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

GitHub Copilot Workspace: Technical Preview

github.blog

Post reply on HN
Print view View original

41–50 of 333 posts

Re: GitHub Copilot Workspace: Technical Preview

#41

Post by api »

sixhobbits wrote: post #20
studenthrow3831 wrote: post #3

Student here: I legitimately cannot understand how senior developers can dismiss these LLM tools when they've gone from barely stringing together a TODO app to structuring and executing large-scale changes in entire repositories in 3 years. I'm not a singulatarian, but this seems like a brutal S-curve we're heading into. I also have a hard time believing that there is enough software need to make such an extreme prod…

Accountants thought spreadsheets would kill their profession, instead demand for them exploded. Compilers made it much easier to code compared to writing everything in Assembly. Python made it much easier to code than writing C. Both increased the demand for coders. Code is a liability, not an asset. The fact that less technical people and people who are not trained engineers can now make useful apps by generating mi…

This is what I've been predicting for over a year: AI-assisted programming will increase demand for programmers.

It may well change how they do their work though, just like spreadsheets did for accountants and compilers did for the earliest generation of hand-code-in-ASM developers. I can imagine a future where we do most of our coding at an even higher level than today and only dive down into the minutia when the AI isn't good enough or we need to fix/optimize something. The same is true for ASM today-- people rarely touch it unless they need to debug a compiler or (more often) to write something extremely optimized or using some CPU-specific feature.

Programming may become more about higher level reasoning than coding lower level algorithms, unless you're doing something really hard or demanding.

Re: GitHub Copilot Workspace: Technical Preview

#42

Post by jl6 »

throwup238 wrote: post #39
esafak wrote: post #33

I don't use Copilot in particular but in the software space, everything is constantly changing, and the models are trained infrequently so they are never familiar with the things I ask about. Reality is outrunning the models for now but that may change.

There's a Black Mirror-esque movie plot idea in there! Everyone starts using bionic implants to filter out offensive parts of reality like people they don't like and stuff, until one day all the filters start falling out of sync with reality... ...and then aliens invade.

They Live... kinda?

Re: GitHub Copilot Workspace: Technical Preview

#43

Post by thaumaturgy »

studenthrow3831 wrote: post #3

Student here: I legitimately cannot understand how senior developers can dismiss these LLM tools when they've gone from barely stringing together a TODO app to structuring and executing large-scale changes in entire repositories in 3 years. I'm not a singulatarian, but this seems like a brutal S-curve we're heading into. I also have a hard time believing that there is enough software need to make such an extreme prod…

> Are there any arguments that could seriously motivate me to continue with this career outside of just blind hope that it will be okay?

FWIW, as an oldish, so far everything that has been significantly impacted by deep learning has undergone a lot of change, but hasn't been destroyed. Chess and Go are a couple of easy examples; the introduction of powerful machine learning there has certainly changed the play, but younger players that have embraced it are doing some really amazing things.

I would guess that a lot of the same will happen in software. A lot of the scut work will evaporate, sure, but younger devs will be able to work on much more interesting stuff at a much faster pace.

That said, I would only recommend computing as a career to youth that are already super passionate about it. There are some pretty significant cultural, institutional, and systemic problems in tech right now that are making it a miserable experience for a lot of people. Getting ahead in the industry (where that means "getting more money and more impressive job titles") requires constantly jumping on to the latest trends, networking constantly for new opportunities, and jumping to new companies (and new processes / tech stacks) every 18 months or so. Companies are still aggressively culling staff, only to hire cheaper replacements, and expectations for productivity are driving some developers into really unhealthy habits.

The happiest people seem to be those that are bringing practical development skills into other industries.

Re: GitHub Copilot Workspace: Technical Preview

#44

Post by skydhash »

cedws wrote: post #34
packetlost wrote: post #26

Earlier quoted context omitted.

Yes? But it's commonly understood that reading code is harder than writing code. So why force yourself into a reading-mostly position when you don't have to? You're more likely to get it wrong. There are other ways to decrease typing time.

