Harvey Leans on GPT-6 Astra for Longer-Context Legal Drafting
OpenAI says legal AI firm Harvey is using GPT-6 Astra to pull more matter context into drafts and encode lawyer preferences. The post offers no benchmarks or error rates.
5 min read
The archive
All articles
Guides and explainers on artificial intelligence, newest first. Pick a topic to narrow things down.
Ringg Says Its Voice Agents Close 65% of Customer Calls — With a Model-Routing Stack Behind It
OpenAI's customer story on Ringg claims 65% call resolution, 7M+ monthly connected calls and ~90% lower model costs. The routing architecture is more instructive than the numbers.
6 min read
OpenAI Says invideo Tripled Color-Grading Success With GPT-6 Astra
OpenAI's customer story claims invideo's agentic video editor tripled its color-grading success rate and built about 50 custom effects in a day. The baselines behind those numbers were not disclosed.
5 min read
OpenAI Releases MentalHealthBench, an Expert-Written Test for Sensitive Conversations
OpenAI says more than 80 licensed clinicians in 22 countries wrote the rubrics for MentalHealthBench, an open evaluation of AI responses across everyday stress, high distress and emergencies.
5 min read
OpenAI Pushes ChatGPT Ads Into Southeast Asia and Taiwan, Passing 60 Countries
OpenAI says ChatGPT Ads is rolling out in seven more Asian markets, bringing the ad product to over 60 countries. Key details on formats, pricing and measurement remain undisclosed.
5 min read
Airbnb Expands Its OpenAI Deal to GPT-6 Astra, With Self-Reported Productivity Gains
OpenAI says Airbnb is broadening access to its frontier models across engineering and product teams. The productivity figures are notable, self-reported, and short on methodology.
5 min read
OpenAI and Grab Will Train 30,000 Gig and Merchant Partners Across Southeast Asia
OpenAI's first regional skilling program with Grab targets 30,000 driver, delivery and merchant partners over two years, starting in Singapore. The details on outcomes are thin.
5 min read
OpenAI Puts Prompt Caching at the Center of GPT-6 Agent Economics
OpenAI says GPT-6 ships with a reworked prompt caching system, a 30-minute reuse window, a caching dashboard and miss diagnostics — with customers reporting hit rates above 90%.
5 min read
A $10M Fellowship Bet on Mental Health Research — and the AI Tools Around It
MIT says a new $10 million gift will fund 50 two-year fellowships for young psychiatric disorders researchers, in a field where AI risk prediction tools are starting to appear.
5 min read
OpenAI Adds GPT-6 Sol and Luna, Pushing the Fight to Cost Per Task
OpenAI says its two new GPT-6 tiers halve API prices versus GPT-5.6 promotional rates and improve prompt caching. The benchmark claims come with fine print worth reading.
6 min read
Parallel says GPT-6 Astra halved its research agent's time and cost — with caveats
OpenAI's customer story credits GPT-6 Astra with cutting Parallel's research agent runtime and cost roughly in half. The numbers are relative, the methodology is thin, and the baseline is unnamed.
5 min read
Guardrails for AI Applications: A Layered Approach to Safer Outputs
No single filter makes an AI application safe. This guide explains how to layer input checks, model instructions, output validation, permissions, and monitoring into a practical defense.
8 min read