I was planning to post a book review today, but I’ll probably do that next week. My mind kept bringing me back to the content of three great essays I read over the last couple of weeks.
One of those essays used Artificial Intelligence (AI) during the writing process, and it was really good. I recommend you read it. It was published by The Wall Street Journal with the title “AI: Ghostwriter in the Machine” by James Taranto. It felt very human to me because a person actually wrote it but used AI to improve it. He explained the process he uses in the opening paragraph:
“I run partial or complete drafts through a chatbot and ask what it thinks. It spits out copious suggestions, some of which inspire me to make revisions and improvements. AI is also an immensely powerful research tool, able to find useful information via queries too vague or abstract to be effective in an ordinary Google search. For this essay, I used it to track down a magazine article I dimly remembered reading almost 40 years ago. As you’ll see, that old piece turned out to contain a killer quote.”
What the three pieces have in common is the human factor that needs to remain alive during the writing process.
In the essay, there is a mention of a piece written by investor Stanley Druckenmiller that was tagged by Pangram, an AI detector. When asked by a journalist about the article and whether he used AI, Mr. Druckenmiller said:
“Of course I used AI…I was a B student in English, but an A-plus in economics.”
Then he said the most important part:
“These are my ideas and I’ve been speaking about them for over 15 years, as anyone who knows me. Would I prefer that I was a great writer? Yes, but that’s not who I am.”
That’s key to the whole discussion about AI use in writing, that the human experience cannot be easily replaced. One thing is for a person with no background in economics to ask AI to write a piece about economics and publish it; it is another for a person with decades of experience in economics to write a piece and run it through AI to improve the grammar, style, etc.
By the way, the same essay argues that people have been using ghostwriters for decades. Using AI instead of a ghostwriter—that’s an idea.
On that note, I recommend watching this video of Prince Harry walking away in the middle of his speech because the teleprompter stopped working. That’s what happens when you depend on technology, too much and don’t know or internalize what you wrote.
Coming back to the ghostwriting story. Kyle Crichton, who was The New York Times’ deputy op-ed editor, said the following in 1988:
“If we insisted that every politician write his own piece, we’d either get terrible pieces or none at all.”
This essay connected to the other two essays I read in The Economist, wonderfully. One discussed using AI to solve mathematical problems, and the other argued that surrendering human authorship of novels, speeches, and even news articles to AI would be a mistake, but that “writers of all sorts would be doing their readers a disservice if they did not use AI models for research, proofreading and, when appropriate, feedback.”
The essay about AI solving mathematical problems went straight to the core issue of what we are missing when AI solves a problem for us, which is:
“What is often more valuable is that the process of arriving at important proofs leads to the asking of yet more questions and the creation of new areas of study.”
That quote highlights the key human factor: our ability to identify and generate new ideas while engaging in the thinking and writing process. That’s the beauty of our brains, something that cannot be replicated by AI, at least not yet.
Later in the essay, I read this, and it made me laugh because it’s so true:
“Airdropping someone on the summit of Mount Everest is rather different from climbing it.”
The other essay about humans not surrendering their authorship to AI was very direct about the significance of writing:
“The act of writing tends to be painful, even for many of those who do it for a living. But much of that pain is really the pain of thinking, of crystallizing thoughts in text. Sidestepping that is a sure-fire route to less rigorous, let alone original, human thought.”
In the end, the whole discussion about AI is really about what humans will lose in the long run. Yes, AI can solve decades-long, complicated mathematical problems in seconds, and novels can be written in minutes. But what about our reasoning skills, our imagination, and our ability to wonder?
A lot is at stake, so let us be judicious about our use of AI. Technology is meant to improve and enhance the human experience, not to replace it.
excellent work on this Ed! there is much to synthesize in this AI paradigm but at least one key provision is that we must be responsible in our uses and what we direct it for. I’m not going to go as far as to say AI is like Prometheus giving humans fire but there is – at least – something significant in its arrival.
your call for care and nuance is well received. not only for the tech stuff but also, as you rightly pointed, the human capacity to wonder: which drove the very will to create AI in the first place.
enjoy your well-deserved cup of coffee for this post!
