ChatGPT is scary.
Besides being scary, ChatGPT is also one of the most talked about things in the world in the last couple of months, so I’m sure most of you are tired of hearing about it.
That’s why I decided to talk about it only when I had something to say other than how scary it is. Because it’s terrifying. Actually, the whole of AI is terrifying, considering the impact it might have (or is already having?) on human society. But this time I want to talk about ChatGPT as the most powerful AI tool that is available to the general population. But the world is already full of bad news and terrifying predictions about the future, so this time I want to talk about something that can give you some hope.
There is at least one thing where ChatGPT will never be able to replace you.
The internet is flooded with videos about how to use ChatGPT to your advantage. Most common videos are about how to make X amount of money per month with it. Then, there are those about how ChatGPT can be your personal assistant, how it can complete certain tasks for you so that you don’t have to, etc. Actually, when I think about it, all those "different" videos come down to how to make (more) money with ChatGPT.
But there is one use of ChatGPT that especially rubbed me the wrong way.
"How you can read any book in 5 minutes with ChatGPT."
ChatGPT can summarize any book for you. In fact, it can summarize any chapter of any book or explain specific ideas and concepts from it. As someone for whom reading is a major part of his life and someone who admittedly romanticizes books and reading, this sounded like the worst dystopian nightmare. And although I understand that there are certain cases where this is something extremely useful and practical, it was worrying to see videos of people asking ChatGPT to "read" a book for them and this AI tool being able to instantly retell it in 3 sentences or 3000 words, depending on what they asked.
"It might be useful in some cases, but this will eventually lead people to stop reading books." That was my initial thought and fear.
There was only one thing to do. Face this potential tragedy and test ChatGPT myself, rather than just think about how scary it is and how it might impact us.
Of course, if I was going to test its ability, I had to do it with the books that I’ve not only read but spent a decent amount of time thinking about and taking notes on.
"Give me a brief summary of the book The Art of Loving by Erich Fromm." I decided to start the test with this book because it is one of my favorite books I’ve read recently.
Sure enough, it gave me a summary in a fraction of a second. But this wasn’t particularly impressive. This was a simple search engine function; the internet is full of summaries of any book.
Then I started asking it to give me a brief summary of specific chapters. This was more impressive. Then I tried asking it something that seemed more difficult.
"What role does discipline play in love, according to Fromm? Write 500 words on that."
This time, it was terrifying. Once again, in a fraction of a second, it replied. And it gave me an almost perfect breakdown of everything Fromm says about discipline in the book.
I had to take my time to read the whole book. More importantly, I have spent thousands of hours writing just to be able to share what I’ve learned from a book in a way that other people would understand and hopefully enjoy reading. This tool did it in the blink of an eye. Terrifying.
However, I felt there was something missing in its answers. I asked the same question about different concepts from the book. Once again, an instant and correct response. But there was still something missing.
Then I decided to try with a different book to see if maybe I was onto something.
Of course, it was time to test it with Nietzsche.
"Give me a brief summary of the book Beyond Good and Evil by Friedrich Nietzsche."
I’ve spent years reading this book and taking notes on it. Other than using the name Friedrich Nietzsche and a couple of buzzwords that are generally used when describing him, the first answer had almost nothing to do with Beyond Good and Evil.
So I replied, "This is not what this book is about."
"Sorry! Here is an improved version of the summary: Beyond Good and Evil is..."
The difference between the first and second response was night and day. This thing really does learn fast. Then I started questioning it about some of Nietzsche’s key ideas. Once again, responses were technically correct. But there was something missing from them. Something important.
Finally, I decided to test it with one last book. Hagakure by Yamamoto Tsunetomo. I’ve read this book more times than I could count, and it has become one of the most important pieces of writing I’ve read in my life.
As always, I started with a summary. The summary wasn’t incorrect like with Beyond Good and Evil. But that’s not what the book was really about, at least not for me. Then I decided to ask it about some core ideas from the book. Once again, not an incorrect answer, but it’s barely scratching the surface.
