It has been nearly a year since ChatGPT’s public debut, and its evolution since then has been remarkable. In just over 11 months, OpenAI’s chatbot has achieved significant milestones such as writing programming code, processing information from multiple modalities, and expanding its reach across the internet with APIs. During OpenAI’s 2023 Dev Day keynote address, CEO Sam Altman and other executives introduced the chatbot’s latest iteration, ChatGPT-4 Turbo, as well as a new way to bring generative AI technology to everyone called GPTs.
GPTs, which are task-specific iterations of ChatGPT, can be compared to the single-purpose apps and features on your phone. Instead of maintaining a timer or transcribing voice instructions, GPTs can be trained to do almost anything. OpenAI offers eight examples of what GPTs can be used for, including a digital kitchen assistant that suggests recipes based on pantry items, a math mentor to help with homework, and even a Sticker Wiz that turns your ideas into custom die-cut stickers.
The introduction of GPTs builds upon OpenAI’s existing Custom Instructions feature, which was released in July. Power users were already recycling and updating their most effective prompts and instruction sets, and GPT-4 Turbo now automates this process by updating seed parameters and focusing on reproducible outputs. This improvement allows users to have greater control in customizing the GPTs according to their specific needs.
One of the significant enhancements in GPT-4 Turbo is its improved code interpretation, retrieval, and function calling capabilities. With its expanded context window size, users no longer need extensive knowledge of JavaScript programming to develop their GPTs. They can devise and develop their GPTs using natural language, making it more accessible and user-friendly.
Related Post
Moreover, any GPT created by the community will be immediately shareable. OpenAI plans to launch a centralized storefront later this month where “verified builders” can post and share their GPTs. These GPTs will be ranked on a leaderboard based on popularity and usage, potentially allowing creators to earn money based on the number of users utilizing their GPTs.
GPTs will be available to both regular users and enterprise accounts, similar to the previously released ChatGPT Enterprise. Enterprise accounts allow institutional users to create their own internal-only, admin-approved mini-chatbots. These enterprise GPTs can be trained on specific tasks, department documentation, or proprietary datasets, providing tailored solutions for businesses. Enterprise GPTs will be available to customers starting on Wednesday.
Privacy remains a key focus for OpenAI, and the company continues to implement additional technical safeguards to prevent the creation of GPTs that violate usage policies. These safeguards work alongside existing moderation systems to ensure responsible and ethical usage. OpenAI is also introducing an identity verification system for developers to enhance transparency and trust, although specifics about the process were not elaborated on.
In conclusion, OpenAI’s ChatGPT has made significant strides since its public debut, culminating in the release of ChatGPT-4 Turbo and the introduction of GPTs. These advancements allow users to develop custom GPTs using natural language, opening up a world of possibilities for both regular users and enterprise accounts. With privacy and security at the forefront of their concerns, OpenAI continues to ensure responsible and ethical usage of their AI technology.
