unmagical

You helping commit a genocide doesn't mean there's not a genocide.

He choking out Smeagol?

There's a difference between getting in cause you love the craft and staying in because of the money and benefits even after your soul's been crushed by the industry. AI is helping some of those devs just postpone becoming carpenters.

I'm a Software Engineer. I don't chose what gets built. I get paid to build what someone else's wants built. I do that so I can eat and stay sheltered.

For the love of all that is good in this world, if you are the developer that thinks this shit is ok, find the tallest building on the planet and jump. There is no reason to make things this screwed up. I hate you just as much as I hate trump, and that is saying something.

There is nothing that this planet would lose if you were gone from it. I am so sick and tired of dealing with shit products.

Until the email is strategically lengthened past the point of my client rendering the thing in full and I have to open it in a new window and scroll all the way to the bottom only to find that they intentionally left "Unsubscribe" as just plain text and not a link.

Or you are lucky enough to have a link, but it demands you login to make changes, but you checked out as a guest and don't have an account.

Or you deliberately unchecked "send me marketing emails" when you bought something and they automatically signed you up for not just their own emails, but those of all their subsidiaries.

Or they just don't have an unsubscribe option at all.

I'd rather just mark it as spam and cause a repeated ping on their reputation every time they send an email.

Cool cool cool. So I don't "own" the games I "buy" therefore if I reproduce them from someone else's copy it isn't "stealing" due to "lost sales," right?

You have got to be kidding me. Why in the world would a character in fiction be beholden to a rule from an unrelated mythology when referencing a different character in fiction. That is incredibly outlandish.

The model has seen images of dragons (likely even multiheaded ones), children, riding, and Mars. Those existing images are what allows it to pick the appropriate patterns and draw an output.

There are significant physiological differences between a child's body and an adult's. If you don't have a collection of naked children to pick from it will be difficult to produce naked children to the extent indistinguishable from actual pictures of abuse. You might can get partway there building off a chain of descriptors based on an adult model, but I don't think you're getting all the way to photorealistic child abuse material without the context trained on child abuse material.

I'm absolutely not willing to test my hypothesis, however, for what I hope are fairly obvious reasons.

It should probably be illegal... Any curtailing of speech cuts the impact of the first amendment. It's something that shouldn't be done lightly.

Free speech and expression should not be curtailed unless absolutely necessary; this goes for joining the Klan to making porn to having died hair. The limits in place for yelling fire in a theater, making credible threats, and producing real csam are there to prevent immediate harm.

With AI generated csam ... no person is harmed in its creation. Consumption of it could lead to harm, if it normalizes the behavior, making it seem alright for someone to harm a real child.

If AI generated csam causes people to - who wouldn't have otherwise - harm children, then it should be added to the list of banned speech. If not, then even though many of us will find it disgusting, it shouldn't be made illegal.

These are all arguments. Non committing doesn't mean you aren't making arguments. I was focusing specially on your 3rd argument as it was the most unfounded in my opinion.

assuming the training model didn't use real csam to begin with

This is a big contingency in your argument, and it's not an assumption we can make.

The stochastic copy and paste machine needs similar input to what it is expected to output. That's why there is an immense effort to collect as much data as possible from everything right now. If the relevant data isn't in the training set then it can't produce a related output.

You can train models on the output of other models, but at some point the original csam was produced and then processed for patterns by the machine. If you just copy those tagged patterns your still basing your output on very real csam.

Historically the US has banned all forms of csam including hand drawn or generated. This is a shift in that paradigm which now opens the flood gates for investigators as it dilutes real cases with facsimiles generated from real cases.

I'm not a free speech absolutist. There are categories of speech that should be protected and categories of speech that should not; and there are actions and products that absolutely should not belong in the "speech" category to begin with. Spending money is not speech, taking and owning pictures is not speech, and generating pictures of children being abused is not speech. There's no reason any of that should be bundled up in our overly broad web of "speech" just cause we want to err on the side of personal liberty over common good and protection of this vulnerable in society.

I can't wait till it's torn down.

Well, he can't have his story of messianic redemption if they have nothing to be redeemed from! /s