Following the release of the new GPT-4 engine and Whisper API in March, OpenAI announced Thursday that he started introducing plugins for ChatGPT. These will allow the chatbot to interact with third-party APIs, tailoring its responses to the specific circumstances set by the developers while expanding the range of actions the bot is capable of.
Suppose you want to develop a chatbot that users can talk sports with. Before the latest GPT-4 update, the chatbot could only discuss games and scores that happened in the past, specifically in 2021, when GPT-3 training data was collected. It wouldn’t pull real-time data or even be aware that the year 2022 existed. With a chatGPT plugin, you will be able to integrate ChatGPT functionality into your existing code stack where it can do everything from fetching real-time news calls (sports scores, stock quotes, breaking news) to extracting a specific knowledge base of information such as your company’s internal documents or your personal cloud. It will even be able to take action on the user’s behalf, like booking a flight or ordering takeout – think of an installable Google Assistant created by the folks at OpenAI.
“The AI model acts as an intelligent API caller. Given an API specification and a natural language description of when to use the API, the model proactively calls the API to perform actions”, the OpenAI team wrote. “For example, if a user asks, ‘Where should I stay in Paris for a few nights?’, the model can choose to call a hotel booking plugin API, receive the response from the API, and generate a user-facing response that combines the data API and its natural language capabilities.”
The company also notes that using plug-ins to bridge the knowledge gap between what the model was trained on and what has happened since should help reduce the AI’s tendency to hallucinate facts when she answers complex questions. “These benchmarks not only enhance the utility of the model, but also allow users to assess the reliability of model results and recheck its accuracy, potentially mitigating the risks of overreliance,” the team wrote.
The additional capabilities and information provided to the model through its plugin also greatly increase the chances of the model returning problematic responses. To avoid the $100 billion hit that Google took over Bard, OpenAI would have subjected these plugins to extensive stress testing. “We conducted red team exercises, both internally and with external collaborators, which revealed a number of possible concerning scenarios,” the team wrote. They plan to use these findings to “inform security mitigations by design” to improve transparency and hinder the plugin against risky behavior.
The plugin itself is still in early alpha with limited availability. OpenAI has granted early access to a handful of partner companies, including Expedia, instagram, KAYAK, open table, Shopify, Soft, WolframAnd Zapier for use in their existing applications. You will need add your name to the waiting list to try it for yourself.
All products recommended by Engadget are selected by our editorial team, independent of our parent company. Some of our stories include affiliate links. If you purchase something through one of these links, we may earn an affiliate commission. All prices correct at time of publication.