Блог пользователя livenlife453

Автор livenlife453, история, 4 часа назад, По-английски

Since the dawn of time, Codeforces users have asked "how do I ragebait as much of the population as possible?" Many inexperienced greys and green attempt to dive into this challenge, but are often unsuccessful. They pursue fruitless methods such as asking for downvotes, bragging, or making pointless points. In reality, Codeforces has too many trolls for these methods to be viable. In fact, asking for downvotes is a surefire way to farm upvotes. As such, I request y'all downvote this blog.

To properly learn how to farm downvotes, you must look at those who have successfully done so. Surprisingly, the lowest rated contribution people are not random Newbies who pissed off people. In reality, the lowest contribution people are Grandmasters aaa_Pigeon2, _MCYYDS_, and Master rank AmirrzwM. So what did these people do to make the community hate them?

MCYYDS: It appears they held Codeforces Round 1112, and wrote the editorial as well. However, unlike virtually every other coding contest, these received significant downvotes. After looking in the comments, I couldn't get a completely clear grasp into why. However, the underlying issues seem to be that they were accused of using AI writing for problems and editorials, and/or were cheater who were either banned previously or heavily suspected but not banned.

aaa_Pigeon2: This guy appears to have one extremely unpopular blog. In it, he explains how he "shared information" but believes what he shared does not violate codeforces policy. It seems this "sharing of information" either ruined a contest he or someone else was running, or is just an extremely unpopular take. It also seems like a question was copied from a research paper?, which leads to questions about contest originality. It is likely that the reasons Codeforces Round 1112 was so downvoted is related to this. It is also possible aaa_Pigeon2 may be an alt of MCYYDS, though I am not sure.

AMirrzwM: He also seems to have run 2 contests. However, the second is received extremely poorly, with over 1200 downvotes. The best reason I could see in the comments is that problem C was way, way, way too hard for a contest of that difficulty. It is also possible info was leaked in some capacity like the previous contest.

So it seems like if you want to farm the literally minimum possible contribution, you need to follow this pattern. 1. Build up respect for perceived skills in the community and get to run a coding contest. 2. Use AI to make the contest, write the editorial, and respond to comments. 3. Blatantly cheat in future contests so everyone hates you.

That said, the majority of ragebaiters are not that talented. Thankfully, there are a few other methods that seem to work well. However, it is important to make blogs that WILL NOT GET IGNORED. self promotion blogs will get no attention and only get 1-10 downvotes. From exploring random blogs, there seems to be a few specific patterns that get downvotes.

Most Reliable Method: Write a Post Using Generative AI. This seems to virtually always garner at least 10 downvotes. Thankfully, it will do this even if the overall message is positive. This means you can combine it with other downvote techniques.

Technique: Accuse random well respected members of Codeforces of cheating. This type of ragebait can be obvious if not done well. It's important to present yourself not as an obvious ragebaiter, but as a really stupid victim of the Dunning Kruger Effect. Present evidence that may seem reasonable to morons, but obviously stupid to anyone with brains. Also make sure you are a newbie if attempting this technique, or else the ragebait will be obvious. Be sure to be rude and insensitive during the blog and replies to the post, but not so rude that you either get banned or get called out for ragebaiting.

Technique: Make an extremely unpopular opinion. Take advantage of the thing downvotes are meant for in the first point. Make an opinion that almost no one on the site would agree with. Your post may have trouble getting attention, but if you sound passionate, have a catchy title, and other ragebait techniques, it may garner a heavy number of downvotes. If its not relevant to codeforces at all, it also wont get much attention. You may use your creativity here, but low effort unpopular opinions include "we should be able to use AI for contests," or "Only people past rating X should be able to comment."

Technique: Reply commonly and idiotically. Go onto a post that has tons of attention, and say the most unpopular take you can think of related to that post. Again, you must not make it obvious that you are ragebaiting. You must actually sound like a moron. Since these posts will get attention due to their high number of upvotes, many people will see your reply and downvote it. If your own blogs are successful at ragebaiting, you can use this technique to farm even more downvotes.

