Artificial intelligence: beyond the hype
What AI means, how models learn, and why good questions matter more than impressive demonstrations.

The ideas behind the technology. The people behind the ideas.
News, context and clear explanations of AI.
What AI means, how models learn, and why good questions matter more than impressive demonstrations.
A new report describes AI-assisted influence activity. Attribution and reach are OpenAI assessments and should be read with care.
The research evaluates multi-step work and adherence to business rules. Its test results do not guarantee reliable automation in every real deployment.
OpenAI models will support agents across Atlassian and Rovo, with an emphasis on connecting enterprise context to specific work.
Materials produced by an internal AI model are available on GitHub. Lean formalizations cover part of the collection rather than every result.
OpenAI describes flashcards and quizzes while preparing an application-planning tool. The initial College Planner experience is intended for US students.
Intelligent UI combines text, graphics and interactive elements. Availability is rolling out gradually and differs by reasoning mode and plan.
How to consider quality, limitations and risk when using AI in a real process.
A look at attention and the research that introduced the Transformer architecture.
We distinguish news from analysis and possibilities from promises. We provide context and links to original sources.
About Super Intelligence ↗