Chris Young
If artificial intelligence were to write the introductory paragraph for this article, it might sound a little like this, “In the digital age, where creativity meets technology, artificial intelligence (AI) is taking the world of content creation by storm! Gone are the days of tedious brainstorming and endless drafts…”
Okay, ChatGPT… we’re good.
When ChatGPT 3.5 first came out, it was astonishing to many in the marketing industry, myself included. It was the first LLM that could actually pass as artificially intelligent. After playing around with it for a few months, the limitations of this technology quickly became evident. The outputs were passably human. However, many shortcomings became evident–not least of which was the blatantly wrong information, and the tiresome use of hyperbole. In fact, many of the outputs I read make me believe that AI uses hyperbole for the same reason that unintelligent people talk loudly–what they lack in quality, they make up for in volume, or in AI’s case, exaggeration; it’s a coping mechanism. When you can’t do something well, you do it poorly and loudly while hoping that there’s enough gas in the tank to get you across the finish line. If you’re not fooled by LLM outputs, congratulations–you’re smarter than an AI, or at least GPT 3.5.
In the AI-world, everything is crucial, of dire importance, or nearly life or death–even pancake recipes and made-up travel itineraries for that trip you probably won’t be taking this year. But does that mean that AI has no place in content creation?
Almost certainly not.
While I have been critical of AI (and not just because my career as a writer depends on my successful deprecation of AI), there are sundry uses for AI in content creation. Let’s also remember that content creation, while largely dependent on the written word, also has visual, audio, and other experiential components where AI might be able to help.
AI for written content creation is by far the most notable application of AI to content. The first manifestation that most of the public was exposed to was a large language model (LLM)--namely ChatGPT 3.5–whose primary application was the written word. Earlier versions of ChatGPT were well-known by those who study artificial intelligence and neural networks. However, Average Joe Marketer was likely not in that first group of scholars. In addition to ChatGPT, Claude has also become another popular LLM that, in my experience, sounds a bit more measured (dare I say, human) than GPT does.
Other popular SaaS companies such as Twitter (I refuse to call it X) and Google debuted their own chatbot-style AI platforms: Grok and Gemini, respectively. When it comes to content creation, Grok may have limited utility, acting mainly as an AI to aid in scientific discovery. Gemini, on the other hand, appears to be a legitimate competitor to other LLMs.
Content creation is not as simple as sitting down at a keyboard and starting to type things out. There are constituent parts to the entire process–brainstorming, research, writing, and editing. While I regard all LLMs as inadequate to create compelling thought leadership content, there are a few areas where they can help move the process along.
First, and perhaps most importantly, LLMs can accelerate the research process. One of the hardest parts about researching for content are the “unknown unknowns”--those things that you don’t know that you don’t know. These often appear at the very beginning of the ideation stage of content creation and leave the creator wondering, “Alright, where do I even start?”
AI can excel at breaking through this by acting almost as a conversational search engine. It’s very easy to start with a prompt such as, “Okay, chatGPT, I’m writing an article that’s generally about [obscure topic here]. What are some interesting takes or bleeding-edge ideas within this space?”
From the AI output, find something that interests you or something that’s especially relevant to the project scope, and simply prompt AI to expound with “Say more about this.” Repeat this same prompt as often as necessary, and you’ll soon find yourself falling down a rabbit hole of ideas (in a good way).
As a writer, one of the most challenging parts of creating written content is coming up with a conceptual outline that coalesces around a single, central theme. The order in which I write the sections and how much emphasis I put in each section are determined by the central “argument” in the content I create. This process can frequently be as frustrating as it is engaging, especially when the piece could branch off in any number of directions or no perceivable direction at all. I recommend creating your own outline first, and then asking the AI to come up with its own outline. Areas where outlines overlap are probably good sections to include. Your mileage may vary, however, I frequently find that the suggestions are valuable and I end up incorporating at least some aspect of the AI outline into my own writing.
Finally, and perhaps most usefully, AI does an excellent job of restarting the engine when writer’s block inevitably shows up. Simply copy and paste the text of your WIP, and precede it with this prompt: “I am working on the following text shown below in quotation marks. I am stuck on the last sentence that appears and I need you to give me 5 ideas of how to move forward.” In some instances you may be able to use a sentence verbatim. In many other cases, the output will simply get the “juices flowing” enough to help you continue to the next main idea. Either way, it eliminates the frustration of staring at a stubborn paragraph for hours on end while the deadline steadily inches forward.
When creating written content, one of the biggest temptations is to use AI to create an entire piece of content. Some projects can be done mostly by AI. Other, simpler projects should be done mostly by AI, but there are some projects–particularly those meant to shape the opinions of others–that are only worthy of a human intellect with true creative capacity. There is an ever-present temptation to use AI for as much as possible. However, in a world of ease defined by DoorDash, self-driving cars, and GLP-1 agonists, don’t allow AI to make you lazy. Writing, at its most fundamental level, is the act of thinking made manifest. Using AI to automate your process of critical thinking can, if nothing else, diminish your capacity as a person.
For visual content creation, artificial intelligence plays somewhat of a similar role as the one played in written content creation. Programs like Midjourney and DALL-E are some of the most popular image generation platforms. In my opinion, the only use for this technology is in the ideation phase as the first step in moving from idea to words, and words to image. I have never encountered a satisfactory iteration of an AI-generated image that I would feel comfortable using in content marketing. I have seen businesses do it much to my chagrin; on a good day it might convince me for a few seconds. However, using AI to create client-facing or consumer-facing visuals is just lazy. It’s better to use AI to create a “rough draft” image that can be passed off to a real designer who can develop the idea.
Your use of AI should be guided by a firm knowledge of AI’s inherent limitations in content creation. Do you need someone to give you information and bounce ideas off of? Great, AI does an excellent job of that. However, for anything that requires a complex analysis of two or more subjects, or any novel opinion/point-of-view, it’s best to leave that to a human SME. AI can be used well to create the foundation for excellent content, and it can be used to create highly generic content to an adequate level. What it cannot do is create excellent content–the very thing most needed by marketers. As the market becomes more and more saturated with generic content that adds scant value, the need for human written content will grow.
AI democratizes low-value add content; where businesses used to have to pay for a human to create content (even low value content), now AI can do it and anyone can access AI. This means that a piece of poorly written content that used to take 2 - 4 hours to produce can now be done in a matter of minutes. When a tool like this is available to everyone, the incentives are aligned such that everyone will begin using AI for content creation. Why not save the time and money? This mentality is flawed because, paradoxically, cheap things usually end up being more expensive in the long term than the genuine article. When everyone is producing low value-add content, those business that decide to truly invest into content creation have an opportunity to stand out from the competition by simply continuing to hire human content creators–something they were already doing two years ago. Low quality videos won’t get views; low quality images won’t help with conversions; low quality SEO articles will continue to rank on page 188 (or thereabouts) of the Google SERPs.
To continue excelling in your content creation strategy, you don’t need to do anything different than what you were doing before AI came to market. In fact, an even easier way to continue producing great content is by working with a content creation agency like Be More Digital. Our content writers, video creators, and more, are real humans with real talent that can help your business fill the space left by bad content. To learn more about working with Be More Digital, get in touch with our customer success team today or take a look at one of our many successful case studies.
.png)