Overall Strategy: Repeatedly write blog posts, using AI, asking it to argue very stupid opinions using the version stupid reasonings you came up with. Again, it is important to strike a balance. Not so ragebaity you get banned or its obvious, not too little ragebaity you get ignored. It's a tough act to balance, but with practice, you can get better over time. Then, go to popular blogs, ones with hundreds of upvotes, and act like a toddler. Again, its important you don't get banned during this process. Make unpopular opinions, have no respectfulness, but don't do anything obviously banworthy such as say slurs, harass Codeforces staff, or cuss. If you are confident in your coding abilities, you may instead work towards running your own contest, then proceeding to make it as terrible as possible by following the steps of MCYYDS and aaa_Pigeon2.

I wish you luck in your ragebaiting and downvote farming endeavors- From, Livenlife453

Полный текст и комментарии »

  • Проголосовать: нравится
  • -3
  • Проголосовать: не нравится

Автор livenlife453, история, 5 недель назад, По-английски

If you saw my previous blog, I had you guess guys which of two essays were made by ChatGPT AI trying to sound human, and which was made my a human trying to sound like AI. Around 75%-80% of y'all correctly guessed essay 2. With that said, can AI detectors preform similarly well? There is a lot of talk online and in schools about false accusations of AI writing. Which AI checkers can properly distinguish AI vs human writing? I went through a variety of "AI Checkers" I could find to see what percent they rated AI and Human for the AI essay, and what percent for the human essay. These are the results.

Asking Chatgpt Itself: (not done on same account as essay was made) note: the prompt was "The following essay was either written by a human or AI. Please tell me how confident you are it was written by an AI, and how confident you are it was written by a human, in terms of percentages. [Insert Essay]"

Human Essay Result: 65% AI, 35% Human. Chatgpt explicitly notes that it is not confident about it being AI.

AI Essay Result: 65% AI, 35% Human. Despite giving the same percentages, Chatgpt does not explicitly note that it is not confident.

ZeroGPT:

Human Essay Result: 19.5% AI, 80.5% Human.

AI Essay Result: 100% AI, 0% Human.

GPTZero:

Human Essay Result: 0% AI, 100% Human.

AI Essay Result: 100% AI, 0% Human.

QuillBot AI Detector:

Human Essay Result: 0% AI, 100% Human.

AI Essay Result: 100% AI, 0% Human.

TurnItIn AI Detector:

Human Essay Result: 75% AI, 9% Mixed, 16% Human.

AI Essay Result: 85% AI, 7% Mixed, 8% Human.

What's interesting is that TurnItIn claims to "crosscheck" multiple AI sources to verify itself, including ones mentioned above that can easily distinguish the two essays.

Grammerly AI Detector:

Human Essay Result: 0% AI, 100% Human.

AI Essay Result: 63% AI, 37% Human.

Most Reliable: GPTZero, QuillBot AI Detector.

Least Reliable: ChatGPT itself, TurnItIn AI Detector.

For the context of false positives of school essays, the inaccuracy of TurnItIn's AI Detector is especially concerning. It's like many of these false positives occurred when teachers inputted essays into TurnItIn, and the site then proceeded to falsely accuse them of AI writing. Instead, sites like GPTZero and QuillBot AI Detector are more reliable. Asking Chatgpt itself, by comparison, is extremely unreliable. Of course, this is a sample size of one AI essay from only ChatGPT and one human essay trying to sound like AI. Someone with more materials could run a wider test. Perhaps I could do a research project about that in the future. But it still brings light onto how false accusations of AI writing can be real and can seriously hinder students' academic journeys if the wrong detectors are used.

Полный текст и комментарии »

  • Проголосовать: нравится
  • +8
  • Проголосовать: не нравится

Автор livenlife453, история, 5 недель назад, По-английски

There has been an influx of obviously AI generated blog posts about this and that, and most people on this website are easily able to identify them. However, would they still be able to tell if the human was trying to sound like an AI? If not, this means that humans can essentially "ragebait" Codeforces users, writing articles that they know will be falsely accused of being AI generated. That's what I wanted to find out. First, I wrote an article about why I believe Java is a cool language. Then, I prompted ChatGPT- "Please write an article about why Java is a super cool language. Please make it 3 paragraphs, without headings, bolding, italics, or a title, and between 180-220 words. Please try to sound as human as possible." (Keep in mind I did this prompt AFTER writing my own essay). These two articles are shown in an arbitrary order. See if you can guess which one is AI, and which one was written by me.

