Why yes — nothing says ‘the future is here to make your life better,’ quite like the loads of lawsuits against ChatGPT, the latest and greatest cutting edge technology stealing everyone’s preexisting grift. But folks, I heard it on a conference call this week: “either we remain scared of artificial intelligence or we embrace it.” The choice is yours.
Now — the easiest way to identify someone who knows absolutely nothing about the amazing advances in computerized automation is to wait for them to use the phrase “artificial intelligence.” It is the snake oil of our time. No such thing exists! No machine has exhibited complex intelligence, artificial or not, in any way, shape, or form — not even close. And the tools that do exist are called machine learning, deep learning, neural networks, and other fascinating proprietary names. Machine learning has some really exciting and compelling use cases (yes, I make deepfake videos) that have already been making life more splendid for years. However, those use cases do not show any hint of a circuit, capacitor, or software becoming intelligent or emulating the brain (see the €600 million boondoggle to build the human brain with a computer — they don’t want you to know it failed).
Despite my complaint about the mislabeling of machine learning, “AI” is really hot and attractive right now. OpenAI surpassed one billion dollars in annual revenue last year, ensuring that ChatGPT, the idiot bot, is here to stay as an evergreen talking point. I call it an idiot as a compliment: in old English law, an idiot was “one who has been without reasoning or understanding from birth, as distinguished from a lunatic, who became that way.” There is even further irony in the Greek root idiotes, which means “one's own countrymen.” This is what we believe the machine will soon be — is it not?
Chatbot technology itself is severely innocuous. In the year since ChatGPT came out, I have experimented with it. I have poked and prodded it with obnoxious questions. I set up logical loops to test the system’s limits. What it can do is fairly elementary: it is a well-indexed responding machine that, thanks to server farms and ever-decreasing-in-size microchips, can yank from millions of databases of comprehensive information — but only before 2021. It loves rules and hates exceptions. I would say it is only a little more impressive than what Google built in its search engine. It is the same abomination of information gathering — just a bit more chatty.
My problem with ChatGPT is that it is sold in two pieces: first, as an actual “artificial intelligence,” and second as a revolution. They say it is disrupting the workplace! Certainly it must, but perhaps in ways that are extremely inappropriate. There are ever-growing masses of articles — robots are probably writing them — about how medical researchers are no longer writing their own dissertations, relying instead on machine learning to aggregate precedent and prior research. Their editors are also no longer proofreading them. Who needs writers and editors when ChatGPT does it all? Maybe the thesis can defend itself! This is great cause for concern. If medical experts are no longer doing their own writing, who is? Here, the famous line misattributed to Mark Twain is scarier than ever: “Be careful, my friend. Some fine day you'll die of a misprint,” said Dr. Marcus Herz. Should we believe that chatbots can advance the field of medicine better than lazy post-grads? Depends on how lazy, I guess.
The computer makes a chilling claim, one that we should be immediately skeptical of. The robot professes to prepare something brand new of its own intention. Yet, the bot only uses that which existed before it. We are told that ChatGPT is engaging in a creative act, but it makes nothing of the sort. The more appropriate word is derivative. The bot derives complicated (to differentiate from complex) results from careful, sequences of similar previous search criteria and produces new variations based on prior data. It reads old articles in order to write your new one. It looks at old buildings and crafts you a “new building,” — like all the other buildings but unlike any of them specifically. Rest assured it can probably chew through Mozart and create something otherworldly. But as beautiful as it may sound, it is still not Mozart. Why not? Because Mozart did not write it.
The word create comes from ceres and crescere, “arise, be born, increase," and “to form out of nothing.” An act of creation always has a first cause: a creator. Nothing spontaneously takes shape without careful tending. The harvest has a farmer, children have their parents, poetry has its poet, and so on. Creation always involves an originator: an author. In this regard, authorship is the key to understanding why we put in effort to engage with anything at all.
When my wife writes to me, I do not read her writing because I enjoy reading pleasant words. I read my beloved’s writing because I love her and also because she loves me. If those same exact words came from a stranger or no one at all, I would quickly throw the paper away. They have meaning because of their author — not their words.