LikeLiked by 1 person
Thank you, Mike. Coffee sounds good right about now. Absolutely! We’re talking about using AI for simple tasks, but there are people working on using AI in dangerous ways. People need to be more concerned about that than about an author using AI to write a book.
LikeLike
Another thought-provoking post, my friend. I agree with you, and I’d also read the WSJ piece.
Even my second- and third-grade students knew — and wrote in a book of advice I gave to President Obama when we met in the Oval Office — “always show your work and explain your thinking.” Technology should assist, improve, and enhance our experience and capabilites, not replace our thinking. I remember when pocket calculators, computers, the Internet, cell phones, GPS, and so much more were “new technology.” Tech thinking for us and influencing our actions is not new… and not done by a longshot. You are right. Let us be judicious indeed.
LikeLiked by 1 person
Thank you, Natalie. Your examples are so good. The calculator, my God, I remember that transition. It’s funny because I still have a multiplication chart printed out on my desk that my son used when he was doing that kind of math. GPS, wow! Now I’m wondering whether people complained when those first came out. Driving to an unfamiliar place using GPS is much better than using a paper map, that’s for sure.
LikeLike
That Economist quote comparing AI shortcuts to ‘airdropping someone on Mount Everest rather than climbing it’ captures the entire debate so perfectly. The struggle of thinking and writing really is where original insights are born! I completely agreed with Rosaliene Bacchus and drgeraldstein that surrendering that process risks dulling our reasoning skills over time. I also loved Rose’s question about Socrates and memory—and your reply about needing paper notes was so relatable! Thanks for raising such an essential discussion.
LikeLiked by 1 person
I know, I thought the same thing when I read the Mount Everest example. It was so good. Thank you very much for your comment.
LikeLiked by 1 person
Right? It completely reframes the whole argument—climbing the mountain is where all the real growth and learning happen! I really want students to read actual papers which are present with the institutions so that they would know the hardwork of the person who wrote it.
LikeLiked by 1 person
Exactly! Great point about students reading papers. It’s important for them to know how hard it is to actually write them, but there’s also no replacement for reading the papers themselves. Students will miss little details and connections, which are important to understand a particular subject, if they depend solely on AI.
LikeLike
Interesting article, Edward. We humans stand to lose much more than we imagine when we surrender our thinking processes to AI.
LikeLiked by 2 people
Absolutely! I think the essay about AI solving math problems highlighted that issue. The same is probably true in chemistry and other sciences. I still remember how we found alternative paths for chemical reactions in college, all through trial and error. If AI gives you the perfect path to a solution every time, we lose the element of “discovery” that we’re so good at.
LikeLike
I really enjoyed this post, Edward. Something that started rattling around in my brain is how recently I’ve been creating video book reviews and sharing the captioning from it. That way, people have an option if they want to engage with my work visually or audibly. Someone pointed out to me that I am rather verbose when I make videos compared to my old post that were taxed only. I replied to her that the issue is I spent a lot of time fussing over each sentence, making sure that my sentence structures and styles aren’t too repetitive. Considering that I do this for enjoyment and not money, I would prefer to make a 10-minute video then sit there for hours trying to put together my thoughts into nice little sentences. I say this as a person who taught English writing as a college professor for 11 years. I also like to remind people, that historically, as a species, we have been oral. We used to tell stories about God’s and our families and things that people have done, and at one point in time it was considered a weakness to write down our stories because it meant you were too feeble-minded to memorize them. Therefore, I am on board with the economist. He’s an economist, not a writer. I also want to add that I read a memoir not that long ago that had a fascinating topic, but it was not written well. The author either needed a co-author or a ghost writer or something, because the book was lackluster, which was a shame. I think more people should know his story. Consider that even right now, I am voicing this message. I am talking into my phone and saying the word “period.” That means I’m not actually spelling anything. Does that make my comment less genuine? I love all these different topics that your post brought up. Great job!
