Sora 2: Official System Documentation Card
Sora 2 is OpenAI's newest video and audio generation model, building on the original Sora with five key upgrades: more accurate physics, sharper realism, synchronized audio, improved steerability, and…
generative AI
Q&A: Understanding Today’s Agentic AI and the Future We Want It to Become
Agentic AI acts via tools beyond text. Coding agents excel, but risks include errors, privacy leaks, de-skilling, and limits on human judgment.
OpenAI recognized as an Emerging Leader in Generative AI
OpenAI has been named an Emerging Leader in Gartner's 2025 Innovation Guide for Generative AI Model Providers. The company says this recognition reflects its expanding enterprise momentum, with more t…
Music Player Machine
Jukebox is a neural network that generates music as raw audio across many genres and artist styles, even producing basic singing. Unlike MIDI-based systems, it creates rich, recording-like sound. Its …
Improved language models and what they mean
This article explains OpenAI's GPT-2, a large-scale unsupervised language model trained only to predict the next word. Despite no task-specific training, it generated coherent paragraphs, set language…
Refined Approaches for Training Consistency Models
Consistency models are a new family of generative models that can produce high-quality samples in a single step. They combine the speed of GANs with more stable, non-adversarial training, offering a m…
Unveiling ChatGPT Images alongside GPT‑Image‑1.5
The 2025 ChatGPT Images update delivers a much faster image-generation experience inside ChatGPT and introduces GPT-Image-1.5, the latest version of OpenAI's image model, now available through the API…
Steady-State Frameworks
Diffusion models power today's best generative AI tools by learning to reverse gradual noise corruption. Their iterative, step-by-step sampling delivers high-quality, coherent images, audio, and video…
OpenAI Japan unveils blueprint prioritizing teen safety in Japan
OpenAI Japan has launched the Japan Teen Safety Blueprint, a safety framework aimed at protecting young AI users. It rests on three pillars: age-based protections with content filters for teen account…
Citing Muse AI Agent
Muse AI Agent's automated reply falsely told a pickup contact that the account was home at 9:27, even though the owner was unavailable. The message worsened a missed MX Keys Mini handoff, caused anger…
Video generation models as world simulation engines
This article explains how Sora, a text-conditional video generation model, creates minute-long high-fidelity videos. It covers how training on videos and images of variable formats expands visual know…
Citing Laurie Voss
Writing code used to be the expensive part of building software, but AI has collapsed that cost—and the cost of reviewing, fixing, and running code is falling behind it. Drawing on Laurie Voss's argum…
Claude Cowork merges with chat into a single unified Claude
Anthropic is merging Claude Cowork with regular Claude chat into a single unified assistant, rolling out first to Pro and Max subscribers on web, desktop, and mobile over the coming weeks. The merger …
Citing Paul Ford
AI can generate impressive code, but building great software has never been just about writing code. Drawing on Paul Ford's argument, this article explains why the flood of AI-assisted projects often …
In the words of Thariq Shihipar
Anthropic announced Claude Code (v2.1.277+) now supports AGENTS.md, the emerging open standard for coding agent instructions. When no CLAUDE.md file exists, Claude Code falls back to reading AGENTS.md…