OpenAI recently hosted its first-ever in-person DevDay conference, where it introduced several significant updates and features. One of the key highlights was the unveiling of the new GPT-4 Turbo AI model, which comes with the capability to provide answers with context up to April 2023, an improvement over previous versions limited to January 2022. The model also allows inputs of up to 300 pages, enabling users to summarise entire books, and supports DALL-E 3 AI-generated images, along with various preset voices for text-to-speech functionality.

In a move to make their AI models more accessible to developers, OpenAI announced price reductions for GPT-4 Turbo's input and output tokens. Additionally, users will now have the option to create customised versions of the popular ChatGPT chatbot, aiming to offer more personalised AI experiences for different applications and websites.

OpenAI also launched the GPT Store, a platform that allows creators of custom GPTs to share their models and earn revenue based on usage numbers. The store will feature different categories, such as productivity, education, and entertainment, showcasing a diverse range of valuable and engaging AI models.

OpenAI also made a commitment to cover legal costs associated with copyright infringement claims related to its AI models, ensuring customer protection and aligning with industry standards.