Humans Are Starting to Write Like AI

A few years ago, it was easy to tell human writing from AI-generated material. Human communication sounded irregular but natural, while AI tended to sound polished, overly meticulous and flawlessly structured. Today, the line is blurring. As AI continues to advance and people increasingly use these tools for work and everyday tasks, many are adopting the same writing patterns that AI systems generate rather than the systems adapting to more humanlike styles.

Recognizing the AI Writing Style

AI writing used to be easily recognizable. It often sounded repetitive and oddly formal in comparison to normal human speech. Now those same patterns are cropping up more in human writing as people spend more time dealing with chatbots and AI applications. In fact, researchers at Penn State found that readers could only distinguish AI-generated text about 53% of the time, illustrating how fuzzy the line has become between human and AI writing. 

The Overuse of Certain Words

One of the easiest ways to recognize AI-influenced work is word choice. AI models tend to over-rely on particular terms that appear to be emotionally evocative, though simpler words would sound more natural. Words such as “delve,” “multifaceted,” and “landscape” have been intimately linked with AI-generated text since they are used significantly more often in chatbot responses than in regular human writing.

Models like GPT-4o utilized phrases like “camaraderie,” “tapestry,” “palpable” and “intricate” at more than 100 times the rate of human authors, research showed. So as more people use AI for brainstorming, editing and writing, some of those words are beginning to show up in human speech as well. 

Perfectly Balanced Formatting

AI writing also tends to be predictable in structure. Typically, a paragraph begins with a clear topic phrase, then provides an explanation, an example and then ties up with a neat conclusion. That pattern is effective for clarity, but it can also make writing feel mechanical when every paragraph has the same rhythm.

There is frequently more variety in human writing. Some thoughts are blunt. Some of them ramble a little, going on side tangents to lengthen the sentence and explore other facets before they get to the point. Those natural irregularities are starting to go away in favor of cleaner, more symmetrical formatting as more people turn to AI to help them compose emails, reports and social postings. 

The Rise of the Numbered List

Another obvious difference is the increasing popularity of numbered lists and a substantially split structure. Chatbot responses are generally broken into steps or bullet points because AI systems are meant to deliver information in a scannable format.

That framework is clearly useful online, but it’s also changing the way individuals write, too. More blog entries and LinkedIn updates are written in that same consumable format. Sometimes the formatting is more important than the personality behind the content. 

A Constantly “Helpful” Tone

Usually, AI-generated text doesn’t seem too biased or untidy. Instead, it attempts for a tone that is even and sounds fair, encouraging and universally agreeable. The outcome is tidy writing, but sometimes without a strong voice.

That tone is becoming increasingly ubiquitous online, and some human writing is beginning to lose the idiosyncrasies and flaws that once made it memorable. Strong opinions get softened. Humor is less risky. Often, disagreement is expressed in a meticulous, extremely courteous style that mimics how AI systems are programmed to respond. 

Why Humans Are Adopting a Machine’s Voice

The turn to AI-like writing is not occurring in a vacuum. People are now frequently encountering AI-generated language in search engines, chatbots, social media postings, marketing emails and workplace apps. The more individuals read and react to that kind of communication, the more natural and normal it starts to feel. Research shows that 39% of all published articles in 2023 were from AI tools. This trend shows the internet has become inundated with machine-generated material.

Part of the appeal is speed. AI writing is designed to present information quickly and clearly, making it easy to digest. Human brains are wired to choose the path of least resistance, so it’s no surprise that individuals start mimicking the same simplified terminology and layout they see online every day. Over time, tight, AI-style communication can become the default way people write professionally.

The writing process itself is also changing with the increased use of AI assistance. Nowadays, many brainstorm using AI, edit drafts with prompts or rewrite sections using chatbot ideas instead of starting from scratch. That back-and-forth process establishes a co-writing connection, resulting in a final product that blends human and computer influence. There is a point where the AI’s style and the persona’s genuine voice can be hard to tell apart if enough iterations are done.

There’s also a social side to it. AI-written text frequently seems calm and structured, making human text seem disordered in contrast. In the workplace, people may start tempering their tone or avoid more daring stylistic decisions because AI-driven communication has become subtly synonymous with clarity and expertise. 

What Are the Downsides?

The increasing impact of AI on human communication is not always a bad thing. AI tools can help individuals organize thoughts, communicate faster and beat writer’s block. Still, there’s a growing fear that prolonged exposure to AI-generated language could slowly erode the distinctiveness that makes human speech personal and memorable.

One of the biggest dangers is the loss of distinctiveness. Good writing has always had personality, experience and perspective, even when it’s not perfect. AI-generated language tends to round out rough edges in favor of more broadly acceptable phrases. The more people use the same polished tone and terminology, the more personal and brand voices can begin to merge. That sameness might make it feel like internet material becomes repetitive and replaceable over time.

There’s also the risk that formulaic writing could lead to more formulaic thinking. Research conducted by the MIT Media Lab found that subjects who used LLMs for essay writing showed decreased brain connectivity and poorer memory recall than participants who wrote essays without AI assistance. If people and AI systems keep moving toward the same form of communication, the internet risks becoming less diverse, less opinionated and eventually less enjoyable to read. 

How to Avoid Sounding Like AI

Even if you avoid using AI, it’s nearly impossible to avoid reading AI-generated text. Similarly, AI technologies are increasingly embedded in everyday operations, and for many people, they may genuinely enhance productivity and creativity. If the goal is not to avoid AI entirely, humans should use it in a way that enhances human thought, rather than replacing the personal voice and perspective: 

• Use AI as a brainstorming partner, not the writer: You can’t avoid sounding like AI if you’re using AI to write. AI can help you brainstorm ideas or organize your research, but the best writing comes from people who actually sit down and do their own writing, rather than copy-pasting enormous chunks of AI-generated material into a manuscript.
• Practice more conscious editing: One of the easiest things you can do to prevent sounding too AI-generated is to consciously look for sentences that sound excessively polished, repetitious or emotionally flat. Hearing a draft read aloud will help authors catch anything that sounds odd or not like the way they usually talk. 
• Spend more time reading human writers: Search engines, social networks and even newsletters are often full of AI-generated content now, making it all the simpler for that style to become mainstream. Reading books, essays and long-form journalism from strong human writers may help you maintain a sense of rhythm and originality. 
• Pay attention to personal writing habits: Every writer has their own sentence rhythms and certain words they like to use, which give their writing a distinctive quality. Instead of trying to get rid of such “defects” altogether, writers can learn to see them as part of their voice rather than something to optimize away all the time. 

Keeping the Human Voice Intact

For years, the discourse surrounding AI has been about making machines talk more like humans. Now, the reverse may be happening just as swiftly. As AI-generated communication becomes more widespread online, the challenge is less about recognizing AI-generated writing and more about ensuring human voices don’t get lost amid it.