Machine learning news — deep learning, reinforcement learning, neural architecture search, diffusion models, and new ML frameworks and libraries.
Outpost VFX achieved 8x faster AI model training for visual effects face replacement by using AWS multi-GPU EC2 P5 instances, overcoming single-GPU bottlenecks that previously caused week-long delays. The collaboration w…
IBS Software used Amazon Bedrock's managed distillation capabilities to build a bilingual named entity recognition system for cargo logistics, achieving 95.085% F1-Score accuracy while reducing operational costs by 14x. …
Karpenter, an open-source node provisioner, directly provisions cloud instances without node groups, while Cluster Autoscaler scales pre-defined node groups. Karpenter is faster, cheaper, and more flexible, making it the…
Amazon Web Services announced that fine-tuning Amazon Nova models using SageMaker AI can achieve up to 94.77% accuracy in extracting data from emails, reducing costs by 50% and latency by over 30%. The solution, develope…
Ancestry has deployed machine learning to accelerate its record digitization pipeline from nine months to under nine days, with its corpus now exceeding 71 billion records. The company adds about 10 million records daily…
A developer built a multi-modal evidence review agent for damage claims using Python, OpenAI GPT-4o, and structured prompting. The system processes text, images, and historical context to produce consistent, explainable …
GPU utilization in production Kubernetes clusters averages just 5%, according to Cast AI's 2026 State of Kubernetes Optimization Report, while hardware costs continue to rise. The GPU Cost Optimization Loop—Measure, Allo…
Three high-performance computing experts—Jack Dongarra, Torsten Hoefler, and Satoshi Matsuoka—are questioning the necessity of GPUs in AI and scientific computing in a forthcoming paper, prompted by the emergence of the …
Schrödinger partnered with Google Cloud to deploy AlphaEvolve, an AI coding agent from Google DeepMind, achieving a 4x speedup in molecular discovery by optimizing machine-learned force field training algorithms. The opt…
Ancestry has spent 42 years digitizing 71 billion family records from 88 countries, a process that was slow and costly. Since 2014, the company has used AI, including machine learning and natural language processing mode…
Harness introduced pipeline-native autonomous AI agents for DevOps automation, running in sandbox containers and integrating with the Harness MCP Server, Worker Agents, LLM Gateway, and Software Delivery Knowledge Graph.…
A developer built an AI agent skill for the Antigravity CLI that migrates applications from AWS to GKE. The agent scans cloud dependencies, spawns parallel subagents to refactor code and infrastructure, and validates cha…
Flexion deployed Reflect v1.0, a robotics intelligence platform integrating mission control, motion planning, whole-body control, and runtime software with natural-language mission prompts. The platform improved humanoid…
A developer exploring AI platform engineering found that the most challenging problems lie in distributed systems and scheduling rather than machine learning itself. The analysis of GPUs, Ray, vLLM, and Kubernetes reveal…
A team rewrote a subset of the TinyGrad deep learning framework in Lean 4, creating TGrad, which outperforms the original in 4 of 5 benchmarks by up to 3x. The project demonstrates Lean 4's viability for everyday softwar…
A 2026 paper by Goldfeder, Wyder, LeCun, and Shwartz-Ziv argues that specialization is inevitable for AI systems, drawing on optimization theory, evolutionary biology, and machine learning. The authors contend that unive…
Jon and Mindy Gray donated $55 million to create the Basser Cancer Interception Institute at the University of Pennsylvania, which will use artificial intelligence and biomarkers to detect and intercept hereditary cancer…
The KV cache accelerates LLM inference on GPUs by storing prior token key/value projections instead of recomputing them, reducing per-step attention cost from quadratic to linear. Decode is memory-bandwidth-bound, so cac…
The tensor4all team, led by Hiroshi Shinaoka of Saitama University, has released tenferro-rs, a Rust-native dense tensor stack for scientific computing that supports linear algebra, autodiff, einsum, FFT, and CPU/CUDA ba…
The European Research Council awarded a €2.5 million Advanced Grant to Professor Mladen Popović of the University of Groningen for a five-year project that will use AI to analyze 250 samples of Dead Sea Scrolls, combinin…