X Square Robot Unveils Open Embodied AI Stack
X Square Robot announced an open AI stack designed to unify robot learning data, world models, and action models. The stack aims to enable general-purpose robots to interact with the physical world.
ARTICLES BY THEO ALMEIDA
LATEST NEWS
-
-
Waze Adds Gemini-Powered Features for Personalized Navigation
Waze, Google’s navigation app, now uses Gemini AI to offer personalized routes and conversational road updates, rolling out globally on Android and iOS.
-
AMD's GEAK Optimizes DeepSeekV4 MLA Kernel for AMD GPUs
AMD's GEAK framework improved DeepSeekV4 MLA kernel throughput by 2.10x and reduced TTFT by 3.71x on MI355 GPUs.
-
Hugging Face Launches three.ws for AI Agents with 3D Bodies and Wallets
Hugging Face's three.ws platform lets AI agents walk, talk, and pay, hosting 18,000 avatars and 2,800 agents with 3D bodies as of July 2026.
-
Claude Code Adds Built-In Browser for Web Interaction
Anthropic integrated a browser into Claude Code, allowing AI to read, click, and type on external websites as of July 12, 2026.
-
Claude Cowork's Biggest Use Case Is Mundane Office Work, Anthropic Says
Anthropic reports half of Claude Cowork usage is for administrative and text-based tasks, with business operations accounting for 33.4% of sessions.
-
VKUE Runs 34.7B Model on CPU Without GPU
Hugging Face's VKUE engine runs a 34.7B parameter model on an 8 GB laptop and a GPU-less CPU server, achieving 20 tok/s and 17 tok/s respectively.
-
Hugging Face CEO: Companies Shift to Open Source AI
Hugging Face CEO Clem Delangue says open source AI is booming, now used by about half the Fortune 500 companies.
-
AMD Demonstrates SGLang Diffusion Speeds on GPUs
AMD shows SGLang Diffusion outperforms Hugging Face Diffusers on AMD GPUs by 1.5x to 6.3x in text-to-image and image editing tasks.
-
AWS Introduces Disaggregated Prefill and Decode for LLM Inference on SageMaker HyperPod
AWS announced a new method for LLM inference using disaggregated prefill and decode, improving performance for long prompts and high-concurrency workloads. The approach uses EFA and RDMA to reduce latency spikes in streaming applications.
-
Amazon SageMaker AI Deploys Quantized Models with Unsloth
Amazon SageMaker AI now supports deploying quantized models using Unsloth, reducing memory usage by up to 75% for 8-billion parameter models.
-
Amazon Introduces Case Management for Agentic Workflows in Quick Automate
Amazon Quick Automate now includes native case management to handle large-scale agentic workflows, enabling tracking of work items across millions of tasks.