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

Coping with Copilot

sigarch.org

Post reply on HN
Print view View original

401–410 of 463 posts

Re: Coping with Copilot

#401

Post by robryk »

morelisp wrote: post #269
throwaway675309 wrote: post #263

Earlier quoted context omitted.

A masters program is generally going to cover deeper theoretical subjects and will also give you stronger mathematical underpinnings in those areas. If you're just looking for a code monkey to do your e-commerce react website then obviously work experience will always trump academia.

Those aspects of a MSc are easily accessible to any motivated undergraduate as well.

Yes (in the field of CS, in other fields it might vary). That said, for me the most important difference was a thing around expectations: I felt fine to spend all the time I wished on the thing that became my thesis, to the exclusion of everything else, while at the same time not feeling the obligation to do so.

Re: Coping with Copilot

#402

Post by peppermint_tea »

grade on explaning the code and not the code itself?

Someone who can explain some code understood it (or memorized the explanations)

from an academic point of view, it is not that bad. even for the people in the "memorized" category, they memorized the explanation... brings them quite close to understanding IMO

Re: Coping with Copilot

#403

Post by comfypotato »

morelisp wrote: post #239
nightpool wrote: post #184

Earlier quoted context omitted.

I'll be honest: has anyone here ever looked at a resume with an MSc in CS and said "Yep, I bet this person knows how to write code well"? Surely there are some research level jobs where a PhD in CS is necessary, but for most coding jobs, an MSc is just 2 years of missed experience where you're learning bad habits and antipatterns by never having your code used by other people in any strenuous way. Certainly there are…

I'll generally rate an MSc lower than BA/BSc. (The exceptions would be if you needed the MSc to meet some visa requirement, or if you had an BSc in some weakly-related field beforehand.) From downvoters: I would genuinely love to hear a defense of an MSc per se. What university has their shit together enough to actually accelerate/compress 2-4y of work experience into 1-2y years of academic project, which is what's a…

I think an accelerated bs/ms might be another exception you should consider. (This being when a bs student can do another year immediately following to get their ms.) It makes sense on ROI as many companies have HR departments more likely to get you into the interview process, and then it’s often a slightly higher starting role. Also, from a knowledge standpoint, some graduate classes help practical abilities only. A 500-level databases course, for example, isn’t research-oriented. It just teaches more advanced topics that might actually matter in the workplace (an intuitive understanding of how/where to use an index, more complicated examples of projects with database connections, etc.). I fall into the case of your first exception, so I’m not actually sure undergraduate courses don’t teach these things. I’ve had a few 500-level classes that had absolutely no research orientation (“object oriented design and development” is another example of a class where the projects are just super-complicated examples of what you experience in industry). Pursuing an accelerated degree shows that the student wasn’t careless in thinking research was something they wanted to do.

Re: Coping with Copilot

#404

Post by comfypotato »

shaky-carrousel wrote: post #323
throwaway675309 wrote: post #263

Earlier quoted context omitted.

A masters program is generally going to cover deeper theoretical subjects and will also give you stronger mathematical underpinnings in those areas. If you're just looking for a code monkey to do your e-commerce react website then obviously work experience will always trump academia.

This comment reeks of elitism.

It’s also stating the obvious to some extent. Extrapolating “master’s” to “PhD”: a company designing and selling hardware verification software would be better off hiring someone with an algorithms PhD to work on their SAT solver. Plenty of self-taught folk are better programmers, but experience parsing the academic literature defining cutting edge SAT techniques is valuable here. This isn’t even really an “R&D” role in which the research experience is necessary. If you care about ad hominem aspects of the discussion, I identify as a code monkey.

Re: Coping with Copilot

#405

Post by danielbln »

argiope wrote: post #390
codeflo wrote: post #381

Earlier quoted context omitted.

I’ve seen many instances where a low-skill programming place either lives with a bug, or alternatively, adopts a solution that involves meticulously copy&pasting a workaround into 300 places. Having said that, I know great, intelligent programmers, sone of who worked on compilers, who are simply bad in a whiteboard setting. Having a filter is necessary, I’m just not convinced a Google-style interview filters exactly…

The google-style interviews are certainly filtering for something though. An unfounded theory of mine is that the unfortunate eagerness of Google to sunset their products [1] might be a side effect of hiring for people that rather want to solve exciting coding problems and furthering their career by using X new language/framework, than maintaining legacy systems and keeping backwards compatibility, which might not so…

Aren't all FAANG using these style of coding interviews? And if you look at Amazon, they have almost the opposite problem of pushing out too many (AWS) products.

Re: Coping with Copilot

#406

Post by comfypotato »

morelisp wrote: post #383
bbor wrote: post #286

Earlier quoted context omitted.

Would you agree that all of the most important aspects of a bachelors in CS are easily accessible to a motivated non-student?

