OpenAI’s New Marketplace for Custom GPTs Raises Concerns about Quality Standards
OpenAI is set to launch its GPT store this week, allowing users to create custom AI assistants using simple text instructions. While this development has been praised for democratizing AI development, concerns about the quality of the custom GPTs available on the marketplace persist.
The store’s original launch date in November 2023 was postponed due to leadership issues within OpenAI. However, those who signed up as GPT Builders recently received an email announcing the imminent launch and urging compliance with the store’s guidelines.
Media outlets and social media platforms have been abuzz with speculation about the impact of this marketplace on AI innovation. Some hail the no-code approach as a significant step towards making AI development more accessible, while others caution against the potential low standards of the GPTs available.
One major concern is the profit motive that may flood the marketplace with countless iterations of popular AI assistant ideas. This saturation could make it challenging for users to discern the genuinely useful GPTs from the mediocre ones.
Jeff Watkins, Chief Product and Technology Officer at digital consultancy xDesign, acknowledges the appeal of the no-code approach but warns that the agents built using the GPT builder lack true customization. According to Watkins, these agents may be limited in comparison to alternative options such as GPT Plugins.
Watkins argues, This marketplace will not be a genuine threat to those building a proper Gen-AI service. They also require a premium subscription and aren’t suitable for a deployed service, the OpenAI APIs are more suitable for that.
On the other hand, Peter van der Putten, AI Lab director at low-code software specialists Pegasystems, sees potential in the concept of custom GPTs. He believes that OpenAI’s marketplace will contribute to the adoption of generative AI agents, particularly in long-tail consumer apps.
Van der Putten suggests that businesses could benefit from a simplified way to build custom AI assistants, but adds that strict governance is necessary to ensure privacy, security, and integration with internal systems.
Despite the excitement surrounding the no-code approach, Watkins predicts that the marketplace may become overwhelmed with hastily built agents targeting popular use cases. He urges OpenAI to implement quality control measures and a ratings system to filter out the genuinely useful GPTs from the rest.
As OpenAI’s GPT store prepares to launch, it remains to be seen how the concerns about quality standards will be addressed. While the marketplace has the potential to democratize AI development and popularize generative AI agents, it will need to strike a balance between accessibility and maintaining high-quality standards.
In the ever-evolving landscape of AI, it is crucial to navigate the delicate balance between innovation and quality, ensuring that users can trust the AI agents they encounter in the marketplace.
