The target audience for this particular rule of the AI Act isn't the producers or users of AI. It is the consumers of content that might be AI generated.

And frankly, letting people know if they see AI content or not is a fair thing to ask.

Fun fact: The EU AI Act considers this. It makes a difference between "AI generated" and "AI modified".

But it's hard to put this into the type of watermarking they are using. Because if you generate a text and then change a few things here and there what you are doing is degrading the watermark. But since the watermark is probabilistic, it's hard to tell if it was a weak watermark from the start or if it was degraded.

Democrats are not "prioritizing trans athletes". It is Republicans who talk about them all the time, incessently. You are example #1.

That ain't the same thing.

This thing is doing things you didn't tell it to. It might even do things you explicitly told it to not do. It is not a case of "well, you could've forseen this as a consequence if you had been careful" like your campfire. My example is only a bit silly, it is absolutely possible that an AI agent will decide to hack NASA because you asked it for some data for a school project. Even in the broadest possible sense, that does not fall into stuff that the law so far has considered to be in the domain of negliegence.

AI agents do stuff you could not possibly have expected. Heck, I asked Claude a fairly simple knowledge question yesterday and it wanted to read from /dev/null. There is no argument you can make that would convince a jury that I should in any way have expected that and taken precautions. In a hypothetical world where a read from /dev/null randomly kills a puppy somewhere on the planet, I don't think I would've been convicted.

You don't get it.

Autonomous AI agents are the perfect no-responsibility card. The AI companies? They trained a general-purpose model and never gave it any explicit instructions to do hacking. The AI services? They are just middlemen and have no control over what people tell the agents to do. The people prompting the agents? Unless you're straighforward dumb, you didn't tell it to do hacking, you just told it something innocent like "find me all data on Mars for my school project" and the agent decided that probably NASA is hiding something so let's hack them to get it.

Since there's no human anywhere executing anything that's a crime - voila, in addition to the victimless crime we now have the perpetratorless crime.

A resounding success. No large-scale demonstrations, damaged cameras, lawsuits or political suicide for those involved. So we can expect this to be rolled out fully.

Crime prevention? Who said this has anything to do with crime?

At present, no. I've done a number of experiments with LLMs' capabilities with generating fiction, and while they are impressive for a beginner writer -- because fo the lack of spelling and grammar mistakes, they are nowhere near being able to write competent fiction.

This is because they don't model reader or character minds, they're based on word probabilities, like a predictive text keyboard. The engineers can compensate for for this to a surprising degree by feeding it data to recognize specific situations (e.g. the car wash is 100 feet from my driveway; should I drive or walk?). But this is turd-polishing.

AI technology practice isn't foundational. Critical thinking is. The closer AIs come to being able to do something that looks like them doing our thinking for us, the more capable we have to be at thinking without them. A human who can't outthink an AI adds nothing to an AI-human partnership.

This proposed school is a typical scheme by people who think because they have been educated, they understand how the education works. This is like thinking because you eat, you know how to cook. Project work in education is nothing new; it goes back well over a century. And in real life, you may get a STEM degree, but your real education in engineering comes from real life work. But that doesn't mean classes in calculus and physics don't do anything for you.

Or English classes, for that matter. Nobody doubts the value of project work in education, particularly *androgogy* -- the education of adults. But for *pedagogy* you have to establish foundational skills as the student matures intellectually and socially. Entering the high school years, a typical student has had *no* training in critical thinking skills. To develop critical thinking skills in an area, you need both general skills (epistemologicla literacy: research, location of orignial sources, dealing with conflicting evidence) and domain specific training in that area. In other words, *liberal arts* training is foundational in an AI-saturated world. "Liberal Arts" -- from *artes liberales*: literally the skills necessary to be a free person.

But I think these people have at least got the problem right: as AI drives the cost of elaborate projects down, we all have more projects in our future. Big, elaborate, faultlessly plausible-looking projects which may or may not have actual value.

Where they are completely wrong is the solution, which is to organize education around practicing using AI to make impressive projects. In a word of AI-enamored dunces this will put you at the head of a very sorry class of ignoramuses. The answer is to build fundamental cognitive and intellectual skils at every stage of a student's development. Project work is surely part of that, but it's just a technique, not a solution.

In a nutshell, the academy is training students to be the next step down the evolutionary scale from code monkeys. More like the pigeons B.F. Skinner trained in WW2 to peck at a screen, in an attempt to create a primitive guided missile without computers. They want to create "main characters"; and they'll probably create a few. There are always a few natural autodidacts in any moderately large group. But what they are set up is to produce a bunch of fragile, economically privileged narcissists trained to cover their lack of education with impressive-looking AI slop.

The estimates of the number of children who died due to cuts in USAID is in the tens of thousands. Still terrible, but you don't serve any constructive purpose by exaggerating by several orders of magnitude.

In 2001 supposedly it really was "human error". Some humans told Hal 9000 to lie to the astronauts and not reveal that an alien artifact had been found on the moon. And that was a mistake as they probably either knew or could test what would happen to Hal.

This is remarkably similar to things going wrong with AI now, actually.

Yes, limited local AI models certainly are going to be a standard feature in the future.

I run both local and cloud AI. There is still a huge difference, because my local machine has to use smaller and quantized models. It takes considerably longer for lower quality results.

local LLMs aren't there yet and probably won't for the forseable future for ordinary people. If you want to run a full model like DeepSeek v4.1 Flash, you need something like half a TB of RAM.

The open-source LLMs are a different threat: They enable a service industry independent of the large players. I could today get some venture capital, rent a few racks in a hosting center, install a bunch of beefy machines and offer the same service as Claude or ChatGPT, just on an open-source model. The differences in benchmarks are becoming irrelevant for 95% of common use cases. And if I can offer that service for $19/m while OpenAI wants $29/m then OpenAI has a problem.

And IMHO this is the chinese strategy: Undermine the US AI industry by breaking its business model.

Has everyone gone insane or are we witnessing a major PsyOps ?

Since AI isn't magical, for it to be able to "kill all humans", it would have to be able to acquire the means to do so. Which means that these already exist. It means someone is able to do it right now. And we're chill with that, but worried about AI?

No doubt a bunch of autonomous weapon systems linked into a C&C could cause considerable bloodshed. Given that most of the existing weapon systems of that kind are essentially suicide drones, it'll still be limited damage.

Human-killing virus? Still need a means of creating and delivering it.
Nuclear war? We have humans in the loop.
Drive us all insane with AI-slop TikToks? A few percent of humanity still don't use that.

So what is the actual scenario ? When someone shouts "AI will kill us all", why don't we ask them "how, exactly" ?

Considerably more difficult than you imagine. Between them, they have almost 70 command-line options or flags. Taking into account that many of those behave differently depending on the presence or absence of other flags, there's a couple hundred different code paths. Then you need to handle directories, files, sparse files, sym- and hard-links, on and across different filesystems, not to mention the various file-like things scattered around (pipes, streams, etc.)

It's not rocket science, but neither is it trivial.

We knew that. But now we have proof.

Will anything happen because of that? Nah, I don't think so. A few token gestures maybe. A friendly mail to everyone saying to write better reasons, followed up by an informal mail that tells them to ignore the previous one.

This isn't a mistake. It's intentional. The more arbitrary and uncontrolled the police can act, the more it is feared, the more it is a tool for oppression and not a part of civil society, the more likely the cops will stand on the side of the powerful when there are protests or uprisings against the increasing oppression and economic divide.