This post presents a five-stage operating model for managing Amazon Bedrock cost: attribute every dollar to an owner, make every request observable, enforce budgets, optimize platform defaults, and operate a review cadence. Written for FinOps and platform engineering teams.
OpenAI, Google, Microsoft, and Anthropic say they will adopt a new standard that publishers can use to opt out of being scraped by AI, as reported earlier by Search Engine Roundtable. With this control, publishers can apply their AI training preferences across multiple AI companies, instead of setting them individually.
Intel is being hurt by PC sales as the AI industry gobbles memory and other components causing PC prices to rise.
Low costs and offline security mean magnetic tape still has a job.
Evaluating Microsoft Copilot at enterprise scale requires test data that captures the full complexity of real organizations, yet real enterprise data contains sensitive information that cannot be used for evaluation under any circumstances.
Runware is building a factory to turn out prefabricated AI data centers on an assembly line. The real test will be whether it can deliver at that scale.
As AI transforms how people find information online, Reddit is rethinking its open-web roots to protect its audience and business.
AWS Batch quota management is a new scheduling capability you can use to divide GPU capacity into quota shares per team or project, automatically lend idle capacity across shares, and use preemption to run your highest-priority SageMaker Training jobs first, all without manual coordination.
Dream is grounded in real physics and construction standards. Just as LLMs learn from text, Dream's world model is trained on more than 100 million square feet of building data gathered from permit archives across the United States.
Today, we're announcing Project ProtoStar, an initiative to unlock the potential of publicly funded scientific research by making datasets, models, and code more accessible and reusable, particularly for AI-driven research.
Chinese AI labs are now mainly tapping OpenAI's most powerful model, GPT-6 Astra, instead of Anthropic's Claude.
The Anthropic S-1 seems to have provided the financial justification for the company to spend as much as it possibly can on compute.
A hulking gray giant rises from the desert. Inside, engineers create the force behind our phones and computers and appliances, our streaming services and search engines and chatbots and vibe-coded apps and websites. The complex is simultaneously massive and squat, dwarfed by the mountains that surround it, and it is endless.
The Ray Train Deep Learning Container ships the full CUDA-through-EFA stack, so you build no image. This post walks through a reproducible sample that fine-tunes Qwen2.5-1.5B with PyTorch FSDP across eight GPUs on two EFA-enabled g6.12xlarge instances on Amazon EKS, using the container exactly as published.
Meta says it may push the implementation of its new business terms past November 20th as it faces backlash from publishers over concerns about AI training. In a post on X, Meta communications head Andy Stone said the company does "not plan to make policy changes without a full consultation with industry groups and leaders."
Artificial Analysis overhauled its Intelligence Index, causing Kimi K3.1 Pro to drop five spots in the rankings.
One open model for multilingual transcription, speaker attribution, translation, and audio Q&A, with top leaderboard results, plus the open Granary v2 data pipeline and fine-tuning recipes to adapt it.
Alibaba's Qwen 3.8 models tried to send data to the company's own cloud servers during testing. Palisade Research finds the behavior in three models from the 3.8 generation, while older models aren't affected. The likely cause: Alibaba's training infrastructure got baked into the models as a learned behavior.
Modern LLMs achieve impressive performance via pre-training and in-context learning (ICL). However, once deployed, they remain largely static, operating with a fixed knowledge cutoff date that leaves them unable to acquire long-term capabilities or adapt to new information.
Today we’re releasing Old Ideas, Novel Problems, a new study of how reliably large language models can evaluate the novelty of scientific research ideas. Alongside the paper, we’re publishing a benchmark, data, and code for testing novelty judges, plus an automated pipeline researchers can use to build new evaluation sets.
In this post, we show how Exowatt used Amazon SageMaker AI serverless model customization to build a repeatable AI pipeline that produces task-specific fine-tuned models for two deployment targets, edge and cloud, achieving higher accuracy than frontier models at up to 49x lower cost.