>it's commonly understood that reading code is harder than writing code I don't know about that. Maybe for kernel code or a codec. But I think most people could read (and understand) a 100 line class for a CRUD backend faster than they could write one.

  php artisan make:controller
There are code generators. Even with dealing with other languages library, I mostly copy-paste previous implementations and editing with Vim motions makes it faster.

Re: GitHub Copilot Workspace: Technical Preview

#45

Post by JTyQZSnP3cQGa8B »

studenthrow3831 wrote: post #28
packetlost wrote: post #10

Earlier quoted context omitted.

The hardest part of software development is not writing code, full stop. It never has been and it never will be. The hard part is designing, understanding, verifying, and repairing complex systems. LLMs do not do this, even a little bit.

I guess my worst fear is not "no more jobs because AI can code" but "no more junior jobs because AI can code under the supervision of a senior". SWE jobs will exist, but only seniors will have them and juniors are never hired. Maybe the occasional "apprentice" will be brought on, but in nowhere near the same amount. Where my blind hope lies more specifically is in networking into one of those "apprentice" roles, or m…

IMHO juniors who rely on AI to write code will never learn. You need to make mistakes to learn, and AI never makes mistakes even when it’s wrong.

As a senior, I write code 25% of the time, and it’s always to understand the intent of what I should fix or develop. This is something that AI will not be able to do for a long time since it cannot speak and understand what customers want.

The last 75% of my time are spent refactoring this "intent" or making sure that the business is running, and I’m accountable for it. AI will never be accountable for anything, again for a long time.

I’m scared for juniors that don’t want to learn, but I work with juniors who outsmart me with their knowledge and curiosity.

Re: GitHub Copilot Workspace: Technical Preview

#46

Post by sqeaky »

sixhobbits wrote: post #20
studenthrow3831 wrote: post #3

Student here: I legitimately cannot understand how senior developers can dismiss these LLM tools when they've gone from barely stringing together a TODO app to structuring and executing large-scale changes in entire repositories in 3 years. I'm not a singulatarian, but this seems like a brutal S-curve we're heading into. I also have a hard time believing that there is enough software need to make such an extreme prod…

Accountants thought spreadsheets would kill their profession, instead demand for them exploded. Compilers made it much easier to code compared to writing everything in Assembly. Python made it much easier to code than writing C. Both increased the demand for coders. Code is a liability, not an asset. The fact that less technical people and people who are not trained engineers can now make useful apps by generating mi…

I strongly agree with your points and sentiment as far as the state of machine intelligence remains non-general.

Currently, LLMs summarize, earlier systems classified, and a new system might do some other narrow piece of intelligence. If system is created that thinks and understands and is creative with philosophy and ideas, that is going to be different. I don't know if that is tomorrow or 100 years from now, but that is going to be very different.

Re: GitHub Copilot Workspace: Technical Preview

#47

Post by cedws »

davidbarker wrote: post #12
rgbrenner wrote: post #4

Earlier quoted context omitted.

Thanks for writing that up. It's what I suspected from my use of CoPilot. I love copilot as an autocomplete tool... but it frequently gets things wrong, and using the chat feature to ask it to complete some task usually just generates code that breaks things. So until that improves, Im skeptical a workspace tool would work. Workspace seems like an awesome idea though.. once the tech is further along.

Have you tried Cursor? ( https://cursor.sh ) It's a fork of VS Code with some AI features sprinkled in. It writes around 80% of my code, these days. It also has a few useful features: - a chat interface where you can @-mention files, folders, and even documentation - if you edit a line of code, it suggests edits around that line that are useful (e.g. you change a variable name and it will suggest updating the other u…

Looks interesting, but I don't really want my code to go via some unknown company. As far as I can tell in "Privacy Mode" code still goes via their servers, they just promise not to store anything (with the caveat that OpenAI retain stuff for 30d).

