11 Words to Avoid So You Don’t Sound Like AI
Generative AI overuses certain words. Check out 11 words and phrases that people see as telltale signs of AI content.
Generative AI is here to stay, if for no other reason than it helps people get work done better and faster than before.
Consider these stats:
- 95% of professionals using AI spend less time on manual work
- 89% of professionals believe AI improves the quality of their content
- 48% of sellers struggle with communicating value to customers
This is a clear pain that generative AI can solve – creating high-quality content that communicates value.
But since the launch of ChatGPT, people have rapidly developed internal ‘AI detectors’.
Here are 11 words and phrases that people consider telltale signs of AI-generated content (which you can use to prevent your content from being seen as written by AI).
On a side note, this list won’t contain words that are frequently considered AI yet are more likely just a different culture’s word preference, such as the word keen.
Delve
Delve is arguably the reddest red flag that something was written by AI.
It’s garnered so much attention that even Y Combinator cofounder Paul Graham has stated on multiple occasions that delve is evidence of AI content.
One chart even shows the use of delve skyrocketing since 2022, the same year ChatGPT launched.
While not concrete proof that AI overuses delve, it does correlate with people’s perception that delve is now an AI indicator.
Tapestry
Tapestry is frequently used as a metaphor to ‘tie’ several elements together, such as “Our skills, experiences, and relationships are like threads in the tapestry of our lives.”
Generative Artificial Intelligence often likes to use tapestry to start sentences, such as “In the tapestry of life” or “Friendship is a tapestry”.
In today’s ever-evolving world
You can swap ever-evolving for another adjective like ever-changing or ever-increasing. Alternatively, you might see rapidly evolving or rapidly changing.
In any case, if you see the construction In today’s [optional adverb + adjective] world, chances are you’re reading either an English-as-a-foreign-language textbook or an AI output.
Which means unless you’re studying for Cambridge English exams, it’s most likely the latter.
In summary | In conclusion | In essence
Here’s a special 3-in-1 combo.
School teaches us that good writing composition requires us to finish essays with a conclusion paragraph. Naturally, generative AI’s going to do the same.
The trick is how to start that paragraph, and AI’s go-to solution is one of these phrases.
You might even find AI using other words like Ultimately, or To conclude,.
It’s important to note
Like most of the phrases that AI uses, there’s nothing inherently wrong with using It’s important to note.
It’s just going to sound… off.
While it’s useful for drawing attention to a certain point, it’s a filler phrase that could be replaced with simpler and more concise versions like Keep in mind or even just Note:.
Cutting-edge
Likely a byproduct of being developed by Silicon Valley-based geniuses, but any new technology or advancement is likely to be described as cutting-edge.
Honestly speaking, the downfall of cutting-edge from cutting-edge (pun intended) marketing jargon to AI red flag is tragic.
Quinoa
This is more of a fun addition because if you use the word quinoa, you’re either:
- A sales rep for a farm or company selling healthy foods
- A person looking for alternatives to wheat and rice
- ChatGPT
Why? People who brainstorm recipe ideas using ChatGPT frequently receive quinoa as a suggestion.
Multifaceted
Multifaceted is a good example of AI’s tendency to present topics as highly complex, if not downright convoluted. When AI drops the word multifaceted, you can rest assured it will be followed by at least 4-5 paragraphs of explanation.
A normal person would probably just say complex or complicated.
Testament
Unless you’re talking about the bible, chances are low you’ll encounter testament in everyday conversation. You might run across it in more formal circumstances where rhetoric has greater importance, such as a speech.
That’s why if you encounter testament somewhere like a cold email, it’s probably AI.
Certainly!
Ask ChatGPT a question and there’s a good chance it will reply with Certainly!
Perfectly normal word, perfectly normal situation, perfectly normal red flag.
After close to two years of ChatGPT responses, you’re likely to be conditioned and triggered to think AI whenever you see Certainly! (In comparison, people commonly respond with Sure,.)
Revolutionary
Go to any person who’s used ChatGPT and you’ll learn that everything is revolutionary in ChatGPT’s eyes.
This is probably more our fault though. Humans have a tendency to use words like revolutionize and revolutionary to make something seem cutting-edge, disruptive, and cool (Note: Despite containing several red flags 🚩, this sentence wasn’t written by AI).
Create humanly awesome AI content with AiSDR
One of ChatGPT’s “fatal flaws” is creating copy that’s adjective-heavy with lots of ‘fluff’ words. That’s why people’s AI detectors have become calibrated to take note of how frequently adjectives and non-everyday words are used.
Our outreach success team constantly fine-tunes our AI to make sure it isn’t plagued by these issues. So that every email sounds humanly natural, humanly awesome, and 100% unique, they instruct AI about which words, phrases, and spellings to use and not use.
The result? Leads can’t tell that the messages were originally by AI.
