Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Frontier AI has broken the open CTF format

kabir.au

Post reply on HN
Print view View original

461–470 of 502 posts

Re: Frontier AI has broken the open CTF format

#461

Post by legacynl »

archi42 wrote: post #314
nektro wrote: post #308

easy, CTFs should ban it. then it'd be more like the chess community

The article addresses this: > Rules that ask people not to use LLMs are ignored and almost impossible to enforce in open online events. It's quite sad to see CTFs dying. I never had the time do seriously participate in CTFs, but I always respected those who did, as well as the people organizing these events.

> Rules that ask people not to use LLMs are ignored and almost impossible to enforce in open online events.

That's such a non-reason. If your competition cannot enforce the rules of the competition, then what's the point? Does the CTFs specifically need to be 'open'?

Re: Frontier AI has broken the open CTF format

#462

Post by nicce »

legacynl wrote: post #460

> The issue was never that AI could help. CTF players have always used tools. [...] Teams that refused to use AI were not just missing a convenience; they were playing a slower version of the competition. So the obvious solution is to fully ban AI and AI generated tools? To destroy your own hobby just because AI can semantically be considered a tool, seems very stupid to me. If the point of these CTFs is to practice…

Many existing challenges are purely based on knowing the existing precise tool. It has been the difference of "very easy" and "insane" challenge, whether you knew the tool or not.

So, I would not start banning the tools. They always been there. We just need to fine tune the challenges where bar goes beyond things and you really need to use AI as tool. Maybe the definition of insane starts to be custom kernel fork with planted bugs and you need to use AI to find the bugs, and use some exploit chains against that kernel with specific web server. That is the real world right now, I guess.

Re: Frontier AI has broken the open CTF format

#463

Post by nicce »

gmm1990 wrote: post #362
hemlock4593 wrote: post #285

I feel the post. For me AI has ruined both, playing CTFs and also building CTFs challenges. The most annoying thing to me is the "yeah idk but here is the flag" mentality. Before when playing CTFs with my mates was usually sitting there for hours tackling a challenge until some other mate joined, had some look together and solved it with you together in 30 minutes which is the most rewarding learning experience. Nowa…

I don’t know like chess engines didn’t kill chess. You could just play with people that don’t use the “engine”

> I don’t know like chess engines didn’t kill chess. You could just play with people that don’t use the “engine”

Impossible to do online, at least. People go beyond things for virtual fame and "owning" others. Even if it is based on cheating.

Re: Frontier AI has broken the open CTF format

#464

Post by electr1cBugaloo »

This article hits really close to home for me. I have just recently started doing CTFs and feel like I am already a somewhat decent intermediate player. Much of the motivation for me is learning but also comparing/succeeding vs. my peers.

Asking AI for learning / explaining purposes is absolutely fine in my book - but I have absolutely no motivation to set up AI to solve challenges. Without AI you can't really compete successfully. So AI is really taking motivation away for me which in turn again prevents me from learning more. I am not sure that this is solvable.

Re: Frontier AI has broken the open CTF format

#465

Post by fullstackchris »

fragmede wrote: post #337
Nifty3929 wrote: post #296

Must I beg to have an acronym spelled out a least once, the first time it's used? Even if you assume 90% of readers already know, the other 10% (including me, in this case) will thank you, it doesn't take much effort, and it expands the reach of your communication or idea. Exceptions for cases where the acronym is just so well known that a lot of people don't even know what it stands for even though they know the con…

We live in the goddammed future. Huamnity's knowledge is at your fingertips. Right clicking the Nth word of the article and putting in any semblance of effort to learn on your own is too much to ask? I don't know everything, there's tons of stuff I don't know about, but when I'm at my web browser, the least I can do about something is ask Google about a word or phrase or subject that isn't familiar instead of being s…

There are also widely accepted standards to written word.

The best example is when an abbreviation can be expanded to more than one phrase, and both are widely used.

Re: Frontier AI has broken the open CTF format

#466

Post by PunchyHamster »

ruszki wrote: post #454
jere wrote: post #438

Earlier quoted context omitted.

Classic of course. The point being: don't make fun of people for not knowing something. In this thread we're making fun of learned helplessness.

