visho33's blog

By visho33, history, 2 days ago, In English

Hi! I'm looking for some career advice :)

I'm currently working as an ML researcher, but I've realised that I don't like (academic) research. I got into research trying to fill the void after my last ICPC. I like solving (algorithmic/mathematical) problems, but research involves a lot of things I really really don't like: writing papers, introductions, related work, conclusions, references, responding to reviewers, reviewing, etc. I recently submitted my first paper, which had a proof about the expressivity of transformers that I found really beautiful. Unfortunately, the reviews were quite negative because they didn't consider the scope relevant to ML (which I disagree with). To be honest, I don't actually care that much about ML itself, I just enjoyed finding and proving the result.

Another thing is the uncertainty of research. I'm just an anxious guy, and this job has been pretty bad for my mental (and physical) health. I'd really prefer a job with more concrete objectives, where I can tell when I'm making progress.

So I've been wondering what other career options might suit someone like me. I've been considering switching to SWE, but as I studied applied maths, I'm not sure how difficult that transition would be. I'm also curious about how SWE jobs are changing with AI.

So my question: have you found any jobs that feel somewhat similar to competitive programming? I mean, in the way you work and solve tasks. Or just any jobs that you genuinely enjoy?

I'd really appreciate any advice or personal experiences. To be honest, I've been feeling very lost about my future career lately, and it's been making me sad.

Thanks!

  • Vote: I like it
  • +215
  • Vote: I do not like it

»
2 days ago, hide # |
 
Vote: I like it +9 Vote: I do not like it

Have you tried looking at quantitative finance?

  • »
    »
    2 days ago, hide # ^ |
     
    Vote: I like it +3 Vote: I do not like it

    I applied to JS but failed the interview :( I've been considering quant finance, but I'd like to get a better idea of whether I'd enjoy the work before applying to more places, preparing for interviews, and thinking about relocating.

    • »
      »
      »
      2 days ago, hide # ^ |
       
      Vote: I like it +3 Vote: I do not like it

      I thoght combination of ML + 2200 codeforces would be enough to get in. It basically covers 80% of what they ask during interviews.

    • »
      »
      »
      2 days ago, hide # ^ |
      ← Rev. 2  
      Vote: I like it 0 Vote: I do not like it

      if you dont mind me asking, what went wrong in the JS interview despite your 2200+ rating?

      • »
        »
        »
        »
        2 days ago, hide # ^ |
         
        Vote: I like it +17 Vote: I do not like it

        My excuse is that my English is so freaking bad hahaha. I think I can solve the problems (more or less), but during the interview it was so hard to explain my thoughts while solving because I needed all my brainpower just to speak English, so I had no brain left for maths XD

      • »
        »
        »
        »
        25 hours ago, hide # ^ |
         
        Vote: I like it +6 Vote: I do not like it

        I think JS interviews are actually quite hard, I got rejected too

    • »
      »
      »
      2 days ago, hide # ^ |
       
      Vote: I like it -14 Vote: I do not like it

      Bro you are a Master, then why did they reject you ? no ways you could be weak at maths..

    • »
      »
      »
      2 days ago, hide # ^ |
       
      Vote: I like it 0 Vote: I do not like it

      I think any job that aligns with CF in terms of the actual work or tasks probably wont have the same fast feedback loop that CF does.

      Umm for example, you can spend 1–2 hours solving a decent problem and get an immediate AC or WA. Even in a job involving similar algorithmic or mathematical problem-solving, I dont think youll get that kind of instant validation.

      Maybe what youre looking for is a job with smaller, more focused tasks, well-defined objectives, and relatively quick feedback loops? which is hard to find in a corporate setting ig since theyre more long term projects ?

      And another question that might arise, do you like the process of thinking more or implementing it ?

      Not entirely sure, just a theory and questions for you lol.

      • »
        »
        »
        »
        2 days ago, hide # ^ |
         
        Vote: I like it 0 Vote: I do not like it

        I agree that expecting feedback every 2h is too much, I'm not such a TikTok kid yet hahaha. Yes, what you said next is what I'd like, at least something more in that direction than academic research.

        I prefer the process of thinking, but implementing is ok for me too when I'm not in an important contest with a timer, bc I try to rush and always get bugs F.

