On AI Assisted Content Creation - Keeping it Real

By G. Sawatzky, embedded-commerce.com

August 15, 2025

There seems to be plenty of negative opinions on using AI to assist with content editing including frequent, often sarcastic, posts about the use of em-dashes and other AI generated stylistic choices. This article is being written to address this subject and to share my perspective and personal approach to using AI for content generation.

If an individual uses a one sentence or high level prompt to request that an LLM service generate an article or research document and then publishes as their own with only a cursory review then I would agree that this lacks integrity and legitimacy; however, I do believe that there is an honest and legitimate way to use AI to assist with content generation.

Let’s consider the world before mass LLM adoption.

Resumes - How many professionals and executives use resume writing services that result in prose that is more in the style of a resume writing professional than that of the resume subject? Do resume writers not push the boundaries of embellishing the experiences and competencies of the resume subject?

Documents, Articles and Papers - How often are professionally published articles and papers entirely written by a single individual without any input from other individuals (i.e. multiple rounds of drafts, reviews and re-edits)? How often do they use other individuals to conduct or assist with research or leverage the research of others (with proper references and citations of course)? If they originate from organizations, publishing houses and well established professionals; one could assume they may benefit from the services of a professional editor.

The difference between what I describe above and the use of LLMs is that in the above cases, the details of one’s work history or the original drafts and ideas for articles and papers are, or were, written or provided with some original non-AI generated form by the individual. Furthermore, AI assisted research and content development is prone to mistakes and “freelance” content generation (i.e. hallucinations) by the AI. I believe that LLMs can be used for content development while maintaining integrity as long as we emulate this traditional approach, with an increased emphasis on thorough and rigorous reviews after any application of AI to the content.

In my own work, I have done my best to remain committed to this approach; however, as a single individual I do count on AI to be productive. I do not consider myself an academic or a professional researcher but as an individual practitioner that is passionate about certain subject matters and I stand by my published content as being faithful to my positions on those subject matters.

This article was 100% written by G. Sawatzky. AI was not used in the generation of this document's content (aside from final html generation and formatting).