TE
科技回声
首页24小时热榜最新最佳问答展示工作
GitHubTwitter
首页

科技回声

基于 Next.js 构建的科技新闻平台,提供全球科技新闻和讨论内容。

GitHubTwitter

首页

首页最新最佳问答展示工作

资源链接

HackerNews API原版 HackerNewsNext.js

© 2025 科技回声. 版权所有。

A coder considers the waning days of the craft

778 点作者 jsomers超过 1 年前

155 条评论

cglan超过 1 年前
Maybe I’m in the minority. I’m definitely extremely impressed with GPT4, but coding to me was never really the point of software development.<p>While GPT4 is incredible, it fails OFTEN. And it fails in ways that aren’t very clear. And it fails harder when there’s clearly not enough training resources on the subject matter.<p>But even hypothetically if it was 20x better, wouldn’t that be a good thing? There’s so much of the world that would be better off if GOOD software was cheaper and easier to make.<p>Idk where I’m going with this but if coding is something you genuinely enjoy, AI isn’t stopping anyone from doing their hobby. I don’t really see it going away any time soon, and even if it is going away it just never really seemed like the point of software engineering
评论 #38257679 未加载
评论 #38259086 未加载
评论 #38257583 未加载
评论 #38257680 未加载
评论 #38258891 未加载
评论 #38259620 未加载
评论 #38257834 未加载
评论 #38257548 未加载
评论 #38259535 未加载
评论 #38257533 未加载
评论 #38257667 未加载
评论 #38257714 未加载
评论 #38267570 未加载
评论 #38260292 未加载
评论 #38259736 未加载
评论 #38259908 未加载
评论 #38259959 未加载
评论 #38269856 未加载
评论 #38267359 未加载
评论 #38259356 未加载
评论 #38260313 未加载
评论 #38263676 未加载
评论 #38260199 未加载
评论 #38257614 未加载
评论 #38260441 未加载
评论 #38259681 未加载
评论 #38267725 未加载
评论 #38270659 未加载
评论 #38267324 未加载
评论 #38259560 未加载
skepticATX超过 1 年前
Am I the only one becoming less impressed by LLMs as time passes?<p>I will admit, when Copilot first became a thing in 2021, I had my own “I’m about to become obsolete” moment.<p>However, it’s become clear to me, both through my own experience and through research that has been conducted, that modern LLMs are fundamentally flawed and are not on the path to general intelligence.<p>We are stuck with ancient (in AI terms) technology. GPT 4 is better than 3.5, but not in a fundamental way. I expect much the same from 5. This technology is incredibly flawed, and in hindsight, once we have actual powerful AI, I think we’ll laugh at how much attention we gave it.
评论 #38259484 未加载
评论 #38261111 未加载
评论 #38259525 未加载
评论 #38260055 未加载
评论 #38259043 未加载
评论 #38260330 未加载
评论 #38260375 未加载
评论 #38261054 未加载
评论 #38259457 未加载
评论 #38274720 未加载
评论 #38267104 未加载
miiiiiike超过 1 年前
I have a simple front-end test that I give to junior devs. Every few months I see if ChatGPT can pass it. It hasn’t. It can’t. It isn’t even close.<p>It answers questions confidently but with subtle inaccuracies. The code that it produces is the same kind of non-sense that you get from recent bootcamp devs who’ve “mastered” the 50 technologies on their eight page résumé.<p>If it’s gotten better, I haven’t noticed.<p>Self-driving trucks were going to upend the trucking industry in ten years, ten years ago. The press around LLMs is identical. It’s neat but how long are these things going to do the equivalent of revving to 100 mph before slamming into a wall every time you ask them to turn left?<p>I’d rather use AI to connect constellations of dots that no human possibly could, have an expect verify the results, and go from there. I have no idea when we’re going to be able to “gpt install &lt;prompt&gt;” to get a new CLI tool or app, but, it’s not going to be soon.
评论 #38260522 未加载
评论 #38261087 未加载
评论 #38262168 未加载
评论 #38261038 未加载
评论 #38260715 未加载
评论 #38261104 未加载
评论 #38263930 未加载
评论 #38260635 未加载
评论 #38262857 未加载
评论 #38262732 未加载
评论 #38262741 未加载
评论 #38261075 未加载
评论 #38261063 未加载
评论 #38260437 未加载
评论 #38265540 未加载
评论 #38263826 未加载
评论 #38261350 未加载
评论 #38266258 未加载
评论 #38260861 未加载
评论 #38270005 未加载
评论 #38262250 未加载
评论 #38260869 未加载
评论 #38266998 未加载
评论 #38262503 未加载
评论 #38265934 未加载
评论 #38261560 未加载
评论 #38263968 未加载
评论 #38260405 未加载
评论 #38263630 未加载
评论 #38264091 未加载
评论 #38260643 未加载
评论 #38260754 未加载
评论 #38265534 未加载
评论 #38260898 未加载
评论 #38276575 未加载
评论 #38263292 未加载
评论 #38261315 未加载
评论 #38261837 未加载
评论 #38260453 未加载
评论 #38260660 未加载
评论 #38261109 未加载
评论 #38262383 未加载
评论 #38260552 未加载
评论 #38263727 未加载
评论 #38262594 未加载
评论 #38261309 未加载
评论 #38261084 未加载
评论 #38267144 未加载
happytiger超过 1 年前
Do people seriously consider this the waning days of the craft? I don’t understand that.<p>My view is that I am about to enter the quantum productivity period of coding.<p>I am incredibly excited about AI assistance on my coding tasks, because it improves not only what I’m writing, but also helps me to learn as I go. I have never had a better time writing software than I have in the last year.<p>I’ve been writing software for a few decades. But now I’m able to overcome places where I get stuck and have almost a coach available to help me understand the choices I’m making and make suggestions constantly. And not just wandering over to a fellow cuders desk to ask them about a problem I am facing, but actually give me some productive solutions that are actually inspirational to the outcome.<p>It’s amazing.<p>So why do people think that coding is coming to some kind of end? I don’t see any evidence that artificial intelligence coding assistants are about to replace coders, unless you… suck badly at building things, so what are people getting on about?<p>I feel like somebody came along and said, “foundations are now free, but you still get to build a house. But the foundations are free.”<p>I still have to build a house, and I get to build an entire house and architect it and design it and create it and socialize it and support it and advocate for it and explain it to people who don’t understand it but… I don’t have to build a foundation anymore so it’s easier.<p>Shoot me down. I’m not relating here at all.
