Your web browser is out of date. Update your browser for more security, speed and the best experience on this site.

Update your browser
CapTech Home Page

Articles February 17, 2023

ChatGPT and the Business Case for AI-Generated Content

Mike Colicchio
Author
Mike Colicchio

Over the past few months, ChatGPT and other generative Artificial Intelligence (AI) programs have been dominating headlines and breakroom discussions, with millions of people signing up for accounts—either diving in headfirst or cautiously dipping their toes in the water. Many businesses and creative professionals may find themselves on the fence, curious about the potential of these tools but concerned about the quality and accuracy of the content they deliver. Even with a basic understanding of the technology, capabilities are advancing so rapidly that it can be difficult to know where to start, how to leverage it effectively, or even whether to use it within your company in the first place. Taking some time to explore AI’s evolving role in content creation and how these new tools are being used may give us some insight.

How AI Can Help Generate Content

The content tools that are available today have come a long way since rudimentary spellcheckers and even more recent tools like Grammarly (introduced in 2009 but now more capable and powered by AI). Those tools use a form of Natural Language Processing (NLP) to look up words and usage patterns in a database to detect errors but have been limited by the storage and processing speed of computers.

Generative AI, specifically Language Learning Models (LLMs) like ChatGPT, goes far beyond what Clippie could do for writing—they’re more conversational, they can perform complex tasks, and they can analyze the context and meaning of words and phrases to generate what it predicts will be the most appropriate response.

To do this, AI models are trained on hundreds of gigabytes of text-based data, with sources from websites like Wikipedia, as well as digitized books, poetry, scholarly articles, and more. They’re trained to analyze and recognize patterns in data, making billions of connections between words, creating a sort of conceptual map based on existing content. To generate a response, a user enters a prompt, which can be a few keywords or several paragraphs of explicit instruction, and the tool rapidly produces text content similar to that traditionally produced by a competent human writer. If you’re not happy with a response, you can rephrase, adjust parameters, or add more context until the tool produces something that works.

While ChatGPT and its peers can generate surprisingly passable imitations of any human-generated content it has enough data to learn from—such as Shakespearean sonnets or Ryan Reynolds Mint Mobile ads—there are more practical examples:

  • Marketing content: For landing pages, email newsletters, pay-per-click ads
  • Web and SEO content: Headlines, meta descriptions, microcopy, blog posts, product summaries, FAQS, and even code (HTML, JSON, etc.)
  • Social media content: Status updates, video descriptions, post captions, hashtags
  • Email and SMS marketing content: Subject lines, body copy, calls to action

New Tools for Your Workflow

Many AI-based content-generating tools are now on the market. Some are free and open source, while others are paid, offering a range of pricing tiers and options depending on the features needed and the type or amount of content you’re looking to generate. Though not all are based on the same underlying model as ChatGPT, here are a few we have incorporated into our content workflow:

  • ChatGPT: Released publicly in November 2022, ChatGPT is the latest iteration of OpenAI's GPT-3 (Generative Pre-trained Transformer 3). It’s one of the most popular AI tools available today because of its ease of use, wide-ranging responses and writing styles, and its ability to assess context. It supports a variety of formats and languages, but with varied results, so just as you would for any editorial process, it’s important to review the content for accuracy.
  • Quillbot: Think of Quillbot as a Swiss Army Knife for writers. It has a co-writer feature that provides writing tips on vocabulary, generates summaries and citations, checks for grammar and plagiarism, and more. It works well for just about any writing application. Try using it in conjunction with ChatGPT or another AI-powered tool to see what works best.
  • Copy.ai: This tool also runs off OpenAI’s GPT-3 and boasts 90+ tools and content templates, a friendly interface, a long-form document editor, and tons of options to tweak and customize the output to your liking. It’s well-equipped to create starter content for several types of copywriting, including blog posts, ad copy, and marketing emails.
  • Grammarly: While it’s a popular tool for fixing typos and grammatical errors, it’s also great at making context-specific recommendations, like changes to tone or phrasing if your writing needs to sound more professional or more casual. Grammarly’s Style Guide feature also helps businesses create and maintain consistent content with a library of word preferences and brand-specific terminology.
  • Namelix: Having trouble coming up with an original name for your new brand, product, or service? Namelix uses a special language model to generate short, creative names based on the keywords you enter. It also features a logo generator and can improve its recommendations over time as it learns your preferences.
  • Surfer: Specifically geared toward supporting SEO efforts, Surfer uses machine learning (ML) and target keywords provided by a user to analyze high-ranking pages and provide on-page and technical SEO recommendations to increase site visibility and organic traffic. It also makes it easy to understand and act on recommendations by generating HTML, JSON, and other code snippets that content contributors can add directly to their CMS.

