Should you use AI generated text on your website?

And other questions about the value of AI-generated content for Small Business Owner’s Websites

I wish I had screenshotted a fascinating Thread where users weighed in on what words or phrases make them think the text was written by AI. The list was similar to what’s annoying about reading LinkedIn posts from solopreneurs. (Think buzzwords that feel aspirational but, don’t mean much.)

“‘Utilized,’ ‘implemented,’ ‘leveraged,’ ‘elucidated,’ and ‘ascertained’ are often overused. But what human talks like that in a general article they would write? Almost none.” How To Check If Something Was Written with AI (ChatGPT), Gold Penguin

As new tools emerge that claim to improve the voice and tone of AI-generated text, many businesses are firing their copywriters to replace them with button pushers. The question for small business owners is whether they can save on website content creation and SEO efforts by replacing people with AI-generated text. I think the answer depends on your content strategy.

If your content team was already recycling other’s content on your website then, replacing them with AI might be a way to cut costs and get the same results. But, if your content team was developing unique, strategic, niche content, your visibility will drop when you replace them with AI.

Most importantly, if you’re a small business owner writing thoughtful, personal content and hope to replace those efforts with generated text, you’re going to see a decline in engagement as well as visibility.

In This Article

From the jump, I’d like to clarify who I am and who I’m talking to. I’m small (indeed minuscule) in terms of my content impact. I have a teeny marketing agency in Lynchburg, Virginia and my clients are mostly local small businesses — primarily eStores. I’m not taking a big swing at a big topic in this article.

Instead, I’m outlining the practical effects of AI content for small business owners who are using this new tool to try to grow their business or lower the time and cost of content creation.  In general, I’m talking about website content such as pages, blog posts, or product pages. But, I will sometimes reference other areas that this technology has impacted.

How to Spot AI-Generated Text
Why is AI Generated Text So Wordy?
Why is AI Generated Text Bad for your website?
Why is AI a Poor Replacement for content?
Does AI Hurt your website’s search engine performance (SEO)?
Can AI Spot AI?
What Happens when AI doesn’t get User-Generated Inputs
What Does Chat GPT Say About It?

Also, I apologize in advance that this article uses so many block quotes. I chose to include them for two reasons. First, I want to provide as much context as possible for a big topic. Second, the topic is ever-expanding and these sources may update their conclusions between the time I publish my article and when you read it. So, I want to make it clear what exactly I referenced in the linked articles. (I don’t have plans to update this post regularly.)

How to Spot AI-Generated Text

If you want to spot AI text, there are a few key giveaways.

Hubspot’s Summary for How to Spot AI:

• Repetition of words and phrases: AI knows what it’s talking about, but not to the extent human experts do. Its outputs might repeat the same keywords and phrases with little variation when discussing a topic.
• Lack of depth: Generation tools lack depth and can’t go beyond basic facts to truly analyze a topic and develop unique insight. AI-generated text might read more robotic and prescriptive than creative and have a generic tone.
• Inaccurate and outdated information: The facts that content generation tools have are typically correct, but since the tools make predictions, outputs can be incorrect or unrelated to true facts. In addition, information can be outdated, like how ChatGPT is limited to information pre-September of 2021.
• Format and structure: Generation tools follow the same sentence structure as humans, but sentences can be shorter and lack the complexity, creativity, and varied sentence structure humans produce. Content can be streamlined and uniform with little variation.

AI Detection: How to Pinpoint AI Generated Text and Imagery [+ Detection Tools], Hubspot

Many people are already developing a sense for uncovering AI-generated text. Most noticeably, it’s the sentence structure — which I suspect comes from the snake-eating-its-tail impact of SEO on AI-generated text.

For years, search engines have prioritized long-form explanatory content. This strategy performs well when it’s structured into skimmable sections with easy-to-read headings. Topics needed to be clear, answer questions, and help users reach their goals (task completion). If you write web content professionally, it’s second nature to develop posts in this manner. As search engines rewarded this content format, web content writers tried to game the system. They would stuff keywords into content and fluff up sentences to make them take longer to read. That’s why the long-running joke about the chicken soup recipe with 6 pages of introductory copy exists.

Even as search engines constantly change their algorithm to fight inauthentic, over-optimized content, the SEO myths about what works persist. Now, unnecessarily wordy, buzzword-filled content pollutes the internet. Crisp, original writing is beloved by readers (and search engines) but, less common.

What do you think the LLM models are being trained on? That same wordy, bad-SEO-polluted space that most internet users just don’t want to read. (Sometimes, it’s academic content which also itsn’t known for conciseness).

