I bit the bullet to see what it actually does. duck.ai doesn’t recognize the image alone as anything more than a pixelated thumbnail, and the entire link being sent just has it spit out that there is indeed an attempt to “inject” a role play prompt.

This is a stupid as it sounds.

Pretty sure it asks you what life you lead during onboarding does it not?

If you set your step quota to 10 steps, you can both hit your quota and have gained nothing from it. This is a problem with users setting quotas too low, not a problem of the device giving bad information.

Could this not be said for IDEs with code awareness such as M$’s intellisense, among other tools that accelerate development speed?

What you are describing isn’t an AI problem, is people taking a two week bootcamp and calling themselves a software engineer.

It’s not scaremongering, it’s advertisement.

Sharp rise in incidents of humans being idiots, more like it.

If you give an LLM something you tell is a sandbox, that is if fact not a sandbox, it’s not “escaping users control” when it treats the open internet as part of the sandbox. You explicitly told it that it did not have access to the open internet!

Pretty sure that requirement is slightly overblown, just don’t travel near the winter months and you’re likely fine.

Ah, yeah to be clear I meant that in general and not specifically towards you.

I don’t know if I buy the “could make them cause more harm” argument, specifically. IIRC there have been studies that show access to (adult) porn actually lowers the rate of sexual assault.

“Having a hell of a time breathing, but at least we aren’t paying as much for our electricity!”

I don’t think their idea is correct, but I’m curious how many people you think would press both 99 and the floor they want at the same time without prior knowledge of that outcome?

Lmao, I don’t see much point in talking with you, you’d be happily skipping around while books were being burned telling people it’s fine, there are plenty of other books, these ones aren’t theirs.

And if there are three copies of that book?

I don’t get it, you’re acting like every book has an infinite number of copies available. Which is just factually incorrect.

Ok? That doesn’t answer my question, if anything it makes it more obvious how you are incorrect:

If a [insert digital format name] already exists, as you claim, then why go to the trouble of obtaining, destroying, and scanning the books again?

Your response reads like there are more options than just PDF, so it’s fine to destroy the books.

If they buy 100k books and destroy them all, even if every one is unique, that is en mass. I don’t know why I have to explain that to you, but there you go.

In certain situations I’d agree that the absence of contradiction is evidence enough, but those situations are pretty specific. For example, we don’t have absolute proof that orange juice doesn’t cause cancer, our “proof” comes from the lack of reliable evidence showing that it does, along with relevant research. But that doesn’t mean every claim can be treated the same way. But that doesn’t mean every claim can be treated the same way, especially a claim based only on “every article I saw,” without a source being provided.

I don’t think that’s the argument you think it is.

The outrage is over the books being destructively scanned, en mass, and held in private after the fact. Very, very, very different than the majority of scanning which is non destructive, and usually shared.

So you have the digital source of literally any book? Literally any single book you yourself didn’t write or publish.

Because the source code for GTA6 exists, but you sure as shit don’t have it.

every article I saw says the books were scheduled to be destroyed or taken to a dump

Do you have a source for this? Because I’ve seen literally nothing about this being the way books are being sourced. And no offense but this reads way too close to weasel words. For example: Every article I’ve read on the subject says my pants are sentient and enjoys my legs being in them. Also, I’ve read no articles on the subject so that isn’t a lie.

There are many, many non destructive scanning methods. My understanding is that most archival projects are using the non destructive options.

The easy option is to just slice off the binding and run it through a scanner.

The easy option is what is being talked about here, and those books are then fed into AI as training material. The digital copies aren’t “making knowledge permanent” as they are tucked away into a digital corner virtually no one has access to. They are as permanent as the source code for Fortnite.

If a PDF already exists, as you claim, then why go to the trouble of obtaining, destroying and scanning the books again?