Re: GitHub Copilot Workspace: Technical Preview

#48

Post by jonahx »

studenthrow3831 wrote: post #3

Student here: I legitimately cannot understand how senior developers can dismiss these LLM tools when they've gone from barely stringing together a TODO app to structuring and executing large-scale changes in entire repositories in 3 years. I'm not a singulatarian, but this seems like a brutal S-curve we're heading into. I also have a hard time believing that there is enough software need to make such an extreme prod…

Serious answer to a legitimate question:

1. Good senior developers are taking the tools seriously, and at least experimenting with them to see what's up. Don't listen to people dismissing them outright. Skepticism and caution is warranted, but dismissal is foolish.

2. I'd summarize the current state of affairs as having access to an amazing assistant that is essentially a much better and faster version of google and StackOverflow combined, which can also often write code for well specified problems. From what I have seen the current capabilities are very far from "specify high-level business requirements, get full, production app". So while your concern is rational, let's not exaggerate where we actually are.

3. These things make logical errors all the time, and (not an expert) my understanding is that we don't, at present, have a clear path to solving this problem. My guess is that until this is solved almost completely human programmers will remain valuable.

Will that problem get solved in the next 5 years, or 10, or 20? That's the million dollar question, and the career bet you'll be making. Nobody can answer with certainty. My best guess is that it's still a good career bet, especially if you are willing to adapt as your career progresses. But adapting has always been required. The true doom scenario of business people firing all or most of the programmers and using the AI directly is (imo) unlikely to come to pass in the next decade, and perhaps much longer.

Re: GitHub Copilot Workspace: Technical Preview

#49

Post by a_t48 »

studenthrow3831 wrote: post #3

Student here: I legitimately cannot understand how senior developers can dismiss these LLM tools when they've gone from barely stringing together a TODO app to structuring and executing large-scale changes in entire repositories in 3 years. I'm not a singulatarian, but this seems like a brutal S-curve we're heading into. I also have a hard time believing that there is enough software need to make such an extreme prod…

For adding and refactoring it can be a great tool. For greenfield development, it's more tricky - yesterday I sat down and started writing something new, with no context to give Copilot. Mid sourcefile, I paused to think about what I wanted to write - it spit out three dozen lines of code that I then had to evaluate for correctness and just ended up throwing away. I could have probably helped the process by writing docs first, but I'm a code first, docs second kind of guy. Totally sold on LLM written unit tests though, they are a drag to write and I do save time not writing them by hand.

It's going to be a bit before LLMs can make an app or library that meets all requirements, is scalable, is secure, handles dependencies correctly, etc, etc. Having an LLM generate a project and having a human check it over and push it in the right direction is not going to be cheaper than just having a senior engineer write it in the first place, for a while. (I could be off-base here - LLMs are getting better and better)

I'm not worried about being replaced, my bigger worry is in the mean time the bottom end falling out of the engineering market. I'm worried about students learning to program now being completely dependent on LLMs and never learning how to build things without it and not knowing the context behind what the LLM is putting out - there's definitely a local maxima there. A whole new "expert beginner" trap.

Re: GitHub Copilot Workspace: Technical Preview

#50

Post by erksa »

studenthrow3831 wrote: post #28
packetlost wrote: post #10

Earlier quoted context omitted.

The hardest part of software development is not writing code, full stop. It never has been and it never will be. The hard part is designing, understanding, verifying, and repairing complex systems. LLMs do not do this, even a little bit.

I guess my worst fear is not "no more jobs because AI can code" but "no more junior jobs because AI can code under the supervision of a senior". SWE jobs will exist, but only seniors will have them and juniors are never hired. Maybe the occasional "apprentice" will be brought on, but in nowhere near the same amount. Where my blind hope lies more specifically is in networking into one of those "apprentice" roles, or m…

You can not have seniors without juniors.

Interacting with a computers (and therefor creating software) will probably soon detach itself from the idea of single chars and the traditional QWERTY keyboard.

Computing is entering a fascinating phase, I'd stick around for it.

Post reply on HN