The answers I was receiving didn’t seem to come from the book that I knew, the book that impacted my life in such a strong way.
I decided to give it one last chance.
"Give me a list of lessons from Hagakure that apply to everyday life even in the modern age."
The response was painfully generic. I tried again, adjusting my question. Different arrangements of words, but the same generic response. I repeated this process a couple of times.
Based on what ChatGPT was telling me about Hagakure, I don’t know if I would even decide to spend my time reading this book.
But maybe its responses weren’t generic. Maybe "the problem" was in me. In reality, if my goal was to read this book as quickly as possible and with enough understanding so that I could share its key concepts with someone, ChatGPT did a perfect job.
I realized what was missing.
In reality, ChatGPT was giving me correct answers about all of these books. But what was missing was me. My life, my personal experiences, and everything that makes me the unique individual that I am.
When you are reading a book, you are not just absorbing words written on paper. You are interacting with the writer’s ideas. You are making connections between their ideas and your thoughts, feelings, and experiences.
The point of a book is not to read all the words in it. What every writer hopes for is that their book will resonate with someone. That the reader will be able to not just understand but also feel what the writer was trying to say. And for that to happen, you have to read the book yourself. You have to go into it with everything that is inside of you and that makes you who you are.
ChatGPT didn’t spend most of its life overthinking and being afraid to take action in life. It cannot possibly find the same value in Hagakure that I did. In just a couple of seconds, it could "read" Hagakure more times than I read it in the last 2 years. But a million more reads wouldn’t be enough to find the lessons that were relevant for me. It couldn’t possibly make the connection between samurai anecdotes from a battlefield and the things I’ve been through in my life.
ChatGPT could "read" all of Nietzsche's works in a second. It could write an essay listing Nietzsche’s core ideas in a couple of seconds while I struggle to find my voice in writing in order to give a beginner-friendly introduction to Nietzsche. But ChatGPT never suffered. ChatGPT never saw its loved ones suffer. Chatgpt was never down on its knees, asking itself how to keep going. It can read all of Nietzsche’s works a billion times, but it can never find in them what I did.
ChatGPT was able to give a breakdown of each chapter of "The Art of Loving" in a matter of seconds. But now I know what was missing. Love. What was missing from those summaries was the experience of loving and being loved. ChatGPT never felt love. It never felt the excitement, joy, and ecstasy of love. It also never felt its pain and anguish. It could give me a comprehensive breakdown of everything that was written in the book. But it could never explain what all of that actually meant. What it meant for me. What it meant for you. What it meant for anyone who experienced love.
The main response to criticism of ChatGPT’s insufficiencies is, "This is just the beginning, and this thing is improving fast. Wait till you see how powerful it is in a couple of years."
Sure, that’s true for a lot of the ways it can be used. Probably most of them. But no matter how much it improves and no matter how fast it learns, this is at least one thing where ChatGPT or any other AI tool can never replace you. They can never replace who you are, what you’ve been through, and your ability to connect those things with the experiences and ideas of other unique humans around the world.
Thank you for reading.
If you get value from my work and you are interested in becoming more consistent, consider checking out my new book, “The Art of Showing Up” by clicking here.
Since I was writing this book in public, if you haven’t already you can read the free parts by clicking here.
If you like my Existential Espresso writing, there is zero-cost support in the form of subscribing, liking this post, commenting if you have any thoughts on it, and of course sharing this with anyone who would find it interesting.
Or you can consider becoming a paid supporter of Existential Espresso for 5$ per month. By doing this you would be helping me to keep investing time into researching and writing all the content on the daily basis.
What you get by becoming a paid supporter is access to the locked essays (such as “Why Having a Price on My Head Didn’t Upset Me”or “Why Living With a Bulletproof Vest is The Best Thing to Ever Happen to Me”), as well as an opportunity to recommend topics for future essays.
However, even taking the time out of your day to read what I have to share with you means more to me than you can imagine. Thank you.
Beautifully written. As AI gets better it could begin to converge to “one true answer” but like you said, the part that’s missing is the unique individual, “YOU”
So good to have you back <3