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

91–100 of 333 posts

Re: GitHub Copilot Workspace: Technical Preview

#91

Post by idan »

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…

How many students / people early in career would benefit from having something to help them explore ideas?

How many don't have the advantages I had, of a four-year university, with professors and TAs and peers to help me stumble through something tricky?

How many have questions they feel embarrassed to ask their peers and mentors because they might make them look stupid?

Don't give up. This is a generational opportunity to lift up new developers. It's not perfect (nothing is). But if we sweat hard enough to make it good, then it is our chance to make a dent in the "why are there not more ______ people in tech" problem.

Re: GitHub Copilot Workspace: Technical Preview

#92

Post by happypumpkin »

> At last count, Copilot had over 1.8 million paying individual and 50,000 enterprise customers.

> Copilot loses an average of $20 a month per user, according to a Wall Street Journal report, with some customers costing GitHub as much as $80 a month.

Presumably at some point they need to actually make money on this? That is a $432 million / yr loss just on individual users.

Re: GitHub Copilot Workspace: Technical Preview

#93

Post by idan »

a_t48 wrote: post #75
idan wrote: post #71

Earlier quoted context omitted.

So, part of the trickiness here is that there's a few different moving pieces that have to cooperate for success to happen. There needs to be a great UX to elicit context from the human. For anything larger than trivial tasks, expecting the AI to read our minds is not a fruitful strategy. Then there needs to be steerability — it's not just enough to get the human to cough up context, you have to get the human to corr…

Yeah to be clear I think Copilot Workspace is a great start. I wonder if the future is multi-modal though. Ignoring how obnoxious it would be to anyone near me, I could foresee narrating my stream of thoughts to the mic while using the keyboard to actually write code. It would still depend on me being able to accurately describe what I want, but it might free me from having to context switch to writing docs to hint t…

I mean we explored that a little with Copilot Voice :D https://githubnext.com/projects/copilot-voice/

But yeah, the important part is capturing your intent, regardless of modality. We're very excited about vision, in particular. Say you paste a screenshot or a sketch into your issue...

Re: GitHub Copilot Workspace: Technical Preview

#94

Post by sdesol »

alfalfasprout wrote: post #87
sdesol wrote: post #84

Interesting. I don't think the AI code generation will live up to developer expectations, but I do see the value in "project management for developers". The value I see with workspace is not the code generation, but the ability to help developers organize their thoughts.

The real problem is a large part of the enormous hype train is investors thinking they can replace talented engineers w/ AI. That's fundamentally doomed for failure. What AI can really do well is take an already competent engineer and suddenly get rid of a lot of the annoying tedium they had to deal with. Whether it's writing boilerplate, doing basic project management, organizing brain dumps/brainstorming, etc.

I do think workspace by GitHub is a smart move. What is clear as day is, we don't have enough data to model how developers think and iterate to solve problems. With workspace, I see it as a way to capture structured data on how developers think.

This is certainly a long game though. I think GitHub with MS money can continue to lose money on Copilot for the next 5 years to gather data. For other VC ventures, I don't think they can wait that long.

Re: GitHub Copilot Workspace: Technical Preview

#95

Post by danenania »

For anyone who might be interested in an open source, terminal-based approach to using AI on larger tasks and real-world projects, I'm building Plandex: https://github.com/plandex-ai/plandex

I've tried to create a very tight feedback loop between the developer and the LLM. I wanted something that feels similar to git.

Apart from the planning and code generation itself, Plandex is also strongly focused on version control. Every interaction with the model is versioned so that it's easy to try out different strategies and backtrack/revise when needed.

I just released a new update (literally a few minutes ago) that includes some major improvements to reliability, as well as support for all kinds of models (previously it had been OpenAI-only). It's been fun trying out Claude Opus and Mixtral especially. I'd love to hear people's thoughts!

