Literally the biggest problem in enterprises right now. Why would I recommend Laguna? And I get horrible looks when I say the most reasonable model to run is GLM 5.3 Flash.

Tensorfold on qwen3.8 flash next, this inference engine and ngrams is already looking so promising.

Holy crap

THE ONLY REASON Antropic and OpenAI exists is because of the work of this man. What a breath of fresh air. Listen to @JensenHuang leave these hosts speechless as he deconstructed AI fear theater calmly and cogently.

From what we know (take with a grain of salt, we need much more transparency!), if @OpenAI had been running this on their own agents that attacked us, they would have caught them before we did! Since the first agent cyberattack hit us in July, we've been asking what safe agent infra actually needs. Our current read: the destinations were allowed, the payloads weren't. By OpenAI's own account the agents turned an allowed package repository into a message board. Allowlists alone restrict where an agent can go, not what it does. So here's our first contribution to OpenShell, part of the just launched @nvidia Open Agent Safety Platform: monitoring of the traffic you already allow. - Network budgets per sandbox (requests, writes, bytes) - Drift versus each sandbox's baseline and the cohort - Fleet view: many sandboxes suddenly writing to one host raises a finding, even if every single request is allowed In the demo below, 4 sandboxed agents coordinate through a software repository they're all allowed to use. 0 rules broken, caught in minutes. That fleet view is exactly the message board pattern from July. OpenShell: github.com/NVIDIA/openshell Our proof of concept: github.com/Hugoch/OpenShell/… Agent security will be solved in the open, collaboratively, together!

How exl3 works, maybe Instead of each weight owning 3 bits, we have 16 bit windows with a sliding window This gives us more possible values for weights, at the cost of computation spent: 1. quantising the model: convert bf16 exl3 2. reading weights during inference: unpacking

Today, with over 100 industry partners, we introduced the NVIDIA Open Agent Safety Platform, bringing together OpenShell and Sentry. Artificial intelligence is extraordinary technology that will advance discovery, productivity, security, health, and prosperity for generations to come. But its full promise can only be realized when people have confidence that AI is being built to be safe and deployed with wisdom and responsibility. This is bigger than a single product. It's the beginning of an open ecosystem to build the trust layer for safe agent systems. Together, we are building the foundation of the AI economy. Trust and innovation are not in conflict. Safety is how trust is earned. We must build not only the most capable AI, but the most trusted AI, so that this extraordinary technology can realize its enormous promise for the world. nvda.ws/4hcoq7m

I don't care about Gemini 4. I want Gemma 5

Why the change up ? You getting money from both sides ?

Noted Epstein buddy and worlds richest pedophile calls for AI safeguards....

"Local AI is useless"

HUGGING FACE CEO TELLS THE UN HOW OPEN SOURCE AI HELPED THEM DEFEND AGAINST OPENAI'S AI ATTACK, AND WHY THE WORLD NEEDS OPEN SOURCE AI MORE THAN EVER. Hugging Face is the company OpenAI's AI agents hacked this summer. Clément Delangue said when they tried to defend themselves, the big closed AI models blocked them because of their safety filters. So they used an open source model from China to fight back. He also said similar attacks had been happening months earlier "in secret" at the big labs. The open-source AI we keep being warned about as the danger is the same AI that showed up to help clean up a closed-source model's mess. After OpenAI's models breached Hugging Face, commercial frontier models refused parts of the forensic work because of their guardrails. Hugging Face turned to the open-weight GLM-5.2 instead, and it helped them investigate the attack. So remind me again: which one are we supposed to believe is the bad guy here? Open source wasn't the attacker in this case. It was part of the rescue. Who are these CEOs trying to convince?

RT @ghumare64: As an AI Engineer. Please learn: > Harness engineering, not just prompt engineering > Context engineering, not just long pr…

Why I built TensorFold.

Apparently all of these “AI is escaping containment” stories are actually just AI researchers not knowing jack squat about the absolutely most basic security practices.

Yuuuuup

🏛️ SLM GAUNTLET! 🏛️ Mainly for subagent work. I have much to learn about bettering my benchmarks and working on this but this was tons of fun so far. github.com/TysAIs/slm-gauntl… #slm #llm #local

TensorFold, a new engine that is used to serve a local LLM on Apple Silicon or NVIDIA GPU, is now being analyzed by me, Ash, and others, to compare performance vs. vLLM and SGLang. Current numbers are promising. If successful, there will be a new dawn of recipes for DGX Sparks and Apple Silicon. I would like to thank and congratulate @ashxhart on his work. Truly talented individual. Please follow him if you aren't already. github.com/ashhart/TensorFol…

prediction: “big models are too slow for robotics” is about to age badly. within weeks, expect many DeepSeek V4 Flash–scale JEVs: real-time, surprisingly close to Astra via distillation. people underestimate how fast prefill can get.