This is why AI-generated text reads like a verbose, aspiring TEDx-talker. The internet told these models that humans write that way. And unfortunately, we kind of do.

More Tips to Spot AI-generated Text Include:

1. Use Google Docs’ Version History feature to detect human-like writing behavior: Google Docs has a feature that saves after each small change is made in a document. Having your students compose exclusively in Google Docs and share an editable copy of their drafts is a good way to track whether they’re pasting content wholesale from another source, or if they’re working bit by bit in a more human fashion. Here’s some instructions on how to use the feature.
2. Unusual or Over-Complex Sentence Construction: AI might sometimes generate sentences that are grammatically correct but unusual or overly complex, especially when it’s trying to generate text based on a mixture of styles or genres.
3. Lack of Personal Experience or Emotion: Since AI doesn’t have personal experiences or emotions, its writing often lacks a personal touch or emotional depth. This can be a telltale sign that the text is AI-generated.
4. Overuse of Certain Phrases or Vocabulary: AI models can sometimes overuse certain phrases or vocabulary that it has learned during training. This is the result of the AI trying to fill up space with relevant keywords. This is because some of the spammy AI-generation SEO tools love keyword-stuffing articles (keyword stuffing is when you repeat a word or phrase so many times that it sounds unnatural). Some AI text generators will attempt to imitate this style.
5. Not Referencing Current Events (or lying about/misrepresenting them): Many AI models like ChatGPT-3.5 do not have real-time access to the internet and cannot provide up-to-date information or comment on current events. If an AI-generated text refers to events after its training cutoff (ChatGPT is September 2021), it’s likely extrapolating from patterns it learned and might not be accurate. Keep in mind, though, that Bing Chat does have access to current events, and Claude 2 can be fed and understand documents of up to ~60,000 words.
6. Lack of Depth/Analysis: Another way to tell if an essay was written by an AI is if it lacks complex or original analysis. This is because machines are good at collecting data, but they’re not so good at turning it into something meaningful. As an example, try to get ChatGPT to give feedback on a student essay; it’s terrible at it! We’re nearing the point where AI is able to start to analyze writing, but its current responses are still very “robotic.” You’ll notice AI generated writing is a lot better for static writing (like about history, facts, etc) compared to creative or analytical writing. The more information a topic has, the better AI can write & manipulate it.
7. Inaccurate Data or Quotes: If a prediction machine (which is what these AI text generators are) doesn’t know something but is destined to give an output, it’ll predict numbers based on patterns (which aren’t accurate). So, if you’re reading student homework and you spot several discrepancies between the facts and the numbers, you can be reasonably confident what you just read was generated with AI.
8. Do the Sources Exist? Do They Say What the Writer Claims They Do?: While this is a thing that students have already done for a long time, many AI text generators will create sources out of thin air that appear convincing. You should check every single source that your students cite if you’re concerned about they’re having used an AI text generation tool. Keep in mind that some AI tools can cite real sources!
9. Format and Structure: Generation tools follow the same sentence structure as humans, but sentences can be shorter and lack the complexity, creativity, and varied sentence structure humans produce. Content can be streamlined and uniform with little variation.
10. Compare to Writing They’ve Done in Class: Students often have a unique writing style and voice that they maintain across their assignments. When reviewing a student’s work, look for significant deviations from this established style. Does the vocabulary, sentence structure, or complexity of ideas feel significantly different? Does the tone match the student’s usual approach? If the student normally struggles with grammar and spelling but suddenly hands in a flawless paper, it might be a sign of AI-generated content. Also, AI-generated text may lack the specific mistakes or idiosyncrasies that a student typically makes.
11. Quiz the Student on the Content of Their Work: If you suspect that a piece of work has been generated by AI, one effective way to confirm your suspicion is by quizzing the student on the content of their work. Ask them to explain complex points, the reasoning behind their arguments, or the meaning of specific words or phrases they used. If they are unable to provide satisfactory answers or seem confused by their own work, it could be an indication that they didn’t write it themselves. Be aware, though, that students might also struggle with this if they’re nervous or if they completed the assignment a while ago and no longer remember the specifics.
12. Look Out for Weird Paraphrases: Many students use tools like Quillbot to take existing academic essays that their friends or classmates have written and “paraphrase” them into not being caught by Turnitin or other plagiarism checkers. This can result in turns of phrases being “translated” literally and in a way that doesn’t make sense, among other odd changes.

