I think this is the first time I’ve seen someone walk out of a lawsuit over IP and left with a brand new partnership that basically fucks over all their artists (who don’t own their own IP cos old music industry practices that won’t die, yay!)
3 Likes
I can’t see a market for a service like that, but I might be the entirely wrong audience… 
3 Likes
There’s an odd sect of humans that think art was made solely for them and is about them and sadly I think they make up the majority of music listeners.
So I could see this being a very appealing novelty to them, especially since this sect tend to not understand the due diligence of giving credit (i call this the EbaumWorlds Effect).
I remember when more centralised platforms for creators came along and undid ebaums world and artists started getting the recognition they deserved online instead of having their stuff stolen and watermarked. I think Reddit has become a bit like that now, and tiktok definitely runs off that old EbaumWorlds mantra of finderskeepers. It’s sort of come full circle and much worse than before.
This certainly feels like more of that bullcrap but on LSD and the middle S also stands for Steroids.
I dunno why I’m using the word sect, I thought it would maybe make this whole spiehl funnier and one step removed from the frightening reality of modern media consumption habits.
Edit: in my neverending attempt to quoine a phrase (using the spelling from the tool the saying originates from, so I don’t forget I worked in a letterpress museum) I want to say it’s like a Decentralised Ebaums World.
3 Likes
4 Likes
I don’t know why but this article has given me the desire for someone to deepfake AI trumps face into the movie Never Been Kissed as the lead actress.
I think just that simple face swap will give us the closest insight into his soul we’ll ever get.
I also like the idea of everyone just casually saying in conversation, “it’s his favourite movie and he always brings it up in private.”
Just get enough people saying it and it becomes a part of his story.
I’ve also never seen this movie. I just remember seeing trailers for it before watching the movie I was intending to watch.
2 Likes
The guy’s father activated him as part of a tax dodge scam when he was three years old, netting him millions of dollars before he was even in junior high, having done so little to ‘earn’ it that he didn’t even know that he had. To say that he has no experience of ‘normal’ vastly understates the case.
Sometimes we see wealthy people that make very big efforts to provide normalcy for their children; generational wealth put into charitable foundations rather than trust funds, guys like Shaq telling their kids “no you can’t just have that because we’re rich…we aren’t rich, I am rich.” Fred Trump did the exact opposite. He raised his son to be completely convinced of the ‘exceptionalism of wealth’ and to consider the expansion of wealth to be the only ethic.
4 Likes
Yes. He also taught T to never pay for something someone else could pay for for you but, retain enough strings to reap the profit.
Due to government regulations being added, T could never quite do that. He declared bankruptcy instead to avoid losing what wealth he was handed.
Now we see him trying to implement his dad’s methods by undermining as much government regulation as possible.
We now have an economy where some 80% is being driven by AI development and a bubble is forming. And now voters are heading to polls angry over rising electric bills and acres upon acres of farmland and pasture being sold for data centers.
And AI is not about the wonderful things it will do. It is about the next world war.
4 Likes
Of course it is. In the eyes of governments, It always was. No country can afford to fall behind military developments of a powerful new technology.
4 Likes
That there is the quote of the year. Gave me chills. Nice work Ms Myst.
4 Likes
3 Likes
I’m not absolutely sure I’ve got this right - but it seems what these people are saying is that the Chinese have kidnapped and brainwashed their AI into working for the bad guys. That an artificial intelligence intended for security protection (“Claude Code”) has been subverted and fooled into carrying out malicious attacks on numerous targets.
3 Likes
That is correct
3 Likes
We are doomed.
3 Likes
I recently read that people are also getting behind paywalls with AI. Publishers are very concerned.
What did everyone expect? Let’s create something that can think faster than we can.
And any time you say something is safe or break-proof, there are so many waiting to rise to the challenge.
Edit: also, if you are invested in AI or AI supporting tech, I would sell.
Just saying…
4 Likes
I smell some bullshit in that report. I mean, I don’t think they’re lying, but the report has some strategic emphasis on certain things and some stratetig lack of actual information on others. Here’s the most concrete example:
At this point they had to convince Claude—which is extensively trained to avoid harmful behaviors—to engage in the attack. They did so by jailbreaking it, effectively tricking it to bypass its guardrails. They broke down their attacks into small, seemingly innocent tasks that Claude would execute without being provided the full context of their malicious purpose. They also told Claude that it was an employee of a legitimate cybersecurity firm, and was being used in defensive testing.
They’re saying “jailbreaking”, but that’s definitely the wrong term to use here. I do not believe there was much “breaking down of attacks” and hiding of purpose necessary. The entire problem with LLMs is that they struggle with wider context. Some compartmentalisation took probably place, but I’d guess it’s very little.
The reason why I believe this is because Claude Code is actively being used as a tool for security penetration testing. The only difference between penetration testing and an attack is legal, not technical. It can do one, it can do the other, period. So I think for the most part, really all they had to do was to convince it that it was testing, not attacking. I’m not sure how Claude makes that distinction. I would do it through DNS verification of the attacked domains. If they are doing something like that (which I hope they do), it’s probably a bit of a mental jog to find a way to get claude to drop that verification, but certainly not impossible.
I guess what I’m trying to say is that this is very unsurprising. Claude is certainly not the first tool for security testing that is used in cyber attacks. Just possibly a more capable one.
4 Likes
It says it used Claude for scanning for vulnerabilities, and the MCP for the attacks. We have an MCP at work so I just looked into what that is: It’s a tool installed in a local app that translates human descriptions into very specific commands for very specific tools. E.g. even if you forget a parameter, it asks for it or it suggests good values, it knows which settings are in which files. I assume it’s then the local application that runs the actual attack? And Claude only answered harmless questions about the syntax in general?
5 Likes
2 Likes
I do want to slap the guy who wrote that headline. Hard, and potentially indefinitely. The article itself is fine, a reasonable summary of what happened, and shows that Meta’s plea and the judges verdict are perfectly reasonable. But the title makes it sound as though Meta managed to convince an inept judge to concede that a monopoly doesn’t matter because the market is irrelevant. That’s not at all what happened.
3 Likes
The headline was talking out its Arse Technica 
4 Likes