeitura-ri…

Mentre voi date la caccia all' "inmingcrato" islamico, vi hanno distrutto la sanità, il welfare, il sistema pensionistico. Senza che voi ve ne siate minimamente accorti.

In questo Paese abbiamo diverse migliaia di persone che guadagnano più di 10.000€ al mese, che difendono a spada tratta persone che guadagnano più di 10.000.000 di euro al mese, credendosi come loro e sputano su chi guadagna meno di 1.000€ al mese. Hey, rattusi avidi cialtroni, siete molto più vicini alla povertà che ai vostri eroi super miliardari.

La mayoría de la gente usa NotebookLM solo para resumir PDFs y se está perdiendo lo mejor. Google tiene la herramienta de estudio mas potente de el mundo. Te dejo 5 usos potentes y poco usados de NotebookLM que realmente marcan la diferencia: Guarda este hilo, te servira de mucho 📒

Invece, è un po' come chiedere a un insegnante nel Ventennio di dichiararsi fedele al Partito Fascista per poter lavorare.

arXiv Just Banned Researchers for AI Hallucinations arXiv hosts 2.4 million scientific papers. On May 14, 2026, it announced it would ban researchers for up to a year if hallucinated citations were found in their work. Then academics lost their minds. A small group of researchers went on the offensive after arXiv clarified that authors are responsible for hallucinated citations found in papers submitted under their names. "So this means you expect every author to check every citation and make sure that every citation is real and accurate?" an economics professor at Smith College replied in shock. Read that sentence one more time. A professor at a named institution expressed shock, genuine shock at the idea that they should verify the citations in their own academic paper before submitting it to be permanently entered into the scientific record. That shock is the story. Not the policy. The reaction. Even in 2026, there are still plenty of researchers who refuse to use AI to publish their research papers. Others use the tech for tasks like sourcing journal articles for references, editing copy, or formatting citations but they face pressure to verify every claim. A vocal minority of academics argue they should be able to use AI to write original research while remaining immune from any hallucinated claims or data that make their way into the final product. Immune. That is the word they used. Immune from responsibility for the contents of their own papers. Here is the mechanism behind the outrage that nobody in the mainstream coverage explained clearly. AI citation hallucination is not rare. The Columbia University study published in The Lancet in May 2026 scanned 2.5 million medical papers and found 4,046 fabricated references across 2,810 published papers with the rate increasing from 4 per 10,000 papers in 2023 to 57 per 10,000 by early 2026. A 12-fold increase. Perfectly correlated with the adoption of AI writing tools. Which means researchers have been submitting papers with hallucinated citations and passing peer review at accelerating rates for three years. And at least some of them knew. Because checking your own citations manually is not difficult. It is tedious. The AI made it easier not to check. arXiv drew a line. Authors are responsible for the contents of papers published under their names. The line should be unremarkable. It is the basic standard of academic integrity that has existed since the first journal published the first paper. The shock at being held to it reveals how thoroughly that standard has already eroded. The scientific record is permanent. The papers published with hallucinated citations are still there. They have already been cited. Those citations have already been pulled into other papers. The contamination is already in the literature. The ban does not undo any of that. It just establishes that going forward, the person who puts their name on a paper is the person responsible for what is in it. That this needed to be announced and that the announcement was met with outrage is the most alarming part of the entire story. Source: arXiv announcement · Thomas Dietterich · May 14, 2026 · Futurism coverage · (Link in the comments)

In questo sistema in cui il denaro è “contato”, chiunque accumuli grandi ricchezze è un ladro, poiché impedisce a milioni di suoi simili, di suoi concittadini, di poter vivere in modo dignitoso. Se non lo capite è perché siete laidi e marci dentro.

We are gonna make sure that Opensource AI wins by having vision beyond theirs, anticipating their moves, playing the long game, and acting for the collective good of humanity. Watch us make it happen. Opensource AI Must Win.