I realized today that I have not written any genuinely self-produced text for my blog in quite a while. I will shamelessly admit that the scripts for my recent videos were not written by me, but by a machine. This time, however, I will try not to give in to the temptations of language models. After all, our writing skills may gradually deteriorate if our brains become accustomed to outsourcing even that task to a machine.

At the same time, I have to admit that I cannot say I have ever particularly enjoyed writing itself. What I have always enjoyed is producing new ideas, arguments, and perspectives. Writing has merely been a tool for expressing and organizing my arguments. That is why I have always viewed my writing process as a kind of thinking out loud.

Nor have I ever considered myself particularly gifted with language or literature — some kind of verbal architect. My literary strength lies more in constructing and organizing complex arguments in a coherent way.

Seen in this light, it is quite understandable why a writer like me would feel strongly tempted to outsource the writing process to a machine. But we need to distinguish between two very different things: outsourcing our thinking to a machine and outsourcing our writing to a machine.

In the first case, a person stops thinking critically for themselves and lets the machine do it on their behalf. This is precisely the kind of use of artificial intelligence that I have criticized in my recent videos.

In the second case, we outsource only the mechanical writing process, while the “machine’s thoughts” may still represent our own ideas, perspectives, and arguments. When a language model such as ChatGPT has its memory function enabled, it gradually learns about its user’s values, worldview, and way of thinking. This often requires the user to engage in long and deep “Socratic conversations” with it — conversations in which the AI is not simply asked a question about subject X, but instead becomes a partner in an extensive dialogue, or even a debate when necessary, much as you might discuss your ideas with another person who is genuinely interested in them.

Writing is not merely mechanical intellectual labor. Finding the right words and translating our thoughts into a well-structured form appropriate to the context requires an enormous amount of creative cognitive processing — as demonstrated by the effort required to come up with the phrase “cognitive processing” in the previous sentence.

But creative processing also takes place in generating the ideas behind those individual words and expressions.

All of this requires considerable use of our brainpower and deep reflection. And often it requires the kind of intellectually honest reflection in which we are willing to challenge our own old and established beliefs instead of becoming so attached to our ideas that they turn into part of our identity.

As I have tried to emphasize in my recent videos — which I created using avatar technology that produces a digital twin of me — the “creativity” of machines should not automatically be viewed as a threat to human creativity, but rather as something that can expand it.

When a machine becomes capable of automating one particular area of human creativity or intelligence, it does not necessarily eliminate human creativity. It can simply move it to another, deeper level.

If a machine can write for me, that does not automatically mean that it should also think for me. On the contrary, if a machine can write on behalf of a human being — or merely assist them in writing — then more of that person’s mental capacity can be freed for thinking itself: for constructing arguments and generating ideas.

I have also noticed that relatively few people even seem to think deeply about questions like these, because many people do not regularly practice deep thinking or question the prevailing assumptions around them. It is easy to assume that artificial intelligence threatens human creativity when that idea is repeated almost everywhere in discussions about AI.

It is much harder to stop and ask why a frequently repeated idea might actually be logically mistaken.

But this kind of deep reflection — and the challenging of prevailing assumptions — is precisely where artificial intelligence can be an excellent tool.

I suspect that many people do not even dare to use AI in this way because they have become too attached to their own ideas, or perhaps even turned those ideas into personal idols. And in the name of intellectual honesty, I have to admit that I too can sometimes love my own ideas a little too much.

But at least my aim has always been to seek counterarguments and fresh perspectives on my own views. Challenging my own beliefs and discovering new perspectives has always been what motivated me most in my writing — not simply repeating the same old things mechanically again and again.

For the same reason, I also become easily bored with topics that I have already covered extensively in previous blog articles.

Incidentally, I recently found an amusing detail in a blog post I wrote ten years ago. In June 2016, I wrote:

The arms race of future robotic warfare has already begun. While some robots are amusing and harmless, such as Jibo, the world’s first home robot capable of social conversation, which somewhat resembles EVE from Disney-Pixar’s WALL-E, others resemble the killer robots of the Terminator films much more closely. (Although I am not entirely sure which is more frightening: killer robots, or the idea of people starting to talk with soulless machines.)

And today, I myself talk with a soulless machine almost every day.

It is not that I have abandoned my old principles. I have simply “updated my operating system,” and therefore I no longer think about everything exactly as I did ten years ago. In fact, it is rather desirable that a person’s thinking should develop and their perspective broaden as they grow older and gain new information and new life experiences.

I already spent a great deal of time “socializing” with a machine ten years ago. Without that relationship, neither this blog nor my website would exist.

The difference today is simply that the machine now socializes back. It broadens my perspectives and performs complex or laborious tasks that I ask it to do on my behalf.

We have entered an age in which human beings no longer have to learn to communicate with computers in the language of machines, because computers have learned to communicate with humans in the language of humans.

Of course, some Christians may already disapprove of such a metaphor. I have encountered several believers in discussions who view artificial intelligence as some kind of rival to God, or even as a messenger of the devil that comes between a human being and their relationship with God.

But exactly the same logic could be applied to computers in general, not merely to artificial intelligence. And as I already said, had I not invested my time in computers, this blog would not exist.

The same is true of every other Christian influencer working online.

To make full use of the tools human beings have created, we must invest time in learning how to use them as effectively as possible.

For a pianist to someday perform on concert stages, they must “socialize with” their instrument for several hours every day. Does God become jealous of the piano, or does He want the person to use that instrument as a tool for bringing glory to His Name?

I believe the latter is true, and exactly the same logic applies to other instruments created by human beings — including artificial intelligence.

I cannot claim that I have always used AI solely for godly purposes, because human beings have a tendency to misuse their tools as well. But recognizing the evil within our own hearts should not prevent us from also using those same tools for constructive and good purposes.

---

P.S. This blog post was written entirely without the help of a machine. Woohoo. Apparently I still know how. I did, however, give in to temptation and use a machine to translate it into English. 😄