Last Updated on August 24, 2024
OpenAI is introducing a new feature that enables corporate clients to tailor their GPT-4o model using their own company data. This update responds to increasing competition among startups for AI business solutions and the mounting pressure on companies to prove the value of their AI investments.
Starting Tuesday (Aug 20), OpenAI will offer this customization option, known in the AI field as fine-tuning. Fine-tuning involves training existing AI models with additional data specific to certain tasks or topics. For instance, a skateboard manufacturer might fine-tune an AI model to serve as a customer service chatbot, equipped to handle inquiries about wheels and board maintenance.
This new capability marks a first for OpenAI’s flagship GPT-4o model, as fine-tuning was not previously available for GPT-4o or its earlier version, GPT-4. However, OpenAI has permitted users to fine-tune several other models, including GPT-4o mini, a more affordable and simplified variant of GPT-4o.
Numerous tech companies provide options for customizing AI models of various sizes. OpenAI aims to streamline the process for customers to modify its most advanced model by working directly with the company, rather than through third-party services or less capable products, according to Olivier Godement, the head of product for OpenAI’s API.
Godement stated,
“We have been intensely focused on reducing the barriers, friction, and effort required to get started.”