Advantages and Limitations of AI-Generated Content

Though each of these content-generating tools offers different features and targeted use cases, they share certain strengths and weaknesses of the current generation of AI-based tools:

Key Advantages


Speed: Because of their speed and wide-ranging capabilities, generative AI can accelerate the process of creating content, saving businesses and writers time and money. Much of the early information-gathering and research done on projects can be expedited, allowing teams to maximize available time and focus on higher-value questions and topics that go deeper than surface-level knowledge. They can also give writers and content teams that initial push to help overcome writer’s block by creating initial drafts or outlines that can be improved upon with unique insights, real-world examples, proprietary brand content, and customization.

Consistency: AI-powered writing tools can help maintain a consistent writing tone, style, and messaging. As noted above, Grammarly’s Style Guide can be loaded with a brand’s preferred vocabulary, terminology, and other content guidelines that are then enforced in the text editor, similar to how spell-check works. Such automated guidance can support new hires or content contributors less familiar with the company’s brand voice or style guide.

Efficiency: AI copywriting tools save time by automating repetitive writing needs that would otherwise need to be done manually. CarMax recently used Microsoft’s Azure OpenAI Service to generate vehicle summaries used by potential purchasers. What would have taken 11 years for a human writer was completed in a few months, allowing the editorial staff to focus on higher-priority content not currently possible using AI. Businesses can leverage AI-generated content and other tools to deliver projects with limited resources and aggressive timelines.


Potential Liabilities


Quality and Accuracy: While ChatGPT and other tools can generate increasingly sophisticated content, it won’t likely be as entertaining or engaging as content created by an experienced writer. One of the concerns about ChatGPT is how confident it sounds in its responses. The responses are declarative and sound authoritative, but sometimes they’re just plain wrong. It’s important to keep in mind that these tools are trained on existing data and are subject to any errors, bias, and even misinformation embedded in those data. ChatGPT responds to prompts with what it “thinks” is correct based on the patterns and correlations it identifies, so it’s still a matter of garbage in, garbage out. Bottom line: as with any copywriting, you should edit and review AI-generated copy to ensure it’s logical, interesting, factually correct, flows as intended, and tells an engaging story.

SEO: While there are many areas in which AI-powered technology can accelerate SEO efforts — keyword research, performance audits, and generating meta titles and descriptions, there’s also the potential for SEO spam and abuse. Some are waiting to see how Google and others reshape their policies around AI-generated content before investing in it. Google’s previous policy was vague but was recently updated, saying that it’s just the “spammy automatically generated content” that should be avoided. Businesses that focus on creating quality content and use AI to accelerate and augment those efforts will outperform those simply looking to use it as a crutch.

Ethical and legal implications: For all the buzz and excitement surrounding generative AI, there’s an equal amount of concern about issues like copyright infringement, intellectual property ownership, safety, and accountability. (OpenAI’s mission is to be a guide and ensure that Artificial General Intelligence, or AGI, “benefits all of humanity,” so at least they’re well-intentioned.) And while these tools may technically generate unique responses, advocates acknowledge that because the information and outputs are derived from existing information, they are often not truly original. For now, we’ll have to wait and see how recent lawsuits pan out and hope that detection tools for AI-generated content get much better.

So, What Does AI-Generated Content Mean for Organizations?

While AI-generated content tools may serve as a crutch for less-seasoned copywriters or researchers, experienced writers will likely lean in and use these tools as time-savers and idea starters, whether to quickly outline a piece of writing, research other ways to frame an argument, or simply to find alternative words and phrasing. 

The emerging consensus is that AI content generators work best as another tool in the writer’s toolbox — assistants to human copywriters, rather than replacements. That means that, depending on the specific needs of an organization or project, augmenting a content development team with an AI copywriting tool can save time and money, allowing the team to focus on more critical or high-value work.

Furthermore, businesses and content creators can differentiate themselves from competitors by learning how to write effective prompts or instructions to get higher-quality output. “Prompt design” is emerging as a highly-valued skill that can maximize the potential of generative AI tools. Anyone can now generate passable, “good enough” content, but those who become true operators and use these tools to creatively enhance their own content will see the most benefit.

The recent leaps and bounds made by generative AI certainly signal a new era. Like most emerging technologies, it will evolve over time, and today’s use case may give way to something completely unforeseen in the future. But at the start of 2023, tools like ChatGPT can serve as undeniable content accelerators for organizations of all shapes and sizes if leveraged judiciously.

Related Content