评论 #38260706 未加载
评论 #38260600 未加载
评论 #38260756 未加载
评论 #38260354 未加载
评论 #38260346 未加载
评论 #38260760 未加载
评论 #38260983 未加载
评论 #38263723 未加载
评论 #38265472 未加载
评论 #38260698 未加载
评论 #38265156 未加载
评论 #38261391 未加载
评论 #38267209 未加载
评论 #38263648 未加载
评论 #38267158 未加载
EZ-E超过 1 年前
I don&#x27;t see it - and by that I don&#x27;t mean I don&#x27;t think AI can write good code and get better over time. I just don&#x27;t see how it would work as a workflow to replace (most) devs by AI.<p>If I take a junior programmer&#x27;s task, say, creating CRUD endpoints. Describing the requirement in a way that matches exactly what I want will probably take more time that doing the coding assisted by something like copilot. Can we really imagine a non technical user using an AI having it do development from A to Z? What if the generated code has a bug, can we really imagine that at no point someone will need to be in the loop? Even if a tech person intervenes in the case of a bug how much time would be lost to investigate what the AI wrote and trying to understand what happened in retrospect - the time or cost saved to write the code would be lost quickly. Writing code is a small part of the job after all. LLMs are good at generating code, but they are fundamentally not problem solvers.<p>The technology is amazing but I think LLMs will just be another tool in the arsenal for devs. It&#x27;s also an amazing tutor. It can avoid having to call a developper for some self contained problems (writing a script to scrape content from a web page for example).
评论 #38259493 未加载
评论 #38259842 未加载
评论 #38259669 未加载
simonw超过 1 年前
&gt; What I learned was that programming is not really about knowledge or skill but simply about patience, or maybe obsession. Programmers are people who can endure an endless parade of tedious obstacles.<p>This captures the reason I&#x27;m optimistic about AI-assisted programming.<p>The learning curve for getting started programming is horribly steep - and it&#x27;s not because it&#x27;s hard, it&#x27;s because it&#x27;s frustrating. You have to sweat through six months of weird error messages and missing semicolons before you get to the point where it feels like you&#x27;re actually building things and making progress.<p>Most people give up. They assume they&#x27;re &quot;not smart enough&quot; to learn to program, when really they aren&#x27;t patient enough to make it through all of that muck.<p>I think LLMs dramatically impact that initial learning curve. I love the idea that many more people will be able to learn basic programming - I think every human being deserves to be able to use computers to automate tedious repetitive tasks in their lives.
评论 #38259269 未加载
评论 #38260049 未加载
评论 #38261020 未加载
评论 #38262099 未加载
评论 #38259276 未加载
decasia超过 1 年前
I have been having the following debate with my friend who does AI and neural network stuff:<p>Him: Coding will soon be obsolete, it will all be replaced by chatgpt-type code gen.<p>Me: OK but the overwhelming majority of my job as a &quot;senior engineer&quot; is about communication, organizational leadership, and actually understanding all the product requirements and how they will interface with our systems. Yes, I write code, but even if most of that were augmented with codegen, that would barely even change most of what I do.
评论 #38258797 未加载
评论 #38257946 未加载
评论 #38259663 未加载
评论 #38258567 未加载
评论 #38257743 未加载
评论 #38260529 未加载
评论 #38258766 未加载
评论 #38259775 未加载
ken47超过 1 年前
It feels like this article was not written by a programmer, and it feels like a number of the commenters are not professional engineers. What part of a programmer jobs can AI realistically replace in the near term?<p>For the sake of argument, let’s say it could replace the coding part <i>cost effectively</i>. Can it still do all the other parts? Take ambiguous requirements and seek clarity from design, product, etc. (instructing an AI to a sufficient degree to build a complex feature could almost be a coding task itself). Code reviews. Deal with random build failures. Properly document the functionality so that other programmers and stakeholders can understand. Debug and fix production issues. And that’s just a subset.<p>Realistically, in the future there will be a long phase of programmers leveraging AI to be more efficient before there’s any chance that AI can effectively replace a decent programmer.<p>This will be an advantage to engineers who think on the more abstract side of the spectrum. The “lower level” programming tasks will be consumed first.
评论 #38260369 未加载
评论 #38261212 未加载
评论 #38260215 未加载
评论 #38260407 未加载
评论 #38267259 未加载
waldohatesyou超过 1 年前
I feel like the headline does not match the article here. The headline implies that programming as a craft is to be replaced but the article ultimately argues that it will change significantly which matches my intuition as well.<p>At the end of the day, the bar is being lowered. Is that a bad thing? From a selfish perspective, yes. From a societal perspective, no. At the risk of digressing, I think one of the issues that my part of the world (Canada and to a lesser extent, America) has been faced with is inequality. I know people who work more &quot;average&quot; service jobs and they make substantially less than engineers do and that&#x27;s something that&#x27;s made me pretty uneasy over the past few years. The societal value of generative AI is in making knowledge work such as law, medicine, and software engineering much more accessible to &quot;average&quot; people.<p>Are there downsides to that? Probably but I think granting power evenly is probably a better path to utopia than misguided elitism. The latter sounds like the path to despotism.
