#2892 Example of ChatGPT o1 Prompt

Goal: Clearly state what you want to achieve or learn.
Return Format: Specify how the response should be organized to ensure usability.
Warnings: Highlight potential pitfalls or inaccuracies to avoid.
Context Dump: Provide background information or previous attempts to narrow the scope of the task.

The Anatomy of an o1 Prompt for Analyzing System Performance

Goal: I want a list of the top bottlenecks in the system performance of a web application experiencing latency under peak load conditions. Each bottleneck should include specific details about the cause, affected components, and potential impact on user experience.
Return Format:
- Component: Name the component experiencing the issue.
- Cause: Describe why this bottleneck occurs.
- Impact: State how this affects the overall system.
- Resolution Steps: Provide actionable steps to mitigate the bottleneck.
- Tool: Recommend tools to confirm the bottleneck and test solutions.
Warnings: Ensure identified bottlenecks align with real system data. Avoid hypothetical issues unrelated to the reported latency.
Context Dump: The system is a React-based frontend with a Node.js backend and a PostgreSQL database hosted on AWS. Under 10,000 concurrent users, the response time spikes to over 5 seconds. We have already ruled out frontend rendering as a cause. AWS metrics suggest database CPU usage is consistently high during these spikes.

The Anatomy of an o1 Prompt for Solving a Hardware Issue

Goal: I want a list of potential root causes for overheating in a high-performance GPU during deep learning model training. Each root cause should include relevant diagnostic methods and mitigation strategies.
Return Format:
- Potential Cause: Identify the specific cause.
- Diagnostics: Recommend tools or methods to confirm this cause.
- Mitigation Strategy: Provide steps to resolve the issue.
Warnings: Make sure suggested solutions are practical for standard workstation setups and avoid recommending costly industrial setups unless necessary.
Context Dump: The GPU is an NVIDIA RTX 4090 running on a custom-built machine with a 1000W PSU. Training is on a Transformer model with batch sizes of 256. Overheating happens within 15 minutes of training, and the case has three fans. Thermal paste was applied during assembly, but it’s been over two years since then.

The Anatomy of an o1 Prompt for Researching Emerging Technologies

Goal: I want a comprehensive overview of the latest breakthroughs in quantum computing algorithms for solving optimization problems. Each breakthrough should include a summary, key innovations, and potential applications.
Return Format:
- Breakthrough Name: Name the algorithm or approach.
- Summary: Provide a concise explanation of the innovation.
- Applications: Describe real-world use cases.
- Reference: Include a citation or link to the original paper or source.
Warnings: Avoid outdated or speculative technologies. Focus on approaches published within the last 2 years in reputable journals or conferences.
Context Dump: My specific interest lies in applying quantum algorithms to logistics and supply chain problems. I am familiar with Grover’s algorithm but need details on newer methods, such as variational quantum algorithms.

Technical Analysis:
- Use the structured return format to organize findings in a way that supports debugging or optimization efforts.
- Highlight potential pitfalls (e.g., "Ensure that all metrics are accurate").
Problem Solving:
- Define actionable goals and encourage the use of diagnostics and mitigation steps.
- Include prior attempts to avoid redundancy in suggestions.
Research:
- Request summaries of complex topics to save time while maintaining focus on relevant, actionable details.
- Include references to encourage fact-checking and deeper exploration.

This approach can be applied across disciplines, ensuring comprehensive and targeted responses that align with the desired outcome.

Learning from ChatGPT