 "Customise GPT to GPT 4 Turbo: Everything OpenAI announced at 'DevDay' event")
US-based technology company OpenAI on November 7 announced customisable GPT for users to create their own tailored versions of the ChatGPT chatbot, and a refined but cost-effective version of GPT-4 model called Turbo. OpenAI said that its custom version of the chatbot would be able to combine instructions, information and combination of skills for a specific use case.
According to OpenAI, users would need no coding experience to create custom GPTs. Moreover, users would be able to share their GPTs with others through an online ‘GPT Store’. Customised chatbots would be available only to ChatGPT Plus subscribers and OpenAI enterprise customers. The company has promised that the creators would be paid based on how much their GPTs are used.
Also Read: 7 best artificial intelligence courses to help you grow in your career
OpenAI said each GPT version would be able to access the web, DALL-E for image generation and OpenAI’s code interpreter tool for writing and executing program codes. There would also be a ‘Knowledge’ section in the builder section for uploading case specific data, said the company. OpenAI would also let custom GPTs access external data like emails, databases, and more with future updates.
On privacy, OpenAI said that users' data would not be shared with the GPT creators and users would have the option to choose what data is to be recorded.
Alongside customisable GPTs, OpenAI lso introduced a new GPT-4 Turbo model in preview that is said to be refined but cost effective version based on the GPT-4 model. The GPT-4 Turbo would be available in two versions, one dedicated to text analysis and another capable of processing both text and visuals. The new model is available in a preview through an API, while the company plans to roll out both versions in coming weeks.