评论 #38258787 未加载
评论 #38257838 未加载
评论 #38259341 未加载
glonq超过 1 年前
As a grayhead who does not code daily anymore, I was impressed with ChatGPT as a programming sidekick. It&#x27;s like having an on-call junior developer who only costs $20&#x2F;mo.<p>I needed a quick and dirty utility developed last month and by breaking the problem down into 4-5 steps myself I got ChatGPT to write functions for those steps and then I bolted them together. Most of it went smooth but one part took far too much coaxing and refinement before it generated the right thing.
sensanaty超过 1 年前
Whenever I see people talking about how they use ChatGPT to write entire software suites, I feel like I&#x27;m taking crazy pills or something. I can&#x27;t even get GPT4 to give me decent basic ass Ruby scripts where it opens a file that doesn&#x27;t have some weird behavioral quirk or footgun, yet people often claim that they&#x27;re doing insane feats with it all the time.<p>My experience, even if I&#x27;m just trying to be lazy and have it generate some super generic code for me, is that I have to spend double the time I&#x27;d have spent if I just wrote the damn script myself with how much I have to scrutinize every single line it generates.<p>And I personally have 0 desire to prompt &quot;engineer&quot; and do the &quot;Open&quot;AI&#x27;s work for it by thinking up the perfect prompt for it. One day for sure these AI models will be really good and able to do some impressive stuff, but for now all I can see is a hype wave from people with vested interests in this stuff pushing it onto ignorants, akin to the crypto BS just with even more money behind it.
评论 #38271275 未加载
评论 #38261782 未加载
Barrin92超过 1 年前
I mean this in the least offensive way possible, but whenever one of these flowery ChatGPT pieces comes out, it&#x27;s always written like this<p><i>&quot;at one point, we wanted a command that would print a hundred random lines from a dictionary file. I thought about the problem for a few minutes, and, when thinking failed, tried Googling. &quot;[...]<p>I returned to the crossword project. Our puzzle generator printed its output in an ugly text format, with lines like &quot;s&quot;&quot;c&quot;&quot;a&quot;&quot;r&quot;&quot;</i>&quot;&quot;k&quot;&quot;u&quot;&quot;n&quot;&quot;i&quot;&quot;s&quot;&quot;<i>&quot; &quot;a&quot;&quot;r&quot;&quot;e&quot;&quot;a&quot;. I wanted to turn output like that into a pretty Web page that allowed me to explore the words in the grid, showing scoring information at a glance. But I knew the task would be tricky[...]</i><p>This man has written software professionally for 20 years? The last part of the article is at least correct. Code generation isn&#x27;t going to replace programmers. Almost all SNES and NES games were written in Assembly. Modern game devs learn Unity and Unreal and visual scripting. Are there now more or fewer game devs? Writing a few lines of code that generate <i>a metric ton</i> of more code is what most of us have been doing for many years now. Abstraction and tooling does not change the nature of the profession and it certainly doesn&#x27;t end it.
yonisto超过 1 年前
I&#x27;m currently re-writing a GPT4 code which is a very interesting experience.<p>The biotech startup I join 3 months ago have a pipeline written by one of the founders, who has experience in programing but not extensive so he used co-pilot and GTP a lot. The thing WORKS, it actually delivers! this is what got the startup going so it is amazing.<p>On the first month of the re-write I was able to cut costs in the 100x mark. And on the third month I was able to uncover some subtle bugs that affect the entire pipeline and fixing them is out of the realm of possibility in the current state of AI.<p>I have no answer whatever employers will continue to value what I bring to the table in 5&#x2F;10&#x2F;15 years time especially after AI will improve. I suspect many will not as they will have &#x27;good enough&#x27; results from the AI.
评论 #38260523 未加载
评论 #38262119 未加载
bruce511超过 1 年前
Obviously we can discuss the merits of GPT4, but actually my first issue is with his root question;<p>&gt;&gt; I would make sure that my kids could program computers. It is among the newer arts but also among the most essential<p>I&#x27;ve got kids. 1 can program. 1 tried it, and didn&#x27;t like it. 1 just rolled her eyes at me.<p>Turns out, for most people, programming is not like reading or writing at all. Got a question? Ask Google (or GPT) or Alexa. 99% of people can&#x27;t code, and will never need to code.<p>You may as well argue that cooking healthy food is a universally needed skill. Or riding a bicycle, or kicking a ball. All are of course really helpful life skills, but lots of people get by in the world just fine without them.<p>So as one parent to another, I can tell you that you don&#x27;t need to worry about this. It&#x27;s good you have plans for your children. But don&#x27;t worry too much about it. Turns out children will learn what they are interested in, and pretty much ignore the rest. They&#x27;ll grow up, change, mature and become functional people in their own right.<p>No doubt you&#x27;ll get plenty of eye rolls along the way.
评论 #38259418 未加载
RayVR超过 1 年前
I think more people in these comments should read the whole article before responding. The author intentionally takes several turns to arrive at a nuanced view, with a final statement that I think most here would agree with: &quot;Hacking is forever&quot;
评论 #38259611 未加载
alexmuro超过 1 年前
I don&#x27;t feel like we are in the waning days of the craft at all. Most of the craft is creating an understanding between people and software and most human programmers are still bad at it. AI might replace some programmers but none who program as a craft.
评论 #38257847 未加载
jelsisi超过 1 年前
The article came to a romantic end that doesn&#x27;t satisfy how I feel about the workplace now. I have to admit a dirty truth, that I joined this field because of the pay and job stability, not simply for a love of programming. Because of this I feel distressed for its future. No longer can someone walk into a job slightly under qualified and grow with time. Companies will opt for smaller teams of 10xers that with the companionship of AI can pump out 10x10 more code then the rest of us. Simply put, as a 20 something out of college, I&#x27;m going to need to figure out how to make as much money as soon as possible.