You are making fun of people who are saying that this article is shit, without addressing that part at all.

consider 2 conversations

"hey what X means?" "X means it"

vs

"I dont know what CTF stands for so I dont know if I am interested in this article or learning anything about it. Maybe I am.

Are you really arguing for not just typing out whatever 3 words this stands for once in the name of clarity?"

The commenter could just say the first instead of deciding his learned helplessness is everyone's else problem

Re: Frontier AI has broken the open CTF format

#467

Post by tardedmeme »

mr_mitm wrote: post #282
tardedmeme wrote: post #279

Earlier quoted context omitted.

That is incorrect. Photons have mass. They have no rest mass. They also cannot rest, so you might wonder how relevant that is.

The concepts of rest mass and relativistic mass are considered outdated. In modern physics, "mass" means what they meant by "rest mass". Here some indication I'm not making this up: https://hsm.stackexchange.com/questions/2465/when-and-why-di... In any case, I never use those concepts, and I know no professional particle physicist that does. By "mass", I mean rest mass.

When you put a photon in a stationary box, the "relativistic mass" of the photon becomes part of the "rest mass" of the photon-box system. You can't ignore it.

Re: Frontier AI has broken the open CTF format

#468

Post by metroholografix »

utopiah wrote: post #220
swiftcoder wrote: post #115

Earlier quoted context omitted.

The only way this actually works is if you move CTF to in-person only. There's no other way to reasonably prevent the whole leaderboard being taken up by whoever spent the most on tokens.

Sure, I don't know how to make it work. I just know that DeepBlue didn't kill competitive chess. We simply have at least 3 different rule sets, namely - no computer assistance, which does also mean no mobile on competition, human only - advanced chess with assistance - computer only, no human assistance and arguably chess itself is not doing worst since.

Official chess competitions are taking place under stringent monitoring conditions and even then, with professional reputations on the line, there have been multiple high profile cheating incidents.

Amateur online chess on the other hand is besieged by cheaters that use engines, even in casual non-ranked games where there's absolutely nothing to gain besides a pat on the ego. This has drastically changed how the game is played today with lot of players gravitating towards speed chess (bullet and blitz) to compensate. That will thin the herd of cheaters but one still runs into engines on a weekly basis.

This is also the tip of the iceberg, with the true scale of the problem being orders of magnitude worse, as someone dedicated enough can use an engine to cheat in a way that's essentially undetectable.

Re: Frontier AI has broken the open CTF format

#469

Post by ajam1507 »

RealCodingOtaku wrote: post #452
Nifty3929 wrote: post #296

Must I beg to have an acronym spelled out a least once, the first time it's used? Even if you assume 90% of readers already know, the other 10% (including me, in this case) will thank you, it doesn't take much effort, and it expands the reach of your communication or idea. Exceptions for cases where the acronym is just so well known that a lot of people don't even know what it stands for even though they know the con…

The comments are annoying. No matter the niche, it is always good to write the abbreviation the first time it is used, in fact, W3C recommends it[0]. Anyone who does not follow this either are not informed well enough, or has ableism. Most replies in this thread shows the latter. https://www.w3.org/WAI/WCAG21/Understanding/abbreviations

This page has 5 abbreviations that are not spelled out before you even get to the body of the article.

Re: Frontier AI has broken the open CTF format

#470

Post by wglb »

Nifty3929 wrote: post #296

Must I beg to have an acronym spelled out a least once, the first time it's used? Even if you assume 90% of readers already know, the other 10% (including me, in this case) will thank you, it doesn't take much effort, and it expands the reach of your communication or idea. Exceptions for cases where the acronym is just so well known that a lot of people don't even know what it stands for even though they know the con…

If you wish to explore a new concept, perhaps ask duckduckgo "what is ctf" an you may well get a response such as:

    CTF stands for Capture The Flag, which is a cybersecurity competition where participants find hidden "flags" in vulnerable programs or systems to test and develop their security skills. There are two main types of CTFs: jeopardy-style, where teams solve various challenges for points, and attack-defense, where teams defend their systems while trying to exploit their opponents'.
Post reply on HN