The First Draft: Your Business’ Starting Point for AI-Generated Content
Generative AI remains all the rage in marketing departments across the global economy. And for good reasons – it saves time and saves money, while augmenting output. But truth be told, a lot of AI-generated content still isn’t up to par, especially early, unedited iterations. So just be careful. It’s not how you start, but how you finish that creates content worth publishing.
By Jared Frank | 9 min read
If your business is using AI to generate content, you might not be seeing the results you were promised. You can’t get rich quick with content, even AI-generated content. And when used properly, AI does not generate content. AI generates ideas.
Even the term “AI-generated content” is a misnomer. “AI-improved content” is a more accurate one. AI as a tool is immensely helpful for brainstorming on the front end of the content creation process and for proofreading on the back end. But outsourcing your content creation 100% to AI in the middle of that process will crater your performance. At the end of the day, the human touch is still, and will always be, what connects fans to your brand.
Humans want to talk to humans. And humans want to appreciate the art of being human. If users can smell the AI on content simply copied from a large language model (LLM) and pasted onto your website, it’s a hard pass. If users know all your graphics and other media are machine-made, they will become blind to them.
The machine simply isn’t intelligent enough yet (and won’t ever be) to replace the uniqueness of original human thought. The AI truth, as truth always is, is in the middle – the machine is meant to complement.
AI is your first draft, NOT your final draft.
Key Takeaways
OLD DOG, NEW TRICK: The term “artificial intelligence” was first used in the 1950s. It’s only Generative AI that is a new application of that technology. AI-generated content refers to any text, speech, image, audio, video, or other media that is created by a computer rather than a person.
PROBABILITY IS NOT CERTAINTY: When an AI engine “generates” content, it is really “predicting” content. AI responses to prompts are “best guesses” based on mathematical probabilities.
ONLY THE FIRST DRAFT: While AI can lift heavy loads, and fully automating your content strategy is tempting, it’s crucial to couple AI with human governance. AI cannot replace your creativity, judgment, and knowledge of brand standards.
The Definition of AI-Generated Content
First, a quick history lesson. AI is not a novel technology. “Artificial intelligence” is a broad term, first coined by John McCarthy in 1955. McCarthy’s hypotheses set in motion decades of research and resulting applications.
As presented by Tim Sharp in his Section School course, there are seven AI “capabilities.” Classic machine learning applications comprise the first six – Prediction, Classification, Recognition, Recommendation, Interaction, and Automation. Marketers know these six well through daily interactions with predictive analytics, lookalike audiences, social media/news feeds, programmatic advertising, and chatbots, while we buy a recommended product on Amazon based on previous purchases and listen to a suggested song on Spotify as determined by listening history.
Generation is the new seventh application of AI. AI-generated content refers to any type of text, speech, image, video, or other media that is created by a computer program rather than a person.
Generative AI has mushroomed the amount of content out there now in the digital sphere. But how much of this content is any good? None of it, if unedited, in this marketer’s opinion. More content does not mean better content. So just be careful to use AI-generated content responsibly.
AI machines do not think. They do not create. They guess. Large language models, for example, spit out one word at a time, crafting their responses based on the probability of the correct next word, using the data set of the entire internet to evaluate “correctness.”
Examples of AI-Generated Content
AI-generated content comes in many forms, including:
- Written Content: LLMs, such as ChatGPT, Perplexity, and Claude, generate articles, blog posts, social media posts, product descriptions, and even full-length books from simple instructions provided by the user. These tools are incredibly helpful… at creating first drafts.
- Digital Art and Design: Text-to-image tools, such as Adobe Firefly, Leonardo, and DALL-E, turn prompts into professional-grade visual content, eliminating the need to hire a full-time graphic designer.
- Music and Audio: AI can compose music by analyzing patterns from existing songs or genres. Platforms like OpenAI’s MuseNet allow users to create original music tracks based on different styles or moods, often used for videos or podcasts. Additional AI tools in market can perform audio engineering functions for spoken word recordings – an incredible use case for DIY podcasters.
- Videos and Animations: AI tools, such as Runway and OpenAI’s highly anticipated Sora, take a user’s text instructions or a still image to generate video illustrations.
- Speech: AI can even clone real voices or generate synthetic ones for use as voiceovers in marketing videos.
These examples showcase the versatility of AI-generated content across multiple creative fields, helping marketers save time and exercise cost containment when managing content projects.
How AI-Generated Content Works
The generative AI process starts with an input from a person, commonly referred to as a prompt. The AI then takes that input and, using its knowledge of language or design patterns, produces new content that matches the prompt. The old adage “garbage in, garbage out” applies here. The better the prompt, the better the content.