评论 #38258617 未加载
评论 #38259169 未加载
评论 #38258860 未加载
cnity超过 1 年前
I like to think of a simple thought experiment when these types of arguments come up: Suppose you want some video game to exist. The most amazing, expansive, ground-breaking video game ever made. There are only two scenarios:<p>1. AI can replace coders altogether. In this case we live in an absolutely amazing world in which anyone can conjure the video game of their dreams instantly.<p>2. AI cannot do this, in which case there is still (clearly) an essentially infinite demand for new software.<p>Obviously, we do not live in world number 1. If we have arrived at the world in which AI can do the job of all coders, then why is there not an infinite supply of high quality software? In fact, this would be a software utopia.
评论 #38261923 未加载
FigurativeVoid超过 1 年前
I may be in the minority, but after the initial “wow” period, I have been underwhelmed with co-pilot.<p>Don’t get me wrong, there are few times it has really really impressed me. And there are a few things it really shines at, but most of the time I find it getting in the way.<p>I’m not so concerned about the craft of programming. But those that make a good living automating trivial tasks should be more worried.
评论 #38257688 未加载
评论 #38258903 未加载
评论 #38257631 未加载
betaby超过 1 年前
Article itself feels like it was written by LLM - unnecessarily long, pretentious and boring without much of a substance. But hey, folks from this forum aren&#x27;t the target audience for &#x27;Newyorker&#x27;!
评论 #38259696 未加载
评论 #38257520 未加载
评论 #38257512 未加载
评论 #38260345 未加载
pdimitar超过 1 年前
It&#x27;s quite amazing how divisive the AI topic turned out to be in HN.<p>My opinion is that AI proponents extrapolate too much and too optimistically; they always assume linear or simply uninterrupted improvement, and they assume they breadth of AI will increase as well. I see no proof for any of those so far.<p>Time will tell, obviously, and they may very well turn out to be correct. I just wonder why people get so worked up about it.<p>Personally I also don&#x27;t understand why people want the programmer profession to cease to exist but oh well. I guess envy and schadenfreude are factors. I wouldn&#x27;t know because I don&#x27;t wish anyone unemployment.
评论 #38259352 未加载
评论 #38269432 未加载
minzi超过 1 年前
It seems like lots of people are pointing out the inadequacies of the current state of the art models. It doesn&#x27;t seem that unlikely that many of these problems will be solved in the coming years. I won&#x27;t make any guesses about timelines, but there is certainly a large number of very well funded smart people working on these problems. I&#x27;m not sure if this current path leads to AGI, which is what I would consider the minimum requirement for truly replacing human jobs. However, it does suggest to me that building AGI is achievable.<p>In my mind there are two major architectural problems that still need to be solved if I am to be convinced AGI is close:<p>1. Medium term memory<p>2. The need to relive an entire conversation just to produce the latest response.<p>Here is another question: Suppose we develop AGI. What will the energy requirements be for that system? The scale at which ChatGPT is used today would be peanuts compared to a system that is supposed to replace millions of knowledge workers.
评论 #38263535 未加载
layman51超过 1 年前
I’m so confused by this article’s comparison of GPT-4 to a powerful Go or chess engine. First, is coding some sort of competition like Go or chess? I have heard of competitive programming or code golf, but have little clue how that compares to “the craft of code” or why using GPT-4 diminishes it. Is it cheating if I use GPT-4’s assistance to write code? On the other hand, I do see why it would be harmful for people who don’t have the knowledge or experience to be relaying answers from an LLM to others on forums.<p>Second, I am seeing a parallel here with the college educators being preoccupied with how to teach calculus. Yesterday, I was reading the preface that Martin Gardner wrote to “Calculus Made Easy” by Silvanus P. Thompson and he wrote something along the lines of how many students and even educators (circa 1998) do not see the point in knowing how to differentiate or integrate by hand. I suppose now more than ever, some people even suggest to stop teaching calculus to high school (and even college students) in favor of some kind of vague statistics, data science, or information literacy course. But most mathematicians would discourage that kind of approach.<p>I think I agree with the last point about how it is not really about the knowledge of code itself but rather it is about being curious and knowing enough to not let yourself be taken advantage of.
评论 #38258187 未加载
GabeIsko超过 1 年前
Do people have problems with Leetcode stuff or something? Aside from generating code I already know how to write before hand quickly, I find AI super useless for coding. It&#x27;s never been able to solve a single problem I have worked on - mostly involving infrastructure deployment and config.
评论 #38259193 未加载
评论 #38259319 未加载
bdamm超过 1 年前
Anyone ought to be able to see the new truth; that AI is multiplying the power and importance of the programming and software engineering discipline, by allowing us to churn out far more software in far less time. Entire genres of code tasks will get wiped out (data adaption layers come to mind.)<p>Software engineers, even if just reduced to AI-bot controllers, will still be essential links between people who have no idea how computers work and the actual machines.
