Artificial Intelligence and Giving Up on Life

Well, it’s been a while, but I am feeling the need to return to my blogging ways, if only to vent my spleen a bit.

Living in the year 2026 is to be assailed by countless stories of gloom and doom day in and day out. Hell, our idiotic and evil president just launched a war with Iran out of nowhere! The horrors are relentless. But there’s one thing that has been wearing on my last nerve for years at this point: so-called “artificial intelligence”.

First, let’s have a little definition of terms: when I talk about AI, I’m specifically talking about the modern fad for so-called “generative” AI, which are basically all LLMs, large language models. The way they works has been steeped in mystery and obfuscated with fancy new ideas like “tasks” and whatnot, but in general they operate by predicting what would be the best response to a prompt, based on various types of data.

Now, these models require reams and reams of data to be useful at all, especially the ones that generate images or even video. So, naturally, the companies developing them simply stole all the material they could get their grubby little hands on from all over the internet. This is obviously bullshit and should be prosecuted, but it isn’t really my primary problem with them. Copyright is… a shaky idea, in the age of the internet, at best.

Like, I’ve sailed the high seas of the web myself, in my day. The fact that you can copy things and leave the original intact changes the whole calculus. Obviously artists deserve to have a living, but that just gets back to the fact that capitalism as a system should be destroyed if we want humanity to flourish. Most of my political beliefs end up straying back in that direction, for some weird reason.

Anyway, the thing that drives me mad with anger regarding modern AI is more… philosophical, but it has a practical aspect as well. To me, using it feels like giving up on life. It is also being used in more sinister ways, to destroy livelihoods and crush the power of labor in society, but even if that weren’t true and we lived in a communistic paradise, I’d still say that it should be outlawed, for the good of humanity.

Breaking the Chain

You may have heard of the famous novel Moby Dick; or, the Whale. I’ve written about it in the past, and one of the main messages of it, in my analysis, is that the tragic lot of whalers is made invisible to those who benefit from their labor.

All that dangerous and painstaking work, the years spent at sea, the men dragged to their watery graves (not just due to the prophecies and madness, but in the course of ordinary business), it’s all completely ignored by both the people who profit from whale oil and the people who use it. They go to the shop and get a little tin to light their lamp and don’t think twice about it.

In some sense, the book serves to elevate this humble profession, and bring these whalers’ plight some small spotlight. But this issue, the idea that the labor behind a product is invisible, is still with us to this very day, and AI perhaps represents a new and insidious form of it.

After all, the idea is that the “intelligence” in question is completely artificial, it’s just the computer generating whatever it is that you’re looking at. You, the user, are asked to elide both the physical realities that enable this, like data centers that may pollute and destroy communities anywhere in the world, and the labor that was used in the creation of the content itself.

After all, every LLM is trained on countless millions of hours of human labor. Almost all of it entirely uncompensated, just shoveled into the slop machine to produce your improved emails or funny meme pictures of Shrek or what-have-you.

AI relies on speed and ease to compltely obfuscate these facts. It all happens right there in your browser or on your phone, in some slick interface. You never have to reckon with anything going on behind the scenes, and it is entirely willing to create a whole solipsistic universe for you, where you never need to interact with other humans at all.

This, I think, is the greatest risk of this technology. That people will, and probably already are, using it as a substitute for everything. As a sort of great filter for the world outside the walls of their home, the new first layer of their brain. They could become entirely dependent on it to form their view of the world, and these models are easily bent and manipulated, as the clumsy flailings of Elon Musk with Grok have demonstrated.

There is a great chain of human ingenuity and artistry that surrounds you at all times. Look at any random object near you right now, and unless it’s a wild animal or plant, then it bears the marks of human care. Your clothes were sewn by human hands, your chair designed by some random designed, meetings were had about the precise shape and tolerances of the lamp on your desk.

What AI purports to do is break that chain. It doesn’t, but it claims to. It seeks to draw a great veil between you and all the millions of lives before you that created the world you live in now. It is an act of profound disrespect.

Giving up

Perhaps the act of creation has become so rote for humanity that some have grown tired of it, and don’t want to bother with it anymore.

This is the thing that galls me on a personal level. It’s the business-brained aspect of it. The thing that is like… people automating their lives away. Deciding that working on things, engaging with the world, is simply too hard. They’re going to leave it up to the machine.

On one level, like… I get it. I do. An LLM is like a magic trick, it can be dazzling, seeming to produce something from nothing before your very eyes. It is similar to what one can accomplish with some knowledge of coding or just… how computers work, really. Which is something that brought me to my current career!

The basic joy of interacting with something on a screen… it captures some shadow of that, I think. The idea of writing prompts as a skill, figuring out the magic words to make the computer do your bidding, is adjacent to real coding, but somehow a lesser act for its imprecision and lack of direct feedback.

The thing that really gives me joy is fiddling with things, which you cannot effectively do with these AI systems. They’re not reliable, they do not learn, indeed they don’t actually know anything, in the conventional sense of the world. They have no personhood, no memory, no personality, only the average of their training set.

But moreover, it’s been demonstrated that excessively using AI for any sort of coding work degrades your skills, and makes you slower, not faster. Cutting out the work of typing or at least integrating code blocks into your existing work reduces the friction, but that also reduces your inherent understanding.

Much of the world of coding is not composing code, but understanding the building blocks of the system you are assembling. Given that, it’s easy to assume that you can just automate away the tedious parts and get the same benefit. But I think with AI coding agents, we’ve reached an important point in the continuity of coding tools.

