Hypersonic ballistics aimed at the Commons

2025-07-18

Imagine this: You aren’t a quantum physicist, the dude in front of you pretends to know everything there is to know about quantum physics (QP). How do you know he is honest? You can’t ask him a question about QP and tell if his answer is correct, you can’t verify them because you don’t know the answer.

Well, the dude seems to know a bunch about QP: he says a lot of things you don’t understand, he seems excited about the same topics as people you know for sure know QP, uses similar words, he can wax lyrical about QP for hours on end, etc. He kinda passes the vibe check. Also, he has a degree.

Ok, but what happens if the dude is pseudonymous? you can’t check his degree. At best, other pseudonymous users in the same online space can say “I like this dude”. But how can you tell the other users are any good at QP, or that they are honest?

Well, aren’t you just left with the vibe check? Does the vibe check work?

To a point, no? The dude is well spoken, his explanations are clear and enlightening. He’s consistent. You can quickly tell if something is easy or not to read. But for consistency, you have to go beyond the words, inspect the implications, the hidden interactions, and find where two different assertions meet, and whether they make sense together.

And what if you find a contradiction? Maybe it’s your misinterpretation, maybe the dude simply didn’t formulate the assertion correctly. In any case, searching for contradiction isn’t the first tool you reach for: it’s very costly in time and attention.

And in truth: people tend to be honest. It’s costly as a human to make shit up, and the potential gains are never that high compared to the potential cost and actual cost.

On the internet though, no one knows that you are a dog. You can only judge someone by their writing. They write a lot, and they write well, that means they spent time on the subject. It’s costly to make shit up. Even more costly to make a lot of shit up. And all of that for what? Some vainglorious fame?

No… And that’s good. That means that you, as a random internet user with no knowledge in QP can just find something someone wrote on the internet and basically trust that they know their subject. After all, why would someone go on the internet and lie? (but like for real)

Until the fire nation LLM slop invaded¹.

Before, you could, as a complete stranger, drop into any phpBB forum and get a vague idea of whom to trust and whom to suspect. And that meant you could learn about anything! QP, pixel art, game design, fanfic authoring, web dev, kernel dev, anything, like really!

You just kinda got the vibes, you could gauge how much energy someone put into a post by how well written it was. Energy spent meant sunk cost, it meant the person was serious and honest. You could use it to evaluate trustworthyness.

But not anymore.

LLMs have been thrust into us. “Thrust” is a very weak word. More like strapped on a billion 3M22 Zircon hypersonic missiles and fired at full speed at the whole world.

Like some sort of voracious mold, colonizing at speed every interface on your machines, chatbot buttons pop up, always just under your cursor when you were about to click. Search results are polluted with complete nonsense that would make Nostradamus frown in dismay. Workers, formerly free from interference, are suddenly told by their boss how to do their job, they must integrate chatbots in their workflow, whatever the cost – notwistanding the cost may be “I can’t do my work anymore” or that the boss doesn’t have the least idea what the job is. Billions, if not trillions of dollars, thrown into the shredder, just to foot the energy bills of those monsters.

It’s an insanely expensive coordinated effort, the like of which we’ve barely seen since the space conquest. Where does the money for this effort come from? It’s money we gave to a very small clique of similar-minded folk. The reasoning was that they were apter at putting it in the correct places than the ones who brought to us washing machines and paid vacations.

They decided to use this money, this influence, to get everybody to rely on chatbot for their work.

Well, but for what? For what!??! screams the entirety of the world. Oh! Great resource allocators, please do tell us what that thing is supposed to do? And they answer: not our job, here it is for free, do whatever, what matters is that you use it.

But what can one do with a machine that only makes shit up?

Remember what I said earlier?

people tend to be honest. It’s costly as a human to make shit up, and the potential gains are never that high compared to the potential cost and actual cost.

It reduced the cost variable of dishonesty!

Now, anyone can make shit up at no cost at all! This means you can’t rely on someone’s attention to spelling, good phrasing, clean explanations, to gauge quickly the quality of a post. You have to sludge a bit (maybe it’s just not a native English speaker) before you figure out you are in a miasma of slope.

It’s annoying I tell you. I see videos of cute cats, I can’t tell whether they are computer-generated or real. I see a picture of a dude holding a possum, the hand looks weird, maybe it’s computer-generated? A critical security issue report? Maybe it’s genuine, but most likely computer-generated. Even if you recognize the slope, you have to engage with it, maybe they were using it in good faith. The fact is, the work you have to do to detect dishonesty is orders of magnitude greater, and reduces drastically your ability to just go on the internet and learn. Or just accept recommendations from strangers.

It’s not sci-fi, it’s Kafka. Or maybe yes, but PKD sci-fi, where reality is as sturdy as margarine.

So the environment changed. But it’s important to note: it wasn’t a fatality, it was a very deliberate choice to develop this specific technology. It was a specific policy to get this choice away from democratic control and into a small clique of people (financial investors). It was a choice to release to the public the technology. It was a choice to frame the technology in this very missleading way. It was a choice to subsidize the usage of the technology (to this day) with hundreds of billions in electricity bills. Currently, investors are threatening companies that do not follow in this death march. It’s still reversible. In fact, it cannot hold long in the current state.

They went and poisoned the well. For what? Not even for their own profit! They are losing money to this day, and with no perspective of recouping.

They saw the Commons, saw it as a threat to their economic models, it doesn’t abid to any alternative, and wanted to destroy it. Not just by releasing LLMs, but by also systematically DDOSing public source code repositories.

They were outraged that anything else than wealth could work as a social marker to distinguish the good from the bad. Shocked at the temerity of an engineer telling them that, no, the laws of physics made their project impossible.

Did they want revenge? After all, what is $100B dollars to a fund manager? They could gift themselves this little treat, it’s not like the lost quantities threatened them in any shape or form. $101B - $100B is still $1B, hundreds of times more than you need your whole life.

Now, we have to clean up behind them. We have to spend time and energy to find ways to work around the damages that their technology inflicted on us. It’s something we didn’t need, and by far. We’ve so many problems at hand, and they keep adding to it.