ChatGPT plugins represent a new capability allowing language models to access real-time data, perform computations, and utilize third-party services securely. OpenAI is gradually introducing these plugins, starting with a limited user base and expanding access as they gather insights into usage and ensure safety.
Initially, plugin development is open to selected developers from the waitlist, who can leverage documentation to create plugins. These plugins enable ChatGPT to interact with diverse third-party services and access up-to-date information beyond its initial training data.
Current plugin development involves companies like Expedia, Instacart, and Wolfram, as well as open-source contributions. By acting as conduits for personalized and specific data, plugins enhance ChatGPT’s functionality, paving the way for a broad spectrum of applications and advancing human-AI interactions in innovative ways.
More details about Code Interpreter by OpenAI
What is the function of the ChatGPT plugin in Code Interpreter?
The ChatGPT plugin in Code Interpreter functions to grant language models access to real-time information, enabling them to execute computations within code. This includes the ability to utilize third-party services, acting as a conduit for the models to interact with recent, personalized, or specialized data beyond their initial training.
Can the plugins perform complex code computations?
Yes, the plugins are capable of facilitating complex code computations. For example, ChatGPT’s code interpreter plugin can handle both quantitative and qualitative mathematical problems, as well as conduct data analysis and visualization tasks.
Can I build my own plugins for ChatGPT?
Yes, developers who have been invited from the ChatGPT plugin waitlist can construct their own plugins. They can utilize comprehensive documentation provided by OpenAI to facilitate this process. These plugins empower ChatGPT to integrate various third-party services or access specific data sources as required.
How secure is the use of plugins in Code Interpreter?
The use of plugins in Code Interpreter prioritizes security. Special precautions are taken, particularly when plugins enable AI to execute Python code, ensuring that executed code is sandboxed to prevent unintended real-world impacts. Strict network controls are in place to prevent external internet access from executed code, and each session operates under stringent resource limits to maintain safety.