Honeypots are used pretty often, sure. They're not enough, though useful.
Behavioral analysis is way harder in practice than it sounds, because most closet cheaters do not give enough signal to stand out, and the clusters are moving pretty fast. The way people play the game always changes. It's not the problem of metric selection as it might appear to an engineer, you need to watch the community dynamics. Currently only humans are able to do that.
>It's quite literally impossible to cheat anymore (in a way that disturbs normal players for more than a few games)
AKA the way that is easiest to detect, and the easiest way to claim that the game doesn't have cheaters. Behavioral analysis doesn't work with closet cheaters, and they corrupt the community and damage the game in much subtler ways. There's nothing worse than to know that the player you've competed with all this time had a slight advantage from the start.
In CS2, the game renders your enemies even though you can't see them (within some close range). The draw calls are theoretically interceptable (either on the software/firmware or other hardware level). Detecting this is essentially impossible because the game trusts that the GPU will render correctly.
if you cheated with wallhacks, post-game analysis can detect it.
And it is possible to silently put you into a cheating game match maker, so that you only ever match with other cheaters. This, to me, is prob. the better outcome than outright banning (which means the cheater just comes back with a new account). Silently moving them to a cheater queue is a good way to slow them down, as well as isolate them.
Not with 100% accuracy. This means some legitimate players would be qualified as potentially cheating.
You don't have to play with wallhacks constantly on, you can toggle. And it doesn't detect cases where you're camping with an AWP and have 150ms response time instead of 200ms. Sometimes people are just having a good day.
> cheating game match maker
This is already a thing. In CS2, you have a Trust Factor. The lower your trust factor is, the bigger the chance you will be queued with/against cheaters.
Overwatch has made the decision that closest cheaters are not a problem and have actually protected a cheater in contenders, although they were forced to leave the competitive scene. None of it ever became public.
Word of mouth, but if you looked at their twitter and proof presented it was undeniable. If you want to go digging check a french contenders player that there are videos of with an instance of where the aimbot bugged out and started aiming directly at the center of a player with perfect reaction time and movements.
Again, you are missing the point, just because something is "likely" to happen doesn't mean it did happen.
What you are basically asking is that we should provide a "negative proof", imagine me going through all the pro matches to prove my point that it did not happen (going in this extreme) when you can just show me a proof that it did happen.
Less skilled players can't distinguish better players from cheaters, and reports are usually abused and used in bad faith. Even a good-faith report really just means "I don't want to see this player for whatever reason". It's used as a signal of something in most systems but never followed outright in good games because players get a ton of useless reports.
Players in some games with custom servers run webs of trust (or rather distrust, shared banlists). They are typically abused to some degree and good players are banned across multiple servers by admins acting in bad faith or just straight up not caring. This rarely ends well.
I used to run popular servers for PvP sandbox games and big communities, and we used votebans/reports to evict good players from casual servers to anarchy ones, where they could compete, but a mod always had to approve the eviction using a pretty non-trivial process. This system was useless for catching cheaters, we got them in other ways. That's for PvP sandboxes - in e-sports grade games reports are useless for anything.
No, I wasn’t saying the human player would report them, I am saying the game itself would. If the game receives an update from another player showing them in a location that is too far away from their previous spot, for example, the client would know the other client is cheating, and would report it automatically.
That's pretty redundant then, and also subject to abuse. Server state is already an authoritative source of truth, and the server itself should be doing behavioral analysis (which many do, it's not enough). In real-life conditions of most games, what you see, what server sees, and what each other client sees are entirely different and unrelated things.
They most likely weren't, despite very dubious claims of Amodei and Altman and a certain twitter influencer running a pretty naive writing benchmark ("slop test") that is wrong in a very obvious manner. The only unambiguous cases of distillation were Gemini 2.0 experimentals being trained on Claude outputs, and GLM-4.7 being trained on Gemini 3.0 Pro. The rest are pretty different from each other.
GLM-4.7 (specifically this version) repeats the guardrail prompt injections from 3.0 Pro, word-by-word, and never follows them, which is consistent with training on a reward-hacked CoT. Gemini 3.0 only discusses snippets from this injection in its native CoT (hidden by default, trivial to uncover), but GLM-4.7 was able to reconstruct it in full during training. The only possible reason for this is direct training on a large amount of examples of Gemini's CoT. Its structure and a lot of replies were identical in GLM too.
Gemini 2.0 Exp 1206 was reported to be indirectly trained on Claude's outputs with humans in between [1], which was pretty consistent with its outputs at the time. No other Gemini versions except two experimental ones were similar to Claude.
If it was, that line is not an indicator. Distillation is done on useful prompts, not on "Who are you?" - "I'm this model of that company".
Name training is always shallow, Claude itself would claim it's GPT-3, GPT-4, or Reddit (heh) when confused. It's just dataset contamination, because the web is full of slop. Never trust self-reported names.
>Somewhere a kid is saving up for this. He has read every review. Watched the introduction video four or five times. Looked up every spec, every benchmark, every footnote. He has probably walked into an Apple Store and interrogated an employee about it ad nauseam. He knows the consensus. He knows it’s probably not the right tool for everything he wants to do.
