How to switch back from GPT-3.5 to ChatGPT-4 when you reached the limit of 25 messages in 3 hours

BlueFx Video Templates
3 May 202306:14

TLDRThe video outlines a workaround for switching back to ChatGPT-4 after reaching the message limit of 25 messages in a 3-hour period, which defaults the user to GPT-3.5. The solution involves using Firefox to inspect the conversation's network activity, locating and altering the model version to 'gpt4', and then running the changes to regain access to ChatGPT-4. Credit is given to a Reddit user, kinky Cusco, for this method.

Takeaways

  • 🚫 GPT-4 switches to GPT-3.5 after 25 messages in a 3-hour period.
  • ⏳ To switch back manually, wait for at least two hours after reaching the limit.
  • 💻 Use Firefox browser for the process of switching back to GPT-4.
  • 🔍 Open the conversation where the switch occurred and inspect the elements.
  • 📝 Write a test message to trigger the network activity.
  • 🔗 Click on 'Network' and use the 'Fetch' option in the console.
  • 🔑 Locate the 'model' in the text and change 'text DaVinci 2.0.2 render' to 'gpt/4'.
  • 🔑 Keep the backslash '/' in 'gpt/4' for the change to work.
  • 🏃 Click 'Run' to execute the change.
  • ♻️ Refresh the page to confirm the switch back to GPT-4.
  • 👤 Credit for this workaround goes to a Reddit user, kinky Cusco.

Q & A

  • What happens when you exceed the 25 message limit in a 3-hour period with GPT-4?

    -When the 25 message limit is exceeded within a 3-hour period, GPT-4 will automatically switch to GPT-3.5 for the user.

  • How can you tell if the switch from GPT-4 to GPT-3.5 has occurred?

    -The switch can be identified by the change in the model version displayed in the chat interface, where it shows the older version, such as 'text DaVinci 2.0.2'.

  • What is the suggested workaround if you want to switch back to GPT-4 after being switched to GPT-3.5?

    -The suggested workaround is to wait for another two hours for the quota to become available again before starting a new conversation with GPT-4.

  • Why might someone want to switch back to GPT-4 after being switched to GPT-3.5?

    -One might want to switch back to GPT-4 to continue a long conversation history and to maintain the trained communication patterns established with GPT-4.

  • What is the recommended method to switch back to GPT-4 as described in the script?

    -The recommended method involves using the Firefox browser, logging into ChatGPT, opening the conversation where the issue occurred, and then using the inspect element feature to modify the model version in the network tab to 'gpt/4'.

  • How does the Reddit user kinky Cusco contribute to the solution?

    -Kinky Cusco from Reddit provided the steps to switch back to GPT-4 by manipulating the model version in the browser's developer console.

  • What should you do after changing the model version in the browser's console?

    -After changing the model version, you should click the 'Run' button in the console to apply the changes, then close the console and refresh the page.

  • What are the potential consequences of not being able to switch back to GPT-4 manually?

    -If unable to switch back manually, users would have to start a new chat, which would result in losing the entire conversation history with GPT-4.

  • Why is it important to keep the backslash in the model version when changing it in the console?

    -The backslash is important as it is part of the model version identifier. Deleting it would result in an incorrect model version, and the switch back to GPT-4 would not work.

  • How can you verify that you have successfully switched back to GPT-4?

    -You can verify the successful switch by checking the icon and observing the response speed and pattern of GPT-4 in the chat.

  • What is the main takeaway from the video script for users who encounter the same issue?

    -The main takeaway is the method to switch back to GPT-4 manually by using the browser's developer console to change the model version, as detailed by the Reddit user kinky Cusco.

Outlines

00:00

🤖 GPT Version Switching Issue and Solution

The paragraph discusses a problem encountered when using Chat GPT 4, where exceeding a limit of 25 messages within a two-hour period results in an automatic switch to GPT 3.5. The speaker describes a situation where a long conversation history with GPT is lost when switched back to GPT 3.5 due to the message limit. They then explain a method to manually switch back to GPT 4 using the Firefox browser, involving steps like logging into chatgpt, using the inspect feature, and modifying a model parameter in the network console. The solution is credited to a user named Kinky Cusco from Reddit.

05:01

🔄 Returning to GPT 4 After Forced Downgrade

This paragraph continues the discussion on the GPT version downgrade issue, focusing on a solution to revert back to GPT 4 after being switched to GPT 3.5. The speaker demonstrates a process involving the Firefox browser, inspecting the conversation element, and modifying the model parameter in the network console to change it from GPT 3.5 back to GPT 4. They emphasize the importance of keeping the backslash in the parameter and running the modified code to switch versions. After applying the fix, the speaker refreshes the page and confirms that they are back in GPT 4, encouraging viewers to follow along for more AI-related content.

