My Approach to AI

I consider myself a cautious user of AI tools. I leverage various AI assistants to enhance my productivity and learning, but I maintain human oversight and ethical boundaries in how I use them.

Tools I Use

ChatGPT

Usage: Personal projects only
I primarily use ChatGPT through its Xcode plugin to assist with iOS app development. It’s been particularly helpful for troubleshooting build issues and implementing platform-specific features.

Claude (Anthropic)

Usage: Both work and personal projects
This is my most frequently used AI tool. I leverage it for:

  • Coding and scripting for various work tasks
  • Website development and maintenance
  • Home automation and media management scripts
  • Elaborating on and refining blog posts
  • General problem-solving across a wide range of topics

GitHub Copilot

Usage: Occasional
I use Copilot primarily for summarizing changes made to my website and Xcode projects hosted on GitHub.

Microsoft Copilot

Usage: Rare
When working with work-specific content, I occasionally use Microsoft Copilot as a comparison to Claude’s output.

How I Use AI

AI assists me with:

  • Writing assistance and editing - Refining blog posts and documentation
  • Code generation and debugging - Creating scripts and troubleshooting issues
  • Brainstorming ideas - Exploring different approaches to problems
  • Research and summarization - Processing information I’ve encountered
  • Image generation - Sparingly, not a common tool I reach for

My Principles and Boundaries

What I Do

  • Maintain human oversight of all AI-generated content
  • Use AI as a tool to enhance, not replace, my own thinking
  • Try to disclose when content is AI-assisted (though not always consistently)

What I Don’t Do

  • Never use AI with personally identifiable information (PII) - Given my role as an endpoint engineer, I have access to sensitive employee data. Using AI with this information is not only against my employer’s policy, but something I strongly oppose personally
  • Blindly accept AI output without review and verification
  • Use AI as a substitute for genuine learning and understanding

My Concerns

  • Training data and copyright - I believe entities that don’t want their intellectual property used in AI training must be honored
  • Environmental impact - I’m concerned about the energy consumption of AI systems and look forward to more efficient on-device solutions, though I recognize we’re not quite there yet
  • Transparency - While I strive to be open about AI usage, it’s not always practical to flag every instance

About This Website

Content on this site may have been created with AI assistance. Rather than explicitly marking every AI-influenced piece of content, consider this page a blanket disclosure: if you’re reading something I’ve written here, there’s a good chance AI tools helped me refine, debug, or develop it.

That said, the ideas, opinions, and decisions are mine. AI is my assistant, not my author.

This page was inspired by the /ai manifesto by Damola Morenikeji and the broader slash pages movement.