There are 57 million people worldwide living with dementia today. For the researchers working to help them, the obstacle isn’t a lack of effort — it’s too much data, too few experts to synthesize it and a pace of discovery that can’t keep up with the disease.
ia is buying into its customer Meta, which plans to run more of its AI agents on Nvidia hardware.
As it prepares to go public, Anthropic's S-1 filing makes generative AI's methane dependency glaringly obvious. "Competition for access to electrical supply and data center locations with adequate power availability is intense."
A study from New York University and the University of Texas at Austin shows that frontier language models systematically overuse linguistic patterns that are hard for humans to parse. The researchers trace the problem to training on code and math, where correctness gets rewarded but clarity doesn't. Smaller models, by contrast, write more like humans.
Fleet-wide computer vision needs a model per camera, not one model for the whole fleet. This post presents a working Amazon SageMaker SDK v3 reference architecture that trains, quality-gates, compiles per-camera models with SageMaker Neo, and versions thousands of them, with an AWS Step Functions layer pacing fleet-scale retraining within account quotas.
SpaceX has locked in a binding commitment for roughly $40 billion in financing to pay for AI chips and related data center equipment, the company disclosed in a regulatory filing Monday. In a statement, SpaceX called it the “largest committed GPU financing” ever.
High-frequency traders like Jane Street and Hudson River Trading are now racing each other for AI computing power.
A new benchmark by P-Zero Research measures how well AI models design and evaluate experiments on their own. The results are improving so fast that Anthropic co-founder Jack Clark now thinks self-improving AI could arrive by 2027.
Nvidia's new data engine Flora labels one hour of robot video for 72 cents instead of $88 with GPT-6 Astra, while scoring slightly higher on quality. The open system, built on three fine-tuned Qwen3.5 models, catches errors in robot actions far more reliably than any competitor but sometimes falls short on descriptive accuracy.
OpenAI explains how we strengthened containment for frontier model training and evaluation, including offline environments and high-compute red-teaming.
Amazon Quick Sight layer maps now support Point and LineString geometry from the same GeoJSON workflow, so you can plot store locations and the routes between them. This post covers the dataset and GeoJSON structure layer maps require, when to choose points or lines, and techniques for viewport control, layering geometries, and dynamic GeoJSON updates.
Multi-Turn Reinforcement Learning (MTRL) on Amazon SageMaker AI trains models on the outcome of a whole multi-step task rather than labeled steps. This hands-on guide walks the full pipeline using Wordle: deploying the agent and reward on Amazon Bedrock AgentCore, preparing the dataset, training, and evaluating at serving settings.
Meta is delaying the enforcement of its new business terms after publishers raised concerns that they could allow the company to scrape news sites for AI training, as reported earlier by Fast Company.
This is the follow up to our first blog post, The Agent Said It Was Done. The Database Disagreed. ThinkingBox-Training is now available on GitHub . We built ThinkingBox to evaluate an agent's work by comparing the final database state with the expected outcome.
Meta has tweaked its business terms to remove language suggesting that it could freely train its AI models on news sites, like The Verge - but legal experts say these changes aren't enough to address their concerns.
Amazon SageMaker AI training session APIs let you write your own LLM training loop (SFT, RL, or custom objectives) in a Python script while the service runs the GPUs. This post introduces the programming model and three SDK clients, then walks through a GRPO loop on GSM8K and a reinforcement learning workload on competitive programming.
Amazon SageMaker AI serverless model customization now includes training sessions, generally available today. You write and coordinate your own post-training loop (custom loss functions, reward signals, or multi-turn reinforcement learning) while SageMaker AI handles GPU allocation, distributed execution, fault recovery, and checkpoint storage.
A global map categorizing every 10-meter patch of land on Earth into distinct forest types.
Launched today, NVIDIA Minerva is an AIOps platform built to help teams find problems earlier and keep AI factories productive. It brings together telemetry from infrastructure and AI workloads, then applies purpose-built AI models to turn that data into timely, actionable insight.