Machine learning news — deep learning, reinforcement learning, neural architecture search, diffusion models, and new ML frameworks and libraries.
Dave Treadwell, a 27-year Microsoft veteran who helped build Windows, is now leading AWS's Compute and Machine Learning Services team at a moment when scaling AI compute is one of the most significant challenges in techn…
A developer published a guide on training a generative AI kick drum model using a diffusion-based approach on a Linux desktop with only 6GB VRAM, demonstrating that high-end hardware is not required for audio AI experime…
Neurologists at the Feinstein Institutes for Medical Research in New York reported in a study published today in Nature Medicine that a man with quadriplegia, Keith Thomas, can now feed himself and drink from a cup using…
The machine translation industry is seeing convergence as translation AI companies launch their own TMS platforms, while TMSes add AI features. DeepL launched Translation Flow to connect enterprise apps like Google Drive…
A new Arduino library, OV7725_CPUfree_Camera, captures images from the OV7725 camera without using any CPU on the RP2350 microcontroller, achieving fixed 90 FPS streaming. By leveraging DMA and PIO automation, both CPU c…
AI agents are autonomously altering marketing budgets and audience targets in real-time, redefining how marketing teams operate by executing strategies that would take humans days. The convergence of machine learning and…
AT&T's AI-driven incident management system prevented 3.1 million unnecessary field dispatches and reduced customer downtime by 12 million hours in 2024, the company announced. The system, known as EEIM, evolved from tra…
AT&T built an end-to-end incident management system (EEIM) that reduced customer downtime by more than 12 million hours. The system, developed starting in 2017 using technologies from MongoDB and Snowflake, combines trad…
A developer forked the archived Kùzu graph database, adding vector search, LLM embeddings, and a full Graph Data Science suite to create Gorgonzola, aiming to make it the central engine for healthcare AI. The project int…
Pasteur Labs released Tesseract Core, a universal component framework for differentiable scientific computing that packages scientific software into self-contained, portable components capable of running across local mac…
A developer built OMR (Omni Data Refinement), an open-source Python library that replaces dozens of lines of boilerplate data quality checks with a single call. The library provides a health score, auto-cleaning, schema …
A three-part book series titled 'The Long Detour' on pre-GPU and low-compute machine learning has been published, drafted by a multi-agent system over a corpus of ~5,000 primary papers and directed by a human author. The…
HNSW (Hierarchical Navigable Small World) is a graph-based algorithm that powers vector search in FAISS, pgvector, Qdrant, Weaviate, and Milvus, enabling approximate nearest neighbor search in milliseconds by using a mul…
Researchers developed APT-RL, a framework enabling quadrupedal robots to learn agile, perceptive multi-skill locomotion in the wild by first learning reusable locomotion representations from trajectory-optimization data …
A KDnuggets article lists 10 YouTube channels for AI engineers, organized into categories covering research, practical building, core concepts, and industry analysis. Featured channels include Two Minute Papers for visua…
A developer on Hugging Face highlights the top AI papers for the week, noting a shift toward agent execution, reinforcement learning for reasoning, next-generation multimodal models, and long-term memory systems. Key pap…
AI and real-world evidence (RWE) are converging to accelerate evidence generation for rare disease therapies, addressing the paradox where patients with the greatest need face the longest waits. By combining AI with RWE …
DharmaOCR, a model specialized for Brazilian Portuguese, outperformed newer architectures including Mistral OCR4 and Unlimited-OCR on extraction quality and stability, according to Dharma-AI. The model's advantage comes …
An engineer at a startup spent $1,847 on multi-agent AI in a single weekend and then implemented a fix that cut costs by 82%. The solution involved optimizing agent orchestration and reducing unnecessary API calls.
Artificial intelligence is transforming video surveillance for businesses, moving beyond passive recording to real-time object detection, behavior analysis, and predictive insights, according to a report by Securitas Tec…