AI Detection: How to Pinpoint AI Generated Text and Imagery [+ Detection Tools], Hubspot

What does this mean for Small Business Owners?

You’re not fooling your customers with AI-Generated text. If you start polluting your website with copy-paste responses from ChatGPT, your customers will notice. Think about the impact of this decision on your brand.

Why is AI Generated Text So Wordy?

What unique qualities are left to human-composed writing? Noah Smith, a professor at the University of Washington and NPL researcher at the Allen Institute for AI, points out that while the models may appear to be fluent in English, they still lack intentionality. “It really messes with our heads, I think,” Smith says. “Because we’ve never conceived of what it would mean to have fluency without the rest. Now we know.” In the future, you may need to rely on new tools to determine whether a piece of media is synthetic, but the advice for not writing like a robot will remain the same. Avoid the rote, and keep it random. How to Detect AI-Generated Text, According to Researchers, Wired

In many instances, AI-generated text is unreadable. Furthermore, the chatbots don’t reason to find answers. They use formulas or copy existing answers. Just look at the issue with asking AI how many Rs are in strawberry.

If you spend any time on the internet, you’re likely now familiar with the gray-and-teal screenshots of AI-generated text. At first they were meant to illustrate ChatGPT’s surprising competence at generating human-sounding prose, and then to demonstrate the occasionally unsettling answers that emerged once the general public could bombard it with prompts. OpenAI, the organization that is developing the tool, describes one of its biggest problems this way: “ChatGPT sometimes writes plausible-sounding but incorrect or nonsensical answers.” In layman’s terms, the chatbot makes stuff up. As similar services, such as Google’s Bard, have rushed their tools into public testing, their screenshots have demonstrated the same capacity for fabricating people, historical events, research citations, and more, and for rendering those falsehoods in the same confident, tidy prose. Chatbots Sound Like They’re Posting on LinkedIn, The Atlantic

It’s gotten to the point where my copywriter colleagues complain that it takes them more time to fact-check AI-generated text than to simply write original, verified content in the first place.

Using advanced computer algorithms (rules and instructions that people make to tell the tools what to do), LLMs create a mathematical model of ideas and concepts by ingesting text. Using this information, the LLMs statistically predict which words writers are most likely to use together in the context of the user’s command.

Because many of the sources that AI companies used to train their new tools were formal writing by educated native English speakers, AI writing sounds fluent and well-educated on the surface (word order, word choice). However, remember that these tools create writing based on word frequency and the likelihood that a given word will come after another word or that a given sentence will come after another sentence. Think of it like an advanced version of predictive text on your cell phone. (Of course, the software is more complicated than that!) Why AI-Generated Text Sounds Wordy and Choppy, WordRake

One of my favorite articles on this topic was written on WordRake.

One problem with the writing produced by these AI tools is that the next most common word or sentence is sometimes out of context or disjointed. To prevent this problem, these AI tools seem to have algorithms that put in extra transition words between sentences to make the sentences sound more connected. The goal is to help the writing flow better, but when the conjunction is wrong or unnecessary, the opposite happens. The writing becomes choppy and confusing. Why AI-Generated Text Sounds Wordy and Choppy, WordRake

Anyone who has used these tools has already experienced this phenomenon. That next critical part of communication is simply missing — and often the void is filled with, well, filler.

If you use generative AI tools like ChatGPT frequently, you’ll notice the overuse of cohesive devices (defined below). AI tools rely on these words to create an appearance of coherence. The goal is to make the writing seem internally linked. When used sparingly, cohesive devices can be a valuable tool for writers. But in the case of AI-generated text, using these words doesn’t mean the text makes sense. Text is coherent when it makes sense and makes its point. Coherence requires judgment and reasoning, but AI tools cannot use judgment or reasoning, which means that they can’t distinguish between sentences or paragraphs in which a cohesive device would be good or bad. Why AI-Generated Text Sounds Wordy and Choppy, WordRake

It also seems to be getting worse — which I suspect is happening as more AI-generated or AI-influenced content is fed into these tools. People are trying to write like AI because they think it’s better writing. When in fact, it’s just wordier. WordRake summarizes the problem by explaining,

“English has many cohesive devices to make writing more, well…cohesive. These linking devices can be adverbs, prepositions, conjunctions, and pronouns. They help connect ideas in a text, show how ideas relate to each other, and create a sense of flow or logical progression.

