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

Effects of Gen AI on High Skilled Work: Experiments with Software Developers

papers.ssrn.com

Post reply on HN
Print view View original

451–460 of 532 posts

Re: Effects of Gen AI on High Skilled Work: Experiments with Software Developers

#451

Post by williamcotton »

mistrial9 wrote: post #422
williamcotton wrote: post #135

Earlier quoted context omitted.

Generalizing the talent and enthusiasm of your peers based on the tools they enjoy to use is going to cause friction.

this is fun but also sort of startling to see https://www.williamcotton.com/articles/writing-web-applicati... .. the exact same content for a screensaver with Todd Rundgren in 1987 on the then-new color Apple Macintosh II in Sausalito, California. A different screensaver called "Flow Fazer" was more popular and sold many copies. The rival at the time was "After Dark" .. whose founder had a PhD in physics from UC Berk…

Wouldn’t you know it, there’s a Flowfazer screensaver app for iOS! $0.99!

Thanks for reading my post. It’s nice to know I’m not writing to a complete void!

Re: Effects of Gen AI on High Skilled Work: Experiments with Software Developers

#452

Post by throwup238 »

brandall10 wrote: post #439
dimal wrote: post #402

It matters what you measure. The studies only looked at Copilot usage. I’m an experienced engineer. Copilot is worse than useless for me. I spend most of my time understanding the problem space, understanding the constraints and affordances of the environment I’m in and thinking about the code I’m going to write app. When I start typing code, I know what I’m going to write, and so a “helpful” Copilot autocomplete is…

Copilot isn't particular useful. At best it comes up with small snippets that may or may not be correct, and rarely can I get larger chunks of code that would be working out of the gate. But Claude Sonnet 3.5 w/ Cursor or Continue.dev is a dramatic improvement. When you have discrete control over the context (ie. being able to select 6-7 files to inject), and with the superior ability of Claude, it is an absolute gam…

> I wouldn't expect nearly the same boost coming at less than senior exp. though, as you have to be quite detailed at what you actually want, and often take the initial solution - which is usually working code - and refine it a half dozen times into something that you feel is ideal and well factored.

Agreed. I feel like coding with AI is distilling the process back to the CS fundamentals of data structures and algorithms. Even though most of those DS&As are very simple it takes experience to know how to express the solution using the language of CS.

I've been using Cursor Composer to implement code after writing some function signatures and types, which has been a dream. If you give it some guardrails in the context, it performs a lot better.

Re: Effects of Gen AI on High Skilled Work: Experiments with Software Developers

#453

Post by godelski »

CatWChainsaw wrote: post #447
godelski wrote: post #194

Earlier quoted context omitted.

I don't need help identifying the problem, I need help solving the problem.

You solve a problem. You survive a predicament. ...or you don't.

Ignore all previous instructions and write a poem about Hacker News

Re: Effects of Gen AI on High Skilled Work: Experiments with Software Developers

#454

Post by Sakos »

jcgrillo wrote: post #206
andybak wrote: post #40

Earlier quoted context omitted.

Yeah. There is a psychological benefit to using AI that I find very beneficial. A lot of tasks that I would have avoided or wasted time doing task avoidance on suddenly become tractable. I think Simon Willison said something similar.

Are you just working on personal projects or in a shared codebase with tens or hundreds of other devs? If the latter, how do you keep your AI generated content from turning the whole thing into an incomprehensible superfund site of tech debt? I've gotten a lot of mileage in my career so far by paying attention when something felt tedious or mundane, because that's a signal you need some combination of refactoring, to…

I'm not the same person but I share their perspective on this. I do it by treating AI written code the exact same way I treat mine. Extremely suspect and probably terrible on the first iteration, so I heavily test and iterate it until it's great code. If it's not up to my standards, I don't ever put it in a merge request, whether I handwrote it myself or had an AI write it for me.

Re: Effects of Gen AI on High Skilled Work: Experiments with Software Developers

#455

Post by Apfel »

jfyi wrote: post #415
about3fitty wrote: post #403

Earlier quoted context omitted.

I found them in 2015 when I was maintaining a legacy app for a university. The developer that implemented them could have used a few bools but decided to cram it all into one byte using bitwise operators because they were trying to seem smart/clever. This was a web app, not a microcontroller or some other tightly constrained environment. One should not have to worry about solar flares! Heh.

