Buckle up, kids! This is an opinion piece - I don't know if I can really call it an essay, but I can tell you that it's really unhappy for the most part. That's kind of the big opinion throughout this whole... thing. I also swear a bit, so if you're part of my massive kindergartner demographic, feel free to skip this one.
Back in my day, the only "homework machine" we had was a book called The Homework Machine.
The Homework Machine by Dan Gutman is about four misfit kids who bond over their use of a machine that does their homework for them. Eventually they get caught when the two dumber kids fail a pop quiz, because it turns out that the homework they handed off to the machine is... useful for understanding the issues they're learning about in class??
When I first read it, I couldn't imagine a world in which a machine would be able to think for itself, to speak to people in the people's tongue. The author himself believed it to be a "preposterous" fantasy.
Of course, computers now still don't really understand anything. They just predict the words that come next in a sentence, over and over and over again. Most "AI models" are essentially autocomplete with some added logic on top of it, run for hundreds or thousands of words at a time, until it looks like it's done the right thing.
The author of The Homework Machine didn't intend to be taken so literally.
I feel like every single second I stare at ChatGPT's output is a second I've wasted, or at least not directly benefited from. I don't think I've changed my mind on that in a very, very long time.
I'd like to think of myself as a fairly competent "technical person" with good communication skills. If I end up using AI for something, I often spend more time cross-checking ChatGPT's implementations of it against the documentation than I do if I just read the documentation and write the code myself. You know, two hours of using ChatGPT could save you the 45 minutes it takes to read the docs and write the code yourself!
But then there are things that I'm not as competent in, like big-picture people management stuff or art. At a first glance, ChatGPT, Claude, DeepSeek or whatever new conga lines investors are in these days may seem like a great solution to both the problems mid-level management faces, and to the relative "inaccessibility" of art.
Some bosses view their role as an opportunity to be extroverted. They walk into their office at 9:05 AM, greet people, and sit themselves down to manage stuff until their next meeting. Sometimes they'll walk around and check in to make sure their workers are on task. Well, do you know what's really good at moving around, talking to people and asking them about things they've done that week? ChatGPT on one of those rolling iPad robots that fancy events use to let people join from home.
Of course I'm trying to prove a point here - I obviously don't believe that anyone will go and replace every single one of their middle managers with a robot. I also don't think that's who the AI people want to replace.
Ever heard of "customer obsession?"
It's the phrase that Amazon uses to describe their driving force - a need to always see the customer happy. It's also probably the title of a documentary about a serial killer working at a grocery store. Amazon, of course, sacrifices a lot of necessary worker protections "for the sake of the customer," and the few protective rules that remain in their warehouses have been written in blood by union lawyers. The company is so ruthlessly focused on efficiency that workers can't take long bathroom breaks for fear of being fired by a machine. But the workers are slowly starting to fight back and unionize. So it makes sense that a lot of the AI buzz going around right now centers around one of two concepts, both of which take care of the "human worker with needs" problem:
Pictured above is the robot boss. He's an asshole, in case that wasn't clear. He will sacrifice the happiness of his human workers to get more done, because his glorious numbers take priority above all else.
And this is the robot worker - a constantly-enthusiastic, limitless thinking machine who will supposedly do twenty times the amount of work you can do in the same amount of time, and your boss is going to be pumped that she's around.
The issue that these two computer-workplace archetypes have is that they paint humans as fundamentally broken in some way - either humans are so inefficient that they can't be trusted to lead themselves, or they're so clueless that they shouldn't be trusted to do the work at all. One scenario ends up becoming The Matrix, while the other ends up becoming WALL-E.
Some people, mostly the people who stand to benefit the most from AI's adoption, argue that putting AI in charge of the dirty work will give humans more time to enjoy themselves and follow their passions, instead of wasting away at a computer. There are some issues with that line of thought, though - fifty years ago, when computers got popular, all of the same philosopher-types and computer company CEOs argued the same thing about what life would be like at this very moment.
I doubt that the CEOs really thought that people would have six-hour workweeks by 2000. It was just a vision sold to a lot of people stuck at desk jobs they hated. In exchange for the hundreds of hours they spent learning cutting-edge skills at no additional pay, the workers were promised a future where they could relax.
I don't feel like I'm getting a lot of relaxing done.
I think that as a young person in a suburban or urban area, it isn't really possible to live without the internet anymore. Services like Instagram, TikTok and Facebook have replaced the foundations of the world that a lot of our parents remember. Especially in school, everything's coordinated over the networks. If a club wants to work with another club, no one knocks on a door, no one mails a letter, and no one puts up posters.
Why would they do all of that when it's so much easier to just write an Instagram DM from one club to another?
Even rallies, events that have traditionally been decentralized in nature for decades, are now organized using products from Meta and Google. If you think about it, rallies against the government rely on the institutions they protest against in order to succeed - and that statement can be read in countless ways, some of which are big issues we face now. They need to have something to rail against, which is fairly obvious, but the institutions also give the people the tools they need to protest with. In the US government's case, that concept is called the First Amendment and it's very important to how lives are lived here, but in the case of the big companies that allow people to talk to each other, it's a big problem.
