Sure, we reasonably know training AI is fair use, although that could get modified in the future. But I'm not talking about training an AI model, I'm talking about the output from the model.

"Open" (without the word "source") originally typically meant "interoperable", which for software frequently meant as little as "documented and with headers". Before most people had ever heard of "open source" there were all kinds of other open things, most commonly standards. If your platform is open to the public, that's a legitimate kind of openness that's also familiar.

I bought a MiniPC from China instead, a GMKtec M5 Plus with 5825U, 32GB DDR4 and a 1TB NVMe SSD — for $295 right before everything went to shit. That one is out of production, but an equivalent machine is now $600. I've had it for over a year now with literally no problems.

The memory is 3.2 GHz, and despite being so cheap it doesn't even report a manufacturer, there doesn't seem to be anything wrong with it. It's got wifi 6 (MT7922) and dual 2.5 GbE (RTL8125) and the only thing that requires even a vaguely recent kernel is the wifi. It came with Windows, and I even shrank that just in case I'm ever desperate for a real Windows machine to flash some firmware or something, but ofc I've never needed or wanted to.

Too bad the combination of AI bullshit and tariff fuckery has doubled the price, I'd buy another similar machine.

You know Iran was becoming more and more tolerant and secular before we fucked it all up for oil profit, right?

Unfortunately we don't really. The standard for what is a derivative work is substantial recognizable elements. It doesn't specify exactly how you get there.

I'm not trying to convince anyone to change their terminology, but the aptness of the term is all on the conceptual side.

Drones are really good at handling gusts. No other kind of active aircraft comes close. A bigger problem is market size

Where will I get a SBC that fails after three months now?

We don't know how the brain works so no it isn't modeled on the brain.

We don't know how learning works but the brain does things with what you've learned that LLMs can't do, so it's surely not working in the same way.

Cheers ;)

Yes, it's easier to get people with empathy riled up.

Courts still have not decided exactly how AI and the idea of derivative works interact. Until they do there is room for lawsuits like this.

I'm not seeing that announced clearly. The PR says KEYTRUDA targets PD-1. It seems as likely that they only needed to do a biopsy and genetic analysis of the cancer to determine whether targeting PD-1 would be effective. This is already fairly common, though it's not done for every patient today.

No propaganda is needed to make the senescent orange child rapist a laughing stock, or to allow him to make our country one either.

In case you're serious, and don't already know, the basis of the joke that wooshed you is that Trump said Biden's pardons weren't valid because (he claimed) they were done with autopen. Then he pardoned all of the J6ers with the autopen.

I know. That's why your reply to my comment was nonsensical, as that's obviously what it was about. You were just using a space below my comment to talk to yourself.