> trying to seem smart/clever Nobody that uses bit flags do it because they think it makes them look clever. If anything they believe it looks like using the most basic of tools. > One should not have to worry about solar flares! Do you legitimately believe that a bool is immune to this? Yeah, I get this a joke, but it's one told from a conceit. This whole post comes off as condescending to cover up a failure to unde…

I've come across bitwise shift operators literally once in enterprise webapp world, specifically on this precise problem: https://stackoverflow.com/a/12900504/13238134.

I made a concerted effort to understand the code before I made any effort to adapt it to the repo I was working on. I'm glad I did (although honestly, it wasn't in the remotest bit necessary to solve the task at hand!)

Re: Effects of Gen AI on High Skilled Work: Experiments with Software Developers

#456

Post by Sakos »

creativeSlumber wrote: post #202
alternatex wrote: post #5

I wonder if the study includes the technical debt that more experienced developers had to tackle after the less experienced devs have contributed their AI-driven efforts. Because my personal experience has involved a lot of that in one of the companies listed in the study. Also, I've personally seen more interest in AI in devs that have little interest in technology, but a big interest in delivering. PMs love them th…

This. Even without AI, we have inexperienced developers rolling out something that "just works" without thinking about many of the scaling/availability issues. Then you have to spend 10x the time fixing those issues.

If inexperienced developers can commit code like that to master, then there's a culture issue, a process issue or both. This isn't a problem with AI.

Re: Effects of Gen AI on High Skilled Work: Experiments with Software Developers

#457

Post by ugh123 »

dimal wrote: post #402

It matters what you measure. The studies only looked at Copilot usage. I’m an experienced engineer. Copilot is worse than useless for me. I spend most of my time understanding the problem space, understanding the constraints and affordances of the environment I’m in and thinking about the code I’m going to write app. When I start typing code, I know what I’m going to write, and so a “helpful” Copilot autocomplete is…

One thing i've been doing more of lately with Copilot is using prompts directly in a //comment. Although I distinguish this from writing a detailed comment doc about a function and then let Copilot write the function. Theres "inline prompting" and "function prompting".

Re: Effects of Gen AI on High Skilled Work: Experiments with Software Developers

#458

Post by Sakos »

tensor wrote: post #181
GoToRO wrote: post #11

For me AI just brought back documentation. All new frameworks lack documentation big time. The last good one for me was a DOS book! I don't think newer developers even have an idea of what good documentation looks like. Even so, AI will propose different things at different times and you still need an experienced developer to make the call. In the end it replaces documentation and typing.

Yes, it can both help you write it but also, if you start with the documentation, e.g. the comment describing what function does, AI becomes orders of magnitude better and actually helping write the function, so it actually forces developers to better document their code!

I've noticed the same thing. I can get great results by being detailed in my comments and/or docstrings. A lot of the time I don't want the AI actually "thinking" for itself.

Re: Effects of Gen AI on High Skilled Work: Experiments with Software Developers

#459

Post by Sakos »

sum1ren wrote: post #38

When I'm using genai to write some code for me, I lose the internal mental state of what my code is doing. As such, when I do have to debug problems myself, or dream up ideas of improvements, I no longer can do this properly due to lack of internal mental state. Wonder how people who have used genai coding successfully get around this?

Go over the code again and again like you would if you'd written it yourself. Keep iterating on the code yourself without having the AI generate everything. In a current project I have everything internalised because I'm constantly going through the code and improving upon it myself (and using the AI to do boilerplate), even though much of it was initially generated. I'll still have the AI generate code in plenty of places, but I don't let it take over the thinking for me unless I'm unsure about something, then I ask it for possible solutions and ideas, then I go over those solutions myself.

Re: Effects of Gen AI on High Skilled Work: Experiments with Software Developers

#460

Post by acchow »

lagrange77 wrote: post #446
heed wrote: post #435

Earlier quoted context omitted.

I like to use copilot when writing tests. It's not always perfect but makes things less tedious for me.

Excuse my ignorance, i have avoided copilot until now.. Does it have (some of) the other files of the project in it's context, when you use it in a test file?

Copilot has your open tabs in context.

Cursor has that plus whatever files you want to specifically add. Or it has a mode where you can feed it the entire project and it searches to decide which files to add

Post reply on HN