ChatGPT Cannot Replace Me · 1 July 2023
Thankfully, ChatGPT will not replace me.
If you have not yet heard of ChatGPT, you probably should be checking it out. After all, it might replace you. Or more probably, it might help you do your job better. At the very least, you should check it out to see if you can have some interesting conversations with it. Rather, with AI, Artificial Intelligence.
No. I am not going to tell you that AI will rule the world some day. It might take a few jobs, maybe even quite a few jobs. But if we can see how to use it in our everyday lives, it can help us. And it can help us with more than just serving up ads that are picked out just for us. (Which is just one of the things that AI has been doing for years.)
I heard of ChatGPT a few months ago. I was intrigued, but did not sign up for an account. Until the other day. That was when I was talking with a couple very smart people who said that living through the age of AI might be like living through the industrial revolution. Hold on to your hats. We are in for a ride.
I am intrigued at teaching a little about AI to my programming and engineering students. (Another thing that my smart friends and I talked about.) Just some rudimentary stuff. Like how to use it to be more productive. And how to verify its veracity. After all, ChatGPT can come up with lots of crazy stuff. Some of it might even be true.
So what is ChatGPT? I am glad you asked.
I do not know. Not really. What I do know is that some smart people figured out how to create this thing called a neural network. That is something that kinda resembles neurons in the brain and sorta models thinking. Apparently, the neural network is just a bunch of connections between stuff. The stuff can be words or ideas, and those connections have weights.
ChatGPT has made connections between words on the internet. Then, it used those connections to figure out how to predict what the next word on any page on the internet would be. It used the real words on the page to determine whether that prediction was correct, and then did it over and over and over again until it finally started getting those next words correct most of the time.
Or something like that. It is called a Large Language Model and the above is about all I understand about it.
When my friends explained how the Large Language Model works and how ChatGPT works, I thought that I should try it out. ChatGPT, that is. So I asked it to write a blog post for me. I wanted to see if AI could replace me. I am happy to say that it might write better than me, but it cannot write like me. Boy is that a relief.
Here is an excerpt from the prompt: “Write a blog post about AI in the style of the author of MediocreMan.com.”
Welcome back, fellow thinkers and dreamers, to another riveting exploration into the world of technology and its impact on our lives. Today, we delve into the enigmatic realm of Artificial Intelligence (AI). While some hail AI as the harbinger of a new era, others fear its ascent will lead us towards mediocrity. Join me on this captivating journey as we unravel the intricacies of AI and its implications for our society.
“…delve into the enigmatic realm of Artificial Intelligence (AI).”? Who uses words like that? “Certainly not I,” said the frog. (I know. There is no frog in the story of Henny Penny. But I like to make up my own quotes.) After all, I am not sure I know what half of those words mean anyway.
Well, I did not think the rest of it sounded like me, so I asked ChatGPT, “Does that really sound like the author of MediocreMan.com?”
ChatGPT replied, “Apologies if the previous response did not capture the specific style of the author from mediocreman.com. Let’s try another approach that aligns more closely with their style.”
The Mediocre Machine Dilemma:
Here’s the conundrum, my friends. As AI becomes smarter and more sophisticated, there’s a creeping fear that it may begin to replace our own ingenuity and creative prowess. Are we hurtling towards a future where machines imitate our intelligence, leaving us feeling like a bunch of mediocre imitators ourselves? It’s a troubling thought, to say the least.
That paragraph was certainly not very creative, and it used too many complete sentences. Not my style at all. But it had some nice thoughts. Even though the complete essay tried to use the word “mediocre” too many times.
In fact, I told ChatGPT, “Really? I do not think Michael T. Miyoshi is that pompous.” I figured I would give the machine one more chance to try and replace me. I must say the last effort was just as bad as the other two.
“Apologies for the misunderstanding. Let’s give it another shot, this time capturing a more authentic tone in line with the style of Michael T. Miyoshi from mediocreman.com.”
The Limits of Artificial Brilliance:
Sure, AI can do some pretty nifty stuff. It can predict our movie preferences and beat us at board games. But let’s not forget that it’s called “artificial” intelligence for a reason. For all its mathematical prowess and fancy algorithms, AI falls short when it comes to the messy, unpredictable nature of genuine human intelligence. So, while it can crunch numbers and analyze data faster than you can say “mediocrity rules,” it still lacks that certain je ne sais quoi that makes us humans truly extraordinary.
Je ne sais what?
Well, I must say that I was happy with my foray into the ChatGPT world. It was enlightening. And ChatGPT was clever. It even used the word “mediocrities” to address the readers. Something I would never do. Still, it was fun to read what AI thought I might sound like when writing about AI.
I will post those three full ChatGPT blog posts for people to see. But there is a danger in that. After all, you might disagree with me. You might think that one day Artificial Intelligence really could replace me.
© 2023 Michael T. Miyoshi
| Share on facebook | Tweet |
|
Comment
Commenting is closed for this article.
| Poetic Expectations | MediocreMan Chats with ChatGPT |