OpenAI tests SearchGPT, its conversational search engine with citations
OpenAI has unveiled SearchGPT, a search prototype that answers in natural language and links to its sources. The limited trial opens a new front in the race for web search.
OpenAI has unveiled SearchGPT, a search prototype that answers in natural language and links to its sources. The limited trial opens a new front in the race for web search.
OpenAI has acquired Rockset, a real-time analytics database. The deal aims to improve how its products find useful information across large volumes of data, a key capability for enterprise customers.
NVIDIA closed the June 18 session as the world's most valuable publicly traded company, ahead of Microsoft and Apple. The market is rewarding its central role in the infrastructure used to train and run artificial intelligence models.
Luma AI has launched Dream Machine, a model that generates video clips from text and images. Its public release brings AI video creation to users who until now could only watch demos of rival models.
Apple has unveiled Apple Intelligence, an AI layer for the iPhone, iPad and Mac that rewrites text, creates images and gives Siri more context. It will also optionally integrate OpenAI’s ChatGPT.
OpenAI has temporarily pulled Sky, one of ChatGPT’s voices, after Scarlett Johansson said it sounded like her. The case puts consent and vocal identity at the heart of the race to build AI assistants.
OpenAI has dissolved the team created to research how to control a future superintelligence. The decision comes as its two leaders, Ilya Sutskever and Jan Leike, leave the company, with Leike questioning OpenAI's safety priorities.
Ilya Sutskever, OpenAI cofounder and chief scientist, is leaving the company after nearly 10 years. His departure closes a chapter shaped by his work on large models and the leadership crisis that briefly ousted Sam Altman in November.
OpenAI launches GPT-4o, a model that natively processes text, images and audio. GPT-4o and several features are reaching free ChatGPT users, while advanced audio will roll out later.
Microsoft Research has unveiled VASA-1, a system that generates a video of a talking face from an image and audio. The advance makes these avatars more expressive, but the company will not release it because of the risk of impersonation.
Humane’s $699 AI Pin, a screenless AI assistant, is drawing harsh reviews in its first tests. Critics question its speed, reliability and ability to replace the smartphone.
NVIDIA expects to bring its Blackwell architecture into production this year with major cloud providers and server makers. Its new B200 chips and GB200 systems aim to lower the cost of training and serving ever-larger AI models.
This website uses cookies to improve the browsing experience. Cookie policy.