Mindmap

Keywords

💡GPT-3.5

GPT-3.5 is an artificial intelligence model developed by OpenAI, which is a predecessor to the more advanced GPT-4. In the context of the video, GPT-3.5 is mentioned as the version users are switched to when they exceed the 25-message limit within a 3-hour period while using GPT-4. It represents an older technology compared to GPT-4 and has certain limitations in terms of functionality and conversational capabilities.

💡ChatGPT-4

ChatGPT-4 is an advanced AI chatbot model that is capable of more complex and nuanced conversations compared to its predecessor, GPT-3.5. The video discusses a scenario where users who have reached the message limit with GPT-4 are switched back to GPT-3.5, and it provides a method to switch back to GPT-4 to continue the conversation without losing the context and history.

💡message limit

The message limit refers to the maximum number of messages a user can send to the AI within a specified time frame. In this case, the limit is 25 messages within 3 hours. Once this limit is reached, the system automatically switches the user from GPT-4 to GPT-3.5, which can be inconvenient if the user has an ongoing conversation with GPT-4.

💡conversation history

Conversation history refers to the record of all previous interactions or messages exchanged between a user and the AI. In the video, maintaining the conversation history is important because it provides context for the AI to understand and respond appropriately to the user's queries. Losing this history could disrupt the flow of the conversation and require the user to start over.

💡switch back

Switching back refers to the process of returning to a previously used version of a system or service. In this context, it means reverting from GPT-3.5 back to GPT-4 after reaching the message limit. The video provides a method to manually switch back to GPT-4, allowing users to continue their conversation without interruption.

💡Firefox

Firefox is a free and open-source web browser developed by the Mozilla Corporation. In the video, Firefox is used as the browser to access the chatbot service and perform the steps necessary to switch back to GPT-4. It is mentioned as a requirement for the method to switch AI models.

💡inspect

In the context of web browsers, 'inspect' refers to the function that allows users to view and modify the underlying HTML and JavaScript code of a webpage. In the video, the 'inspect' feature is used to access the developer tools in Firefox, which is a crucial step in the process of switching back to GPT-4.

💡network

In the context of web development and browser functionality, 'network' refers to the monitoring of data being sent and received by the browser. In the video, by clicking on 'network', the user can observe the network activity related to the chatbot conversation, which is necessary for identifying the model being used and changing it back to GPT-4.

💡model

In the context of AI and machine learning, a 'model' refers to the algorithmic representation of a system or process that is trained to make predictions or decisions. The video discusses changing the model from GPT-3.5 to GPT-4 by altering the model parameter in the network request.

💡Reddit

Reddit is a social news aggregation, web content rating, and discussion website. In the video, a user named 'kinky Cusco' from Reddit is credited with providing the method to switch back to GPT-4. This highlights the collaborative nature of online communities in solving technical issues and sharing knowledge.

💡coding

Coding refers to the process of creating computer programs by writing and manipulating source code. Although the video script mentions that the speaker has no idea about codes, the process of switching back to GPT-4 involves a basic understanding of web development tools and the structure of network requests, which is a part of coding.

Highlights

Switching from GPT-3.5 back to ChatGPT-4 after reaching the message limit is possible with a workaround.

The default limit is 25 messages within a 3-hour period before switching to GPT-3.5.

Waiting for another two hours can reset the limit, but this may disrupt long conversations.

A manual switch back to GPT-4 is not straightforward once the conversation has started.

Starting a new chat would result in losing the entire conversation history.

The solution involves using the Firefox browser and its developer tools.

After waiting for the quota to reset, open the conversation and inspect the elements.

Utilize the 'Network' tab to identify the conversation's data packet.

Modify the model version in the data packet from GPT-3.5 to GPT-4.

Ensure the backslash is included when changing the model version in the code.

Run the modified code to switch back to GPT-4 within the same conversation thread.

The method was discovered by a user named Kinky Cusco from Reddit.

This workaround preserves the conversation history while upgrading back to GPT-4.

The process is user-friendly and does not require extensive coding knowledge.

The video provides a step-by-step guide on how to execute the workaround.

The success of the method is demonstrated by the return to GPT-4 and the continuation of the conversation.

This solution is particularly useful for those who have trained GPT-4 for specific communication needs.

The video encourages viewers to follow along for more AI-related content.

Turn Your Ideas to Images for Free

Generate professional visuals with text or images in seconds.

Turn Your Ideas to Images for Free Try It Free