Archive, page 20
Meta launched Muse Wonder and Muse Wonder Pro, two models from its revamped Superintelligence Labs, claiming they compete with the industry's best. Zuckerberg says Meta has reached the AI frontier.
A Startup Fortune piece offers guidance on structuring an AI agent chargeback model without internal conflicts.
Cerebras stock dropped sharply amid broader AI market fears, according to Barron's.
Crosswalk.com discusses what Christians should know about privacy implications of AI glasses.
DeepSeek is set to lead an investment round in Winverse AI, a startup founded by former Tencent AI scientists, with a valuation that could reach $400 million.
Proba NTree is an AI agent that produces traceable fertilizer emission data, letting users track and verify agricultural emissions figures through an automated workflow.
A new agentic model from Z.ai built for legal tasks such as document review and drafting, automating multi-step legal workflows.
The UK Information Commissioner's Office is investigating AI agents at OpenAI, Meta and Anthropic, reportedly examining how the agents handle personal data.
A new model called Muse Wonder 1.0 has surfaced on Product Hunt and Reddit with benchmarks, though details on capabilities and availability remain limited.
Anthropic has reportedly committed to roughly $530 billion in data-center leases, driven by the economics of running large-scale inference.
GLM-Legal is a Product Hunt listing for Z.ai's legal AI model aimed at contract review and compliance. No model size, license, pricing or public availability details are given.
A UK minister said a cyberattack similar to the OpenAI-linked Medicare breach seems likely in the UK, warning about vulnerability of health systems to AI-enabled attacks.
Microsoft researchers found autonomous AI agents get trapped in 'idea basins' from growing chat context. Their Fork-and-Flush method clears chat history while keeping the workspace, improving results by up to 66 percent.
Cisco is working with Nvidia on pluggable power technology and deepening its alignment with the open-source SONiC network operating system.
Meta is expanding its Muse AI agent while projecting 2026 capital spending of $130 billion to $145 billion to support its AI ambitions.
California is moving on AI safety with a kill switch provision, part of state efforts to give regulators the ability to shut down dangerous AI systems.
Nvidia researchers propose ReplaySSM, a way to rethink state storage so hybrid models mixing SSM and attention layers decode faster. Applies to Nemotron 3, Qwen3.5-3.8, Kimi-K3 and GLM-5.3-Flash.
Phasis is building a platform that orients AI agents using graphs and 'reality layers' to measure culture, values and relations between people and their beliefs.
A technical guide for deploying decision models with llama.cpp on Amazon SageMaker AI, serving OpenAI-compatible chat completions and extending runtime with SemIf to return probabilities over declared options on CPU or GPU.
AMD introduced Helion racks aimed at AI agent workloads, with CPU-only and mixed CPU/GPU configurations due from partners in November.
A guide to deploying one large language model across two GPU nodes on Amazon EKS using Ray LLM Deep Learning Containers, with separate prefill and decode nodes behind one OpenAI-compatible endpoint.
Honeyjar is pitching AI tools aimed at fixing the PR and communications process for journalists, entrepreneurs and public relations professionals.
AWS' channel chief said there is no 'one model to rule them all,' saying the cloud unit backs both OpenAI and Anthropic models rather than a single dominant provider.
Ondas and Palantir are partnering to launch Regna AI, a new company building AI for defense missions. Details of the structure and offerings are limited so far.
OpenAI plans to add 200 jobs in Dublin, expanding its Irish presence.
Boole-1.7B is a 1-bit version of Qwen3-1.7B with 247 MB packed weights, 13.9x smaller than BF16 and scoring 85.9% of the BF16 model across ten tasks. It can be run locally.
Microsoft president Brad Smith called for urgent AI regulation, arguing milk and washing machines face more oversight than AI companies and that the UK is well positioned to lead.
Google is adding AI controls to Android Enterprise, giving IT teams new management settings over how AI features and apps behave on managed Android devices.
Modders say DLSS Multi-Frame Generation and DLSS 5 Neural Rendering can now run on hardware never intended for them, and discuss the work behind the community ports.
ASUS debuted the XA P2N-E2 server powered by dual Nvidia Vera CPUs, targeting agentic AI workloads, at OCP 2026.
Santa Monica startup Blueprints AI, bootstrapped, says its software produces code-compliant construction blueprints from a photo, PDF or sketch in minutes.
Microsoft will make in-house Azure Maia chips its main AI inference system starting 2027 at multi-gigawatt scale; Maia 300 should deliver 40% more tokens per megawatt than leading GPU systems.
At OCP 2026, Cisco announced G302 silicon and USFP optics developed with Nvidia for AI networking.
The newsletter Import AI's 476th issue, with Clark flagging China's open-model advantage, covers the latest AI developments for 2026.
Pallix added Grok and Claude tracking, query fan-out, and ready-to-paste fixes to its AI visibility monitoring platform.
Agent Role Contracts v0.5.0 is a tool for checking AI-agent workflows against defined role contracts to catch misbehavior before deployment.
GUDEA raised $7 million in seed funding for AI narrative intelligence, technology that analyzes how stories and narratives spread.
Samsung is reportedly considering AI-powered widget design tools for its One UI 9.5 software update.
Nvidia's Arm-based Vera CPU has started shipping, targeting the server CPU market as part of its Vera Rubin platform for AI datacenters.
Nvidia added support for Google's Falcon transport protocol in its networking stack, aiming to address growing diversity in datacenter transport demands for AI workloads.