comfortabledoug超过 1 年前
I&#x27;m seeing a lot of concerns about the future of programmers and their employability. Remember that nobody is doing you a favor. People get paid what they&#x27;re worth, including programmers. The thing about imposter syndrome is it sneaks up on you when you&#x27;re surrounded by other talented coders all the time. You start forgetting that what you do isn&#x27;t common; not everyone can whip up code like you can. When AI starts shaking up the world of programming, that’s when you know it’s really game over for all other white-collar jobs.<p>Today&#x27;s big tech companies might seem invincible, but they’re not. Their edge? It&#x27;s going to wear thin, especially once everyone else starts jumping in with custom solutions. Instagram with just 13 people gave Facebook a run for its money. That was before the era of LLMs. Imagine now – anyone could potentially build the next big thing in an afternoon.<p>And don’t get me started on the legal stuff. There are so many laws, no one person can know them all. But LLMs? They can handle it. What happens when the legal system gets flooded with cases about loopholes no one knew existed? Or when everyone can have the world&#x27;s best lawyer for free? It&#x27;s wild to think about.<p>ChatGPT came out of nowhere for most of us, and robotics is going to hit us harder and sooner than we realize. No politician or law can keep up with this tech freight train we&#x27;re on. We’re talking about more progress in the next 20 years than in the last 20,000. Jobs and whether we can stay employed? That&#x27;s going to be the least of our worries.<p>Or AI will hit a plateau and we&#x27;ll all get a little bit more productive.
blauditore超过 1 年前
Wake me up when the current AI hype is over. Eventually people will realize that getting the last few percent right is extremely hard, just like the last few times where we got just so close to AGI.<p>What will stay are some nice tools for whatever tasks, maybe programming too. But bots won&#x27;t replace software engineers. If they could do the current tasks, we&#x27;d just engineer one level higher, just like when high-level language compilers made low-level code largely obsolete (yes yes, not completely of course).
lysecret超过 1 年前
So, a few months ago I wanted to lose some weight and wanted to count my calories. As a coder would, I thought why not build a swift app that uses the ChatGPT API to count them (instead of doing it the boring way).<p>I&#x27;m a fairly experienced backend dev I built a few toy websites and I am casually interested in frontend stuff (like watching Fireship) so I thought I&#x27;d give it a try.<p>Long story short, it worked well. I have my app it has all the features I need I used it every day (till I lost 10kg) and I didn&#x27;t write a single line of code.<p>However, the coding itself was unlike anything I had done before. I felt much more like a manager&#x2F;Slave driver.<p>The core thing was structure structure structure. As soon as the task gets too large, GPT will produce something kind of right that is difficult to understand and impossible to extend. So your task is to split it up, and split it up. Also to apply judgment (it started out using objective-c haha). You have to ask it for alternative libraries. For reference. I relied a lot on something like: &quot;In Python&#x2F;C#&#x2F;Rust I would do it like this&quot;.<p>But yes, my overall experience is that I think we are in for an exciting new time, but we are lacking a new coding paradigm.
评论 #38261788 未加载
uticus超过 1 年前
ChatGPT is another layer. Human -&gt; ChatGPT -&gt; high level programming language -&gt; lower level language &#x2F; IL &#x2F; assembly -&gt; microcode -&gt; circuit.<p>ChatGPT will not only survive, it will get bigger as long as all those layers exist. Unless web design is ready to move to assembly level instructions via RPC (ChatGPT can do assembly), the layers will stay, the complexity will stay, and people who get (and attempt to sell) value from recombining the puzzle pieces will stay.<p>The real point is not that ChatGPT is so good, it is that everything we program has already been done before. We humans just have trouble finding the pieces and gluing them together.
karolist超过 1 年前
&gt; At one point, we wanted a command that would print a hundred random lines from a dictionary file. I thought about the problem for a few minutes, and, when thinking failed, tried Googling.<p>No offence to the author but given that this stopped them in their tracks I choose to ignore their programming related opinions.
armchairhacker超过 1 年前
I think people underestimate the level of general intelligence required to write even seemingly basic &quot;glue-code&quot; (which btw, 90+% of software consists of).<p>If we ever get an AI which can write complex enough &quot;glue code&quot; with a low enough failure rate to replace even junior developers, it&#x27;s going to cause much, much bigger impacts than people losing jobs. The &quot;boilerplate&quot; and &quot;bureaucracy&quot; which is currently handled by &quot;average&quot; folk...that&#x27;s what runs the world. If we ever develop AI which can handle that kind of work, barring some constraint like extreme resource usage or disaster like climate change; we&#x27;ll either be brought into a utopia due to productivity being multiplied, or a dystopia due to a small minority getting control over the world to a precise, extreme extent no small minority has control now.
评论 #38258034 未加载
评论 #38259710 未加载
karaterobot超过 1 年前
If coding goes away as a profession or as a craft, it would be a fitting end to a story arc that entailed the end of so many other people&#x27;s jobs. We&#x27;ve had a good, high-flying run for fifty years, not really thinking too hard about what progress for us meant for other people. I&#x27;m not convinced yet that it will, but I&#x27;m also not sure how many people outside the profession would weep for the loss.
fzeroracer超过 1 年前
Much like how many people predicted we&#x27;d all be driving flying cars, the people predicting that coding will be replaced by AI just isn&#x27;t realistic. Primarily because these AI models can literally only exist as long as there&#x27;s humans constantly creating code for it to read (see: steal) in the first place.<p>AI cannot sustain itself trained on AI work. If new languages, engines etc pop up it cannot synthesize new forms of coding without that code having existed in the first place. And most importantly, it cannot fundamentally rationalize about what code does or how it functions.<p>The more you use it or try to integrate it into your workflow (or worse, have others try to integrate it on their own) the more the inherent flaws of the LLMs come into play.
评论 #38257928 未加载
zwieback超过 1 年前
FYI - many public libraries let you auto-checkout the New Yorker to your free Overdrive account and you can read online or on your phone. I was surprised how easy it has become.
gemstones超过 1 年前
Imagine a doctor 100 years post-Galen &quot;considering the waning days of medicine&quot;. We are a young, young discipline.
zwieback超过 1 年前
The stuff he describes should rightly be replaced by an AI. However, I think we can still level up. After initial excitement about GPT-generated code I realize that the danger is that we will very quickly generate a lot of code with subtle bugs that will take even longer to debug. If I write the code myself I (or whoever I have look over my shoulder) will be quicker finding the bugs. If it&#x27;s GPT code I&#x27;ll first have to dig through the idiosyncrasies, then find the logic error.<p>Also: writing multi-threaded or otherwise concurrent code is hard to imagine even with GPT 6 or 7.
gtirloni超过 1 年前
Today I already have trouble doing git archeology to understand why a piece of code was done in a certain way.<p>Now imagine debugging more and more code that was created by LLMs.<p>Debugging will become more of a voodoo witchcraft kind of thing. I pity anyone who decides to make a living out of that. Probably will earn a lot and die young.
