Andrej Karpathy, former OpenAI researcher and head of AI at Tesla, explains that when people "ask an AI," they're actually interacting with averaged responses from human data labelers, not a magical ...

A new simulator from Epoch AI shows that training models on the scale of GPT-4 would have been possible with older hardware, but at a significantly higher cost. The simulator analyzes how efficiently ...

A new study from University College London shows that large language models predict scientific research results more accurately than human experts. Published in Nature, the study found AI models ...

Alibaba has released QwQ-32B-Preview, a new AI model that focuses on logical reasoning and problem-solving capabilities. The model appears to match and sometimes outperform OpenAI's latest offerings ...

A hacker successfully manipulated an AI chatbot called Freysa through clever text prompting, winning a $47,000 prize pool after 482 attempts. The experiment was simple: participants could try to ...

Five major Canadian media organizations filed a lawsuit against OpenAI, claiming the AI company used their content without permission to train ChatGPT. The lawsuit, filed by Torstar, Postmedia, The ...

Mistral AI adds web search and image generation to its Le Chat AI assistant, while introducing a new visual model that performs well on industry benchmarks. Le Chat users can now access current web ...

AI startup /dev/agents has secured $56 million in funding to create an operating system for AI agents. The company aims to enable computers to collaborate like humans do, which requires developing new ...

A federal court in Massachusetts has backed a school's decision to penalize a student who used AI for homework without disclosure. The case could set an important precedent for how schools handle AI ...

Amazon has rolled out automatic prompt optimization for its Bedrock AI service, promising better performance across various AI tasks with minimal user effort. The new feature lets users optimize ...

Microsoft has released guidance for system administrators to prevent its Copilot AI from accessing and sharing too much information. The problem occurs when Copilot can see data beyond a user's ...

Microsoft has released TinyTroupe, an open-source library that creates AI-powered simulations of human behavior in virtual settings. The system could help companies test products and ideas without ...