What's Happening?
OpenAI is retiring custom GPTs, the personalized ChatGPT assistants introduced in November 2023, and is providing a migration path to plugins. The specific retirement dates vary depending on the user's plan. For Enterprise workspaces, an admin notice
was issued on September 11, with a migration experience targeted for September 17. New custom GPT creation will end on September 25, and full retirement is scheduled for December 11, when custom GPTs will cease to function. For Free, Plus, and Pro plans, the transition timeline may be similar, with in-product notices to announce specific dates. Existing GPTs will remain usable until their retirement date but will become read-only after migration. Not all features will transfer directly; custom actions that called external APIs will need to be rebuilt, and conversation starters, previous chats, and selected models may not copy over. OpenAI has also replaced its App Directory with a Plugin Directory in July, coinciding with the launch of ChatGPT Work.
Why It's Important?
This transition marks a significant shift in OpenAI's strategy for custom AI solutions, impacting developers, businesses, and individual users who have built or relied on custom GPTs. The move to plugins, which combine reusable instructions with connected apps, suggests a more integrated and potentially more powerful ecosystem for AI functionalities. However, the requirement to rebuild custom actions and the potential for migrated plugins to behave differently could lead to disruption and additional development work for many users. For Enterprise clients, the firm deadlines necessitate immediate action to ensure continuity of their AI-powered workflows. The change also highlights OpenAI's focus on consolidating its offerings and streamlining its platform, which could lead to more robust and scalable solutions in the long term, but also requires users to adapt to a new paradigm for AI customization and integration.
What's Next?
Users of custom GPTs, particularly those in Enterprise workspaces, must begin migrating their existing GPTs to the new plugin architecture before the December 11 deadline. This involves rebuilding custom actions and thoroughly testing migrated plugins to ensure they function as expected. For Free, Plus, and Pro users, monitoring in-product notices for their specific transition timelines will be crucial. OpenAI advises users to document their GPTs' instructions, reference files, integrations, and prompts to facilitate the migration process and testing. The company's emphasis on plugins suggests that future AI development and integration will increasingly leverage this framework, requiring developers to familiarize themselves with its capabilities and limitations. The retirement of custom GPTs also implies a shift towards a more standardized approach to AI customization, potentially influencing how third-party developers build and offer AI-powered services.
Beyond the Headlines
The retirement of custom GPTs and the shift to plugins reflect a broader trend in the AI industry towards more structured and integrated AI application development. This move could enhance security and reliability by standardizing how AI models interact with external services, but it might also limit the flexibility and experimental nature that custom GPTs initially offered. The need for users to rebuild integrations highlights the ongoing challenge of interoperability and the rapid evolution of AI platforms. This transition could also lead to a more centralized control over AI functionalities by OpenAI, potentially influencing the diversity of AI applications available. Furthermore, the consolidation of features into plugins could foster a more robust marketplace for AI tools, but also create a steeper learning curve for new developers entering the ecosystem. The emphasis on 'skills, apps and app templates' within plugins suggests a future where AI functionalities are modular and easily deployable across different platforms.