评论 #38257911 未加载
评论 #38269461 未加载
评论 #38259328 未加载
lionkor超过 1 年前
LLMs can be very impressive, but &quot;coding&quot; can be a lot more than traversing a btree or sorting an array.<p>There is software engineering, which is probably what most well paid devs do a little bit too much of, which is requirements analysis, planning, talking to the customer, estimating, etc. Sure an AI could help with all of that, but leaving that entirely up to an AI probably wont work for a very long time. You&#x27;d need a real AGI to take over work that has that many social aspects and requires that much planning and consideration.<p>There is algorithms and datastructures, which is important for developers to learn, but usually not important to write yourself unless thats literally your job. There, LLMs could help with understanding, possibly visualizing these algorithms, and providing example use cases.<p>There are hobby projects, which generally wont be done with any AI by most people I know, since the point is to program for fun. AI can help with that, e.g. for inspiration or example implementations of little weird algorithms.<p>There is maintenance. Once a product, especially a B2B product, reaches maturity, a lot of the work is bugfixing, smaller features, updating dependencies, porting to new platforms, etc. I could see LLMs taking some of this work away, which everyone would probably appreciate.<p>In the end, I believe that AI and in particular LLMs just help programmers, and will maybe replace the 6€&#x2F;hour outsourced copy paste programmers, but that is... not very scary.
unclebucknasty超过 1 年前
When using ChatGPT for some tedious challenges—say, select coding problems and infrastructure setup issues on AWS—I have more than once had the feeling that people weren&#x27;t built for this kind of work anyway, and that it&#x27;s exactly the kind for which computers were intended.<p>It wasn&#x27;t always this way. Sure, programming has always been an exercise in precision and, thus, sometimes tedium. But, there&#x27;s something in the way it has changed over the last decade or so. When I first started my career, I spent long hours solving problems of the algorithmic, logic, or control flow type. It was challenging mind work and called for creativity. There was real joy in making something work. It felt like creation.<p>Now, programming consists largely of wiring together pieces of code that you didn&#x27;t write, and consulting StackOverflow to figure out why it doesn&#x27;t work. It&#x27;s following someone else&#x27;s decisions in an opinionated framework. It&#x27;s transpiling and tooling and configuration. And, now that programmers are expected to do much more, it&#x27;s also using someone else&#x27;s software for devops and tedious infrastructure buildouts, etc.<p>I&#x27;d say the joy in the software engineering role faded long before the arrival of ChatGPT. I&#x27;m actually hopeful that AI will help lighten the load of mundane tedium and help us return to the days when programming was fun.
slowhadoken超过 1 年前
Programming, like math, is its own reward. I think being ignorant of the two is a luxury with a hidden cost.
gwoolhurme超过 1 年前
As someone who bounces back and forth between I will be okay to anxiety that I am doomed. I have no idea how to prepare for this future...
ryukoposting超过 1 年前
GPT-4, copilot, etc are <i>absolutely worthless</i> for firmware. Copilot shows no meaningful capacity for understanding the context of a firmware codebase, and forget generating anything coherent based on your instructions. I don&#x27;t need help writing function declarations, but that&#x27;s the only thing it&#x27;s useful for (and even that is heavily limited).
delbronski超过 1 年前
I used chatgpt4 recently for two projects. One was a success, the other a complete failure.<p>In the first project I asked it to help me build a one page ui with html and css, it did great! I know a lot about html and css, so I was able to ask for what I wanted and help it debug.<p>For the second project I wanted what I thought was a simple threejs animation of an object following a line in a scene. I know very little about threejs and graphics in general. I spent hours wrestling with chatgpt to get this working, but I just got what looked like meaningless non-working code to me. So I took time to learn about threejs, did a few chapters of a course, and came back to the problem. And all of the sudden I had new vocabulary I could use to ask chatgpt what I wanted. And I got pretty close to what I wanted, but not quite. I suspect if I keep learning about threejs and how graphics work I’ll get there.<p>I don’t think this is the death of our craft, but definitely a big turning point for how we build and use software.
dimgl超过 1 年前
I don&#x27;t think it&#x27;s that black and white. AI will get progressively better and sure, people will be faster to deliver things. But there still needs to be someone to be able to decipher it all. I&#x27;ve been using GPT-4 tons and it&#x27;s great to help get to a certain destination, but it isn&#x27;t able to arrive at that destination by itself a majority of times.<p>Additionally I think that a lot of people will assume that once they get GPT-4 to do something for them that they&#x27;re done. That&#x27;s not necessarily true. There&#x27;s a lot of complexity to navigate everywhere. And AI can help you navigate it! But I don&#x27;t think it means that I, a software engineer, can now pivot to being a lawyer, for instance, solely because of GPT-4.<p>Sure, it&#x27;s able to do a lot already. Maybe I&#x27;m being naive. But I see it more as a tool for the future rather than something that is going to automate people out of existence.<p>Edit: hm, why downvotes? If I&#x27;m wrong, help me improve my viewpoint on AI.
