MacNCheezus

Anything more complicated than a static website is going to have a significant amount of server-side code.

Also, the article explains that it's not just the website, but ALL of their repos, which would include their smartphone apps, backend tools, etc.

It means it's the kind of stuff that law enforcement would require a warrant in order to obtain.

I think that's just regular Copilot (without the plus). This is a newer version, at least that's what this quote from the article leads me to believe:

I got ahold [sic] of the Copilot+ software and got it working on a system without an NPU about a week ago

The regular Copilot (without the plus) that sits in the taskbar was rolled out in an update about a month or two ago.

Also, this part of the article gives a method to check if it's running:

Q. How do you obtain the database files?

A. They’re just files in AppData, in the new CoreAIPlatform folder.

Unfortunately there are at least two AppData folders (three to be exact, but one of them is rarely used), and it doesn't specify whether it's %APPDATA% or %LOCALAPPDATA%, but I just checked on my Windows machine (Win11 with all updates installed, including Copilot), and I can find no such folder in either of these paths.

EDIT: the video in this toot clearly shows the location of the database folder, and it's in %LOCALAPPDATA%, which makes sense given that it's stuff that's not supposed to leave your device.

EDIT2: this tweet seems to confirm that this is indeed a feature that's only shipped on certain new devices, which need to be specially certified because Copilot+ requires hardware support.

Always has been.

I like how it calls the captcha an "IQ test".

Like if you steal out of necessity, and get caught once, you then just starve?

I mean... you could try getting on food stamps or whatever sort of government assistance is available in your country for this purpose?

In pretty much all civilized western countries, you don't HAVE to resort to becoming a criminal simply to get enough food to survive. It's really more of a sign of antisocial behavior, i.e. a complete rejection of the system combined with a desire to actively cause harm to it.

Or it could be a pride issue, i.e. people not wanting to admit to themselves that they are incapable of taking care of themselves on their own and having to go to a government office in order to "beg" for help (or panhandle outside the supermarket instead).

New SkyNet origin story just dropped

Yeah, I mean that’s basically what GPT4Chan did, which someone else already mentioned ITT.

Basically, this guy took a dataset of several gigabytes worth of archived posts from /pol/ and trained a model on that, then hooked it up to a chatbot and let it loose on the board. You can see the results in this video.

Yes, I understand that. But I'm fairly certain the quality of the data will still have a massive influence over how much and how egregiously that happens.

Basically, what I'm saying is, training your AI on a corpus on shitposts instead of factual information seems like a good way to increase the frequency and magnitude of such hallucinations.

Well yes, I've seen those examples of ChatGPT citing scientific research papers that turned out to be completely made up, but at least it seems to be a step up from straight up shitposting, which is what you get when you train it on a dataset full of shitposts.

Sure, the AI is never going to understand what it's doing or why, but training it on better datasets certain WILL improve the results.

Garbage in, garbage out.