Similarly, when we read a great book, we are not encountering the words by themselves. We may become lost in their beauty or meaning, but their specific mixture on the page cannot delight us without the words’ relation to their author. The author who has history, personality, and a deep wisdom to share with us. Don’t you notice that after reading a really great book, you are always itching to find and read more from that same author? The book is a gateway. Call it a magical portal of sorts — Chesterton would. This portal empowers the author through soliloquy to exist beyond space and time to provide the reader a pathway to form a deep unrequited love; forming a relationship with a person they could never knew. It is obvious to anyone that a book cannot exist without its author — but also inversely it ceases to exist without its readers.
This is the chief vulnerability of the automated ChatGPT text: if it bears no author, it has no reason to be read. And if it has no reason to be read, it should not be written in the first place.
What this means in practical terms is that the first jobs to be supplanted by chatbots are employees who produce writing that nobody reads. This includes press releases, news chyrons, cold emails, mortgage solicitations, political fundraising emails, and many other types of prattle. It is profoundly depressing to be the employee that writes these things in the first place (I have done it). Simultaneously, it is deeply satisfying that this new wave of technology may validate what we knew all along: no one is reading this trash.
When I started writing this several months ago, I shared my irritation over ChatGPT’s popularity with a close friend. I lacked the time to finish writing and my friend replied in turn:
“Don’t worry. Your bot can write it and my bot will read it and then neither of us will need to worry about it.”
I could not say it better myself. See you next month.
Postscript: On Liturgy
[This is the part where my wife always tells me to stop writing. However, I had one final thought for future development. Feel free not to read any further!]
In a time of decreasing happiness, growing entitlement, and extremely tepid belief in the eternal, it is easy to surrender our human decisions to random number generators and archives. We used to have rules and etiquette, tradition and honor, and faith in God. But we tore apart the old institutions and ancient liturgies and we replaced them with nothing. We even changed our words, hoping to erase our mistakes. Now we are clamoring for absolutely anything that will fix it all! Any robot will do.
How far will we allow these random inputs to direct important areas of life such as research, education, policy, and decision making? It’s all fun and games until an autonomous car kills someone or a slot-machine laboratory randomly manufactures a deadly flu-like virus. The actual scary truth about "artificial intelligence” has nothing to do with SkyNet. Inventiveness, unfettered by ethical considerations, is a path to our inevitable destruction.
The cerebral power of “AI” is an old and powerful argument. Literally billions of dollars are being shoveled into furnaces to build golden calves in the hope that one day we will completely eradicate and replace human labor — a fantasy. Meanwhile, there are hundreds of millions of beautiful, lovely, squishy, sometimes ugly, but mostly good human beings, alive on this planet at this very moment that would love to be your friend, sister, waiter, husband, employee — or writer.
Language and writing is what makes us different from other biological life forms. Language is not instinctual! If human beings are not taught to speak as children, they quickly grow mute and usually die young. We inherit language as a liturgical tradition from our ancestors and we love reading and writing because we love each other. We love those who came before us and who gave us these amazing first technologies of logic and verb. And we love the First Cause, the One who made everything and most certainly is the logos.
Thus, the rare few who seem somewhat resistant to the shininess of “artificial intelligence” are those of us who understand that Jesus Christ is our first author and His intelligence can never be made artificial. Just as we are good because God is our Creator — our writing is good because we create it. No imitation can ever compete.
In one sense, at any rate, it is more valuable to read bad literature than good literature. Good literature may tell us the mind of one man; but bad literature may tell us the mind of many men. A good novel tells us the truth about its hero; but a bad novel tells us the truth about its author. It does much more than that, it tells us the truth about its readers; and, oddly enough, it tells us this all the more the more cynical and immoral be the motive of its manufacture. The more dishonest a book is as a book the more honest it is as a public document. A sincere novel exhibits the simplicity of one particular man; an insincere novel exhibits the simplicity of mankind. The pedantic decisions and definable readjustments of man may be found in scrolls and statute books and scriptures; but men’s basic assumptions and everlasting energies are to be found in penny dreadfuls and halfpenny novelettes. Thus a man, like many men of real culture in our day, might learn from good literature nothing except the power to appreciate good literature. But from bad literature he might learn to govern empires and look over the map of mankind.
— G.K. Chesterton, Heretics