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

Command line functions around OpenAI

kadekillary.work

Post reply on HN
Print view View original

41–50 of 316 posts

Re: Command line functions around OpenAI

#41

Post by Lewton »

nolawi89 wrote: post #21
blagie wrote: post #12

Earlier quoted context omitted.

I find ChatGPT is pretty helpful for debugging. A big chunk of debugging is missing stupid things, and for debugging, it's like having a second set of eyes. It's almost useless for high-level design and architecture. Those tasks do take more time than writing code, but they're super-charged by rapid prototyping, which ChatGPT excels at. For example, right now, if I'm picking between three libraries, I can have workin…

Not yet; I spend 60% my day reviewing PRs. I don't have to if chat GPT gets integrated as a PR reviewer..

ChatGPT will turn your day into spending 90% of your time reviewing code...

Re: Command line functions around OpenAI

#42

Post by BenoitP »

roenxi wrote: post #10

Being efficient at writing code can barely make someone a 5x engineer. A team of people who write code effectively is easy to do. Hyperproductivity in software is all about deciding what problems to tackle. Richard Hipp isn't worried about chipotle restaurant orders in golang, he's worried about how to store data reliably. That isn't a coding puzzle that ChatGTP is likely to help with. Either ChatGTP can do the whole…

I'll add that programmers don't produce code. They produce business guarantees. And their main tool to do so are:

* Understanding the value proposition as a human, often as through a conversation; often with hidden criteria that you'll have to uncover

* Making sure the code respects the value proposition. As reading other people's code takes more time than writing it, ChatGPT won't help that much here.

----

One development to this story I'm anticipating for though is connecting ChatGPT to formal proof systems, and making sure it has sufficient coverage with 'internal thoughts' that are logic-verified (by opposition to killing the problem big-data-style). Sort of like the internal monologue that we have sometimes.

I say short and medium term we're sort of ok; but I don't know about 10 years from now.

Re: Command line functions around OpenAI

#43

Post by austin-cheney »

roenxi wrote: post #10

Being efficient at writing code can barely make someone a 5x engineer. A team of people who write code effectively is easy to do. Hyperproductivity in software is all about deciding what problems to tackle. Richard Hipp isn't worried about chipotle restaurant orders in golang, he's worried about how to store data reliably. That isn't a coding puzzle that ChatGTP is likely to help with. Either ChatGTP can do the whole…

Define “efficient at writing code”. I find that for many people this means something primitive and absurd like typing faster or remembering things better.

Being efficient in the perspective of writing new automation so you don’t have to do the same manual things again can easily make someone a 10x (or greater) engineer. This typically means writing your own tools and not hoping some third party package or framework will do it for you.

Conversely the same is true. If a 10x engineer is in an environment where they are prohibited from writing original tools they will just be average or much worse when their motivation is destroyed.

Re: Command line functions around OpenAI

#44

Post by solomatov »

krsdcbl wrote: post #7
solomatov wrote: post #2

What the articles like this miss is that we understand pretty well how ChatGPT writes code. But what about maintenance, and particularly debugging? How would it make these tasks. And we know that these things take much much more time than writing code.

Plain and simple: Feed the code it produced back into it and ask it to modify the sections needing maintenance or changes, or just tell it the issues you're having and ask it to explain/debug them.

IMO, the problem is that it doesn't work. It could introduce other bugs, make incorrect changes, or mess up the code. If a good developer does this, it converges, with GPT I am not so sure.

We need to find a way for the system to iterate, debug, write unit tests, and repeat the loop. In theory it's doable, but we are yet to see the system which does this.

Re: Command line functions around OpenAI

#46

Post by superb-owl »

There seems to be a lot of dissenting opinion on how these tools will affect software engineering jobs.

Those of us that have deep architectural skills, are able to deal with ambiguous requirements, balance UX with technical limitations, etc, will be safe for a good while.

Those of us that were trained in bootcamps to crank out MEAN/MERN/etc apps might be in a bit of trouble.

I say this not to disparage the latter group--many are great engineers who do great, reliable work. But it's ultimately rote work, something LLMs excel at.

There is a huge tranche of SWEs who are going to become redundant soon. Lots of them gave up careers as teachers etc for a more lucrative career in tech. It's unfortunate.

Re: Command line functions around OpenAI

#47

Post by cod1r »

roenxi wrote: post #10

Being efficient at writing code can barely make someone a 5x engineer. A team of people who write code effectively is easy to do. Hyperproductivity in software is all about deciding what problems to tackle. Richard Hipp isn't worried about chipotle restaurant orders in golang, he's worried about how to store data reliably. That isn't a coding puzzle that ChatGTP is likely to help with. Either ChatGTP can do the whole…

is it me or did he just misspell chatGPT 3 times and nobody noticed?

Re: Command line functions around OpenAI

#48

Post by lordnacho »

superb-owl wrote: post #46

There seems to be a lot of dissenting opinion on how these tools will affect software engineering jobs. Those of us that have deep architectural skills, are able to deal with ambiguous requirements, balance UX with technical limitations, etc, will be safe for a good while. Those of us that were trained in bootcamps to crank out MEAN/MERN/etc apps might be in a bit of trouble. I say this not to disparage the latter gr…

Question is whether you have to go through the bootcamp stage to become an architect? Like many skills, you often start with the simple stuff.

Re: Command line functions around OpenAI

#49

Post by lubujackson »

classified wrote: post #38
Zach_the_Lizard wrote: post #17

Earlier quoted context omitted.

Writing code quickly may not help the mythical 10x engineer, but it might help a 1x engineer become a 5x engineer. In relative terms, the unaided 10x engineer is now only a 2x engineer as his peers are now vastly more productive. That's huge.

... if churning out boilerplate is your definition of "productive".

If churning out boilerplate is what your employers sees as product, does it really matter? Coding udeals are valuable in a vacuum but most companies want those PRs merged, technical debt be damned.

Re: Command line functions around OpenAI

#50

Post by noobermin »

classified wrote: post #38
Zach_the_Lizard wrote: post #17

Earlier quoted context omitted.

Writing code quickly may not help the mythical 10x engineer, but it might help a 1x engineer become a 5x engineer. In relative terms, the unaided 10x engineer is now only a 2x engineer as his peers are now vastly more productive. That's huge.

... if churning out boilerplate is your definition of "productive".

Isn't most code boilerplate belts out the choir.
Post reply on HN