Expert takes on AI tools, productivity, writing, and coding.
Inflect-v2 releases two text-to-speech models at 3.9M and 9.3M parameters, raising the question of how little compute speech synthesis actually needs.
August 7, 2026OpenAI has updated GPT-5.6 Sol for better accuracy and opened GPT-5.6 Luna to free-tier users with unlimited everyday chat access.
August 7, 2026Anthropic's Claude Opus 5 has climbed to first place on the Artificial Analysis Intelligence Leaderboard, signaling a meaningful shift in how frontier models stack up against each other.
August 7, 2026Anthropic has updated the biology-related safeguards in its latest model release. Here is what that shift signals for developers building on the platform.
August 7, 2026OpenAI is disputing Apple's lawsuit point by point, releasing internal messages to counter what it calls factually wrong accusations about its staff.
August 6, 2026Circles tapped the OpenAI API and Codex to rebuild how a telecom company talks to its customers, and the business numbers behind it are hard to ignore.
August 6, 2026OpenAI built a voice system that drops the awkward back-and-forth of traditional AI chat. Here is what that shift actually means for builders and creators.
August 6, 2026A single AI-supervised exam collapsed badly enough that tens of thousands of students now face a retake. The story is a stress test for automated oversight at scale.
August 6, 2026A new benchmark project called Computer Anthology aims to keep pace with fast-moving AI agent development by evolving its tests continuously rather than freezing them at a single point in time.
August 6, 2026As AI agents take on longer-horizon tasks, a quiet debate is surfacing: should engineers factor in model welfare when designing agentic systems?
August 6, 2026