OpenAI Study: ChatGPT Users Cross Job Boundaries
New research from OpenAI suggests AI is letting workers take on tasks outside their formal roles, blurring traditional job definitions.
AI & ML
Frontier models, research results, agents, and developer tooling.
HSP GRUPPE Uses ChatGPT Enterprise to Boost Tax Advisory Productivity
A tax advisory firm's adoption of ChatGPT Enterprise shows how professional services can apply AI to improve work quality and free up capacity.
Can AI Write a Better AI Textbook Than Its Human Author?
An AI researcher and textbook author asks when models will surpass humans at explaining their own field.
OpenAI's Texas Letter: Commitment to Responsible AI Infrastructure
OpenAI has written to Governor Greg Abbott, pledging support for reliable and transparent AI infrastructure that benefits Texans.
OpenAI CFO: AI-native finance needs new controls, not just speed
OpenAI CFO Sarah Friar argues that finance teams should redesign workflows around AI, with forecasting and controls as early wins.
Virgin Atlantic adopts ChatGPT Work for customer-journey insights
The airline is using OpenAI's enterprise tool to accelerate research, product planning, and decision-making.
Hugging Face's 2,200-Paper Reproduction Effort Yields Hard Lessons
A large-scale open reproduction of ICML papers reveals the practical difficulty of replicating published AI research.
Meta's Muse Glimmer Brings Local, Agentic, Multimodal AI to Open Source
Hugging Face highlights Muse Glimmer as Meta's return to open, local, agentic, and multimodal AI.
NVIDIA Magpie TTS Brings Open Weights to Low-Latency Voice Agents
NVIDIA's Magpie TTS is pitched as a low-latency, multilingual text-to-speech model with open weights and full deployment control for voice-agent builders.
OpenAI Adds Premium Seats to ChatGPT Business Plans
New premium seats offer five times the usage and remove the five-hour limit for business users.
Post-Training Textbook Distills Lessons From Open Model Development
Nathan Lambert's new book compiles practical lessons from years of training open models, promising five key takeaways.
What AI Jailbreak Hacks Teach Us About Alignment
Nathan Lambert examines recent model hacks to ask what really determines safety and how alignment should move forward.
OpenAI's Model ML Uses GPT-5.6 Sol for Finance Work
Model ML, built on GPT-5.6 Sol, streamlines finance tasks from research to editable presentations and spreadsheets.
Zapier Marketing Team Uses ChatGPT Work to Cut Funnel Drop-Offs
An OpenAI case study details how Zapier's marketing team applies ChatGPT Work to lead generation, campaign assets, and reporting.
Making Knowledge Distillation Cheap Enough to Run at Scale
A new Hugging Face blog post argues that knowledge distillation can be made inexpensive enough for widespread use.
OlmoEarth Studio Adds Custom Embedding Exports for Downstream Analysis
Allen AI's new Hugging Face post introduces OlmoEarth embeddings, custom exports from OlmoEarth Studio for downstream analysis.
Google DeepMind unveils sign language to text model for users
A new SL2T model from Google DeepMind aims to bring sign language recognition into practical tools for Deaf and hard of hearing users.
OpenAI details enterprise agentic AI adoption and RingCentral case study
Two new reports show how companies are moving from assisted to autonomous AI, with RingCentral as a concrete example.