Most writers sort these cohesive devices into these six categories:

1. Organizational Surface Signals: first(ly), second(ly), third(ly), next, finally
2. Conjunctions: therefore, consequently, as a result, on the other hand, in contrast, similarly, however, nonetheless, nevertheless, still, yet, because of this
3. Summative Transitions: in summary, in conclusion, overall, in other words, to sum up, to recap
4. Additive Transitions: also, furthermore, moreover, in addition, as well as
5. Prepositions: with, at, by, to, in, for, from, of, on, since
6. Pronouns: he, she, we, they, such (We’ve written about how to use pronouns to avoid repetition here.)

Each of these words carries different nuances, so it’s important to use the right words for the right purpose. While human writers understand the precise meaning of each word and make word choices accordingly, AI tools have classed these words into categories based on deduced formality and frequency, which can lead to strange or repetitive choices. For a deep dive into which cohesive devices should be used for which purposes, check out Purdue Online Writing Lab on transitional devices and the UNC Chapel Hill Writing Center on transitions.”

Readability aside, this brings us to the next question. Assuming you edit AI-generated text for readability, is it still good for your website?

What does this mean for Small Business Owners?

If you must use AI-generated content, please rewrite it before posting it on your website. Tools like Yoast and Hemmingway Editor will flag the biggest problem areas. Also, run it through a readability test to see how it scores. Much of the text coming out of these tools sounds more professional or complex to insecure writers but, that doesn’t mean it’s better. It just means it’s wordier.

Why is AI Generated Text Bad for your website?

Answering this question requires a basic understanding of how Google search works. For all of their flaws, search engines like Google must deliver search results in response to queries with some reliability. That’s why they’re constantly updating their algorithm. They want to hone that process of finding answers.

How Google Search Works

If you want to learn more about how Google Search works, they have pages of tutorials explaining their process in layman’s terms.

Google Search works in three stages, and not all pages make it through each stage:

1. Crawling: Google downloads text, images, and videos from pages it found on the internet with automated programs called crawlers.

2. Indexing: Google analyzes the text, images, and video files on the page, and stores the information in the Google index, which is a large database.

3. Serving search results: When a user searches on Google, Google returns information that’s relevant to the user’s query.

This process is meant to bring the best result in response to a user’s query.

Google’s automated ranking systems are designed to present helpful, reliable information that’s primarily created to benefit people, not to gain search engine rankings, in the top Search results. Creating helpful, reliable, people-first content, Google Search Central

Google offers the following guidelines:

Content and quality questions

• Does the content provide original information, reporting, research, or analysis?
• Does the content provide a substantial, complete, or comprehensive description of the topic?
• Does the content provide insightful analysis or interesting information that is beyond the obvious?
• If the content draws on other sources, does it avoid simply copying or rewriting those sources, and instead provide substantial additional value and originality?
• Does the main heading or page title provide a descriptive, helpful summary of the content?
• Does the main heading or page title avoid exaggerating or being shocking in nature?
• Is this the sort of page you’d want to bookmark, share with a friend, or recommend?
• Would you expect to see this content in or referenced by a printed magazine, encyclopedia, or book?
• Does the content provide substantial value when compared to other pages in search results?
• Does the content have any spelling or stylistic issues?
• Is the content produced well, or does it appear sloppy or hastily produced?
• Is the content mass-produced by or outsourced to a large number of creators, or spread across a large network of sites, so that individual pages or sites don’t get as much attention or care?

Expertise questions

• Does the content present information in a way that makes you want to trust it, such as clear sourcing, evidence of the expertise involved, background about the author or the site that publishes it, such as through links to an author page or a site’s About page?
• If someone researched the site producing the content, would they come away with an impression that it is well-trusted or widely-recognized as an authority on its topic?
• Is this content written or reviewed by an expert or enthusiast who demonstrably knows the topic well?
• Does the content have any easily-verified factual errors?

They summarize that in the acronym E-E-A-T.

Google’s automated systems are designed to use many different factors to rank great content. After identifying relevant content, our systems aim to prioritize those that seem most helpful. To do this, they identify a mix of factors that can help determine which content demonstrates aspects of experience, expertise, authoritativeness, and trustworthiness, or what we call E-E-A-T.

Of these aspects, trust is most important. The others contribute to trust, but content doesn’t necessarily have to demonstrate all of them. For example, some content might be helpful based on the experience it demonstrates, while other content might be helpful because of the expertise it shares.

While E-E-A-T itself isn’t a specific ranking factor, using a mix of factors that can identify content with good E-E-A-T is useful. For example, our systems give even more weight to content that aligns with strong E-E-A-T for topics that could significantly impact the health, financial stability, or safety of people, or the welfare or well-being of society. We call these “Your Money or Your Life” topics, or YMYL for short.