    • »
      »
      »
      2 days ago, hide # ^ |
       
      Vote: I like it 0 Vote: I do not like it

      Hey, can I message you about this pls?

»
2 days ago, hide # |
 
Vote: I like it +72 Vote: I do not like it

This post resonated with me surprisingly hard lol. Although I’m not in a position to give you advice I wish you the best!

»
36 hours ago, hide # |
 
Vote: I like it +106 Vote: I do not like it

CF Coordinator.

Or just give up, like me.

It is very funny to read grays "how did you fail the interview, you are 2200+". Yeah, your rating has no importance for jobs, and it was a mistake to spend time on CP if you just wanted to get a job. I applied to Jane Street twice and didn't even get an interview.

  • »
    »
    29 hours ago, hide # ^ |
     
    Vote: I like it +20 Vote: I do not like it

    Speaking as a hiring manager, the most reliable way into these companies is a direct referral from someone in your network who already works there. Large companies typically pay referral bonuses, I've seen anywhere from $1,000 to $20,000 for a successful hire. So unless someone really doesn’t want to work with you or doesn’t think your resume is competitive, they’ll usually agree. A referral moves you into a higher priority queue and makes it much more likely a recruiter actually reaches out. Without one, your resume is dropped in the general pool, at the mercy of automated filters and whatever ad hoc screening the recruiter applies. I realize that this is much easier said than done, specially if you're applying to a foreign company or have a small network to work with.

    Once you DO land an interview though, you’re absolutely right that a high rating doesn’t guarantee success. These interviews also test skills that competitive programming doesn’t exercise, like system design, clear communication, conflict resolution, and company-specific values like being security-minded or bringing “positive energy”. For better (or worse) these soft skills are often easier to "fake" or do "well-enough" for the short duration of the interview.

    • »
      »
      »
      27 hours ago, hide # ^ |
       
      Vote: I like it +13 Vote: I do not like it

      That's the 2nd most reliable. #1 is having rich parents.

      • »
        »
        »
        »
        9 hours ago, hide # ^ |
         
        Vote: I like it 0 Vote: I do not like it

        Can you explain? I can't imagine what you mean other than parents paying for you to be educated and go to a big-name university?

        • »
          »
          »
          »
          »
          8 hours ago, hide # ^ |
           
          Vote: I like it +16 Vote: I do not like it

          most of the rich people are rich because of the connections they got

          basically the biggest factors behind wealth are connections, relationships, and access

          • »
            »
            »
            »
            »
            »
            8 hours ago, hide # ^ |
             
            Vote: I like it 0 Vote: I do not like it

            Absolutely not how it works at Jane Street or pretty much any company that's good, you always go through the regular interview pipeline, it's around equivalent to getting a strong referral, which is what Franktastic mentioned

    • »
      »
      »
      15 hours ago, hide # ^ |
       
      Vote: I like it 0 Vote: I do not like it

      Franktastic, your words are so clear and realistic, and even algorithmic, lol. I really like your style.

  • »
    »
    29 hours ago, hide # ^ |
     
    Vote: I like it -7 Vote: I do not like it

    if it's a waste to spend time on CP, what is not a waste then?

    Is LeetCode any better?

»
25 hours ago, hide # |
 
Vote: I like it +8 Vote: I do not like it

If you find out, let me know please

»
25 hours ago, hide # |
 
Vote: I like it +11 Vote: I do not like it

Just get insanely good at competitive programming and farm prize money, according to this website tourist has made almost $450K from coding competitions, and that only includes big contests, not things like CF contests with prizes

If that doesn't cover the rent alone then do some weird automated dropshipping slop alongside it

(/j)

  • »
    »
    13 hours ago, hide # ^ |
     
    Vote: I like it +9 Vote: I do not like it

    In the old times without AI I had earnings on the heuristic contests (such as huawei challenge), which was a lot for a student. Sadly now I can't get good places in that competitions F

»
19 hours ago, hide # |
 
Vote: I like it +44 Vote: I do not like it

I also have a lot of similar feelings, for now I just postpone my thoughts so my future self can deal with them. I dont think this is much help, but wish you all the good luck!

»
18 hours ago, hide # |
← Rev. 3  
Vote: I like it +22 Vote: I do not like it

The harsh truth is that there isn't one. Anyone that says otherwise is probably not thinking that hard about the question.