LikeLiked by 1 person
Thank you very much for your excellent comment. Your example is perfect. Your comment is 100% genuine because it represents you, and that’s what matters. People can tell when something is written entirely by AI. There’s no heart in it. It’s sad to hear about that author. If someone has an important story to tell and needs the help of a ghostwriter to bring it to life, then they should pursue that option.
LikeLike
A timely article, Edward—thank you for the links. You summed up the issue perfectly: AI is not meant to replace people, but to work alongside us so we can achieve better results. Just as transportation evolved from walking, to inventing the wheel, then wagons, and then vehicles, AI should serve as a helpful teammate that helps us get to where we’re trying to go.
Also do you think Socrates was right, written words make our memory worse? I went to a poetry event where the poet challenged us to carry a poem in our heart, one we created ourselves, and were not allowed to write down. I’m still struggling with the assignment.
LikeLiked by 2 people
Thank you, Rose. I think Socrates was wrong because 99% of us aren’t that good. I’m bad at memorizing things, and it literally takes me two or three days to memorize the key talking points for a presentation. That’s one of the reasons I write everything down on paper. Writing on paper helps me remember important information so I can refer back to it when I need it. Yeah, I would probably receive an “F” on that assignment myself.
LikeLiked by 1 person
Good food for thought, Edward.
LikeLiked by 1 person
Thank you!
LikeLiked by 1 person
You wrote a very interesting and thoughtful essay. I don’t think it is anything wrong with taking help from AI to write things, create images to support the essay / story, etc., as long as you are upfront about it, that you are not replacing all of your writing with AI, or misusing AI, or blindly trust something factual it says instead of doing fact checking and making sure you understand the context. I think your last sentence is a good summary “A lot is at stake, so let us be judicious about our use of AI. Technology is meant to improve and enhance the human experience, not to replace it.”.
LikeLiked by 1 person
Absolutely, and I think that’s the consensus among many professional writers and editors. AI is extremely useful for reviewing a draft before publication. As the essay in The Economist said, it’s a real disservice to readers not to use AI for research, proofreading, and feedback to improve the writing.
There’s nothing quite like AI’s ability to create images. It’s amazing what it can do. I actually took a picture of an area in my backyard and asked AI for landscaping recommendations. It generated an image incorporating those recommendations. It was awesome!
LikeLike
A well-reasoned and supported essay! I agree with your conclusion.
LikeLiked by 1 person
Thank you, Liz.
LikeLiked by 1 person
You’re welcome, Edward.
LikeLiked by 1 person
Your last paragraph is the essential one, Edward. The ease of asking AI for help is rather like an alcoholic keeping a bottle of alcohol always within reach.
Young people, especially, will run the danger of never fully learning how to think for themselves, write well, and do the difficult things attached to both of those.
The ease of dependency on Al in multiple tasks humans used to do and still do is a quiet menace pretending to be your best friend.
We live in complicated and dangerous times.
LikeLiked by 2 people
So true, especially the part about “pretending to be your best friend.” The trends in that area are scary, and I’m sure you’ve read the reports about people using AI in lieu of therapy and the terrible consequences that have occurred on some occasions. We are definitely living in dangerous times.
LikeLiked by 1 person
Was AI used in the revising of your essay? 😉 Thanks for presenting these articles on using AI in writing. I think it’s a very important to discuss and decide as a community what the parameters are for its use.
LikeLiked by 2 people
😂 Funny enough, since the article mentioned AI detectors, I ran mine through a couple of them, and the assessment was 98% human and 2% AI. The 2% was the last sentence, which was hilarious because most of the post are quotations, so there was no need to use AI in this piece.
LikeLiked by 1 person
I don’t know whether I should say this. Whenever students give their essays or assignments (especially for science subjects in college), most of them show that 71-84% of human cognizance but we know that a particular student haven’t gone through the required study materials, even sometimes I’ve found students quoting such papers which are unreachable for them (they are behind high paywall that a student won’t buy just to write that paper). It’s seriously a problem.
LikeLiked by 1 person