My first take on Sora
A first reflection on Sora, the AI video model, following its public unveiling. The team has spent the month collaborating closely with artists to explore how Sora can support and enrich their creativ…
Hands-on reviews of AI apps, assistants and image, video and voice tools: pros, cons, pricing and who each one is for.
A first reflection on Sora, the AI video model, following its public unveiling. The team has spent the month collaborating closely with artists to explore how Sora can support and enrich their creativ…
OpenAI tested Voice Engine with about 20 trusted partners, a tool that creates human-like voices from just 15 seconds of audio. Findings show synthetic voices are convincing enough to enable fraud and…
OpenAI has launched GPT-4o, its newest flagship omni-capable model, and is bringing it to free ChatGPT users alongside expanded tool access. The unified model natively handles text, voice, and other i…
OpenAI's new flagship model, GPT-4 Omni (GPT-4o), natively reasons across audio, vision, and text in real time. Unlike older voice AI pipelines that lost tone and added lag, GPT-4o enables natural, in…
AGI refers to AI systems matching or exceeding human abilities across many tasks. Because such technology could transform healthcare, education, work, and science, responsible development is essential…
Apple and OpenAI have announced a partnership integrating ChatGPT into Apple's ecosystem. ChatGPT will be embedded into Apple Intelligence, Siri, and systemwide Writing Tools, with user permission req…
OpenAI introduced CriticGPT, a GPT-4-based model that critiques ChatGPT's responses to help human trainers spot errors they might miss. As AI outputs grow longer and more complex, mistakes increasingl…
SearchGPT is a temporary prototype testing AI-powered search that combines conversational answers with live web information and clear source citations. Unlike traditional search engines that list link…
Economist Tyler Cowen praises OpenAI's o1 for its reasoning-first design, which lets the model think through problems internally before answering. He found it unusually capable of following long chain…
OpenAI unveils a new multimodal moderation model built on GPT-4o, offering improved detection of harmful text and images. This upgrade to the Moderation API provides developers with a stronger foundat…
OpenAI banned accounts linked to Iran-based threat actor STORM-0817 after discovering they used its AI models to debug Android malware, scrape social media content, and translate malicious tools. The …
Hearst has signed a content partnership with OpenAI to bring its lifestyle content and local news reporting into ChatGPT and other AI products. The deal covers expert-reviewed guidance from Hearst mag…
SimpleQA is a factuality benchmark that tests whether large language models can correctly answer short questions with a single verifiable answer. It measures accuracy, hallucination rates, and calibra…
Rox is placing its full bet on OpenAI by blending real-world commercial experience with deep expertise in large language models. By leveraging OpenAI's models, Rox aims to empower every salesperson to…
OpenAI and Future, a global platform behind 200+ specialist media brands, announced a strategic partnership. Content from Future's portfolio spanning technology, gaming, entertainment, and lifestyle w…
Sora is OpenAI's video generation model that transforms text, images, or video clips into new videos. Built on insights from DALL-E and GPT, it combines visual generation with deep language understand…
OpenAI's video generation model Sora is now publicly available at sora.com, letting users turn text descriptions into finished video clips without cameras or editing software. It supports up to 1080p …
OpenAI announced expanded access to its o1 reasoning model for paid-tier developers, enabling step-by-step thinking for coding, math, and scientific tasks. The Realtime API received reliability and co…
OpenAI's partnerships with Axios and hundreds of publishers let ChatGPT surface licensed news with links and attribution, while grant programs help smaller newsrooms adopt AI tools for summarizing, tr…
This article explains the safety testing behind OpenAI's o3-mini model in plain language. It covers three layers of scrutiny: internal safety evaluations checking whether the model refuses harmful req…