评论 #38258801 未加载
interpol_p超过 1 年前
I use ChatGPT daily and have been coding for over two decades. I get it to write all the tedious stuff, and it basically writes all my internal tools, scripts and so on. The last big things it wrote for me were:<p>A macOS SwiftUI app to go through all the localization strings in my apps and show them in a table where I could see missing translations, and interactively translate them… by sending the, back to ChatGPT<p>Another macOS app to graph Google ngram data for word lists. It was great because I didn’t even need to check whether there was an API for ngrams, I just asked it for the function, then for the code to plot it in Swift Charts<p>Sometimes I consult it on refactoring problems (eg., this struct has new responsibilities, what would be a more appropriate name? How do you see it fitting into this section of the code?). It’s a great tool and easily worth the money<p>It almost never writes any “interesting” code, however
feoren超过 1 年前
&gt; At one point, we wanted a command that would print a hundred random lines from a dictionary file. I thought about the problem for a few minutes, and, when thinking failed, tried Googling. I made some false starts using what I could gather, and while I did my thing—programming—Ben told GPT-4 what he wanted and got code that ran perfectly.<p>This is pretty basic stuff. It should not even require Googling unless you don&#x27;t really know the language. This supports my belief that ChatGPT is simply going to replace the <i>very bad</i> programmers. But we never really needed them to begin with. I have no faith that GPT-4&#x27;s code actually &quot;ran perfectly&quot;; this programmer clearly does not have the chops to tell whether or not the code was correct. And we&#x27;re talking about, like, 100-level CS homework here.
m463超过 1 年前
I&#x27;ve noticed that over the years, coding has gotten easier.<p>I&#x27;ve been able to express myself better with each year.<p>I see AI as another higher-level way of expressing myself.<p>Also, people don&#x27;t write as much new code as they think they do.<p>Still, I have many many... MANY projects that I have on my wishlist. Maybe AI will help, but my list won&#x27;t go away.
评论 #38260733 未加载
jmull超过 1 年前
I’m disappointed, but not really surprised I guess, with how poorly so many programmers seem to understand what chatgpt is and does.<p>It’s great and amazing tech, but it’s very much a kind of search engine, and indeed, writes just the kind of code you find on the internet… sometimes great, often flawed, very often incomplete. Its breadth of knowledge is stunning, but it’s not like it’s actually useful to be able to write code in all the languages.<p>It really doesn’t even touch harder problems. You need to know, at the level of fine details often, what you want from it and what you don’t. That’s hard.<p>…It save’s time. We will perhaps need fewer programmers as people spend less time on grunt work. But that’s hardly the end of coding, and seems like clearly a good thing in the long run.
评论 #38262783 未加载
notjoemama超过 1 年前
ChatGTP cannot provide an actual solution to a multi-tenant application framework for a SaaS product...because I haven&#x27;t found that in the stack I work in. So I created it but it is closed source. If LLMs rely on open source, it will never provide the value that exists from experienced closed source engineers. I suppose LLM&#x2F;AI providers can snaekily suck in closed source code for long enough time that once they&#x27;re caught the cat is out of the bag and the &quot;industry&quot; collapses. But so far, LLMs are cool, but not AGI, nor are they likely to mature into AGI alone.<p>As far as generalization goes, I&#x27;m still waiting for an answer from Tolman-Eichenbaum machines. Or, an integration of LLMs and cognitive maps.<p><a href="https:&#x2F;&#x2F;www.youtube.com&#x2F;watch?v=9qOaII_PzGY">https:&#x2F;&#x2F;www.youtube.com&#x2F;watch?v=9qOaII_PzGY</a><p><a href="http:&#x2F;&#x2F;brilliant.org&#x2F;ArtemKirsanov&#x2F;" rel="nofollow noreferrer">http:&#x2F;&#x2F;brilliant.org&#x2F;ArtemKirsanov&#x2F;</a><p>And his references from the YouTube video:<p>REFERENCES (in no particular order): 1. Behrens, T. E. J. et al. What Is a Cognitive Map? Organizing Knowledge for Flexible Behavior. Neuron 100, 490–509 (2018). 2. Constantinescu, A. O., O’Reilly, J. X. &amp; Behrens, T. E. J. Organizing conceptual knowledge in humans with a gridlike code. Science 352, 1464–1468 (2016). 3. Aronov, D., Nevers, R. &amp; Tank, D. W. Mapping of a non-spatial dimension by the hippocampal–entorhinal circuit. Nature 543, 719–722 (2017). 4. Whittington, J. C. R., McCaffary, D., Bakermans, J. J. W. &amp; Behrens, T. E. J. How to build a cognitive map. Nat Neurosci 25, 1257–1272 (2022). 5. Whittington, J., Muller, T., Mark, S., Barry, C. &amp; Behrens, T. Generalisation of structural knowledge in the hippocampal-entorhinal system.<p>CREDITS: Icons by biorender.com and freepik.com Brain 3D models were created with Blender software using publicly available BrainGlobe atlases (<a href="https:&#x2F;&#x2F;brainglobe.info&#x2F;atlas-api" rel="nofollow noreferrer">https:&#x2F;&#x2F;brainglobe.info&#x2F;atlas-api</a>)<p>This video was sponsored by Brilliant
andsoitis超过 1 年前
&gt; But he had no idea what a pain it is to make an iPhone app. I’d tried a few times and never got beyond something that half worked. I found Apple’s programming environment forbidding.<p>A seasoned programmer who struggles with this?
tabtab超过 1 年前
&gt; <i>had no idea what a pain it is to make an iPhone app. I’d tried a few times and never got beyond something that half worked. I found Apple’s programming environment forbidding.</i><p>What&#x27;s missing these days is domain-specific languages (DSL). Maybe domain-specific API&#x27;s (libraries) can be made better, but I have not seen it done well yet.<p>Domain specific languages made it easy and DIRECT to do things you wanted done in that domain, no weird-ass reverse dependency factory visitor injection design patterns or whatnot, you only had domain-oriented commands and the basics of functions, loops, and conditionals to glue them together.<p>Sure, they didn&#x27;t scale well to &quot;enterprise&quot;, but if that&#x27;s not what you want, so be it! Global variables didn&#x27;t kill kittens at that scale like &quot;real&quot; software engineers warn against. I barely had to read the manual when learning VB classic, Microsoft Access (pre XML), Paradox, etc. it was so intuitive and almost everything related to the domain (desktop GUI&#x27;s and small-ish CRUD).<p>It was almost like pseudo-code! That&#x27;s the secret, the closer it is to domain-specific-pseudo-code, the easier it is to learn and the less bloated the code.<p>Wanna make a successful tool for non-giant orgs or groups? Focus on the domain and only the domain; don&#x27;t let fads and buzzwords and mega-scaling distract you.
