Upgrading existing conversations to the latest GPT4.0

ChatGPT Feature requests
1

It would be very useful to be able to upgrade conversations that started in the older version to the new one.
This will ensure conversation continuity as oppose to having to start a new conversation on the same topic.

1 Like
2

Yes, this is important. Are GPT4 conversations automatically updated to include new GPT4 features such as when GPT4-Turbo becomes live? Some clarity on this would be appreciated.

3

Here is the Answer: Here’s a more precise breakdown:

  1. Older Conversations and Updates:
  • It is correct that older conversations with GPT-4 or any of its predecessors are not automatically updated to newer versions. Each conversation is a snapshot of the model’s capabilities and knowledge at that specific point in time.
  • There is indeed no feature to “upgrade” a past conversation to a newer version of the model. To access updated capabilities or knowledge, a new conversation must be initiated.
  1. Automatic Updates for GPTs:
  • The GPT models are periodically updated by the developers at OpenAI. These updates are not automatic for end-users.
  • Updates can involve various improvements, such as refining the model’s understanding, reducing biases, and increasing its overall performance and reliability. Major version updates, like from GPT-3 to GPT-4, often bring significant enhancements in capabilities and functionalities.
  1. Reasons for Separate Conversations:
  • Keeping conversations separate is beneficial for several reasons:
    • It helps in monitoring and evaluating the model’s performance across different versions. This is essential for the continuous development and refinement of the models.
    • From a user perspective, it maintains clarity and consistency. Users can refer back to a conversation and see the information as it was provided at that time, unaffected by subsequent updates.

In summary, your understanding of how GPT-4 and its updates operate is fundamentally correct. The conversations are static and reflective of the model’s state at the time, updates to the model are managed and implemented by developers, and separate conversations serve both analytical and user experience purposes.