hey are gpt 4 can remember memory i gave in the it's voice chat?
#voice gpt and memories
agile mountain
lean steppeBOT
Hi @agile mountain. I'm kapa.ai, your friendly support bot. I will try to answer your question by searching through the knowledge base. :wave:
Based on the information provided in the knowledge sources, OpenAI's GPT models, including GPT-4, currently do not have the capability to remember information from previous interactions in voice chats or any other form of communication. The memory feature has been rolled out for ChatGPT Plus and Team users, but it's not yet available for GPT models.
The memory feature allows ChatGPT to remember details between chats, making it more helpful over time as it remembers details and preferences from your conversations. You can prompt it to remember something by saying "Remember ~" or "Record ~ in memory". However, this feature is not yet available for GPT models.
Please note that each instance of ChatGPT and GPT models have their own unique memories based on your interactions with them. Memory is not shared between different GPT models or between GPT models and ChatGPT.
For more information, you can refer to these articles:
- Memory FAQ (https://help.openai.com/en/articles/8590148-memory-faq)
- Does memory function with GPTs? (https://help.openai.com/en/articles/8983148-does-memory-function-with-gpts)
- Memory for ChatGPT Team - FAQ (https://help.openai.com/en/articles/9295112-memory-for-chatgpt-team-faq)
Note: you can ask me a follow up question by @ mentioning me again :speech_balloon:
agile mountain
@lean steppe so voice gpt is different? than chat gpt?
lean steppeBOT
Let me check the knowledge base for you. One moment! :books:
Yes, Voice GPT and Chat GPT are different in terms of their functionalities.
ChatGPT is primarily a text-based model that can generate human-like text based on the prompts given to it. It can be used within its own website or app, and it's operational environment is located in ChatGPT. It has a built-in user interface and the ability to share GPT with others. It also comes with built-in tools like Browsing, DALL·E, Code Interpreter, Retrieval, and Custom Actions. The pricing for ChatGPT is included in Plus/Team/Enterprise plans of OpenAI. The creation process for ChatGPT does not require any coding (source (https://help.openai.com/en/articles/8673914-gpts-vs-assistants#h_7569ffca53)).
On the other hand, Voice GPT is built on AI models that are available by the OpenAI API. It involves three main steps: transcriptions (convert speech audio recordings to plain text), language AI (chat with user), and speech (convert text to spoken audio). However, a web page doesn’t make for a good “alexa”, always waiting for commands. You can’t use advanced audio processing to detect wake words, speech, and silence. Thus you only see this in the app, or with a “record audio” button in the completions playground (source (https://community.openai.com/t/727720)).
Moreover, GPTs have their own voice option named, Shimmer, that is distinctly different from the 5 output voices available to use when having voice conversations with ChatGPT (source (https://help.openai.com/en/articles/8400625-voice-chat-faq#h_32cee17169)).
So, while both Voice GPT and Chat GPT are AI models developed by OpenAI, they are designed for different use cases and have different functionalities.
Note: you can ask me a follow up question by @ mentioning me again :speech_balloon:
agile mountain
@lean steppe how much character gpt can handle in single messagge?
lean steppeBOT
Let me check the knowledge base for you. One moment! :books:
The maximum character limit for a single message in GPT-4 is 4096 characters. This limit is mentioned in the OpenAI documentation and was confirmed by a user in a forum post who was initially confused about the source of an error they were encountering. They realized that the error was due to them exceeding this character limit when passing text from the GPT API into OpenAI's Text-to-Speech (TTS) API (source (https://community.openai.com/t/716252)).