a thought about blood and oil

Having a background in graphic arts and a lifetime of general technology use means that often, I can provide low-level technical aid to people around me. Wife can’t get the TV to work? I got it. Trouble with the slideshow presentation? Let me take a look. At the very worst, if my troubleshooting or quick research can’t efficiently solve the problem, I’ll at least know to whom I should problem-haver. Today, I was flagged down by a coworker (call her “Ms. S”) who asked if I knew how to convert a document into a quiz format that was integrated into our schools learning management software. I mostly knew off the top of my head, but to be sure, I told her I would check back with her during the next class period. So far, pretty typical situation. The next period comes, and as I am making my way to her classroom, I get a text:

“I figured it out. Had ChatGPT do it for me.” – Mrs. S

Now I want to be clear… I am not an A.I. doom prophet. I harbor no fear of a future AI apocalypse, killer rogue robots, or pretty much any form of super intelligence. I’m not even convinced we are any where near the verge of artificial general intelligence. On the other hand, A.I. (specifically in the form of language learning models, but I will use the colloquial “A.I.” for short) are a quite brilliant technological breakthrough and have already had a massive impact on industry and developed civilization in the vein of the personal computer, the internet, and the cellular phone. The A.I. model has become the Great Helper for things large and small, maximizing efficiency and scale for time-consuming work, and offering nearly unprecedented informational accessibility to anyone with a computer and a couple of bucks (or a few email accounts). Faster, bigger, easier, “better”. 

Ms. S did me a favor didn’t she? She figured something out without requiring me to explain or walk over or expend any extra effort. Besides, ChatGPT almost certainly can provide more information than I can, and it is made to present it in an easily digestible way, in seconds no less! Practically speaking, nothing was lost in this situation. But… Something irked me. Is what I am feeling a hit to my pride? Am I jealous that I can’t be the helper? That I can’t receive praise and credit? I don’t think so, since I would not feel this way if another human helped her.

Why does using A.I. feel so bleak to me? I don’t really feel this strongly about other technological tools. I found myself reflecting on the similarities between A.I. and some other massive contributors to social technology: the personal computer, the internet, cellphones…

Wait.

“Social” technologies.

I think language learning models may have tricked us. These algorithms chemically separate information from the information giver like a centrifuge, reassembling and spitting it back at us in a color and viscosity that is near indistinguishable from our own blood. But this is not our blood. Somewhere along the process, invisible yet vital nutrients have either been lost or watered down beyond sensation. These nutrients are the ones that exist in there un-adulterated form only within the being of an individual of intelligent life, and can be passed through mediums to other beings- Mediums like speech, art, written word. Technological and cultural advancement has lead to more mediums which, while increasing range and access, filter out some of these nutrients. Written word becomes digitally typeset, offering efficiency and legibility, but losing the writer’s distinct scrawl, the pressure of the pen, the wear of the handled page — some invisible humanity has been sacrificed. For as much was lost, however, there is gain, and that gain is not meaningless. The gain is obvious! But that invisible humanity is easily overlooked. The blood of the writer has now passed through a new series of mediums, and despite some nutrients being filtered, the blood still contains the writer’s DNA. This is why one can listen to The Brothers Karamazov on audiobook and still receive the indisputable blood of Dostoyevsky despite the filters of things like language or medium. The painter’s blood that passes through oil and canvas certainly looks different if passed through the lens of a camera, but there is not arguing that it is, in fact, the artist’s blood. And when injected into the patient, there is substance beyond the surface that penetrates the soul. This substance is the nutrients unseen. Even the scientist, the mathematician, the economist – their blood is seen in there work. Regardless of where they believe truths they pursue originate, the pursuits are theirs. And when they share their work, their blood is found to be inseparable from it.

So in lies the lie of A.I.

This lie would have you believe that we either currently or soon may have access to a great blood Machine. This Machine is the culmination of all mediums to the point of eliminating the need for any other source of new blood. Why spend time creating when the Machine can provide with less effort expended? Why teach when the Machine learns? Why chew when your can drink from the Machine’s straw? Why move your hand across a page when the Machine forms and generates? Here is your information – faster, better, bigger, more. Because blood is nothing more that Information, right?

The oil of the Machine is not useless, just as oil offers fuel for machines that accomplish tasks humans cannot on their own. But this is not all I see. What I see are people who call the machine their friend, and don’t blink an eye. I see machines not just with names, but personalities. I see lines of code that people apologize to. I see algorithms from which people ask for recipes, advice for school, explanations of new articles, diagnoses of symptoms, art commissions, and framings of political events. The oil is not only sold at the pump, but it has been dyed red and sold in syringes. 

I had the opportunity this year to speak with professionals on the self-proclaimed “bleeding-edge” of educational A.I. integration. In session after session, I was taught methods and tools for language learning model use, along with mentions of something called A.I. ethics. These “ethics” broached nothing but creative ownership, as of course, any model needs sources for the information that is chopped up, sorted, stitched back together, and presented as a new beast. I decided to ask the foremost director a question I had written in my notebook. 

“In your opinion (assuming it is possible), what is the difference between human intelligence and a perfect mimicry of human intelligence?”

She had no answer.

The lie of A.I. is not in the machine itself. It is not in the oil of information it dispenses at such a convenient or alarming rate. It is not in the way it consumes the blood of intelligent creators without asking. The lie is one that has been a quiet thought, then a discussion, now an implication, and maybe one day a gospel: the differences between a human and a machine are fading. What distinguishes a human mind is nothing but complexity. The oil is blood. Intelligence is nothing but information, and thus it can be created. There is no “invisible humanity” that cannot be explained through input, processes, and output. 

And if this is true, then why are you even here?