I will scream about AI writing.
I have had it with the breathless articles. ChatAI will write articles for you. I call bull. I have repeatedly put them to the test, and none of the AI writing is anything I would publish (except to show how well the AI writers don’t write).
My plan today was to create a piece of content around the different types of sentences compared to the different kinds of sentences. Then another email arrived about the advantages of using AI writing to create content.
And it sent me over the edge. I am sick and tired of the articles and pitches. So instead of an article about sentences, I will rant. Language Lovers is filled with informative articles. And it is a monetized site. As such, I typically do the things website owners do–look for keywords and use SEO techniques so my content moves up in the rankings.
Not this post. No keywords, no SEO analysis. Just ranting.
Why People are Wrong About AI? (or why the sky is not falling)
There is no such thing as a generative language AI that can detect errors. It is not possible with current technology.
Full stop.
You cannot have a conversation with Siri or Alexa, because they retrieve information. You can ask Alexa to tell you a joke, and it will retrieve one, but do you think Alexa knows why it is funny?
A language generator engine tries to determine what you are asking for and then generate a response that it thinks will match your intent. That’s what it does. It does not fact check. It cannot.
I am not convinced there is much “learning” happening either. For that to occur, it would need a feedback loop that “teaches” it what type of response you want. How can it learn the tone you want? How can it learn to match your writing style?
Does everything you write sound like a Wikipedia entry?
Okay, So Why Not Let the AI Create the Rough Draft and I Fix the Factual Errors?
Simple answer: it is cheating. Either you wrote it, or the AI did, or you wrote it together.
So if you are not the sole author, you should let it me known.
Written by John Smith and AI assistant Gerald (you are going to give your AI assistant a name, right?)
If you aren’t going to be honest and upfront, that’s deception.
If I’m wrong, please tell me why it is okay for you to claim you wrote something when you didn’t.
Before You Tell Me I Am Wrong
I have tried nearly half a dozen on AI writing. These are a few articles with examples.
- Are Exclamation Points Flirty? I Ask Chat.OpenAI
- An A.I. Writer’s Gift Idea for Grammar Geeks
- 5 Fun Facts About Verbs According to AI Writers
- What Another AI Writer Got Wrong
- 5 Interesting Facts About Exclamation Marks
Please don’t tell me that these AI writers work well or that Chat.OpenAI is better until you have read several of these posts.
The Truth
My pocketbook would be happier if I found these AI writers do what they advertise. Then I could write positive reviews and place ads for their products on my pages.
But the content on this website will be written by me. If I recommend a similar product, like Quillbot, it is because it is NOT an AI writer but a tool that helps with research, grammar, summarizing, and paraphrasing.
And Thus Ends This Rant
I feel better now. And more reassured after I ran across this Stephen Heard article, ChatGPT Did Not Write this Post.
As always, please share.
Chat AI is sadly lacking at the moment.
One important aspect that it lacks is experience (And that is something that it will never have).
Exactly. Humans use retrieval and generative language process and then evaluate what they read and say. AI language generators cannot.
Great rant . From what I can see, what it’s producing is soulless
Yes. Completely. As to literary writing, such as poetry and fiction, I don’t see the point.