First the direct answer: you can’t.

Some more context: there’s usually ways to deal with this issue. There’s voir dire for example, where you can do a practice run of the questions and answers for a specific witness, or the party can profer (idk if it’s spelled correctly) the testimony, giving the judge a general idea of the direct or cross-examination.

This usually only happens if something goes wrong. It can be things like the lawyer overstepping a boundary by accident, but it’s very frowned upon and they can be admonished and sanctioned by the judge.

For why to do about the “can’t put the shit back in the horse”, usually the judge gives the jury an admonishment and maybe even a curative instruction right before deliberation, but best case scenario it’s been handled in advance and doesn’t even happen.

Basically the idea is to keep the jury “as clean as possible in their decision”. Best case they only hear the stuff that’s relevant, but any other case you try to tell them to disregard things or remind them what to consider, and you hope they feel duty-bound to uphold this.

Generally it is understood that juries take their obligation extremely seriously and most of the time they genuinely make an effort to rule out evidence that shouldn’t have come in. Of course there are outliers.

Soooo… Pics?

Mein Skalp schmerzt wegen deiner Satzstruktur.

Weird looking owl but I’ll take it

It really depends on if you think this is a genuine exploit or not. If it is, look up Microsoft’s track record on complying with disclosure and make yourself familiar with responsible disclosure guidelines.

If it’s not then you just need to look into the EULA of the specific windows parts affected, so they don’t pull your repository. Because that’s usually the biggest thing they do if you don’t specifically start fights with these companies.

Generally, you’re probably fine. If you want me to look over it I can give you my discord or fluxer or whatever. Not an expert, just a Cybersecurity enthusiast.

Guess it takes a shit post to learn biology

Sponge? Bacteria? Viruses?

I didn’t know it was about that Nvidia presentation. I just didn’t connect this article to it, but I know of the Nvidia presentation. That explains why this alternative interpretation didn’t cross your mind.

Yeah I’m 100% sure they are doing this A) for good PR (blender is a beloved app and funding them is awesome) b) to retain communication with the team for easier agent integration later c) to lobby for their AI.

I can totally understand that it doesn’t look great if blender accepts the money.

I am sure though that blender could probably need the money; open source software is criminally underfunded and those devs need money to survive at the end of the day.

That’s unrealistic. In the grant scheme of things Lemmy is small. There’s not enough activity for bigger criminal activity to occur here yet. Generally, there’s a bunch of three letter agencies who could probably get search warrants or permission to go after instance owners, but they’d need to be aware of Lemmy and they’d need a good enough reason to convince a judge. It’ll be a while until then. Maybe on mastodon first, we’ll see but that’s all speculation.

And should we be concerned? Maybe about those 3 letter agencies, but not about platforms like these, and most definitely not about Lemmy instances.

As long as we keep defederating from actually scary ones with csam on there, we’re more than good.

Also if this is a rhetoric question with a strong insinuation to get angry at lemmy.world then get bent, I like my Lemmy instances drama-free, thank you.

To me the title suggests the dev made the DLSS design intentionally worse, to generate controversy and now he is happy with the result.

The text actually explains that the devs just heard about the DLSS 5 alterations and were happy people loved the old design so much. Nothing controversial from him, not even intentional new designs.

Sounds like you got it right at first glance and I didn’t. However, often these titles are intentionally misleading as to generate more outrage “you should be mad at the developer for creating controversies and being smug about the outcry”. Not a fan of those manipulative tactics in modern journalism or whatever it calls itself currently.

And then you add the clickbait back in: ‘we got the design right’, Resident Evil Requiem Dev backs fans in their backlash against DLSS redesigns

Yeah that’s not even misleading, this is insinuating the opposite of what the dev means. That’s manipulative af. Whoever decided to use this title is a massive dick.

https://www.tosdr.org/

I think you messed up the link OP

“back”

After looking up the Wikipedia definition I kind of disagree; my definition is more narrow but I also think it’s more akin to what we currently understand AI to be.

That said, even with the very loose Wikipedia definition of AI I don’t know how the hell you conclude that the internet / the web is AI. Yes, specific parts of it are AI, but the Internet in my understanding does not fit the definition.

A statistical model built on neural networks solving black box problems.

I don’t hate AI but 99% of applications currently are just vanity, there are big energy problems with it, monopoly consolidation is getting exacerbated by it, identity theft is becoming easier than ever and then stolen training data and information clusters will be the next big issues.

What I’m saying is, a lot of experts (those who survived from the older field of machine learning and its various applications like genetic algorithms and classifiers) are critical of AI, and most people should be. It’s a big opportunity but we could destroy our society with it if we don’t come up with ethical constraints codified in law.

Omfg can we go a few weeks without a big leak pls?

I second syncthing as a solution.

I personally use an smb server and tail scale client + Headscale and then those smb files are locally backed up to a different drive / different PCs that remain in the network, but that doesn’t automatically sync and instead works by connecting to the server directly.

What you’re describing sounds like a solution that automatically resynchronizes on connection, and that means you’re looking for versioning / sync, thus probably syncthing is the easiest.

Hell yeah brother