zlacker

Is it necessarily a bad bias to assume OpenAI is still behaving as it's been behaving during its entire history: recklessly taking other people's IP?

>>ranger+(OP)
I think the way I would split the difference here is that your point should inform how we think about regulation and investigation. How we write rules, how we decide to proceed in terms of investigating things.

But when it comes to specific questions that hinge on evidence, I think you have to maintain the typical presumption of innocence, just to balance out the possibilities of mob psychology getting out of control.

>>ranger+(OP)
Yes, because the courts have yet to decide whether OpenAI has been "recklessly taking other people's IP" in an illegal way. Right now, it's only something believed by people who wish it to be true; legally, it's not clear just yet. In contrast, actually doing SJ impersonation here would be a much clearer violation. There's a huge gap between the two deeds, and I don't see the reason to just assume OpenAI crossed it.

It's like, the people dropping leaflets in your physical mailbox are delivering spam, but you wouldn't automatically assume those same people are also trying to scam you and your neighbors by delivering you physical letters meant to trick people into parting with their savings. In both cases, the messages are spam, but one is legal, other is not, and there's a huge gap between them.

>>TeMPOr+No
Exactly wrong; it's the job of the law to "be careful," not of the people.

Those of us accusing and talking about it have no power -- thus there is literally no harm, and possible good in, putting them on the defense about this.

edit: In fact, the First Amendment of the Constitution essentially directly upholds the idea of "people saying whatever they want" in this regard.