If you want something else (like a job) to give you the same level of joy, you need to consciously put effort to find joy in that thing. That's ultimately how the people that left CP to be happy doing something else did it, no matter how similar the job looks to you on a surface level.

»
14 hours ago, hide # |
 
Vote: I like it +15 Vote: I do not like it

What part of your dream job involves communicating with people? Most of what you said you don't like around publishing a paper falls into that category.

If it's no part, you don't want a job, you want to be idle rich. A job is inevitably someone wanting something from you and paying you for it — even if you like doing it, that's secondary and you don't get to decide what will be done with the output of your work. Dealing with other people's demands, often incomplete or contradictory (it won't be just one person) is part of any job. Sometimes you're the one who will need to figure out the concrete objectives you mention, as part of the job, before you even start actually doing anything.

It's absolutely possible to self-publish a paper. The fundamental purpose of academia is being an exclusive club; your ideas don't have to be accepted regardless of how correct they are. You can still get them out there, unfortunately that doesn't mean you'll be compensated for the time and effort you put in.

Competitive programming-like job is different from a job about algorithms. The latter is more like problemsetting, where you're the one designing problems that might or might not have new algorithmic solutions, and need to figure out how to figure out what directions to look in more than just looking in those directions. Any real job will be open-ended. There's no job in which you're going to be given solved problems to also solve.

For comparison: I'm currently working to make identity verification more secure and private. There are many technical tradeoffs, different groups bring different constraints for different use cases, sometimes I chase around external bugs with colleagues, sometimes I chase around colleagues. Implementation is a minor part. My objective is the final result; if I didn't get to write a single line of code, it'd be boring and I'll probably give it away to someone else, but figuring out when and what to write is a much larger part of the job.

  • »
    »
    11 hours ago, hide # ^ |
     
    Vote: I like it 0 Vote: I do not like it

    I've been thinking about this, and I don't think communication is the problem. Maybe I just mentioned some superficial stuff, there are also deeper things I don't like, but I didn't want to turn this post into a discussion about academia. Maybe I shouldn't have mentioned it at all (now I think it's irrelevant), but I wrote the blog while having a bit of a crisis and thinking "I'll quit, I'll quit".

    I agree that a job is open-ended, but I think open-ended problems can still feel like CP. For example, a Huawei Challenge is also open-ended, and it feels like CP. I don't know whether working at Huawei would feel like that, and the point of this blog was to hear about other people's experiences.

    Thx for sharing your experience :)

»
14 hours ago, hide # |
 
Vote: I like it +8 Vote: I do not like it

Competitive programming led me to software engineering, especially web development. Surprisingly, large-scale web systems have many challenging problems around performance, scalability, search, and architecture. It's a different kind of problem solving, but I found it equally enjoyable.

»
9 hours ago, hide # |
 
Vote: I like it +3 Vote: I do not like it

In the past there were no jobs that were like competitive programming. Especially not SWE. Reading the docs and checking other people's code all day to figure out how to piece together 3 frameworks in a way that makes it run was horrible. Thank god we have AI for that though. Now you get to do systems design and tokens go brr. If you pick your own projects outside of formal subordinate work you will get to do unstructured work and explore extremely interesting things. I think it's more fun than competitive programming now.

Don't worry about it. In the worst case scenario you can still just do SWE work with AI and it will be so much less painful than it was just three years ago. Although SWE work might not exist in a year.

»
6 hours ago, hide # |
 
Vote: I like it 0 Vote: I do not like it

Honestly, I think you should give SWE a try. You already have strong problem-solving skills, and I don't think switching from applied math would be as difficult as you imagine. Maybe you'd enjoy working on real-world problems with clear goals more than academic research. Sometimes you just need to find the right environment. Good luck!

»
5 hours ago, hide # |
← Rev. 2  
Vote: I like it +3 Vote: I do not like it

Your experience is a lot like mine. Unfortunately, I don’t know the way out either. Lately, I've been really worried about my future career, as I didn't expect myself to struggle with projects. The sad part is that academic success doesn’t result in being good at projects and research. You can surely solve any problem you’re given in a course because of its logical structure, but projects without clear goals are usually what you get in real life. And there are no beautiful solutions to them like the ones in olympiads. But I guess we should just leave the past behind and move forward.