My Journey with AI Tools: Practical Tips from a Recent Discussion
I recently had an enlightening conversation with Nohra about my evolving journey with AI tools, particularly focusing on how I use them in my daily work. In my previous post about AI usage, I shared my initial skepticism and outlined how I approach using AI as an assistant. This follow-up post shares what I've learned along the way and what's working for me.
With many tech professionals currently facing career transitions, I wanted to share these experiences in case they might help others who are exploring AI tools in their job search. I've found some approaches that work well for me in crafting resumes, preparing applications, and developing content - while maintaining authenticity and honesty.
Starting with the Basics: Key Tips for Working with AI
Let's begin with some fundamental tips that can dramatically improve your AI interactions:
The "Perfect Prompt" Myth
First, I need to address something important: there's no such thing as a "perfect prompt." This belief that there's some magical combination of words that will make AI produce exactly what you want is not only wrong – it's potentially harmful. Instead of searching for perfect prompts, focus on building a conversation and providing clear context.
Debug Through Fresh Starts
When AI starts producing unexpected or wrong results, many people try to correct it within the same conversation. I've found this rarely works well. Instead, it's more effective to:
- Go back to your original prompt
- Edit the prompt itself to be clearer
- Start fresh with the corrected prompt
This approach works better because it prevents the AI from getting stuck in circular reasoning or building upon incorrect assumptions.
Context is King
The key to getting good results isn't about finding perfect prompts – it's about providing good context. I've found that treating AI like a "dumb assistant" (as I sometimes call it) helps me communicate more effectively. To be clear - I don't believe AI actually understands or thinks. It's just a system responding to inputs. But speaking to it naturally helps ME organize my thoughts and communicate more clearly. I explain what I want and provide necessary background information in a conversational way.
The Power of Project Knowledge
One of the most significant improvements in my AI usage came from discovering Claude's projects feature. This tool allows you to:
- Build a knowledge base for the AI
- Maintain consistent context across conversations
- Create and refine frameworks for specific tasks
What makes this particularly interesting is how closely it mirrors actual AI product development. When you're using project knowledge, you're essentially creating a small-scale version of what happens in professional AI development with Retrieval Augmented Generation (RAG):
- Your documents act as context, similar to how a vector database provides context in RAG systems
- Your instructions shape how the AI processes and uses this context
- Your frameworks evolve through real usage, helping you understand what works and what doesn't
Building Better AI Interactions
Through this approach, I've developed several practical methods:
Using AI to Create Frameworks
The idea of frameworks came from a practical need - I was tired of repeating the same instructions and having the same corrections in every conversation. Instead of starting from scratch each time, I let AI help create frameworks based on our successful interactions. What I call a "framework" is really a formalization of my conversation patterns with the AI - what I consistently ask it to do or not do, how I want it to approach tasks, and what context to consider.
For example, after tailoring several resumes and repeatedly telling the AI "don't embellish my experience" or "don't invent technologies I haven't used," I asked it to analyze our interactions and create a framework that captured all these recurring requirements and preferences. This framework, combined with the existing context, then became part of the project knowledge, making future attempts more efficient and consistent. I no longer had to repeat the same corrections - the framework already contained my preferences.
Tone Guides and Context Building
Similarly, I use AI to analyze successful interactions and create tone guides. This helps ensure consistency across different types of content while maintaining my authentic voice. The beauty of this approach is that the tone guide evolves naturally from actual conversations rather than arbitrary rules.
Strategic Tool Selection
I've found that different AI tools have different strengths:
- ChatGPT excels at research and analysis
- Claude shines in writing and framework development
Understanding these differences helps me choose the right tool for each task.
Moving Forward
This is just a glimpse into what we discussed in the video. I've shared the approaches and concepts that have worked well for me - your mileage may vary, and you might find different methods that work better for your needs. If you're currently navigating a job search or career transition, I hope these experiences give you some ideas to explore while staying true to your authentic experience.
For more context about how I think about AI usage, check out my previous post: AI as an Assistant: A Practical Approach to Responsible AI Usage. The full video conversation goes even deeper into specific examples and how I apply these ideas.
I encourage you to watch the complete video for more details. More importantly, I'd love to hear about your experiences and questions. How are you using AI tools in your job search or daily work? What approaches have you found helpful?
Let's continue this conversation. Feel free to reach out with questions or share your own experiences – I'm always interested in learning how others are navigating this space.
You can watch the full video discussion here: Using AI Tools: A Practical Discussion
Add new comment