You can’t skip learning
Published by marco on
I have some time off and I’ve been working through a backlog of writing that I’ve wanted to copy-edit and finish for a long time. I have hundreds of pages of book citations, half-written book reviews, nearly draft movie reviews, and hundreds of articles in varied fields all “mostly” ready to publish but lacking what I consider to be a final polish.
About five years ago, I partly addressed this by inaugurating my weekly notes, a place where the flurry of writing, notes, thoughts, responses, and ideas would have a home. These articles have the character of “notes” and thus allow me to convince myself that they can slip through onto my published web site with only “light”—or even “no”—editing.
But the other mass remains, much of it having been published in “notes” form but not in “final” form. That’s OK. It’s nice to have a backlog. It’s better than the yawning emptiness of “no idea what to do”, I suppose, though there’s value in that, as well.
Anyway, that’s the context for a thought I had this morning, which is that, should a friend read what I’ve written here, and should this be a casual acquaintance, who doesn’t know me very well, they might wonder why I don’t avail myself of “AI” to help bring these articles over the finish line.
Can you guess? Is it because I hate AI? No, that’s not it at all.[1]
It is because getting the work published in any form isn’t at all the point. The point is for me to go back over what I’ve written, which is a reflection of what a past self has learned. The repetition is the point. The re-learning is the point. The anchoring of that knowledge in the firmament of my own, personal context is the point.
Publication is one step on that journey. The first step was reading the material. The next was writing about it, thinking about it, contending with it, firing ripostes at it. All of this serves to bring this new information further into the web of my existing knowledge. The next step is publication.
Another, possible step is when I find this information, years from now, in a search for a vague memory. Seeing what I’d written—what I’d already experienced three times—lights up those old registers, heats up those old tubes, and brings a section of knowledge online that had lain dormant for lack of use, but will then be quickly ready for use.
This is the work of learning. You’re loading your own mind with knowledge. You’re building your self, your own sense of morality, ethics, and justice.
If the point of my web site were monetization, if I cared more about turning my firehose of thoughts into money rather than wisdom, then of course an “AI” would help me produce reams of content per week.
It would do so, diligently smoothing away all of the rough edges of my writing until I could no longer tell what my voice was really like, until the suggestions of the “AI” seemed naturally better than what I’d written originally.
My interest in ever-more-efficient monetization would carry me naturally and easily toward publishing corrected drafts without even really reading them again. They might even take me to a place where I’d have “AI” summarize what I’d read for me. Or perhaps I’d even get to a place where “AI” would summarize what I’d planned to read or watch or hear, to save me the trouble of doing so.
Now do you see how counterproductive it would be to use an “AI” to “finish” this job? There is no way anyone can help me finish this because getting help defeats the point. The point is not to publish, but to learn.
Learning doesn’t happen in one attempt. It takes repeated layers of learning to finally know something, to nestle it amongst all of the other things you know. Sometimes it accretes gently, dislodging nothing. Sometimes it knocks other things loose, or moves them a bit. Sometimes a whole cliff face comes sliding down. But the idea is to emerge with an amalgamation of what I’ve seen, heard, and experienced, which serves to represent what I’ve learned.
The point of my web site is not to publish; it’s a tool for learning, for growing. An “AI” cannot offer a shortcut, because there is none.
It would be like getting a machine to exercise for you.
I have been accused of hating AI but that’s nonsense. Those accusations come from simpletons who are incapable of accepting a spectrum of nuanced opinion between their unquestioned devotion and faith in their masters’ ability—and desire—to deliver to them tools from on high, and an unthinking refusal to engage with anything new.
“AI” is Ok. I use tools that are useful. If they are not useful, then I try to improve them, or I use them less, or not at all.
↩