Anywhere in the world, the kind of kid that does all this and installs Blender on it is WAY more likely to save up for any janky terrible half-working PC laptop with a bit better specs (memory in particular), or a desktop computer if possible, because A. games B. Linux C. piracy and more software D. he does not care about it being Apple or "just working", in the words of the author himself. I don't know how the US kids in particular feel about this since the reality distortion field is so strong, but anywhere else it's like this.
Depends on the kid’s hobby and purpose of the computer. I’ve always known that MacBooks are better for music making (especially driver hassle, audio signal reliability, etc.). Could I afford a MacBook as my first laptop? No. Did I buy a second-hand MacBook as soon as I could afford it, and have been happy with it since? Yes. As a teenager, I would’ve loved to buy a new MacBook for 500/600€.
I’m sure most other applications are less Mac-optimized, though (software development, 3d/graphics editing, gaming, …).
That's very true, and it's a great tragedy. Kids get skilled in troubleshooting awful computers instead of getting skilled in creative things which actually have future value.
As for piracy, it's just as easy on a Mac, and MacOS has more quality software than any other platform - unless you're talking about software used in factories and such.
But how many kids actually "save up" for a computer vs being given one by their parents or getting a hand-me-down from relatives? I would suspect that many parents would be more than happy to buy a Macbook for their kid if they showed that kind of interest.
I think you're making it sound like they're forced to build Linux from scratch while walking knee deep in the snow, uphill both ways. That's way too detached from reality and certainly it's not a "tragedy". Kids are not doing specialized things that only have future value, they tinker with everything. Usually on what's available, yes.
>and MacOS has more quality software than any other platform
This is simply untrue, and not something a tinkerer cares about on a general-purpose machine anyway (with my niece and son as n=2).
Windows have also always required a lot of tinkering and trouble shooting to make things work in a pleasant way.
But for Linux, the creative software simply isn't there in many cases for a kid to start learning. Unless it's programming, which is not everybody's talent.
A kid tinkering with any kind of creative software learns and absorbs important skills which they can build on later if they want to. These things are much more valuable than system troubleshooting or becoming skilled in a game.
Perhaps I wasn't clear, I'm not saying kids rush to install Linux to run creative software, they're perfectly fine with Windows. I'm saying that nowadays Linux is another cool thing to try, as a father of one of those tinkerers. There's no way in hell he wants a Mac because it's so limited. Same thing about my niece, she's into electronic music since 14, and the only way she got into music making was pirating a ton of shitty and non-shitty VSTs and hosts and experimenting with synths on a desktop PC. She's an adult and using a macbook and legit software on her gigs now, but only because it's cool, everybody's got one, and she now has money to afford serious gear and dedicated machines for music making (another is a PC). Ironically enough, she always had an iPhone because "only smelly incels use Android" (her words).
I genuinely mean no offence, but you sound as an old serious man when talking about useful skills. Kids and teens are not doing this for their careers, they're doing this for fun.
> There's no way in hell he wants a Mac because it's so limited. Same thing about my niece, she's into electronic music since 14, and the only way she got into music making was pirating a ton of shitty and non-shitty VSTs and hosts and experimenting with synths on a desktop PC.
Garage Band is free with a Mac...
And Logic Pro is a cheap subscription or very fair price when somebody like her is ready to take the next step. Maybe they even have a nice uncle who'd like to pay for it?
> I genuinely mean no offence, but you sound as an old serious man when talking about useful skills. Kids and teens are not doing this for their careers, they're doing this for fun.
Yes, but depending on what you put in front of your kids, they're going to get different experiences when doing things for fun and curiosity. So why not help them a little on the way with the right tools and right toys? They'll find a way to get into the weeds and explore their curiosity one way or the other - isn't it much better that they learn about music or audio producing, or photo editing, digital art, or 3D modeling, instead of learning about how to fix the Windows registry or change a configuration file?
A kid will make a spreadsheet about dinosaurs if they only get to play with an office computer. Parents can help them by giving them something better to play with.
In reality you give the same programmer an update to the existing spec, and they change the code to implement the difference. Which is exactly what the thing in OP is doing, and exactly what should be done. There's simply no reason to regenerate the result.
The entire thing about determinism is a red herring, because 1) it's not determinism but prompt instability, and 2) prompt instability doesn't matter because of the above. Intelligence (both human and machine) is not a formal domain, your inputs lack formal syntax, and that's fine. For some reason this basic concept creates endless confusion everywhere.
I had to check the current year twice to make sure I'm not reading this in 2010. Most people moved to messaging a decade ago, and anyone younger than 25 finds the entire idea of direct voice calls bizarre and would be annoyed by getting one from you, unless you're old or have some special circumstances. Did something suddenly change?
For various people, some of that flipped back, and they now record voice messages to send them through e.g. WhatsApp.
My percieved explanation for this is that most of them can't be bothered to type a message but are still very happy to waste the recipient's time. Yes, I'm not thrilled about it.
OTOH, at least these messages are still more asynchronous and less interrupting than a phone call. But the inability to skim them still bothers me to no end.
reply