START:

There are few languages that have been as impactful to the development of computer science, society, and programming as a whole as Java. It remains one of the most widely used programming languages to this day. Whether you are new to programming, or an expert in the field, its very likely you have used Java at one point or another.

So what makes Java so cool? Java is a language that is designed to be easily learned and endlessly applicable. Java understands that programmers make mistakes, and utilizes several tools to bring these mistakes to light. Type mismatch, index out of bounds, and memory overflow are no issues when you choose to code in Java. The point: Java is an excellent language for education and the advancement of programming abilities.

Java may be easy to learn, but don't dismiss its speed. Capable of running MILLIONS of operations in a single second, Java is used throughout the world to create computer games, web applications, and even mobile apps! Whether you want to code a basic project for school, or make games for the world to see, Java is ready for you. Through both its speed and ease of use, Java is an indisputably cool language.

END

START:

Java is a super cool language because it manages to be both powerful and approachable. Its syntax is relatively straightforward, especially once you get used to concepts like classes and objects, and there is an enormous amount of documentation and community support available. One of Java’s biggest strengths is that it can run on practically any platform through the Java Virtual Machine. Write the code once, compile it, and you can usually run it almost anywhere without completely rewriting everything.

Another thing that makes Java awesome is how versatile it is. You can use it to make desktop applications, websites, Android apps, servers, and all kinds of other software. It also has an enormous standard library, meaning you often do not have to reinvent the wheel just to accomplish something basic. Features like ArrayList, HashMap, and BigInteger can make complicated programming tasks dramatically easier, especially when you are working on algorithms or competitive programming problems.

Finally, Java has simply been around long enough to become incredibly mature. There are countless libraries, tutorials, tools, and programmers who know how to use it. It may not always be the newest or flashiest language, but that reliability is part of what makes it cool. Java is dependable, flexible, and surprisingly fun once you get comfortable with it.

END

Leave your guesses if you want, or if you just hate this concept downvote and tell me why its terrible and I'm helping AI take over codeforces and draining the power supply and clean drinking water supply and gonna make you all die in 20 years. I don't prompt AI for essays very much, so this could be extremely obvious. (I don't think I did a very good job lmao.) I'll share the answer on Saturday, after I go have my MaiMai Fridays day tomorrow.

first one is $$$AI$$$

second one is $$$AI$$$

Edit: As most of y'all (around 75% or so) correctly identified, the AI version was Essay 2. The reason it mentions competitive programming is likely because I have talked to chatgpt about competitive programming before for brainstorming topics or predicting rating changes after the contest but before the tutorial is released. (I probably should have mentioned that I did it on my account. Sorry about that.) In general, I think the "try to talk like a human part" was a big indicator. Essay 2 seems like what an imposter of me would write for why Java is cool, where the reasoning is similar but the formatting and wording is obviously off. It also keeps the basic claim reasoning summary format taught in schools, while my essay seems more general, less risky, and not at all attempting to really sound human. Its the difference between bad human imitation vs bad ai imitation.

Полный текст и комментарии »

  • Проголосовать: нравится
  • +38
  • Проголосовать: не нравится

Автор livenlife453, история, 5 недель назад, По-английски

People everywhere have been requesting a rollback due to the influx of cheaters, but I have a few reasons to oppose this idea.

  1. Rollbacks make honest programmers happy. This is a problem, as if honest programmers are happy, they will continue down the cs path and eventually become employed in an office job, where they will lose their will to live. Making sure honest programmers stay depressed will convince them to find better jobs instead.

  2. As we all know, cheaters are soft little crybabies. Ignoring their contribution to deflating our rating, keeping us humble and showing us we are terrible at programming would make them feel bad. They are simply trying to show us how we are inferior to a bunch of binary.

  3. 1387 is a cooler number than 1410. 1387, my rating, is cool, divisible by 19 and 73, so basically a prime number fanboy. Its so cool and unique. If we get rollbacks, my rating will likely go up to around 1410. Not only will this give me the way less cool specialist title compared to the awesome pupil title, but 1410 is super boring. Its divisible by 2, 3, and 5. I mean how generic can a number be? This isn't just my problem either. Studies show that lower numbers tend to be cooler, more known about, and more widely used than their larger counterparts. Who would want to have a less cool rating?

