My AI Use Policy for Copy and Content

In the spirit of transparency and honesty, you need to know my AI Use Policy.

Some clients ask if I use AI tools when writing sales and marketing copy and content. And if so, how?

(A few ask outright. For others, it’s a nagging question in the back of their minds.)

So, let’s lay all our cards on the table, shall we?

Here’s my personal AI use policy.

When I use an AI, I’m the HITL (Human In The Loop). Artificial Intelligence isn’t necessarily a bad thing… if used responsibly, guided by a human being.

And honestly, it’s a good idea to learn and lean into the technology.

(I’ve spent time and money learning the ins and outs of AI use.)

But AI can’t be trusted on it’s own. It can and does make “mistakes.”

No, it doesn’t outright lie to you. It’s not that smart… so I’m told.

But it’s not dependable enough to have the final say in anything copy or content related.

If you think I’m being harsh, don’t trust my judgment.

Ask the AI! (I know, I know… LOL)

It will tell you the same thing.

I asked ChatGPT-4 if everything AI outputs is trustworthy. Here was the response from the algo-naut itself (the emphases are mine):

1. Training Limitations: AI models are trained on vast datasets containing human language data, which can include inaccuracies, biases, and outdated information. The model’s responses are predictions based on patterns it has seen during training, not guaranteed truths.
2. Context Understanding: While AI can handle a wide range of topics, its understanding of context and nuance can be limited. It might not fully grasp complex situations or the subtleties of human emotions and cultural contexts.
3. Data Currency: AI models are not connected to the internet for real-time updates (beyond specific tools that access the internet). They rely on knowledge up to the last update, which in my case is December 2023. Therefore, the model might lack information on recent developments or changes.
4. Error Potential: AI can make mistakes, such as misinterpreting the user’s questions or generating incorrect or irrelevant responses. Users should critically evaluate AI-generated content, especially in professional or critical contexts.
5. Bias and Fairness: AI models can inadvertently perpetuate biases present in their training data. This can affect the fairness and impartiality of the responses. (Garbage in, garbage out… simple computer reality)

Perplexity.ai, another bot I use primarily for investigation and research, was similar in its response. (And even cited its sources, which is why it makes a great research tool… fact checking made simple. Again, some of the emphases are mine.):

No, not everything that AI outputs can be considered trustworthy. The search results highlight several key issues with the accuracy and reliability of AI-generated information:

 AI systems like ChatGPT and Google Bard have been known to generate inaccurate or made-up information, a phenomenon referred to as “hallucinations”. These AI models can base their analyses and creations on inaccurate or fraudulent information, leading to unreliable outputs.

 The (cited) article notes that while AI has progressed significantly, it still largely lacks the empathy, ethics, and bigger-picture understanding that goes into human decision-making. AI is based on algorithms that respond to models and data, often missing important intangible factors.

 The (cited) article emphasizes that the biggest mistake is assuming content created by AI is free from faults. It advises always verifying the source, facts, and claims, even if the output appears trustworthy and convincing. Generative AI tools can produce hallucinations that appear legitimate but are inaccurate.

AI is missing one important element… EI

What’s missing in most, really, all responses generated by AI is the element of EI, or emotional intelligence. And even the AI bots understand this limitation. According to ChatGPT:

AI doesn’t possess emotional intelligence in the human sense. It doesn’t experience emotions, empathy, or self-awareness. Instead, what might appear as “emotional intelligence” in AI responses is actually the result of complex algorithms processing vast amounts of data, including text that reflects human emotional expressions.

It goes on to explain its process, which includes Pattern Recognition, Response Generation, and Tone and Style Adaptation. And then it finishes with this statement:

While these capabilities can mimic aspects of emotional intelligence, it’s crucial to remember that AI’s “empathy” is based on statistical correlations and learned representations of text, not genuine emotional understanding. The AI doesn’t “feel” emotions; it merely replicates the expression of them based on its training.