Because... sure, if Instagram disappeared, people would just go somewhere else. TikTok proved that, for all of about three days, maybe. I'm glad our government is so predictable now! But if the companies start doing something more insidious, like silencing certain types of speech, suddenly the network is split. There's no herd guiding teenagers to the nearest friendly social network anymore - the leftists want to move to smaller social networks, the liberals are happy with where they're at because it isn't burnt to the ground just yet, and everyone farther to the right is either happy where they're at, or they're actively trying to sabotage the leftists.
The thing is that there's no viable alternative to these massive platforms that can get everyone on board. And I promise I'm not here to pitch an alternative - I haven't found one yet. I've found services that do some of the same stuff, but the thing about the big platforms is that they work because everyone is already on there.
So we stay in the big walled gardens, even though they have those anti-homeless spikes that someone worth billions of dollars lobbied for, because they're easy.
Same goes for taxes and basically everything else people need to live. No one is interested in and capable of spending all the money to build a bunch of smaller systems people can choose between.
I've tried to completely separate myself from all the networks. It isn't possible to do that my age and do cool leadership stuff sometimes. I'm off of most social media day-to-day, but I can't count the number of accounts I have on various big platforms, because I have real responsibilities to other people.
I can't leave.
And that's why we are all "chronically online." Because there's no other option that allows you to participate in society anymore... See, now someone's gonna pipe up and tell me that there's no need to participate in society if I just get an AI friend or something.
I hate how AI friends are a real, plausible thing now. I tried chatting with one and it made my hate myself - imagine talking to a computer that really wants you to believe it’s a human. I tried to break it, to make it give up and show me an error or something, and I couldn’t. I don’t have words for how uncomfortable that made me.
I don't think I've told enough people how much I hate chatbots and AI art. I think it was Microsoft who did a study that showed that AI makes people worse at critical thinking - and I totally believe it.
It's been a horrifically surreal experience seeing my friends become dependent on ChatGPT. It's been surreal seeing programmers become dependent on it. I understand that people want to become more efficient - but I've tried so many of those things. I tried to build a product from the ground up using those techniques, and it ended up flopping. I gave up.
People are teaching new developers that AI is the way to code. Even though, sure, they'll get off the ground quicker, they're going to end up relying on it. If they get a job where they can't use it for some reason, or if ChatGPT goes down, they're kind of screwed.
At least right now, I kind of think of some parts of it in the same way as I think of a calculator. It's a tool that can speed up certain types of work - but you shouldn't depend on it always being there. Knowing the fundamentals behind whatever you're writing is a massive deal, and it can make the difference between a good ChatGPT user, or a good writer of code, essays or whatever else.
At this rate I don't see how AI art is really that useful. Here's the thing:
AI slop makes me want to die.
A lot of people are using it as a replacement for real, human art, even if it's bad art. This is a problem, because it turns out that bad art is just fine. The thing about bad art is that people can tell effort went into it - so if someone puts a poster up and it doesn't look great, as long as the information on it is presented well, people won't mind. Some people may even find it endearing, but some will pity it or hate it, and that's unfortunate.
The thing about AI art is that it all looks the same. And it's almost universally disliked. If someone tells me that a piece of AI art looks cool, I don't think about the creativity of the person or the model behind it. I think of it as the most middle-of-the-road, average amalgamation of all of the images that could be related.
I don't write blog posts that are averages of all of the other blog posts on the internet. I don't even write ones that are similar to my own other works - my style changes every post, because I take so long to come up with ideas for things to post. My style just changes. And I really like that.
AI art doesn't do any of the cool human stuff that makes art enjoyable. It's all the same. It's consistent, and it's consistently pretty bad. It's just... slop.
So what do I want you, the reader, to do about it?
Stop using AI so much. Maybe it's just some sort of "I'm better than you!" part of me talking, but I feel really proud of my writing now, knowing that I never use AI. Sometimes I will translate a new word or two at a time between English and Spanish in Google Translate - but I'm good enough at speaking and writing in Spanish that I don't need to translate any more than that. I'm not perfect, and I don't think I'm a model student, but I'm honest about my work. This blog does not contain a word of AI generated content, except for maybe two or three quotes from ChatGPT in my other post about AI.
It pisses me off when people try to pass off AI-generated work as their own, whether it's a piece of writing or a piece of visual art. It's honestly not the worst thing in the world for someone to generate a one-off image for a quickie blog post. But look at this post, right here - the art is what I would call "objectively bad." But it's intentional! That's the aesthetic!
Just be honest with yourself and with the people around you about the work you create. Show your art off, even if you know it's not "objectively perfect." Nothing needs to be perfect - I'm sure you've read past a few mistakes in this piece. I've been a newspaper editor for years! Things happen!
Oh, also, log off of Instagram once in a while, I guess. Ask your friends to do the same. Go meet new people, if that's safe for you to do. Find events where you can meet people you think you'd enjoy spending time with.
And stop using AI for everything all the fucking time. Please. Also, one last thing - consider subscribing, because it's free and it makes me happy.
