I still dont understand why u cant distribute a large llm over many different processors each holding a section of the parameters in memory.
Because each weight in a layer influences each weight in the next layer, which means the bandwidth requirements are enormous and regular networking solutions are insufficient for that.
Until you have concrete evidence, speculating like this in public lowers you to the same level as Trump's supporters when they were claiming the previous election was stolen.
Based on this article, that makes you the asshole in this scenario. TFRs exist for a very good reason, it's not just big guberment trying to make your life harder.
I'm a way that's a downside too, it's now easier for shit regimes to block content they disagree with, without blocking the larger services. Less people will be disrupted by it, so less people will care. See: blocking Facebook,YouTube or WhatsApp vs. blocking a single activist site.
There's a few, some of them have a free tier, most of them are paid. They are often geared towards role play or storytelling. If you search for uncensored roleplaying AI I'm sure you'll find some.
I don't need an agreement from my burger place to sell a burger they made for me to someone else without being obligated to tell them it came from my burger place. There's a lot of things that are iffy with this approach, but rebranding agreements aren't one of them.
Yeah we love reading this but in the end it's just someone barely meeting the definition of journalist pulling some rando comments from Twitter because it suits the narrative that helps them get clicks. What is the news here, really?
They are usually driven from their hotel to the airport by someone else, arranged by the airline.