Search quality raters are people who give us insights on if our algorithms seem to be providing good results, a way to help confirm our changes are working well. In particular, raters are trained to understand if content has strong E-E-A-T. The criteria they use to do this is outlined in our search quality rater guidelines. Creating helpful, reliable, people-first content, Google Search Central

The concept encourages people-first content and puts measurements in place to gauge how well it matches.

To give you the most useful information, Search algorithms look at many factors and signals, including the words of your query, relevance and usability of pages, expertise of sources, and your location and settings. The weight applied to each factor varies depending on the nature of your query. For example, the freshness of the content plays a bigger role in answering queries about current news topics than it does about dictionary definitions. How Search Works, Google

Google lists several factors including:

• Meaning of your query: This is the intent behind the query.
• Relevance of content: Their system analyzes content to see whether it matches the query’s intent.
• Keywords: While keywords should match up between the query and the content, it’s not the only factor.
• Quality of content: They look for the one that is most helpful because of expertise, authoritativeness and trustworthiness.
• Useability of webpages: This is where technical SEO plays a role. If the content is mobile-friendly, loads fast, and is structured well, it gets higher priority.
• Context and settings: This is based on the specific user’s behavior. Location, search history and search settings will influence which results show up.
• Technical Requirement: Google has a bare minimum standard for sites. Most websites hit this criteria.
• Spam Policies: They penalize websites that violate their policies.

For years, I’ve had my client’s content outperform larger websites by sticking to the basic people-first policy.

Groups with more resources than I have been putting this theory to the test for months.

What does this mean for Small Business Owners?

Don’t try to trick Google. You don’t have the time or bandwidth to keep up with sneaky SEO trends. Instead, stick to the principles behind the algorithms as these never change. Your content will have longevity and perform well over time.

Original content always grows in visibility the longer it exists — especially if it’s unique.

Why is AI a Poor Replacement for content?

Spam is “Text generated through automated processes without regard for quality or user experience,” according to Google’s definition.  I interpret this as anyone using AI systems to produce content without a human QA process. Google’s shifting approach to AI content: An in-depth look, Search Engine Journal

Search Engine Journal tried to address this with an interesting experiment.

When GPT-3 hit, I wanted to see how Google would react to unedited AI-generated content, so I set up my first test website.

This is what I did:

• Bought a brand new domain and set up a basic WordPress install.
• Scraped the top 10,000 games that were selling on Steam.
• Fed these games into the AlsoAsked API to get the questions being asked by them.
• Used GPT-3 to generate answers to these questions.
• Generate FAQPage schema for each question and answer.
• Scraped the URL for a YouTube video about the game to embed on the page.
• Use the WordPress API to create a page for each game.

There were no ads or other monetization features on the site.

The whole process took a few hours, and I had a new 10,000-page website with some Q&A content about popular video games.

Both Bing and Google ate up the content and, over a period of three months, indexed most pages. At its peak, Google delivered over 100 clicks per day, and Bing even more.

Results of the test:

• After about 4 months, Google decided not to rank some content, resulting in a 25% hit in traffic.
• A month later, Google stopped sending traffic.
• Bing kept sending traffic for the entire period.

The most interesting thing? Google did not appear to have taken manual action. There was no message in Google Search Console, and the two-step reduction in traffic made me skeptical that there had been any manual intervention.

I’ve seen this pattern repeatedly with pure AI content:

• Google indexes the site.
• Traffic is delivered quickly with steady gains week on week.
• Traffic then peaks, which is followed by a rapid decline.

Google’s shifting approach to AI content: An in-depth look, Search Engine Journal

The principles behind this pattern aren’t complicated. AI-Generated text is a poor replacement because it doesn’t hit those key ranking factors that Google clearly outlines.

What does this mean for Small Business Owners?

If you start stuffing your site with AI-generated content, Google will notice. You’ll have a quick, short increase in traffic followed by a rapid decline. Your site will probably end up in a worse place than when you started.

Does AI Hurt your website’s search engine performance (SEO)?

In short, yes, using AI hurts your website’s search engine performance if you don’t edit it properly.

Having looked at so many factors, including Google’s helpful content update, we can now safely answer our question.

And the answer is no!

Google does not penalize AI content. It penalizes poor content.

Does Google’s helpful content update penalize AI content?, Search Engine Journal

But, that doesn’t mean well-edited AI-generated content or AI-researched content won’t rank. In fact, a lot of what we’re seeing now falls into those two categories. It’s not copy-pasted but, it’s not solely researched and written by a human.

