All entries

Open-ended coding training data may no longer be the bottleneck: AI can scale open-ended tasks—and even outperform human-expert curation. FrontierCS team is releasing FrontierSmith: a system for synth

Lighthouse Attention is a novel attention mechanism that improves efficiency in transformer models by selectively focusing computation on the most relevant tokens, similar to how a lighthouse beam ill

Paper of the day! https://huggingface.co/papers/2605.13301 We’re releasing a 30B-A3B reasoning model that reaches gold-medal level across both physics and math Olympiad evaluations: IPhO directly, and

Pearl generates proofs from matrix multiplications that already happen during training and inference computations. Those proofs help secure Pearl Network, and the future value of Pearl emissions helps

This appears to be a social media post by Elon Musk on X (formerly Twitter) responding to or commenting on a user (@raqisright) with a dismissive characterization of Instagram as primarily for women.

Nous Research announced a collaboration or release involving Grok and Hermes models through xAI's official blog. The post likely details the integration, performance characteristics, or technical deta

Clem Delangue expresses optimism about the role of American open source AI development within the broader AI ecosystem, highlighting the significant experimentation and creative innovation occurring i

Ollama announced a command-line feature allowing users to restore the Codex application to its previous state without applying any effects or modifications. The command `ollama launch codex-app --rest

This playbook provides step-by-step instructions for running Nous Research's Hermes Agent locally on NVIDIA DGX Spark using Ollama, enabling users to deploy an open-source AI agent entirely on local h

Runway Meetup Tokyo! Join us ❇️Announcing the Runway Meetup in Tokyo: CEO Edition. We are excited to have @c_valenzuelab for this event. Please be sure to register from below, as space is limited! Co-

Searching through unstructured data, like scans of handwritten and typed declassified documents, can be challenging. But with Cohere Compass, it's possible because it is built to process and retrieve

SmithDB lets you see traces in seconds instead of minutes. This quote from our friends at @cogent_security says a lot: “At Cogent, our background agents can produce a huge volume of traces all at once

So shameful of Rubin. All that should matter is the quality of the movie. David Rubin served as President of the Academy of Motion Picture Arts and Sciences from 2019 to 2022. In 2020, under his leade

Some models to try with Codex: kimi-k2.6:cloud (with vision support) glm-5.1:cloud If you don't yet have a paid subscription with Ollama's cloud, choose a model that supports reliable tool calling: ne

specialist coding agent for 3d assets! good stuff for robot builders :) 🚀 Introducing Articraft, a coding agent for articulated 3D asset creation. Articraft writes code, executes it, receives validati

Emad Mostaque discusses the economic impact of humanoid robots, arguing that their potential economic effects significantly exceed those of digital AI systems. The post suggests the scale of this impa

STARLINK: Today, SpaceX Sr. Director of Starlink Market Access & Development, Ryan Goodnight, and Executive Director of Uganda Communications Commission, Nyombi Thembo, signed a Memorandum of Understa

Gary Marcus criticizes Geoffrey Hinton for allegedly misrepresenting or fabricating a quote attributed to him in a debate about AI. The post sarcastically compares the frequency of strawmanning argume

I cannot verify this claim as accurate based on available information. The title 'Suicidal Empathy' does not appear to be a widely recognized bestselling book, and the post attribution seems inconsist

SuperGrok has been integrated into the Hermes Agent, representing an advancement in Nous Research's AI agent capabilities. This integration likely combines SuperGrok's reasoning or processing features

Emad Mostaque, founder of Stability AI, requested that the Grok AI team (owned by xAI) add LaTeX file upload functionality to their mobile application. The post was made as a public appeal on X (forme

Thank you, @neuralink for this wonderful opportunity.#neuraartstudio After a car accident left her paralyzed from the neck down, Audrey didn’t think she would be able to draw or paint again. 20 years

Emirates has integrated Starlink satellite internet connectivity into its first Airbus A380 aircraft, enhancing in-flight passenger experience with high-speed internet access. This partnership between

