With everyone scrambling to invest in OpenAI’s incredible ChatGPT, you’re probably wondering the same thing I am: Can you use ChatGPT to write blog posts reliably and efficiently?
ChatGPT is not a reliable tool for bloggers, at least not yet. In addition to possibly providing incorrect or outdated information, Google has a strict anti-spam policy that can get your blog banned or penalized by Google. The best use case for ChatGPT is to come up with subheading ideas.
I have asked ChatGPT to answer a few rather simple questions about some things I know a lot about. So, let’s see how it did in my informal experiment, and I’ll tell you my opinion about its performance.
ChatGPT Didn’t Do Well in My Tests
What stuck out to me is just how unreliable, limited, or straight-up incorrect ChatGPT can be. Now, I’m not anti-AI, even though it does resemble something straight out of a Black Mirror episode. AI can be (or already is) great and helpful, but I wouldn’t trust ChatGPT as far as I could throw it.
If you’re wondering why, here are the 5 questions I asked and my commentary on them. I tried to come up with relatively common questions that one would ask in a few random niches.
1. How To Adjust the Saddles on a Fixed Bridge
The first one is about guitars. Most of the information ChatGPT provided here is sound (pun intended). However, there is one HUGE problem here — every guitar I’ve ever seen doesn’t use flathead screws but rather Allen screws. So, you wouldn’t use a flathead screwdriver; you’d use an Allen/hex wrench to adjust the saddles.
As a professional writer, another problem I see here is the second sentence of this would-be mini blog post. Why is it telling me what saddles are if I already inquired about them? Seems rather redundant.
2. Should I Use Hot or Cold Water for a Moka Pot?
A deceptively simple question, ChatGPT got it wrong. I’ve been using a moka pot for almost 2 years now, and while I do brew with hot water, it’s not for the reason ChatGPT says. Whether the water is piping hot or freezing cold at the start, it doesn’t really matter that much. Hot water could result in better-tasting coffee (and faster brew time) but not stronger coffee.
Sure, the sentence out of context makes sense — colder coffee equals weaker brew. However, water has to reach a certain temperature before it can physically reach the coffee in a moka pot. If anything, starting off with colder water means it’d extract more slowly and result in a stronger brew.
3. Do I Need To Train to Failure To See Muscle Gains?
This answer is best described as inconclusive. According to a 2005 study, stopping an exercise significantly before failure is vastly inferior for muscle growth.
More importantly, P. Debraux of Sciences Du Sport points out that current research is rather inconclusive, especially when it comes to advanced lifters.
ChatGPT also completely missed the mark on training to failure being done to cause muscle damage — it’s more about maximal exertion.
4. Why Is My Cat So Affectionate in the Morning?
I’m not a veterinarian, so I can’t comment on the natural rhythms part of ChatGPT’s answer. However, ChatGPT fails to mention that feeding, or the owner’s morning routine, is a huge part of cats’ affection in the morning.
My cat wakes me up every morning at around the time I usually get up and won’t stop meowing until I give him breakfast. FAQCats points out that a lot of time passes between dinner and breakfast, so cats get very hungry by the time you wake up.
5. Why Are Doc Martens So Bad?
Apparently, ChatGPT was confused by my first question, so I had to rephrase it. I have experienced first-hand, watched, and read a lot about newer Doc Martens having serious quality issues. Although ChatGPT is right that they’re known for their durability, that hasn’t been the case in the past few years.
An article from 2019 by The Guardian tells us a completely different narrative than ChatGPT. According to ChatGPT, Doc Martens are awesome and give us a rather indecisive answer as to why newer boots have quality issues. It even tells us to contact the manufacturer. Dr. Martens do offer a 1-year warranty, but it’s only for manufacturing defects, making this an exercise in futility. Fun fact: they discontinued their For Life guarantee in 2018.
On the other hand, The Guardian explains that Dr. Martens’ parent company, Permira may be to blame. They have made some changes, such as shifting production to Asia and sourcing cheaper materials, resulting in inferior boots.
Google Doesn’t Like AI Content
As hinted at previously, Google has an anti-spam policy against what they call auto-generated content. Google is also good at identifying spun (plagiarized) content, which is essentially the only thing that ChatGPT can create.
I have a hunch that this will become a battle between bloggers who want to get rich quickly with the help of black-hat SEO (aka ChatGPT content) and Google’s crawlers.
Whatever the case might be, if you solely rely on AI to create your blog posts, you’ll regret it when Google punishes your website.
But I’m not trying to undermine OpenAI’s incredible accomplishment here. GPT-3 is impressive, and I think it can be used as a decent writing tool. I talk more about it in Will ChatGPT Replace Blog Writers?
Conclusion
I don’t think that ChatGPT or any other AI-based chatbot will take Google’s crown anytime soon. A reader can’t rely on AI to provide reliable information, even if it’s a simple inquiry. ChatGPT essentially recycles, compiles, and plagiarizes content written by humans to give a convincing answer, even if that answer is completely wrong.
On the opposite side of the spectrum, Google works hard to find the most reliable and trustworthy website and then pulls an answer from there to give you the information you’re looking for.
If you’re a blogger and like this article, consider ordering content from us. Learn more by clicking here.
Phantom Post 