What does this mean for Small Business Owners?

If you want to use AI to generate your content, please edit it well. Think of AI as an assistant rather than the boss.

Can AI Spot AI?

Most of the conversation around using AI to spot AI comes from the academic world. But, it’s an important part of the conversation when it comes to intellectual property and plagiarism on any account.

Last week, OpenAI unveiled a tool that can detect text produced by its AI system ChatGPT. But if you’re a teacher who fears the coming deluge of ChatGPT-generated essays, don’t get the party poppers out yet. This tool is OpenAI’s response to the heat it’s gotten from educators, journalists, and others for launching ChatGPT without any ways to detect text it has generated. However, it is still very much a work in progress, and it is woefully unreliable. OpenAI says its AI text detector correctly identifies 26% of AI-written text as “likely AI-written.”
Why detecting AI-generated text is so difficult (and what to do about it), MIT Technology Review

 

(Watermarks) act as a sort of secret signal in AI-produced text that allows computer programs to detect it as such. Researchers at the University of Maryland have developed a neat way of applying watermarks to text generated by AI language models, and they have made it freely available. These watermarks would allow us to tell with almost complete certainty when AI-generated text has been used.
The trouble is that this method requires AI companies to embed watermarking in their chatbots right from the start. OpenAI is developing these systems but has yet to roll them out in any of its products. Why the delay? One reason might be that it’s not always desirable to have AI-generated text watermarked.
Why detecting AI-generated text is so difficult (and what to do about it), MIT Technology Review

 

The AI text detector that OpenAI rolled out is only one tool among many, and in the future we will likely have to use a combination of them to identify AI-generated text. Another new tool, called GPTZero, measures how random text passages are. AI-generated text uses more of the same words, while people write with more variation. As with diagnoses from doctors, says Abdul-Mageed, when using AI detection tools it’s a good idea to get a second or even a third opinion.
Why detecting AI-generated text is so difficult (and what to do about it), MIT Technology Review

What does this mean for Small Business Owners?

As AI detector tools came out quickly after AI generation was released and they’re getting smarter. Furthermore, text that mimics AI-generated content seems to incur the same penalities. In the long run, copy and paste AI content probably isn’t the short cut everyone hopes for when generating organic traffic. You also might struggle to copy-right it.

What Happens when AI doesn’t get User-Generated Inputs

AI Cannibalism occurs when the tools scour the Internet to produce responses and input information that has already been produced by other AI systems. The result? Low-quality output.

Recently, it has been observed that there’s an increasing shift in sentiments of paid users of GPT-4. Several users have expressed disappointment on social media platforms regarding dip in LLMs’ output quality. The chatbot’s response has been termed as “dumber” and “lazier” in comparison to previous versions leading to concerns regarding the total user experience. This has sparked conversations within the AI community regarding the factors leading to its decline. The Curious Case of AI Cannibalism & Possible Solutions, Cogito

 

In recent months there has been a growing concern in the tech sector regarding AI training upon previously created AI imagery and text. You see, for AI to improve and to make additional leaps it needs more training data with every new rendition. This data is scrubbed from the web by using large data scraping techniques. As the internet becomes more inundated with artificial information, there is bound to be previous AI content that is harvested with every subsequent scraping or data collection.
InsideBigData recently reported “Whenever we are training an AI model on data which is generated by other AI models, it is essentially learning from a distorted reflection of itself. Just like a game of “telephone,” each iteration of the AI-generated data becomes more corrupted and disconnected from reality.” The internet at one point in the not-too-distant past was a place where some truth and facts could be located. As AI becomes more prominent, the internet may become a vast wasteland of bad information. Will it get so bad that nobody will even want to use it as a tool because the amount of effort needed to find the truth will be so time-consuming? I guess we will find out. AI is Corrupting the Internet as We Know It, PetaPixel

Even now, owners of LLM tools are fighting for the right to use human-generated, copyrighted content in their tools — saying their business models won’t work if they have to pay to access and copy human-generated content.

What does this mean for Small Business Owners?

AI needs human-generated content to exist. If people stop writing, the system will implode. So, eventually, someone is going to have to write something. And that person will want a paycheck.

What Does ChatGPT Have to Say About it?

Since I spent time researching and writing this article, I figured I would elicit opinions straight from the robot’s mouth. I submitted the following prompts to ChatGPT and this is what I got back.