So there you have it, straight from the droid’s mouth, not mine.

AI doesn’t “feel” emotion.

I do.

I can explain the tasty sensation of a freezing, sweet mint-chocolaty-chip ice cream as it melts across my tongue and slides down my gullet on a hot summer day in great emotional detail.

I’ve eaten one.

(Okay, maybe more than one. Or two.)

But stick an ice cream cone in an AI’s mouth, and it would literally fry its electronic brain circuits!

Poof.

What I do now (and did before AI was even a thing)

I specialize in writing copy that is original and effective. To accomplish that, I maintain the role of primary writer — doing the important thinking, strategizing, and persuasive writing myself — especially when composing an initial first full draft.

I might let my little robuddy do the UFD (ugly first draft)… if I’m stuck win writer’s block land. Maybe just the first paragraph or two.

But never the official full first draft.

Nor the final. Ever.

That’s the main point of having an AI Use Policy. To let you know that I’m still the scribe in charge.

Always.

Not only am I the Human In The Loop, but I’m also the only one in the room who possesses the innate and crucial quality of emotional intelligence.

To put it philosophically… if you cut me, do I not bleed?

And cry.

I do.

(But never cuss… well… rarely ever.)

So again, I specialize in writing copy that is original and effective and real. And that takes a human touch, from first full draft to final product.

If you were to “dust” my finished work, you would find my human fingerprints all over the scene of the copy. I do, however, use AI as a kind of “writer’s assistant.” For example, I might get an AI to:

• Transcribe an interview.
• Summarize an article, web page, or video.
• Research a topic and find reputable sources.
• Provide insights into a target market; particularly what motivates them to buy.
• Generate ideas for creative concepts or content topics. (AI can be a great brainstorming buddy.)
• Come up with fresh examples, scenarios, stories, analogies, metaphors, turns-of-phrases, etc. to use in the copy.
• Review a draft and recommend edits and improvements.
• Proofread for typos, grammar, and other errors.
• Play the role of your customer and tell me how persuasive the copy is from that POV. (The next best thing to testing the copy live!)
• Make my coffee and pour me a cup. (Well… still workin’ on that one. Getting close.)

My AI Use Policy: Protecting Your Trade Secrets

Here’s another important part of my AI use policy. One that’s designed to protect your information.

If you provide me with information that is a trade secret or otherwise confidential, I will never input that information into an AI without your approval.

We can put that in writing if necessary. But that’s my AI Use policy, regardless.

(Just let me know it’s a trade secret, okay? Don’t make me guess. I’m no mind reader.)

And if you don’t want me to use AI at all for your project, say so! We can put that in writing as well. But I’m able to spend more human-powered QUALITY time on your project by using AI for the grunt work. Sounds counterintuitive, I know.

But that’s a fact, Jack.

At the end of the day, it’s your copy you’re paying for, not mine. It needs to be done to your expectations and specifications. And regardless, it will always be written from the human perspective… mine, not AI’s.

Bottom line: Regardless of how I use AI tools, my work for you will always be original and leverage my 14+ years of experience writing engaging, high-response copy. I guarantee it.

Hey, I even put it in writing in this AI Use Policy so you could see it. The majority of copywriters don’t do that, especially the amateurs and hobbyists. Maybe they’re afraid of full transparency in their work.

But it’s what I do…

It’s what you deserve.

Shazbot!

Steve Maurer, IME

Contact me today

P.S. (They always read the P.S.)

AI-bots I have known (and use):

• Jasper (for campaign creation)
• ChatGPT-4 (for quick outlines, brainstorming, and easy cake recipes)
• Perplexity (for research when citation and verification is vital… and it always is)
• Claude 3 (not too shabby to converse with and level up your skills with a little back and forth)

And these are not the free versions. I pay monthly and yearly for these tools.

I’ve got skin in the game.

Just like you.

Contact me today