OpenAI postpones launch of custom GPT store, continues enhancements based on feedback

San Francisco: OpenAI, the developer behind ChatGPT, has decided to delay the launch of its custom GPT store until early 2024, as per an internal memo.

Initially announced during OpenAI's inaugural developer conference in November, the custom GPTs and their associated store were initially slated for a later release within the same month.

However, the company has opted to postpone the launch to early 2024, citing ongoing efforts to refine the GPTs based on customer feedback, as outlined in the internal memo, reported Reuters.

This delay coincides with recent internal changes at OpenAI, notably the temporary removal and subsequent reinstatement of CEO Sam Altman, following internal workforce concerns.

The custom GPTs represent early-stage AI assistants designed to undertake practical tasks, such as flight bookings on behalf of users. The platform's concept also includes the prospect for users to share their GPTs and potentially earn income based on user engagement.

In a recent announcement, OpenAI disclosed plans to collaborate with various organisations to curate both public and private datasets, aiming to train more sophisticated artificial intelligence (AI) models.

The ChatGPT, a widely known chatbot capable of generating poems and text from basic prompts, is built on extensive language models trained solely on publicly available internet data.

OpenAI's latest initiative seeks to diversify and enhance the available training data, specifically focusing on capturing more natural and conversational language styles.

The company has expressed its intent to procure datasets that authentically express human intentions across diverse languages, topics, and formats, as detailed in a recent blog post.

Moreover, OpenAI has extended invitations for potential partnerships to aid in the development of an open-source dataset tailored for training language models, aiming for widespread accessibility in AI model training.

Simultaneously, the company is preparing private datasets aimed at training proprietary AI models, aligning with its commitment to advancing AI capabilities and fostering a robust AI ecosystem.