I use "crap" in the most mild, generalistic sense here. The primary impetus for this post comes from this video.
The reason that I watched the video was because I'm subscribed to the creator. I like the creator of the video and the point of this post isn't to throw shade. There were some aspects of the video that I agreed with. Still, it motivated me to write this one up. And most of the points in the video strike me as, at best, misguided.
Back in my Clean Harbors days, a recurring theme was that I kept getting tagged to support the beleaguered "team" for the King County Solid Waste contract. I started typing up a bit of a rant as to why I was doing this work, but it felt too much like airing someone else's dirty laundry. Also, it's too much of a tangent. So skip that. There were reasons why. I'm cognizant of them. Doesn't really matter. I was stuck working on these sites a lot.
On one particular day, I was running the Auburn Supermall HHW site with two technicians from a different branch of the company to help me. This was moderately unusual both because usually this particular site was one that our own people staffed and also because the company would usually short-staff the site on days when I was sent there and only give me one other person. Still, this would have been utterly forgettable if not for the fact that one of these technicians had no real experience with the sort of work we were doing, while the other one had considerable experience. So throughout the day, the new guy was effectively getting on-the-job training from both of us. There were a lot of questions, and eventually the conversation turned toward queries on the nature of a "labpack."
Labpacks were, at the time, my primary job function. My actual job title was, I think, field chemist. But I could more specifically have been labeled a "labpacker" (if that wasn't a term more often used for a cardboard box than for a person). Anyway, I was a labpacker. The experienced technician working with me that day was not, but he'd worked on enough sites where he was assisting someone who was labpacking and he'd seen enough of it to have a general idea of the concept. He was also a generally knowledgeable guy, so in cases where labpack staff were new or incompetent, he might actually have known more than they did. This led the conversation to his someone exaggerated claim, "I have an app on my phone that can do your job."
He did not. There is no phone app that can comprehensively provide the information to prepare labpacks, especially not to the degree of expertise I have (or had). There still isn't, and there certainly wasn't back in 2015, which would have been when this little story took place. Even without playing around with the phone app, I knew that this guy was both underestimating the complexity of labpacks on the sorts of job sites where he normally wouldn't be around to see them (university research labs, for instance) and overestimating the capabilities of the app that he'd played around with. It's not that he was stupid: like I said, this was a generally knowledgeable guy. He knew way more about labpacks than the average person. He also didn't know enough to know what he didn't know. He wouldn't have been in a position to stress-test the app. Also, I think that his limited experience with the app misled him as to just how much human input was needed for it to function. So, it was maybe cool software with some potential, but the problem was that anyone with only cursory familiarity with the subject matter would be likely to drastically underestimate the gulf between the capabilities of the software and the hypothetical capabilities that would be needed for it to "do your job." Not only could the app not do my job, it couldn't even come close. This got me wondering, "But what if it could?" And almost immediately, the obvious rejoinder to that occurred to me, "Then I'd lose my job."
I don't know how to make this clear without coming across as protesting too much, but I wasn't worried about this. At this point, I'd probably only relatively recently seen CGPGrey's "Humans Need Not Apply" video. So the general concept of humans losing their jobs due to new technology was relatively fresh in my mind. And almost immediately, I found myself arriving a some corollaries. We started with "Technology cannot replace my function at work, and really it's a long way off from being able to." From there, I moved on to "But what if it could?" followed by "Then I'd lose my job." Next came "And no one would care." Finally, there was "Not even me."
No one thinks that I am somehow owed the right to eke out my existence through my proficiency in labpacking. If I machine could do it better, then every customer would gladly use the machine. I'd have to find some other line of work, and it would be silly of me to think otherwise. If human labpackers were to raise a fuss at being replaced by machines, they'd be the subject of disdain and derision, the laughing stock of the entire hazardous waste industry.
Ever since the hullabaloo regarding "AI art" started, I think back to that day in 2015. I think about it a lot. I've been characterized as "smart." Sometimes, that was due to distorted perspectives on the part of others. But at least I've maybe been mistaken for being "smart" or "intelligent" or "brainy." I've never been mistaken for being "artistically talented." And maybe that's for the best. It sure seems like the so-called "intelligent" kids all eventually get their lessons in humility. Think too highly of your own cognitive abilities and the world will slam you face-first into a reality check. But art is subjective. Once enamored with ones own "creative" powers, how can one be brought to earth?
Perhaps the biggest story of the 2020's will be the gargantuan, protracted hissy fit thrown by spoiled brats who grew up being told that they were "creatives" and that their artistic talents were wonderful. The video I watched today wasn't the first one to blather on about some notion that AI art is soulless or to bemoan the loss of genuine artistic genius in a sea of "AI." No, I could have written most of this in response to numerous other videos or posts I found on the internet. It's a common sentiment. And sure, technically I never asked any of these artists whether they'd care if an app could replace labpack field chemists. But I don't need to. I already know the answer. And no one would care. Not even me.
The people who produce this content won't actually come out and say that of course it's fine if technology replaces other jobs, but that creative jobs are different because they're sacred or something. But they'll dance right up to that line. I guess that's what I like so much about the CGPGrey video. He never does any of that. He contemplates the topic from as many angles as he can. He tries to present it in a balanced way.
Anyway, that's my message to all the "creatives" out there. You are a beautiful and unique snowflake. But also, you know, nobody owes you lunch for that. So suck it up. The mundane commonfolk around you, lacking your magical artistic spark, have been dealing with this stuff the whole time. Welcome to the club.
An aspect of the CircleToons video that I hadn't really noticed in previous screeds against "AI" art was the drastic overestimation of "AI" capabilities in non-artistic matters. It's an echo of that conversation from back in 2015. Artists know art, and they are keen on all the flaws in "AI" art that I might not notice so easily. I haven't looked for examples, but I'd imagine that musicians might similarly deride the obvious-to-them flaws in "AI" musical compositions. And just like that environmental technician knew enough about labpacks to have some idea, but not enough to realize how deficient his phone app really was when it came to its capacity to do my job, it seems that most artists severely overestimate the capabilities of "AI" to write longform content, or accurately provide information on technical topics, or really to do anything outside their sphere of art. The know that the machine is bad at making art, but they assume that it is good at doing other things (it's actually still pretty bad).