It’s important to note that the AI isn’t creating anything. When an AI engine “generates” content, it is really “predicting” content. AI responses to prompts are “best guesses” based on mathematical probabilities. Wow, read that sentence again. AI machines do not think. They do not create. They guess.
LLMs, for example, spit out one word at a time, crafting their responses based on the probability of the correct next word, using the data set of the entire internet to evaluate “correctness.”
“Probability does not equal certainty,” as Sharp warns in his Section School course.
Again, just tread carefully and responsibly.
When paired with human oversight, AI-generated content can be a powerful tool that enhances efficiency, creativity, and audience reach.
The Benefits of AI-Generated Content
Generative AI provides immense benefits to any content strategy and workflow, starting with saving time and boosting productivity. AI can:
- Create or summarize large bodies of text in the blink of an eye.
- Analyze large spreadsheets.
- Create high-quality photos, videos, and other media.
- Contextualize brand guidelines (tone, manner, style, etc.).
- Localize different languages, geographies, cultures, etc.
Additional benefits include scalable production with consistent output. AI doesn’t tire or make random mistakes the way people might when writing or designing for hours. It also kills writer’s block by jumpstarting the creative process with suggestions for topics, outlines, and first drafts. Additionally, AI helps optimize content for search engines. When writing, AI suggests keywords, ensures grammar rules are followed, and even structures your content for better readability. Overall, when paired with human oversight, AI-generated content can be a powerful tool that enhances efficiency, creativity, and audience reach.
Challenges Related to AI-Generated Content
With great power comes great responsibility. The use of AI-generated content comes with the responsibility to use it the right way. Challenges abound, including:
- Quality Concerns: AI content often feels generic and lacks the creative flair and personalization that a human might add.
- Lack of Originality: AI relies on existing, and sometimes outdated, data, which means it will not come up with truly original ideas.
- Hallucinations and Bias: Because it is governed by mathematical probabilities and not certainties, AI can sometimes get it wrong.
- Copyright Issues: AI models are often trained on content, including copyrighted material, without explicit permission from the creators. Another gray area is determining who owns the copyright of AI-generated content.
The current Generative AI mania is akin to a gold rush. Tools by the thousands have gone to market without a manual for how to use them or even a complete understanding for what all they can do. Many have come seeking fortunes that few will uncover.
Where Your Business Can Quickly Get Started with AI-Generated Content
Any new technology is daunting when first getting started. But remember, AI isn’t new, so there’s nothing to be afraid of. And you can, and should, start small and start free – you’re not going to “mess up.”
If you haven’t already, the best way to start is to simply hop on the free version of ChatGPT and start talking with the machine. Talk about anything. The point isn’t to publish content at this point. It’s only to get comfortable with the user interface, process, and capabilities.
Don’t try to write the next Great American Novel. Use AI to write one blog. Just one. See how it feels. See what you learn. And if the thought of writing gives you a panic attack, use the AI tool to write the first draft for you. The AI will not get everything perfect, but it can give you a helpful framework to build on. You can then add your own insights, refine the language, and make sure the content reflects your brand’s unique voice and personality.
If your business has an ecommerce site, a good starting point might be to use AI to quickly create your product descriptions. Again, it’s just a draft. Nothing has to get published if you hate it.
Look Beyond the Hype to See Beyond the Overwhelm
The current Generative AI mania is akin to a gold rush. Tools by the thousands have gone to market without a manual for how to use them or even a complete understanding for what all they can do. Many have come seeking fortunes that few will uncover.
We tend to overestimate what emerging technologies can do. Fat will be cut. Markets will be right-sized.
Until then, I encourage you to avoid being swept away in the current flush of exuberance. Don’t miss this opportunity to scale beyond your competitors, but take your time to do it right. “Be quick, but don’t hurry” as the legendary basketball coach John Wooden used to say.
And make sure you, the human, takes the time to write the final draft of your story. 
Are you ready to start creating AI-generated content? Let’s chat.
Write to Jared at [email protected].
Disclosure: ChatGPT helped ideate the first draft of this article. The author revised subsequent drafts and contributed original copy to better reflect the intended message and voice.
RELATED CONTENT
An Overview of the Best AI Marketing Tools for Your Small Business
Hundreds of AI marketing tools now exist to assist marketers with their craft. With the incredible volume of tools available, it can be overwhelming to shop for the right ones. Ditch the anxiety with this buyer’s guide of top tools.
Why Long-Form Content Remains a Pillar for Successful Marketing
Most marketers prioritize short-form content to reach customers amidst the noise in their inboxes and social feeds. But counterintuitively, the best content strategies set a foundation for success through long-form content.