AI: The Next Generation

Thinking about: Artificial Intelligence

In my previous post, I considered the possibility that we are in the midst of an AI honeymoon. I suggested that although AI is likely to become increasingly powerful in the future, it is currently as useful as it ever will be. This is because in the coming years, AI will reveal its dark side.

This happened with respect to social media. Initially, platforms like MySpace and then Facebook held great promise. They had the power to bring humanity closer together. We could get in touch with old friends and make new friends. Although these platforms accomplished this objective, their dark side subsequently emerged. We started living our lives to impress others with our social media posts. We thereby made ourselves the target of attacks which, because they could be made from behind a cloak of anonymity, were quite vicious and maybe even life-threatening. Social media also made it easy to propagate disinformation and conspiracy theories.

Something similar is happening with respect to AI. It has given rise to disinformation in the form of deepfakes. Those behind AI platforms can, with a few system prompts, use them to propagate disinformation at scale. And not to be forgotten, the quality of the data provided by these platforms is likely to deteriorate as AI starts incorporating its own slop into its large language models (LLMs). My advice: Make use of AI, but do so in a judicious manner and keep in mind that the current AI honeymoon isn’t going to last forever.

It is easy to imagine ways in which the continued development of AI can jeopardize the well-being of humanity. It can lead to economic disruption as an ever-growing number of people lose their jobs to AI chatbots and robots. Even worse, AI-controlled weapon systems could result in innocent people being killed and might even trigger wars. In this post, however, I will focus my attention on a less obvious threat: People can lose their ability to think—more precisely, to engage in the sort of open-minded critical thinking that I advocate. Indeed, it is likely that this is already happening. Allow me to explain.

Novice writers assume that writing a document is a two-step process. In the first step, you think about what you are going to write about, and in the second, you write down your thoughts. If only it were so simple!

You obviously have to think before you write. In particular, you have to know what you are writing about and have a general idea of what you are going to say. If you are a wordsmith, however, this thinking is simply a prelude to the extensive thinking that will be required to produce a thoughtfully crafted document. For one thing, you will take time to consider the exact right words to use, the best sentence and paragraph structure, and how to most effectively reach your audience. In the process of doing this, you might realize that there is a better way to organize your document, requiring you to rewrite it. You might also become aware of exceptions to a generalization you have made or to inconsistencies in the views you are expressing. In the end, the amount of thinking you do in the process of writing a document might vastly exceed the amount of thinking you did beforehand.

Playing the role of writer obviously requires you to develop the ability to write grammatical sentences and properly punctuate them. Playing the role of wordsmith, however, requires you to develop a second ability as well, namely the ability to concentrate for extended periods of time. In much the same way as you can, with effort and self-discipline, develop the stamina necessary to run—or in the case of some people, walk—a mile, you can develop the ability to concentrate on your writing.

For starters, you need to work in an environment that is free of distractions—which probably won’t be the case if your cell phone is on or if you are writing in a public place. And once you are in a suitable writing environment, you must, in the memorable words of Pulitzer Prize-winning author Richard Rhodes, “apply ass to chair”—and keep it there until your document is complete or the time you have set aside for writing has expired.

David McCullough, the Pulitzer Prize-winning historian and author, was clearly a wordsmith. He understood how thought-intensive the writing process is; indeed, he went so far as to say that “Writing is thinking. To write well is to think clearly. That’s why it’s so hard.” Along these lines, author William Zinsser tells us that “Writing is thinking on paper,” Joan Didion tells us that “I write entirely to find out what I’m thinking...,” and Flannery O’Connor reveals that “I write because I don’t know what I think until I read what I say.”

Although I do not regard myself as being in the same literary league as these writers, I am in complete agreement with them regarding the connection between thinking and writing. I am willing to spend the time and energy necessary to write down my beliefs because I want to optimize my mind. I don’t want it filled with half-baked opinions; I want it filled with nuanced beliefs that I am capable of defending. One excellent—albeit labor- and thought-intensive—way to do this is by “documenting” my beliefs, the way I am doing in these substack posts.

Let me end this post—and explain its title—by turning my attention to the next generation of potential Thinkers. Consider, for example, someone who is now in fourth grade. There is a very good chance that as they grow older, they will rely on AI to write on their behalf. In doing so, however, they will deprive themselves of an opportunity to develop both their ability to concentrate on a task and their ability to deal with the problems that arise in the process of carrying it out. These abilities will serve them well in life, even if they don’t grow up to become authors.

It would be one thing if, as AI got “smarter,” we remained as smart as we currently are. A case can be made, though, that as AI gets smarter, we will rely more heavily on it to think on our behalf and will thereby become less willing and able to think for ourselves. The time might come when, with respect to AI, we are like kindergarteners. We will simply accept what we are told—even though what AI tells us is woefully out of touch with reality.

Humanity has such potential. What a sad ending to our story this would be!