And this isn't even getting into how the "privilege" we're talking about is getting categorized by society as ultra-gay failed men who failed so bad we should let them become women. "You are privileged because you are actually an ultra-gay failed man rather than having to pretend to be one" ???

It's just bizarre to think I'm privileged because some trans women with the same or similar sexuality as me had a slightly easier time navigating gatekeeping that would have kept me away from care and has not even been the standard of care in my country for 25+ years.

It's also such an oversimplification of the medical situation. Stereotype gatekeeping applied to straight trans women, too. I would not likely have been able to persuade the docs that I was exclusively into men, and even if I did, they'd have found another excuse to deny me care.

very little, my ex wife bought the DNA kit because she wanted to know

paid for an ancestry service, but also known ancestor lineages besides my maternal grandfather's (his family were German immigrants) are people who were in the South before 1850. People with unspecified "American" lineage in the South are overwhelmingly of Anglo descent.

I am an American of 75%+ English ancestry and you're damn right I am ashamed of it

I'M 37 I'M NOT OLD

the average person probably does not know the difference between an oxymoron and a pun that evokes an oxymoron

Yes, this is one of many areas you can arbitrage dumb people's prejudices and stupid preferences to your own advantage. Feels a little weird to say that's a good thing though lol

It's striking how core inability to conceive the actual scale of society is to NIMBYism.

Yes, you and every other well off person with 200k+ equity combine to form a massive privileged economic block. There are millions of you my man

yes I saw you post it, and as explained above it's wrong about core things necessary to argue the model itself contains copied material

The Anthropic case is really the limit of what the current law gets you: your theory needs to be based on actual copying, but good luck proving economic damages; if you've registered you copyright though you might be able to get statutory damages (not a lot).

I understand that's not good enough if you want to shut down AI because you think they are inherently made of or always produce copyrighted work, but that's the reality.

What in copyright law makes this a distinction of merit?

The only real distinction relevant to the law is that a human can read a book or look at a painting without "copying" it (or without someone else copying it for them without permission), while a computer model cannot.

And it's important to understand even if a lawyer has high confidence in what the outcome *should* be under the applicable law, the difficulty raises the odds that the court will nonetheless go the other way.

But if it were enforced as written, the internet as we know it would not work, and people would be outraged if things everyone intuits as fair use were no longer possible. Wanting that to happen but only against LLMs seems both unlikely to work and potentially have bad consequences.

Have to disagree with you here, it is actually true that copyright law prohibits huge swaths of extremely common activity on the internet. People hand wave and stretch fair use far beyond its actual applicability, but really this all works b/c enforcing copyright law is just too hard and expensive.