Readers familiar with the work of horror writer H.P. Lovecraft will no doubt recognize the name “Azathoth.” Within Lovecraft’s pantheon, Azathoth is the oldest and greatest of the gods. He dreams the universe into existence but is condemned to sleep while it exists.
He is a chaotic mass of power and knowledge, and yet he is utterly unaware of his creation. These attributes earn him the moniker “ The Blind Idiot God.”
In recent years, the meteoric rise in potency and capability undergone by Large Language Model artificial intelligence has had a dramatic impact on academics. We may ask Gemini or ChatGPT to write us a summary of Don Quixote, quiz us on the Krebs Cycle, or (in a less savory act) simply request that it complete our Spanish homework. It will probably do a reasonably passable job at any of these.
Despite its utility in terms of schoolwork, there is a reason I will never utilize AI for academic purposes. Namely, I have always seen AI as a sort of “Blind Idiot God.”
Most people know this, but AI does not have a thought process in the same way a human does. It simply draws from the incredibly vast library of human works that it has been fed and predicts what a human author would likely write. The process is unconscious, as AI cannot think independently. It knows only that this word is most likely to follow that one.
Due to the insentient nature of the process, we occasionally receive puzzling outputs which no sane human would venture. It will gleefully invent false information, obsequiously support a blatantly untrue premise, or seem entirely unaware of the most basic facts.
As an earsplitting cacophony of human voices, it contains all the same dishonesty, foolishness and confusion we find in ourselves. A youth pastor once told me, “You can’t crowdsource wisdom,”and that is exactly what AI does.
If you feel the need to use ChatGPT to study for your biology quiz, no judgment here. I only urge caution in dealing with this fickle “Blind Idiot God.”