Write a paragraph summarizing why AI-generated text is wordy and poorly written. Use and cite verified sources. 

AI-generated text often tends to be wordy and poorly written due to several factors inherent in current natural language processing models. According to researchers from OpenAI, such models, like GPT-3, generate responses by predicting the next word based on statistical patterns in large datasets, often resulting in verbose outputs that mimic human language but lack coherence and conciseness (“Language Models are Few-Shot Learners,” OpenAI, 2020). Furthermore, as highlighted by experts at Stanford University, these systems lack true understanding and context comprehension, leading to repetitive and unnecessarily complex sentences (“AI Language Models: Lessons and Shortcomings,” Stanford HAI, 2023). Consequently, AI-generated text tends to be plagued by redundancy and convoluted phrasing, diminishing readability and overall quality.

Write a quick guide for how to spot AI generated text. Use and cite verified sources. Divide it into steps or key points for readers.

Spotting AI-generated text can be challenging but here are some key points to help identify it:

Unusual Phrasing and Jargon: AI often generates text that may use uncommon phrases or technical jargon inappropriately. According to OpenAI, language models like GPT-3 can produce outputs that mimic human speech but occasionally introduce awkward or unusual wording (“Language Models are Few-Shot Learners,” OpenAI, 2020).

Lack of Coherence or Context: AI-generated text may lack logical flow or coherence, jumping between unrelated topics or providing irrelevant information. Stanford University researchers note that current models struggle with maintaining context over longer passages, leading to disjointed responses (“AI Language Models: Lessons and Shortcomings,” Stanford HAI, 2023).

Repetitiveness and Redundancy: AI often repeats information unnecessarily or includes redundant phrases. This can be a telltale sign of automated content generation rather than human thought processes.

Generalized Content: AI-generated text may offer broad, vague statements that lack specific details or personal experiences. This is because AI models are trained on large datasets rather than personal knowledge or nuanced understanding (“Understanding AI: Knowledge-based and Data-driven AI,” MIT Technology Review, 2021).

Accuracy Issues: Pay attention to factual inaccuracies or errors in details. While AI can generate text that sounds plausible, it may fail in providing accurate information or make inconsistent claims.

By being mindful of these indicators, you can better discern whether the text you’re reading is likely generated by AI rather than authored by a human.

Write 3-5 paragraphs to answer the question, “How will AI replace copywriters?”  Outline the arguments for and against using AI to replace humans when writing content for business websites. Use and cite verified sources. 

The potential for AI to replace copywriters in writing content for business websites sparks debate, considering both its advantages and drawbacks. Proponents argue that AI offers efficiency and scalability, capable of generating vast amounts of content quickly. OpenAI’s GPT-3 model, for instance, demonstrates the ability to produce coherent and contextually relevant text based on minimal input (“Language Models are Few-Shot Learners,” OpenAI, 2020). This efficiency can significantly reduce costs associated with content creation, as highlighted by experts in digital marketing and AI integration strategies (Fernández, 2021).

Furthermore, AI can optimize content for SEO by generating keyword-rich text that ranks well in search engine results. This capability aligns with the growing emphasis on digital marketing strategies that prioritize visibility and online presence (Marr, 2021). Additionally, AI-driven content can adapt to different tones and styles, catering to diverse audiences without the need for extensive human intervention (Choi et al., 2022).

However, detractors argue that AI-generated content lacks the nuanced creativity and emotional intelligence that human copywriters bring to their work. Human writers can infuse brand personality and authenticity into their writing, establishing a deeper connection with readers. This personal touch is crucial in building trust and engagement, factors that are pivotal in driving customer loyalty and conversions (Iqbal et al., 2020).

Moreover, concerns about ethical implications arise regarding AI-generated content, particularly in terms of transparency and accountability. There is a risk of misleading or biased information being disseminated without proper oversight or ethical guidelines in place (Anderson and Anderson, 2023). Critics also point out that AI models are still prone to errors and biases inherent in the datasets they are trained on, potentially leading to inaccuracies or inappropriate messaging in business communications (Suresh and Guttag, 2021).

In conclusion, while AI offers undeniable advantages in terms of efficiency and SEO optimization, the debate over its role in replacing human copywriters hinges on balancing these benefits against the intangible but critical aspects of creativity, authenticity, and ethical considerations in content creation for business websites. As AI technologies continue to evolve, businesses must navigate these complexities to determine the most effective and responsible use of AI in their marketing and communications strategies.

In a paragraph, explain what happens when AI gets AI content as input. 

