Technology is great until it isn't
As a science fiction nerd and sometimes SF author, I am anything but a Luddite. I’m not an early adopter, but part of that is because I’m also cheap and I want the prices to drop before I buy in.
It’s also because I’ve seen too many wrong choices get made in early adoption, such as buying Betamax instead of VHS or HD DVD instead of Blu-ray. I wait until there’s a clear winner, all so I don’t have to start over.
The truth is that I love technology, and while I don’t want my entire world dependent upon fragile tech, I think it’s wonderful that we’ve got so many things today that I didn’t even imagine in my youth.
Hell, my entire career is built upon something that was DARPA’s baby when I graduated high school. None of us knew anything about the internet at the time, and now look at us.
But technology has a problem. See, it’s great…right up until it’s not.
Scanners at the grocery store are really fast until the moment they won’t scan the product. Because of how the system is set up, the cashier can’t just put in the price and move along, either, so it takes even longer.
Credit card readers make it so much faster to make a purchase as opposed to the whole putting a card in a thing, putting a slip of several pages with carbon paper in between, and running an impression. It also makes it a lot easier for someone to steal your information and run up thousands of dollars of purchases before it’s caught.
Then we have AI.
Artificial intelligence is actually pretty cool. It’s an interesting way to search for information that might not be easily obtained via a regular search engine. Today, I used it to look at natural pigments for furniture paint that I could find in my area, rather than buying. It’s pretty clear that if I work at it, I can make almost any color I want, to some degree, with things I pull out of the earth around me, and a regular search on Google might not have found it as easily.
I’ve used it to search for all kinds of things, as well as to evaluate the impact of things I’ve written.
What I won’t use it for, though, is to write for me. Why would I? AI can do a lot, but it doesn’t really know what I know. It can’t write what I would write because it’s not me.
And that’s where AI writing, particularly on subjects that require expertise of some sort, will always fall.
I bring this up because recently, I came across a book on Amazon. It was a woodworking book, and it had a cover that was clearly AI-generated. I understand how a lot of people feel about those, but I’ve moved past it myself. I think it’s a tool, and some people let the tool do too much work, but others know how to use it and build from what it gives them.
The problem with this book, though, seems to be that it doesn’t stop with an AI cover. The whole thing is AI. The “author” has a number of books, all AI-written, apparently, and that’s where we get a problem.
Woodworking isn’t new. We, as a species, were working wood before we learned how to knap flint to make stone tools. It was rudimentary, I’m sure, with lashed-together sticks or taking advantage of natural shapes to create mortise and tenon joints, but it was there.
Because of its eons-long history, there’s not necessarily anything new under the sun.
The problem, though, is that experience hits people a little differently, and while millions of people have worked with wood, only a fraction of them recorded their own experiences, which means the idiosyncrasies of the individual craftsman can be uniquely beneficial.
I don’t need an AI chatbot to write a book based on 50 years of public Fine Woodworking articles. I can look at those myself. I want a book that tells me what an individual craftsman does and experiences.
Recently, Mike Pekovich had his third woodworking book published. It was amazing, the best of his three, and it’s amazing because of Pekovich’s unique perspective as a craftsman.
Christopher Schwarz, who is undoubtedly my favorite woodworking author, has his own perspectives that often go against the mainstream consensus in the craft. I may not always follow his lead, but it’s good to know that he’s telling me his experiences. I trust him because he’s not going with the flow.
This is what AI can’t give you, and we live in a world where people are letting it dominate to such an insane degree.
Teachers use it to grade students’ work, but that’s fine because the students are using AI to do the work. No one is doing anything on their own, and the AI is grading its own work.
It’s one thing to use a tool. It’s another to let the tool do everything.
And because the technology is still so relatively new, we still haven’t figured out where the line is exactly.
But it shouldn’t be difficult for Amazon to say, “You know what? People don’t want AI slop, and because of that, it’s running customers away from products we sell; we’re going to delete AI books.”
Especially when, if you really want an AI book, you can just have AI write you a woodworking book of your very own.
See, the issue with AI isn’t when it stops working as designed. It’s when it is, and we, the weakness in any system, start asking it to do things we shouldn’t. It’s when large corporations allow AI to infest areas it shouldn’t be and think of short-term gains instead of the long-term ramifications to its brand.
AI is a useful tool, right up until it’s not.
Unlike the scanner or the tool with a burned-out motor, though, AI doesn’t have to break to be a problem.
And Amazon, my go-to for a lot of things, has continued to disappoint me because of this, coupled with its strong desire to do nothing about knock-off products.
Honestly, if I had the money, I’d build my own site where you can buy pretty much anything, but everything would be vetted both to last and to be repairable. Appliances that would outlast you, even if they need some repairs from time to time. Books that could be rebound in a century or two, if needed. Tools made by people who give a damn and that are meant to work and to last. Clothing that doesn’t dissolve in a few years.
And where books written by an AI chatbot are forbidden for all time.
Unfortunately, I don’t have it, and I’m probably not the ideal person to do this, anyway, so if someone else has the means to give it a try, go for it. I’d like you to mention me when discussing the genesis of the idea, and I wouldn’t cry about a permanent discount code or something; the truth is that I’m not holding out for any of that. I just want to see technology used to make lives better, not make things harder than they used to be.
Technology always works great, right up until it doesn’t.
AI is part of that. It’s the scariest part because it doesn’t have to break to not work great.
Tilting at Windmills is 100% reader-supported. If you enjoyed this article, please consider upgrading to a paid subscription for 15% off the first year or making a one-time donation here. Either would be greatly appreciated at a time when I know good and well I’m about to have some serious medical expenses coming.
"And where books written by an AI chatbot are forbidden for all time."
And all you have to do is devise a 100% reliable detector of "books written by an AI chatbot."
Spoiler: You can't. Because that AI chatbot is trained by reading all the published works in a given genre, and assigning probabilities that a given sequence of words will be used. The human authors who have presumably read that genre can likely pick the same ones.
We, as a species, are churning out children like little bots that cannot think for themselves and with the advent of AI, we're encouraging them, not just to ignore critical thought, but letting machine learning do *all* thinking for them. It will eventually destroy us if nothing else does.
We have to remember the reason we crawled out of caves and overcame predators as the top of the food chain was because we could reason and had opposable thumbs. That's all. We weren't the strongest; we didn't have the most stamina and our instincts were less than the saber tooth tigers eating us. Our only advantage was reason and now we're teaching students they don't need to do that anymore, because AI will (ignoring the fact it just makes things up occasionally). One day, possibly centuries from now or even just decades, a solar flare, a nuclear attack or even just some idiot pushing the wrong button, will see our species thrown back into a time before AI and it will see us either learning on curve so steep, we nearly lose our dominance over our planet or we'll be the last of Homo Sapiens Sapiens and we might even deserve it.