A client-side blocklist stops certain inputs before they reach an AI model. This saves computing costs. It also improves the user experience by giving immediate, gentle feedback.
By Pawan Nayar · from the LLOS AI Learning record
A client-side blocklist stops certain inputs before they reach an AI model.
This saves computing costs. It also improves the user experience by giving immediate, gentle feedback.
A hard block stops specific inputs from being sent to an AI model. It uses a list of forbidden words or names. This check happens on the user's computer, not on the server. This method prevents the AI model from processing sensitive or unwanted topics. It saves money because no request is sent to the AI model. The user gets instant feedback without waiting for a server response. The user experience is smooth. There are no harsh error messages. The input field might show a soft suggestion. A 'send' button might become inactive with a helpful message.
A developer creates an AI model for creative writing. The developer wants to avoid generating content about current political leaders. The developer adds a list of these leaders' names to a client-side blocklist. When a user types a blocked name, the input field shows a message like, 'Please try a general topic instead of a person's name.' The 'generate' button becomes gray. This prevents the AI model from receiving the blocked name.
Resolved from this lesson's own record — real definitions, not invented for this page.
More from the same record, pulled by shared topic — not written new for this page.
The 'sacred workflow' is a set of mandatory rules for an AI model. It starts with 'sacred input detection'.
A developer must treat mobile and desktop as two separate apps. They should use media queries to define distinct CSS layouts for each.
A card must never change its size. The content inside the card must fit the card.
An introductory screen should be completely removed from the active layout. This means using a method like `display: none` in web development.
A client-side blocklist stops certain inputs before they reach an AI model. This saves computing costs. It also improves the user experience by giving immediate, gentle feedback.