There are still important differences between our current empirical setting and the ultimate problem of fitting superhuman models. For example, it may be easier for future models to mimic weak human errors than for current strong models to mimic current weak model errors, which could make future generalization more difficult. Nevertheless, we believe that our…

Agentic AI systems—artificial intelligence systems that can pursue complex goals with limited direct supervision—are likely to be widely useful if we can responsibly integrate them into our society. While such systems have considerable potential to help people achieve their goals more efficiently and successfully, they also create risks of harm. In this white paper, we…

Our discussions with The New York Times appeared to be progressing constructively through our last communication on December 19th. Negotiations focused on a high-value partnership around real-time display with attribution in ChatGPT, in which The New York Times would gain a new way to connect with its existing and new readers, and our users would…

We're launching the GPT Store to help you find useful and popular custom versions of ChatGPT. Source link

Integrating artificial intelligence into everyday organizational workflows can make your team more productive. In a recent Harvard Business School study, Boston Consulting Group employees who were given access to GPT-4 reported completing tasks 25% faster and achieving 40% higher quality in their work compared to colleagues who did not have access.(^study) Connor O'Brien, VP of…

Protecting the integrity of elections requires cooperation from all parts of the democratic process, and we want to ensure that our technology is not used in a way that could undermine this process. Our tools empower people to improve their everyday lives and solve complex problems—from using artificial intelligence to improve government services to simplification…

We funded 10 teams from around the world to come up with ideas and tools for collaborative AI management. We summarize the innovations, describe our learnings, and invite researchers and engineers to join us as we continue this work. Source link

We're launching a new generation of embedding models, new GPT-4 Turbo and moderation models, new tools to manage API usage, and soon lower prices for GPT-3.5 Turbo. Source link

Note: As part of our Preparedness framework, we are investing in the development of improved security risk assessment methods enabled by artificial intelligence. We believe that these efforts would benefit from wider input and that sharing methods could also be of value to the AI ​​risk research community. To that end, we present some of…

For Enterprise and Team users, memory can be useful when using ChatGPT for work. It can learn your style and preferences and build on previous interactions. This saves you time and leads to more relevant and insightful answers. For example: ChatGPT can remember your tone, voice, and format settings and automatically apply them to draft…