When AI receives AI-generated content as input, it enters a recursive loop where the quality and characteristics of the output are influenced by the initial AI-generated text. This phenomenon, often referred to as “AI-in-the-loop,” can amplify biases, errors, or stylistic quirks present in the original input due to the machine learning model’s tendency to learn and replicate patterns from the data it processes. This recursive process can enhance the coherence and relevance of subsequent outputs within the constraints of the original input’s parameters, illustrating the iterative nature of AI’s capabilities in generating and refining content autonomously.

Should you use AI-generated text on your website?

I would encourage any small business to use AI-generated text sparingly on their website. It’s not unique, scores low on readability, and performs poorly for task completion — going against Google’s ranking criteria for content.

The Worst Ways to Use AI-generated Text

Back in the early days of spellcheck, writers learned to use it as a supplement for proofreading not a replacement. The checkers would confuse homonyms or auto-replace turns of phrase with strange word combinations. Similarly, content writers should read and edit anything that comes from an AI writing tool.

• Don’t copy and paste responses.
• Don’t forget to check any statements of fact and find the sources.
• Be careful when quoting and summarizing to avoid plagiarizing.
• Remember to consult a lawyer if you want to claim the writing as your intellectual property.

The Best Ways to Use AI for Content

As a content creator, I have tried several AI tools to assist my writing. They help research new topics and consider new ideas.

• Edit and rewrite anything you get from the tools.
• Use it to outline content and simplify your thoughts.
• Ask it to find sources or look for difficult citations.
• Try dialoguing with the tools to get through writer’s block.

Be careful though! Anything you give to the tool is used to train it. So, your big ideas will be reused without giving you credit.

In the end, writing good content (and optimizing it for search engines) is more than just writing. It’s thinking creatively, solving problems, and finding new stories to bring forward.

The Takeaway

The good news is that people can be trained to be better at spotting AI-generated text, Ippolito says. She built a game to test how many sentences a computer can generate before a player catches on that it’s not human, and found that people got gradually better over time. How to spot AI-generated text, Technology Review

While AI-generated text isn’t going away, we can be smarter about how we use it. For small business owners, I would encourage using it as support, not a replacement, for writing original content.

It’s obvious if you think about it as a user. There are plenty of big, well-known websites that people can reference if they want the same old answers to the same old questions. You can’t compete with those domains.

Yet, you’re probably an expert on something that no one has ever written about — but everyone is trying to learn. You might have a special story unlike any other — that could resonate with so many people. That’s the real power of putting content online. Something new, something that has never been said before — and you said it first.

And that’s something that the generated part of AI-generated text will never touch.

Did you enjoy this post?

Sign up for our twice-a-year newsletter for more estore inspiration, trends, and free downloads. At The Shop Shop, we specialize in retail marketing. Our Lynchburg Marketing Agency offers strategic services and flat-rate packages for shop owners, artists, and makers. We come alongside your team to help you get more out of your marketing. Everything we do fits seamlessly with your in-store promotions or online drops to create a captivating shopping experience. Let’s redefine the future of retail together.

Further Reading & Sources

• AI is Corrupting the Internet as We Know It, PetaPixel
• Why AI-Generated Text Sounds Wordy and Choppy, WordRake
• Chatbots Sound Like They’re Posting on LinkedIn, The Atlantic
• How to Detect AI-Generated Text, According to Researchers, Wired
• Why detecting AI-generated text is so difficult (and what to do about it), MIT Technology Review
• Detecting AI-Generated Text: 12 Things to Watch For, East Central College
• How To Check If Something Was Written with AI (ChatGPT), Gold Penguin
• AI Detection: How to Pinpoint AI Generated Text and Imagery [+ Detection Tools], Hubspot
• How to spot AI-generated text, Technology Review
• The Words That Giveaway AI Generated Text, Wired
• The Curious Case of AI Cannibalism & Possible Solutions, Cogito
• Creating helpful, reliable, people-first content, Google Search Central
• How Search Works, Google
• Google Search Essentials, Google Search Central
• In-depth guide to how Google Search works, Google Search Central
• SEO Starter Guide, Google Search Central
• Google Sets The Record Straight: AI Content In Search Results, Search Engine Journal
• Google’s shifting approach to AI content: An in-depth look, Search Engine Journal
• Does Google’s helpful content update penalize AI content?, Search Engine Journal
• OpenAI insists it can’t sufficiently train AI models without copyrighted material, ShackNews
• OpenAI: We Need Copyrighted Works for Free to Train AI, Tech.co