Currently, if you connect to typingmind.com on a non-firewalled network and reconnect to a firewalled network, the web UI is still available, but no chats can be made due to AI/ML blocks.
Firewalls that block AI/ML applications at the network level (like ZScaler) do not allow REST requests to AI/ML API endpoints. With ChatGPT, you can circumvent this block by:
Go to ChatGPT.com on a network that does not have the described firewall and leave the tab open.
Reconnect to firewalled network.
What I believe is happening in the backend is that ChatGPT uses a persistent open socket connection once it's opened. This circumvents the firewall, since there are no new requests being made to the ChatGPT.com domain or its endpoints.
The above is a hypothesis, but it would be great if Typing Mind could do something similar. Currently, if you connect to typingmind.com on a non-firewalled network and connect to a firewalled network, the web UI is still available, but no chats can be made.
Please authenticate to join the conversation.
Open
Feature Request
Chat Management/Interactions
Over 1 year ago
Get notified by email when there are changes.
Open
Feature Request
Chat Management/Interactions
Over 1 year ago
Get notified by email when there are changes.