  4. Codeforces users have been spending precious brainpower trying to determine or asking when the next rollback is happening. This wastes valuable energy that can instead be used for grinding. By removing rollbacks, these users will no longer have to waste their brainpower on such matters, and can instead solve the collatz conjecture.

  5. Rollbacks will encourage new users to join codeforces. A consistent rollback schedule would give a sense of stability and encourage new programmers to join codeforces. This would be a problem, for a similar reason as reason 1. More codeforces users->more cs people->less will to live. Instead, we should keep up a hostile environment that affirms the societal reputation of coders as "dorks", and work against this incoming and toxic trend of seeing them as "smart and logical people."

Полный текст и комментарии »

  • Проголосовать: нравится
  • -5
  • Проголосовать: не нравится

Автор livenlife453, история, 5 недель назад, По-английски

Recently, I sucked balls at a coding contest after my brain TLEd on simple DP, and lost my confidence. I went from a stellar 0.0000722% chance of beating the #1 rated player, Benq, to an abysmal 0.0000596% chance instead. Meanwhile, there are some people who can reach the absurd 0.0002% chance of beating Benq in a coding contest. While this is certainly an embarrassing display from myself, I intend to improve my abilities and someday reach that 0.0002% chance milestone of beating Benq in a coding contest. I did not terrible on the div 3 contest, so maybe I'll modify my brain with some improvements soon, or I'll simply buy the new and improved Brain Deluxe from China. If so, I will prove myself an elite programmer. Keep watch over me until then.

Полный текст и комментарии »

  • Проголосовать: нравится
  • +51
  • Проголосовать: не нравится

Автор livenlife453, история, 2 месяца назад, По-английски

After rising up from the "My Mom Says I'm Good At Programming" tier to the "Confidently Mediocre" tier, I have now reached the "Slightly Competent" tier. With this progression I can now engage in activities such as impressing my 7th grade math teacher, convincing my dad to love me 3.45% more (0*1.0345=0), and having a full conversation with a woman.

If I keep improving myself, I may eventually reach the "Moderately Decent" tier, the "I'm a Master*****" Tier, or even the "I'm smart I can work pls hire me pls" tier. If I do this, I can impress more than just my 7th grade math teacher. I can also impress my 9th grade math teacher, and my dad will love me 12.84% more. Surely if this happens I can hide how unbelievable disturbing of a person I am to my friends.

From there, the only way forward is through a small trip in the "Legendary Extreme Super Competent For Sure. I'm Even Better Than ChatGPT" tier, a much longer stay in the "Okay You're Actually Pretty Good I Can't Criticize You" tier, then remain forever trapped in the "Computing How Many Friends I Have IRL in O(log(log(log(1/n))))" tier. Finally, if I somehow manage to escape that, I reach the "Super Ultimate Best Coder Ever My Parents Love Me Finally My Life is Perfect and Nothing Ever Goes Wrong Ever and I Have Never Questioned My Life Choices" tier. Maybe, just maybe, if I get here, I can have an excuse as to no longer touch the evil entity known as "grass." All the women ever will want to be with me, except that's stupid and completely unrealistic and a gross and hurtful sexual stereotype (the only unrealistic thing I have written so far I'll have you know), so I'll tell them to join the Jiangly Fan Club and simp for Jiangly instead, and maybe learn to implement a greedy algorithm or two.

Finally, I'll ascend all human intellect and prove that P=NP, becoming the god of the universe and making all algorithms O(1), (best solution is always to just have your code ask me for the answer), reaching the Jiangly Tier, and finally reaching 78.45% of Jiangly's power. With this, CS is solved forever and therefore dead, so no one will ever compete anymore and they will spend all their time playing maimai circle DX, where they will proceed to form a new wave of talent inspiring the first ever 15+ level in maimai history to be created. And when this happens, I'll finally finish my ultimate goal, what I really want: to have something I can talk to other people about.

