Anthropic is treating its new Claude Opus 4 language model as safety-critical after tests revealed some troubling behavior, including escape attempts, blackmail, and autonomous whistleblowing. After ...

A research group at Arizona State University cautions that the intermediate steps produced by language models are simply statistically generated text, not signs of human-like thinking or understanding ...

A recent study reveals that people struggle to differentiate between therapeutic responses from ChatGPT and human therapists, with the AI's answers often rated as more empathetic than those from ...

Perplexity has launched "Labs," a new feature for Pro users that can run complex projects like reports, tables, dashboards, or basic web apps. Unlike the faster "Research" mode, Labs works ...

The New York Times has signed a multi-year deal with Amazon, allowing the tech company to display content from the Times, NYT Cooking, and The Athletic on Alexa and use it to train its AI models. In ...

Nvidia delivered another record quarter, but the ongoing US export ban on AI chips to China is cutting deep into its bottom line. CEO Jensen Huang is warning that the policy could have serious ...

Odyssey has released a research preview of a new kind of video: fully AI-generated sequences that respond to your input in real time. The demo features what Odyssey calls interactive videos—AI-created ...

OpenAI has started rolling out an expanded memory feature for free ChatGPT users. According to OpenAI, this group gets a "lightweight" version of memory designed to provide short-term continuity ...

A new study by Apple researchers, including renowned AI scientist Samy Bengio, calls into question the logical capabilities of today's large language models - even OpenAI's new "reasoning model" o1.

OpenAI has unveiled GPT-4o, its latest large multimodal model that sets new standards in performance and efficiency by combining text, image, and audio processing in a single neural network. The "o" ...

The company defines such systems as those that match or exceed human capabilities in most cognitive tasks. Deepmind expects current machine learning methods—particularly neural networks—to remain the ...

Mistral AI has rolled out a major update to its AI assistant Le Chat, adding new features and mobile apps while revamping its pricing structure. One of the most notable additions is "Flash Answers," ...