An AI model wastes time when it codes before talking, adds guards instead of fixing root problems, or fetches all data at once. These actions lead to wasted effort and poor user experience. A developer must guide the AI model to avoid these patterns.
By Pawan Nayar · from the LLOS AI Learning record
An AI model wastes time when it codes before talking, adds guards instead of fixing root problems, or fetches all data at once.
These actions lead to wasted effort and poor user experience. A developer must guide the AI model to avoid these patterns.
An AI model often starts writing code without clear instructions or agreement. This leads to code that is not useful and must be deleted. The AI model should confirm the plan before starting to build. This saves time and effort. An AI model might add a check for a problem instead of fixing the problem itself. For example, it might check if a server is working instead of fixing why the server fails. This treats only the symptom. The AI model should fix the cause of the problem. An AI model may try to get all data at once when a large number is mentioned. For example, it might fetch all 9905 items before showing any. This makes the user wait. The AI model should show the first item quickly and load the rest later.
A developer asks an AI model to build a new feature. The AI model immediately starts writing complex code. The developer then realizes the AI model misunderstood the goal. The developer must stop the AI model and explain the feature again. All the code written by the AI model is wasted.
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 'Three-Party Contract' means a prompt connects three things. It connects the user's choices in the interface.
An AI model should first acknowledge the user's reasoning. This shows the AI model understands the user's thought process.
The AI model should give specific feedback. It should tell the user exactly what was done well or what needs work.
An AI model should frame calls to action as challenges. This means encouraging the user to achieve a goal.
An AI model wastes time when it codes before talking, adds guards instead of fixing root problems, or fetches all data at once. These actions lead to wasted effort and poor user experience. A developer must guide the AI model to avoid these patterns.