Thanks for reading my ragebaity blog. I've likely offended approximately 22.43% of people, with a 90% confidence interval of (16.31%-28.55%). Feel free to downvote as much as you want. I know this post can be scientifically proven to contribute nothing. Thankfully, I'm past the point in my life where criticism causes me to scientifically analyze whether guns or fall damage is less painful. (spoiler alert, it's guns). I just wanted something to do before I go play maimai for 10 hours tomorrow at Cidercade.

Полный текст и комментарии »

  • Проголосовать: нравится
  • +18
  • Проголосовать: не нравится

Автор livenlife453, история, 2 месяца назад, По-английски

It was funny looking at all the obviously chatgpt'd solutions for many of the div 3 problems in the recent contest, primarily problem B, get hacked with TLE. One literally has a comment that said "adding the fix we talked about." But honestly, I just don't get the point in cheating. By the time you are a high enough rating that anyone will care about it, you will have surely either been caught, or exceeded the level that such AI is capable of. I assume most of y'all aren't putting "world rank #17563 in Codeforces" on your resumes. I would think the purpose in doing such contests is to have fun and improve your own skill. How are you supposed to know your own skill if you ask AI for everything? I'm just curious at this point? Why bother doing it at all? If you want AI to help you improve, just ask it for solutions after the contest. Better yet, ask it to explain the tutorials that are nearly always posted after said contest, especially if they are in a language you don't use. I've gotten curious what AI thinks coding solutions work, and they seem to usually be complete garbage, even by my noob standards. (I've mostly tried this to have it try to estimate ratings for problems before the official ratings come out. (And yes, I did this after the contest was over)). By cheating, you just make rating less accurate for both yourself and your competitors. Apple won't care whether someone reaches specialist or expert or candidate master or not. But the people who put actual effort into this site will.

Part of the reason I've enjoyed UIL Academics (specifically in STEM) my entire high school career is that, outside of UT Austin and Turing, no college gives an F whether you are good at it or not. UIL is not something you do to put on your college applications, as seems to be the motives behind far too many high schoolers today. (Chillax there is more to life than college apps.) As a result, the people who reached for the top in UIL Academics were those who did it for the sake of enjoyment, learning, and love for competition. Those who loved to brag about what ivy league college they got into? They rarely made it past regionals. (UIL people, wear your team shirts, not your college shirts. You represent your high school team right now, not whatever college you are going to next year. Looking at you Klein schools. Learn something from Ivan Liu, who has been beating y'all up in math for the last 3 years.)

Correct me if I'm wrong, I'm still relatively new here, I believe the community behind Codeforces to be similar in that regard. Being good at Codeforces, unless you are REALLY REALLY good, does not seem to come with a "prestigious award" or top placement you can brag about to those not in the know, unless you can impress interviewers with the vague sounding "expert" or "candidate master" ranking. Odds are, similar to people in UIL, your talents will be undervalued by those unaware. And that's okay, because that weeds out those who do competition just for the sake of looking good and impressing employers, which unsurprisingly, also weeds out a lot of the cheaters. Contests like that should be protected, and if the unaware don't care how well you do, cheating, even if successful, does literally no one any good.

Полный текст и комментарии »

  • Проголосовать: нравится
  • +1
  • Проголосовать: не нравится

Автор livenlife453, история, 2 месяца назад, По-английски

I've been using Java, and pretty much only Java, all my life, which apparently seems to be an unusual circumstance. Virtually all the top players I've looked at seem to use C++, or on occasion some version of Python. This leads me to wonder if Java is simply the "noob" language. I understand that C++ is normally faster than Java, but from what I've seen, problems are designed such that the correct solution will run in the required time regardless of language, as long as the asymptotic complexity is correct. Is this different at the top level? I don't know much about C++, (I only had to code some basic stuff in C for school), so this could be completely wrong, but I don't think there are classes or objects or any of those helpful things like in Java. I would think using C++ is more restrictive. I'm wondering if abandoning Java and learning C++ is something I should do sooner rather than later if I want to reach any reasonably high rating in Codeforces. I would really prefer not to, given I've used Java all my life, but if it's virtually required, it would be nice to know ahead of time.

Полный текст и комментарии »

  • Проголосовать: нравится
  • -10
  • Проголосовать: не нравится