OpenAI announced new changes to its File Search system last week, allowing more control to developers when asking the artificial intelligence (AI) chatbots to pick responses. The improvement has been made to the ChatGPT’s application programming interface (API) and will let developers not only check the behaviour of the chatbot’s response retrieval method, it also let them fine-tune its behaviour. This way, developers can ensure that only the desirable responses are picked. Notably, an earlier report claimed that the company is planning to launch another AI model dubbed ‘Strawberry’ that can improve ChatGPT’s mathematics and logical reasoning.
OpenAI Improves ChatGPT API for Developers
The AI firm announced the changes to the API in a post on X (formerly known as Twitter). The upgrade significantly improves file search capabilities in the Help API. This allows developers to review the products selected by the chatbot and modify other settings to suit their needs.
APIs are different from the web and customer-facing ChatGPT applications. Although the user interface that end users see is well organized by OpenAI, and is meant to be the same, developers are creating internal tools for businesses or installing a chatbot into various applications and systems, requiring greater freedom.
This is probably because the public version of ChatGPT is configured for general purposes, and the API version is used for a specific function. To be better, users want AI to make no mistakes and return answers that are perfect.
Until now, developers haven’t been able to tailor the API to generate chat responses for specific use cases, but that will change with the new control options. OpenAI has shown how it works on its support page.
Developers can review file search responses. The file search tool in the Helper API selects the answers it believes are relevant to a given search. However, developers can check the answers chosen by AI and test what it has done in previous runs. This information is meant to help them better understand how the device works.
In addition, developers can adjust the product rating settings that are used to search for information to generate responses. By choosing a level between 0.0 and 1.0, they can control what information AI collects and what information it ignores.