Re: GitHub Copilot Workspace: Technical Preview

#96

Post by trashface »

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…

Either way you're going to want to have a backup career plan. By 40 if not earlier you could be forced out of tech by ageism or something else. Unless you transition into management, but even then. I don't think middle management is going to be immune to AI-enabled employment destruction. So you basically should plan to make most of your money from software in the first decade or two. Live cheap and save/invest it.

Best to plan and train early because its super hard to switch careers mid life. Trust me, I'm failing at it right now.

Re: GitHub Copilot Workspace: Technical Preview

#97

Post by troupo »

michaelmior wrote: post #67
skydhash wrote: post #58

Earlier quoted context omitted.

SQL and python are arguably the languages closest to English, and even then getting someone to understand recursion is difficult. How do you specify that some values should be long lived? How do you specify exponential retries. Legalese tries to be as specific as possible without being formal and even then you need a judge on a case. Maybe when everyone has today’s datacenter compute power in their laptop.

> arguably the languages closest to English Yes, but they're not English. All the concerns that you mention are ones that I think LLM development tools are aiming to eliminate from explicit consideration. Ideally, a user of such a tool shouldn't even have to have ever heard of recursion. I think we're a long way off from that future, but it does feel possible.

Have you ever actually tried getting proper non-contradictory requirements in pain natural language from anyone?

Good luck

Re: GitHub Copilot Workspace: Technical Preview

#98

Post by 999900000999 »

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…

Simple.

It's the 90%, 10% theory.

LLMs will do the 90% that is easy, the final 10% it'll get wrong and will insist on it's solutions being correct.

If anything this is horrible for junior level developers. A senior dev now has a restless junior developer at their whim.

As far as your own career, I'd argue to finish your degree, but be aware things are about to get really rough. Companies don't like headcount. Even if it's not true today, in the future AI + 1 senior engineer will be faster than 4 juniors + 1 senior.

Re: GitHub Copilot Workspace: Technical Preview

#99

Post by troupo »

Menu_Overview wrote: post #37
skydhash wrote: post #29

Earlier quoted context omitted.

Software Engineering is not a subset of computer science, they just intersect. And as a software engineer, your job can be summarized as gathering requirements and designing a solution, implementing and verifying said solution, and maintaining the solution in regards to changes. And the only thing AI does now is generating code snippets. In The Mythical Man Month, Brooks recommend to spend 1/3 of the schedule to plan…

> and it would have been faster to just do it. False. Obviously this depends on the work, but an LLM is going to get you 80-90% of the way there. It can get you 100% of the way there, but I wouldn't trust it, and you still need to proof read. In the best of times, it is about as good as a junior engineer. If you approach it like you're pair programming with a junior dev that costs <$20/mo then you're approaching it c…

> Obviously this depends on the work, but an LLM is going to get you 80-90% of the way there.

No. No it can't.

However amazing they are (and they are unbelievably amazing), they are trained on existing data sets. Anything that doesn't exist on StackOverflow, or is written in a language slightly more "esoteric" than Javascript, and LLMs start vividly hallucinating non-existent libraries, functions, method call and patterns.

And even for "non-esoteric" languages it they will wildly hallucinate at every turn apart from some heavily trodden paths.

Re: GitHub Copilot Workspace: Technical Preview

#100

Post by godzillabrennus »

alfalfasprout wrote: post #87
sdesol wrote: post #84

Interesting. I don't think the AI code generation will live up to developer expectations, but I do see the value in "project management for developers". The value I see with workspace is not the code generation, but the ability to help developers organize their thoughts.

The real problem is a large part of the enormous hype train is investors thinking they can replace talented engineers w/ AI. That's fundamentally doomed for failure. What AI can really do well is take an already competent engineer and suddenly get rid of a lot of the annoying tedium they had to deal with. Whether it's writing boilerplate, doing basic project management, organizing brain dumps/brainstorming, etc.

These AI coding assistants already do what you are describing. At some point in the future, AI should be capable enough to do the job of a competent developer. This is not a short-term milestone, IMHO.
Post reply on HN