How Crimson Scribbles Uses AI

Why This Page Exists

Crimson Scribbles uses AI responsibly and effectively. My clients deserve to know how I use it, where it fits into my work, and where it doesn't.

Core Beliefs

Human judgment is irreplaceable

I bring the experience, knowledge, and professional judgment. AI supports my work; it does not replace it.

AI is a tool

AI can take on substantial research, organizational, analytical, and support tasks when those tasks are appropriate to delegate.

Responsibility Cannot Be Delegated

I am responsible for the work, which includes:

  • What I choose to delegate

  • How I direct the work

  • What comes back

  • What I accept

  • What ultimately represents Crimson Scribbles

Better questions produce better work

I use AI to challenge ideas, explore alternatives, identify problems, and make recommendations.

AI in Practice

Grant Development — Direct Client Benefit

What Crimson Scribbles does:

Crimson Scribbles learns about the client—their organization, goals, vision, and voice—and uses that understanding to develop the grant strategy and narrative. I work through the funder's requirements, determine what the proposal needs to accomplish, identify the story we need to tell, and decide how the client's information supports that story.

How AI supports that work:

AI can help organize funder requirements, develop a framework, work through client-supplied information, identify gaps, explore questions, and develop or refine narrative ideas.

Research & Analysis — Indirect Client Benefit

By way of example: Crimson Scribbles gave AI an industry to research. I provided the data I wanted collected, the instructions it needed to follow, and how I wanted the information organized into a spreadsheet.

AI then worked through those requirements to gather and organize the data. Once the information was collected, I started asking questions. AI sifted through the data to help answer them, allowing me to explore the information from different angles without having to manually work through every piece of it myself.

This is work I could have done myself. In a traditional team environment, I might have divided those tasks among several people. Instead, AI handled much of the research and organization while I directed the process, asked the questions, and worked with the results. What might have taken me weeks was accomplished in a few hours, giving me more time to focus on the analysis and decisions that require my experience.

Information Filtering — Managing My Attention

I receive an enormous amount of current information and news about AI and technology. The problem wasn't finding information—it was sorting through it.

AI helps identify what's new, eliminate duplicate stories, and surface developments that are actually relevant to Crimson Scribbles and the work I do. And the best part? I have it delivered to me twice a week. Not daily or hourly.

The benefit isn't simply saving time. It's being able to spend my attention where it matters.

How We Don't Use AI

AI can produce an accurate, grammatically sound proposal. That doesn't necessarily make it impactful.

Where is the client's voice, the story, the strategy, and the human understanding that make the proposal compelling?

AI-generated content has to earn its way into Crimson Scribbles work.

What This Means for Clients

  •  I take the time to understand your organization, your goals, and the story behind the work.

  • I don't hand your information to AI and simply accept what comes back.

  • AI may help me work more efficiently and explore more possibilities.

  • You still work with me.

  • The work still reflects your organization and your voice.

  • I remain responsible for what I deliver.

Protecting Your Information

Protecting client information is part of using AI responsibly. I take care with the information I provide to AI tools and, whenever possible, remove unnecessary identifying details before using it.

I also consider whether information is confidential, publicly available, or necessary for the task before including it in an AI-assisted workflow.

I have turned off "Improve the model for everyone" in ChatGPT because protecting client information was an important consideration in how I use the tool. With that setting disabled, new conversations are not used to improve OpenAI's models.

These practices work alongside the broader privacy and security measures described in the Crimson Scribbles Privacy Policy.