AI TODAYEmail launch listFollow new AI models, software tools and product releases, with original reporting from the publishers covering them.
Feeds refreshed every 15 minutes
AI agents have expanded inference from single-turn interactions into multi-step workflows that reason, invoke tools, coordinate subagents, and carry growing...
Read at NVIDIA Technical Blog (opens in a new tab)
AI factories are interconnected systems where fleet economics depend on how efficiently the entire stack converts power and capital into completed agent tasks....
Read at NVIDIA Technical Blog (opens in a new tab)
A frontier language model is only one component of an AI agent. The surrounding agent system—often called a harness—determines how the model receives...
Read at NVIDIA Technical Blog (opens in a new tab)
As AI agents become more capable and operate over longer horizons, building security and trust into the applications they power becomes increasingly important....
Read at NVIDIA Technical Blog (opens in a new tab)OpenAI reaffirms Zero Data Retention for eligible API customers and previews Private Safety Processing for advanced AI safety without compromising data privacy.
Read at OpenAI (opens in a new tab)
Modern vision-language models (VLMs) can support tasks such as visual question answering, captioning, and image-text reasoning. In practice, however, the data...
Read at NVIDIA Technical Blog (opens in a new tab)
AI agents are only as effective as the context they receive. Even with capable models and well-documented NVIDIA libraries, agents can spend extra steps finding...
Read at NVIDIA Technical Blog (opens in a new tab)OpenAI launches an initiative to strengthen democratic oversight of AI in national security, supporting government institutions with tools, training, and expertise.
Read at OpenAI (opens in a new tab)
Atomistic simulation requires three things: knowledge of the science, compute-efficient implementation of simulations, and accessible interfaces to the...
Read at NVIDIA Technical Blog (opens in a new tab)
A new method for surgically removing training examples from a model reveals that as datasets grow, the link between what a model learns and what it produces dissolves.
Read at MIT News (opens in a new tab)A practical plan for an AI-assisted service, including a clearly hypothetical project budget, delivery checklist and ways to test customer demand.
A repeatable evaluation worksheet for usefulness, error rates, review effort, privacy and total cost, with a comparison against your existing workflow.
Headlines link to the original reporting. Article dates show publication time; the refresh time shows when feeds were last checked.