Writing binary was easier than hand crafting punch cards. Assembly was easier than binary. Low-level languages were easier than assembly. High-level languages were easier than low-level languages. The further you get from the metal, so to speak, the more abstract the work that you do is. One could argue, and it has been argued to me, that this is simply the next step in that journey. No longer need one write individual commands, just tell the machine what you want to happen and it will do the hard work.

In reality, though, I think this seriously reduces the capacity to hold these systems in your mind. Already there has been a degradation of performance to an extreme level in modern computer programs, just in general. You see it all over the place, since the power of our machines has vastly outpaced our ability to utilize them effectively.

This is the old saw that NASA sent men to the moon on some insanely small about of RAM, but a modern browser will eat up 10,000,000,000 times as much to display a simple web page. Writing efficient programs is no longer a useful skill… because it literally doesn’t matter. You can overcome any deficiencies with sheer raw power.

This problem is only going to get worse, until hardware stops improving. Which… is a point we may have hit, actually. I believe the use of AI for code generation will only exacerbate it, and we’ll see some truly disastrous performance and a new age of horrible bugs and unusable programs and websites.

Everything AI does is like this. It’s not a real replacement, but it has the aspect of the magic trick, that it does it for free and zero effort or labor, and that it’s good enough. It’s a way of giving up on the world, and giving up on yourself, while you satisfy your own ego.

Fantasy vs Reality

Not even going to touch on agentic AI, I wouldn’t touch that with a ten-foot pole. Give some hallucinating slop machine access to my computer? Absolutely not. Anyone doing so is completely deranged.

It comes down to the same thing I mentioned above: it’s a magic trick. When it works, it seems great! You don’t have to do tiresome things like uh… setting alarms or… buying things on online stores, yourself anymore. You can leave it all to the machine and kick back.

But like… I am an inveterate computer-toucher. Have been since I was a young child. Maybe that’s why this particular thing holds no appeal to me, I already have this fantasy of ease. Doing these things is already basically no effort, and I don’t do it, so why would a magic AI assistant make any difference?

There is simply not a large array of tasks in the sweet spot for me, where tasks are too difficult or complicated to do, but I still wish them done, and would rely on a manifestly unreliable machine to do them for me. I’d sooner hire a real human assistant to help manage my life than rely on one of these things.

Ah, but that would require more money! And this is where the rubber meets the road: what AI is selling is the fantasy of being wealthy. It offers to make you The Boss! The person ordering others around, that has people fawning over you and deferring to your judgment. You can have that rush of power, the sensation of having others do your bidding, in a private and safe way.

The privacy is really important. For all the people showing their asses on social media over leaving their lives up to AI and reaping disastrous consequences, I’m sure there are many more quietly doing so and just… living with it. After all, you can talk to it, and it will tell you everything is fine, and nothing you’re doing is wrong. It is a self-justifying bubble, in that way.

The Apotheosis of Capitalism

In many ways, generative AI is a sort of apotheosis of capitalism. A perfect machine with tons of horrendous externalities, which twists the financial system and whole supply chains around it, and its only output is the destruction of art and craft and the flattering of ego.

It sucks up resources in return for nothing but itself. It has solved no great problems, it has not improved the human condition one bit. It gives the illusion of control and power, but all you have power over is a shard of a machine that has no mind or will of its own. It replaces the vibrancy of life and human connection with the cold, dead mush of a machine that was built by people who hate you.

Ask yourself this: if AI is really so powerful, so dangerous, so amazing, why are the companies that make it always asking for money? Why do they not simply use their own product to generate amazing new ideas and products to sell to the world? Why are they selling shovels and not digging up the gold themselves?

Because it is fool’s gold, my friends. LLMs will never be AGI, because that’s simply not how intelligence works. It doesn’t improve work efficiency. The best you can hope for is to sell it to some other poor sap or fall into a solipsistic nightmare where it may or may not convince you to kill yourself and others.

I Simply Don’t Want To

Okay, that was a lot of high-falutin’ philosophizing and vague arguing, but what it comes down to is: I hate this shit. I think it sucks. I don’t wanna use it, and I’m tired of hearing about it. It’s actively making the world worse in many very tangible ways, which will get worse before they get better.

Recently, I’ve been getting into the hobby of Warhammer 40,000. It’s a war game played with miniatures, set in the far distant future. It has all kinds of fiddly rules, and the models have to be glued together and painted by hand.

If I could simply go on a computer and say “assemble and paint my models for me, also play the game for me”, I would not be engaging with the hobby at all. Asking a computer to do all the work defeats the entire purpose of it. It’s a kind of work, doing those things, but that’s also what the thing is. Work is not a pure drudgery, it is not nothing but pain, it is engaging with the world itself.

Human skill and intelligence are an infinite fount of wonders! Why would I want to stifle my own creativity by automating these parts of the process?

I feel the same way about writing emails. It’s a skill. Can it be hard sometimes? Sure, but that’s how you develop a skill. Every little thing you give up comes with a cost.

The essence of life is struggle! Struggle on, o noble strugglers!!

---

Ahhhhhhhhh, man it felt good to write all that.

I tried to form a more cohesive argument at first, but this whole topic gets me too worked up. It’s all over the news all the time and on social media, it’s destroying everything. Can’t even get a damn computer these days, RAM and SSD prices are all jacked up because of nonexistent data centers. Ugh.

I decided to illustrate this post with PC-98 screenshots, it felt appropriate. I want to demonstrate that like… I am not opposed to The Computer. It is my friend, and has been for a long time. I think it is a medium for human expression. But there is a real dividing line between past innovations and this new world of slop.

Also they just look nice.

Until next time, shipmates!