Introduction
In a groundbreaking move, OpenAI has recently announced major upgrades to their popular AI models – the GPT-3.5-turbo and GPT-4, with the introduction of innovative features like function calling, and more. The company also announced significant reductions in the pricing of their models, ensuring wider access to AI for developers across the globe. This move, aptly named the “ChatGPT Pricing Update”, is set to make big waves in industries reliant on AI for various operations.
Table of Contents
Function Calling: A Game-Changer
One of the most significant introductions to the GPT-3.5-turbo and GPT-4
models is the function calling feature. This feature enables developers to describe programming functions to the AI, which in turn, generates code to execute these functions. This innovation is instrumental in creating highly efficient chatbots that can utilize external tools to answer questions. Moreover, this feature facilitates conversion of natural language into specific database queries and extracting structured data from text, significantly increasing its applicability in various business scenarios.
Structured Data from Unstructured Text
The function calling capability has the potential to revolutionize the way developers handle large volumes of unstructured data. By enabling the extraction of structured data from text, it simplifies analysis and utilization processes. This can be particularly useful for businesses that need to sift through vast amounts of customer feedback or social media posts. Consequently, OpenAI’s GPT-3.5-turbo and GPT-4 can deliver structured data back to developers with greater reliability, allowing AI-driven applications to reach new performance and usability heights.
Expanded Context Window
OpenAI has significantly improved its models by expanding the context window size, which is essentially the amount of text a model considers before generating additional text. This enhancement helps the models to remember more of their recent input, allowing them to stay on topic and prevent off-topic generation. In its updated version, the GPT-3.5-turbo offers a context window of 16,000 tokens, four times the size of the previous version. While the update comes at double the price of the older model, it can process about 20 pages of text at once, a significant improvement in processing extensive documents or maintaining longer conversations.
Making AI Accessible
Along with adding new features, OpenAI has also taken substantial steps towards making its AI models more accessible. The company has significantly reduced the pricing for the original GPT-3.5-turbo model and the text-embedding-ada-002 model. For GPT-3.5-turbo, developers can now use the model for approximately $0.0015 per 1,000 input tokens and $0.002 per 1,000 output tokens, translating to about 700 pages per dollar. The text-embedding-ada-002 model, used for measuring text relatedness and essential for search functions and recommendations, has seen a whopping 75% price reduction.
Impact on AI Field
The “ChatGPT Pricing update” is a significant step towards affordability, without compromising on the quality of service. It allows developers and businesses of all sizes to access powerful AI tools that might have been beyond their budget constraints. Moreover, the release of GPT-4 signals OpenAI’s strategy of making incremental updates to existing models instead of focusing on creating entirely new ones. This approach could facilitate rapid improvements and enhance user experiences with each update.
Conclusion
OpenAI’s latest updates and pricing adjustments are significant steps towards making advanced text-generating models more accessible to developers and businesses. The integration of function calling and the expansion of the context window could change how businesses interact with AI models and unstructured data. As OpenAI continues to invest in research and development and infrastructure, it’s crucial to keep an eye on their progress and future updates to these exciting models.
Source: OpenAI