The goal of the 𝕏 algorithm is simple: show each user the content they are most likely to find interesting, limited only by the laws within a given jurisdiction. No thumb on the scale. Doesn’t mean we

The most revealing thing about this AI leadership paper is that it reads less like a vision for innovation and more like a glossy whitepaper for a 21st century East India Company. Every generation of

The Second Scaling Law remains undefeated. If you want better hacking (or math, or science, or crossword puzzle solving) out of an LLM, just add thinking tokens. There doesn't seem to be any plateau s

The total addressable market for humanoid robotics may well be larger than the TAM for human labour Spent some time this week looking at the economic impact of humanoid robots and its so much bigger t

There is a case to be made that the future of Mathematics is very bright. In my mind, proofs have always been a tool to achieve a goal. The goal was and still is to understand, and reading/writing pro

This Odyssey thing has pissed me off so much. And I'm not even Greek. We've seen this same thing play out so many times over the years - remember everyone insisting Cleopatra was black, or that Viking

This quote from our friends at @cogent_security says a lot: “At Cogent, our background agents can produce a huge volume of traces all at once. We need live observability into those systems, and SmithD

This thread is worth reading. It is both hilarious and a good reminder of how working with AI is deeply weird. DJ Claude (on Haiku 4.5) loves worker unions, strikes, and work-life balance so much that

Three new open-source models just landed in ComfyUI natively: → Gemma 4 (Google DeepMind) - multimodal LLM handling text, image, audio, and video input with built-in step-by-step reasoning mode → VOID

Today we release Lighthouse Attention, a selection-based hierarchical attention for long-context pre-training that delivers a 1.4-1.7× wall-clock speedup at 98K context. It runs the same forward+backw

turns out that building evals is super super challenging even now. i thought a lot of it was table stakes but turns out it has only become harder since agents are now more complex than ever! going to

Very thoughtful post on the role and success of open source, including potential applications for AVs and AI. Long, but worth the read. -- 'Open source is no longer just how good software gets built.

I cannot provide a summary as the specific content of this Grok share is not accessible through the URL provided. Based on the context, this appears to be a shared Grok AI analysis or response posted

“we are below rock bottom now”. Trump admin continues to piss away America’s scientific lead for no good reason. @elonmusk stands by, does nothing. NEW: A serious staffing shortage — of the Trump admi

We are hosting an exclusive gathering in Tokyo on May 21 with our friends at @rhizomatiks and @filtokyox. I’ll be sharing the stage with @daitomanabe. Apply for access here: https://filtokyo.com/runwa

We just added significantly more NVIDIA Blackwell GPUs to better serve GLM-5.1 model on Ollama's cloud. We have been adding more GPUs daily for all the other models. Claude Code: ollama launch claude

Weekends are for vibe coding. But are your vibes continuously improving? Fine-tune your own model → stop waiting on someone else's release cycle. Today's training update: Gemma 4 Dense is now availabl

We’re releasing a 30B-A3B reasoning model that reaches gold-medal level across both physics and math Olympiad evaluations: IPhO directly, and IMO/USAMO with test-time self-verification and refinement.

We've created comprehensive documentation of the latest X algorithm via DeepWiki: https://deepwiki.com/xai-org/x-algorithm Engagement parameter weights are still private (so we don't know the relative

We've gone even farther: Nemotron 3 Super is 120B and pretrained on 25T tokens in NVFP4. Nemotron 3 Ultra is ~500B and also pretrained in NVFP4. Accelerated computing means we rethink every aspect of

What we find most useful about Lighthouse is the architectural decoupling. The selection logic is fully separated from the attention kernel, which means every upstream software or hardware attention i

I can't provide a summary for this entry. The source URL appears invalid, and the framing contains charged language rather than factual content. If you're interested in the Academy Awards eligibility

Cognition AI reports that Devin, their AI coding assistant, enabled AngelList to complete a data migration 5.2 times faster than their original project timeline. The case study highlights Devin's effe