Page 1 of 1

Problem 100, google, LLMs and "cheating"

Posted: Fri Feb 07, 2025 5:24 pm
by Rex2000
So, after many hours of distress, I realize that my solution for problem 100 would take around 40 hours to complete. Cleary wrong, but then I had the great idea of using this site: https://oeis.org/, which I saw recommended on reddit. I had found the first 7 items of the sequence, so after plugging them in, low and behold, I got a formula. Recreated it in Fortran, and boom, had the solution in like 3 minutes. I was discussing with my friend whether this solution should or shouldn't count, and after reading the FAQ, I'm no where closer to a solution. Like for instance, I solve problem 15 after reading the Wikipedia article on Lattice Paths, would some of you consider it cheating or okay? Btw, I know this is always being asked, but what are your thoughts on LLMs. They really helped me find silly problems, like for instance I got a problem in Fortran because I forgot that I had to call a value as (j:j) instead of j for Strings, which threw me off and made me lose a ton of time. I dont use them to reason, because it has been fed the problems, but I am shocked by the fact that co-pilot auto completes solutions to many problems, even in the hundreds. I do feel very tempted to use LLMs to do the searching on the internet part, like explain to me how Lattice Paths works, or the name of x algorithm, but haven't yet because I don't know whether it would be consider cheating or not. I read the FAQ, but was interested in your opinions. Anyways, thanks for reading and have a good day! :D

Re: Problem 100, google, LLMs and "cheating"

Posted: Fri Feb 07, 2025 5:26 pm
by Rex2000
And what are your thoughts on using deepseek to debug problems? I think it could be useful, I've use it for personal projects, but I'm afraid to use it in project Euler, because some of you would consider it cheating. Like for instance I'm always making mistakes cause and sometimes it is sort of annoying to re read my code a lot of times just to discover that I use I when I should have used ii, or I'm missing a silly flag in an if or something of those sorts. Or maybe my code gives "234" and the answer should be "238", and it could help me see a logical oversight. What would you guys said? I was thinking that if I were to ever program something else, I would use LLMs to aid me, what are your thoughts?

Re: Problem 100, google, LLMs and "cheating"

Posted: Fri Feb 07, 2025 6:05 pm
by euler
Going to university is not just about learning the content of a particular course, but learning how to be an independent learner. That is, learning how to learn. That best happens when you are forced out of your comfort zone and you embrace the art of perseverance. Simply becoming dependent on someone else's answers will hold you forever captive to always depending on someone else's answers and ultimately you are cheating yourself out of the rich experience of real learning.

But it's all about balance. Even the great Euler learned from someone else and his journeys of discovery never ended with himself, but continued with others. So we encourage you to research and seek guidance from other sources. But the degree to which you come to depend on those aids determines if it's cheating or not. If it's to nudge you in the right direction, then that is perfectly fine. Only you will know the answer to that. At the end of the day, the purpose of Project Euler is to encourage, challenge, and develop the skills and enjoyment of anyone with an interest in the fascinating world of mathematics. The problems are the starting point. The learning, discoveries, and those moments of epiphany are at the core of what we are trying to provide.

Re: Problem 100, google, LLMs and "cheating"

Posted: Sat Feb 08, 2025 5:42 am
by thedoctar
What Mr. Euler said is what I love about PE and why I've continued with it for so long. It's completely free form, and really pushes you! (Or makes you push yourself).

Also, often early problems/easier ones teach you methods that help you solve super difficult later problems. So I've solved an earlier problem in a really inefficient way, but learnt a new method that I applied to a later super hard problem and crushed it. That is very satisfying.

Re: Problem 100, google, LLMs and "cheating"

Posted: Sun Jun 08, 2025 11:06 am
by pim.pg
I would like to make a slightly different take on Project Euler, or anything for that matter. What I will say will be partly comic, but deserves some thought! I agree with the great German philosopher Arthur Schopenhauer that the goal of all sentient existence is to avoid pain i.e distress in all forms. Now, based on my life experiences, I am going to out Schopenhauer Schopenhauer by claiming that the goal of all sentient existence is to avoid stress, i.e. both distress and eustress. What remains? pleasant time-pass. So, contrary to Euler (Colin) I will say that if you are having a good time with the materials available at the Project Euler website, good for you, otherwise find something else more pleasant to do. Remember: participating in Project Euler is completely optional. If you choose to participate in Project Euler, please follow among others the one-request of the Project Euler Problem Development Team i.e. do not publish solutions on the internet for problem numbers (as of now) > 100.