Overview
OpenAI has launched updated versions of its GPT-3.5 Turbo and GPT-4 large language models, as well as a new “function calling” feature that enables developers to describe a program operation conversationally to the AI, which will then generate computer code. This opens doors for conversational generative AI chatbots, allowing developers to retrieve information beyond immediate databases and access approved tools from outside the company. Additionally, OpenAI has cut the cost of its APIs, with the smaller model now 25% cheaper. The new Turbo model has a larger context window of 16,000 tokens, quadrupling the previous version’s capacity.