No, we think it’s great, keep going.

If sufficiently advanced malware can break out of a VM, then it’s only a matter of time before an AI breaks out of a measly container 🍿

You’ll be hoping for a while. This ain’t 2005 anymore.

It’s an arms race, the arms just keep moving deeper into the stack system. Used to happen entirely in usermode, one process poking in and reading/writing memory of the game, so anti-cheat started keeping an eye out for malicious processes. Then at some point someone patched their kernel to cheat in a way the game couldn’t possibly detect from usermode, so someone made an anti-cheat that ran at the kernel level too.

Modern KLA is basically a fully fledged rootkit, living in your system from boot, doing absolutely anything they can to try and make sure nothing has been tampered with. Validating signatures on bins, hooking memory mappings, watching for anything that might try to read/write the kernel or game’s memory space unexpectedly.

Casuals stop playing games when cheaters prevent them having fun, and it’s the casuals they need to keep happy to keep their game alive.

IMO the answer is to internally maintain a “fun to play with” metric. It would be specific to the game, but each player’s actions and interactions with other players would be evaluated to determine how “fun” they are to play with (might need to be multidimensional, since different players like having different types of interactions). It doesn’t matter if they’re cheating, or if they’re just really good, or if they use cheesy strategies, etc, if the person isn’t fun to play with, then match them with other people who are similarly unfun to play with.

This would cover your point that, if there’s a cheater in the lobby, and their behavior somehow makes everyone have more fun, then who cares?

We have memory security, virtualization and antitampering features

As someone who games entirely on Linux and wants multiplayer to work out, the features you’re referring to are for keeping the application contained by the kernel, not the other way around. On a system where the user has full autonomy, no application should be able to know what is going on outside of its user space, and I don’t want it to.

It’d be nice if it was a solved problem, but it’s not. From consoles to phones to windows, currently the industry relies on you not having autonomy over your device for anti-cheat to work. Every other solution is either expensive (obfuscation arms race), or untenable (real time, high resolution server side validation of every property of every player).

If it just looks like a stream of TLS packets, so the content is encrypted, what would DPI be able to see? I feel like if it could detect it as a VPN, that’s just a bug that needs fixing, not an inherent weakness in the protocols involved.

But on closed source drivers, right?

by sharing that you use it

aka “promoting”. They were specifically talking about using and not promoting.

People or LLMs?

Hold up, what did I read into it? I directly quoted you and asked for clarification on whether you currently believe that is the state of AI, or whether you’re saying that’s what automation used to be.

If you’re saying that’s what automation used to be, then we agree. But if you believe that modern AI can only do the “tedious bullshit no one wants to do”, that’s literally not the case.

Sora 2 is generating realistic video of anything you want given just a text prompt, rivaling the best VFX artists.

Hollywood is currently clamoring to “work with” AI celebrities who don’t exist, with a synthetic voice, singing songs no one composed with lyrics generated by an LLM. Why give a cut to a pop artist or band if you can synthesize it from nothing?

The education system has been completely upturned because every assignment can be completed by an AI, and there’s no way for the teacher to detect it. And it’s having a measurably damaging effect on students’ intellect.

A popular quote floating around right now is, “I want AI to do my laundry and dishes so that I can do art and writing, not for AI to do my art and writing so that I can do my laundry and dishes.”

And right now I literally can’t know if someone is running an AI with the prompt: “respond to this comment as though you are an out of touch older American who still thinks the capabilities of generative AI are limited to simple automation of tedious tasks no one wants to do anyway.” And you don’t know if I’m an AI with the prompt, “respond to this comment like a condescending tech literate young adult who is afraid of the impact that generative AI owned and funded by an oligarchy is going to have on every aspect of their future.”

I honestly feel stupid even bothering to type any of this out. I’m surely being had.

And it’s worth noting that you can’t automate the interesting parts of a job, as those are creative. All you can tackle is the rote, the tedious, the structured bullshit that no one wants to do in the first place.

Are you saying that this used to be the case and acknowledging that it’s no longer true with modern AI? Because it’s demonstrably not true for modern AI and is the entire reason people are fearful.

Honestly, this post is so far out of the loop, part of me is wondering if it’s AI generated.

Cool, I didn’t know it was smart enough to undo the copy, that’s good to know/hear.

If coding is the means to an end they want, they will learn it.

I started learning how to program because I wanted to mod Halo 20y ago. Gaming is often a motivator. I had a co-worker who started in the 80s, whose only option to play games on his C64 was to type up a bunch of BASIC from a magazine. He had to take care not to make any typos, then play the game, and then didn’t have any persistent tape to save it to, so he just lost it all on a reboot. Turns out, if you’re “forced” to type code in all the time, you start to figure out which bits do what, and you start changing it to behave how you want.

“Hacking” could probably work as a motivator, though with great power comes great responsibility.

But yeah, a kid won’t be interested in programming unless they see it as their only option to do what they want to do. PICO8 might be a good entry. Or something like Minecraft modding.

In BIOS/UEFI you will likely see multiple bootloader options, one for windows and one for Linux (I think mint’s is called “ubuntu” by default). Choose the Linux one.

Merging the space after your Data partition is easy. Merging the space before it is slightly less trivial, but doable. GParted is your friend. It has the ability to grow an NTFS to the right, as well as slide it to the left. The slide is copying everything over though, so it will take time.

(Note that if you are mounting Data in your fstab using the string /dev/sda4 and you delete the partitions before it, you will likely need to update your fstab.)

Personally, I don’t think you need to go as far as unhooking your Linux disk and live booting, but I understand being unsure about it. If any data on these drives is your only copy, that’s your first mistake. Back up your data elsewhere (rule of 3, ideally). Then just use gparted carefully.

Afterwards, you’ll need to regenerate grub to get the extra boot options to go away. Should be straight forward on mint.

It’s gonna feels so good deleting all those nonsense windows partitions.

Edit: I glossed right over your links to your updates saying you had already done all of this lol. GG glad it went smooth for you! Also, I am surprised canceling the NTFS slide mid-copy didn’t break anything lol. You might want to back that up and format the whole drive just to be safe. Never know when you’ll find the files that were corrupted by that…maybe run an fsck on it.

That is literally the opposite of Musk’s goal.

If it’s blocking AUR updates, it could be an attempt to keep some patches to certain exploits from going out? But it seems unlikely that the cost of a ddos is worth the tiny number of possibly vulnerable AUR users out there…

I have a friend who was trying out endeavor with kde. He uses a trackball mouse, and configuring the acceleration curve has been a nightmare for him. Apparently it’s the wayland compositor’s job to expose the ability to configure libinput, and only certain ones do it (KDE being one of them), but configuration isn’t as straight forward as in windows.

He was more able to configure it when using X11, but kept hitting a bug when using a custom acceleration curve where the cursor would shoot to the top left of the screen (I think it triggered when moving the cursor while clicking).

I haven’t looked into it much myself, but it sounds like it has been one of those unfortunate sticking points for him right out of the gate.

Behold, the master race.