Yes, and I hire lots of people without CS degrees. But I do see guided study accelerating the process. People who come in with no CS background do need some years to backfill missing knowledge, or they've some years in self-study on top of a 4y degree. I don't see an MSc accelerating anything; in general one year of MSc seems worse than one year of work experience or one year of motivated self-study, and the students…

As my previous reply to your comments says, I disagree here. It depends on the nature of the courses. I think some graduate coursework falls into the category you’ve mentioned of “accelerating learning” of stuff that’s very applicable in industry. Sure, a research-oriented course about neural networks isn’t going to help someone write SQL. You could even say that an academic introduction to AI is a waste of time, but a survey of software engineering methodologies and the resources surrounding them (SWEBOK, CMMI 1.3) is going to help a student in industry over someone who hasn’t invested that time (and being guided in an academic setting is only going to help).

Re: Coping with Copilot

#407

Post by morelisp »

comfypotato wrote: post #406
morelisp wrote: post #383

Earlier quoted context omitted.

Yes, and I hire lots of people without CS degrees. But I do see guided study accelerating the process. People who come in with no CS background do need some years to backfill missing knowledge, or they've some years in self-study on top of a 4y degree. I don't see an MSc accelerating anything; in general one year of MSc seems worse than one year of work experience or one year of motivated self-study, and the students…

As my previous reply to your comments says, I disagree here. It depends on the nature of the courses. I think some graduate coursework falls into the category you’ve mentioned of “accelerating learning” of stuff that’s very applicable in industry. Sure, a research-oriented course about neural networks isn’t going to help someone write SQL. You could even say that an academic introduction to AI is a waste of time, but…

> a survey of software engineering methodologies and the resources surrounding them (SWEBOK, CMMI 1.3) is going to help a student in industry

Yeah, no. This kind of nonsense filling it up is precisely why an MSc in SE is useless.

Re: Coping with Copilot

#408

Post by morelisp »

comfypotato wrote: post #403
morelisp wrote: post #239

Earlier quoted context omitted.

I'll generally rate an MSc lower than BA/BSc. (The exceptions would be if you needed the MSc to meet some visa requirement, or if you had an BSc in some weakly-related field beforehand.) From downvoters: I would genuinely love to hear a defense of an MSc per se. What university has their shit together enough to actually accelerate/compress 2-4y of work experience into 1-2y years of academic project, which is what's a…

I think an accelerated bs/ms might be another exception you should consider. (This being when a bs student can do another year immediately following to get their ms.) It makes sense on ROI as many companies have HR departments more likely to get you into the interview process, and then it’s often a slightly higher starting role. Also, from a knowledge standpoint, some graduate classes help practical abilities only. A…

Undergraduates in CS can fill their senior year with 500 level courses easily. Motivated ones also much of their junior year. Why would I take the applicant who did it five over the ones who did it in four? Worrying about the “MSc” label at the end for the same work is just credentialism.

Re: Coping with Copilot

#409

Post by int_19h »

nebulousthree wrote: post #380
zelphirkalt wrote: post #324

Earlier quoted context omitted.

Lets face it: There is so much theory crammed into some CS degrees, bachelor and master degrees, that students do not get to code much. Real coding is often learned by oneself in side projects or on the job. If students do not engage in side projects, how do we expect them to be proficient computer programmers out of university? However, I see this as a failure of CS education. Our work is with computers. If we do no…

Your argument is akin to saying "there's so much theory crammed into Astronomy that students do not get to look through a telescope much" CS fundamentally is not about computers. It's about math and the flow of information between theoretical machines. Computers just happen to be a physical implement. The idea that universities should teach to develop industry-fit individuals is just an extension of commodifying univ…

We don't do it because it means trade schools for developers. Which, in truth, is exactly what 90% of what we do in the industry needs. But it's not glamorous; it's not a degree. Trade schools are for proles, and we aren't proles, now, how could we possibly be?

Re: Coping with Copilot

#410

Post by matthewmacleod »

morelisp wrote: post #382
julian37 wrote: post #300

Earlier quoted context omitted.

Agreed about saturation and the choice of variable name, but the code would trigger a division by zero and not result in NaN: https://go.dev/play/p/vYm4tSNEJ7M (Also, in--say--Ruby and JavaScript 1.0/0.0 is Infinity and not NaN.)

Your playground link shows a build error. 0.0/0.0 at runtime will be NaN. And in basically every language, 1.0/0.0 is Infinity. But we're talking about 0.0/0.0. Both good examples of coding where you should be thinking instead, though.

I think if anything there is far too much thinking going on here, for the tiny example I copied from the window I literally already had open with the function I was working on.

For what it's worth, Copilot (correctly) inferred a loop variable called "detection", I imagine based on similar usage earlier in the function. And there is already a conditional in place to prevent invalid operations; if I remove it I see a new suggestion:

  if len(filtered) > 0 {
This tool is far from perfect, but it very much sounds like you folks haven't used it. If that's the case, I would encourage you to research it like all tooling and draw some informed conclusions about it's applicability instead of making assumptions.
Post reply on HN