Dave3of5超过 1 年前
&gt; Ben told GPT-4 what he wanted and got code that ran perfectly<p>&quot;perfectly&quot;? I mean this sounds dubious. ChatGPT gives interesting suggestions and advice when it comes to coding but I&#x27;ve never been able to get any close to a perfect solution.<p>I don&#x27;t think coding AI assistant&#x27;s are meant to be used in that way anyway...<p>I think this is just another fluff piece about how coding is now dead (long live coding). That&#x27;s basically not true. LLM have gave us a leg up but they haven&#x27;t removed the need for a competent programmer.
cantSpellSober超过 1 年前
It&#x27;s funny, back in the days of IE6, Firebug, jQuery, it was so frustrating just getting something to <i>work</i>.<p>Looking back, that&#x27;s what made it fun. You could convince yourself you were a mad scientist building the next Turing machine.<p>Having stability in the JS ecosystem now (hold for laughter) is better for users, and the developments in AI are interesting, but it&#x27;s not as challenging to ship things. Maybe embracing the <i>new</i> challenges is the &quot;craft&quot; we should focus on to bring the fun back.
tj-teej超过 1 年前
&quot;At one point, we wanted a command that would print a hundred random lines from a dictionary file. I thought about the problem for a few minutes, and, when thinking failed, tried Googling. I made some false starts using what I could gather, and while I did my thing—programming—Ben told GPT-4 what he wanted and got code that ran perfectly.&quot;<p>Sure ChatGPT can do this, but let&#x27;s not pretend this is a problem requiring much &quot;craft&quot;.
zacksiri超过 1 年前
I think it depends on what you do, if you are doing repetitive work between projects, you&#x27;re going to find that yes GPT can do most of the work you&#x27;re doing. However if you&#x27;re creating new systems, with multiple components and parts. You&#x27;re going to find that GPT is not that much help.<p>I pay for GPT-4 and Github Copilot, all I can say is when I&#x27;m trying to do repetitive work like writing tests, copilot swoops in and gives me &#x27;mostly&#x27; working code. But when I go in to the real creative stuff, that needs real thought and is non-obvious, I don&#x27;t get much help from copilot it either just doesn&#x27;t suggest anything or suggests completely the wrong thing.<p>I think at least for me, the comparison I would use is, construction work. There are many different kinds of constructions, anything simple from making ikea furniture to building bridges. If the coding you do is the equivalent of making ikea furniture. Then yes that stuff is going away, because it&#x27;s going to be replaced by machines that can produce those parts in bulk.<p>But if you&#x27;re creating something special like the Burj Khalifa, or the Golden gate bridge equivalent but in code, no AI is not going to do that for you. It can help with some parts, but no way AI is going to do that sort of engineering, at least not in it&#x27;s current state.<p>I think it&#x27;s worth noting &quot;Coding&quot; spans so many kinds of systems and scale. Not all coding work is the same. Even if you&#x27;re talking about things which are well-known like &quot;e-commerce&quot; in that industry there are so many variations &#x2F; many different scales &#x2F; many different ways to look at the problem.
zubairq超过 1 年前
As a developer many people ask me if I am worried about AI making my job obsolete. I always answer truthfully, &quot;I CAN&#x27;T WAIT for AI to make my job obsolete!&quot;.<p>After trying several AI tools in the hope of replacing my coding job I find that the tools are good at the bike-shedding tasks like generating some initial code (the easy stuff for me) and terrible at the real work (maintenance and adding features to an existing codebase).
todd3834超过 1 年前
&gt; At one point, we wanted a command that would print a hundred random lines from a dictionary file. I thought about the problem for a few minutes, and, when thinking failed, tried Googling. I made some false starts using what I could gather,<p>If this is the skill level of the coder then I’m not surprised that they are worried about GPT. However they should be stoked GPT can help them code at a much higher level
benreesman超过 1 年前
I was trying to get whisper.cpp built under buck2 this weekend. Unsurprisingly nothing from GPT-4 to valiant to vanilla llama knew how to do that. I got it done but no model (search or LLM) was any help.<p>I only need help when it’s hard. I’ve already got a BERT-style model completing my queries on Google.<p>If your entire job could be done by riffing on stackoverflow answers? Yeah, I’d try to climb the value chain.<p>But most people’s job isn’t to copy paste shit.
type4超过 1 年前
&gt; But taking a bird’s-eye view of what happened that day? A table got a new header. It’s hard to imagine anything more mundane. For me, the pleasure was entirely in the process, not the product.<p>I think that&#x27;s where I differ from a lot of the artful programmers, I&#x27;ve never found pleasure in a perfect, beautiful solution. I get annoyed when I have to dump hours and hours into something that &#x27;should&#x27; be simple! I don&#x27;t want to spend my time fiddling with layouts, CSS, Oauth handshakes, etc. I want to build stuff and get paid for it, that&#x27;s how I view my job as well. Less logician and more mechanic.<p>I use ChatGPT as much as I can, to do all I can and then fix the output when it&#x27;s needed. I view it as a higher level programming language, that spares me from the burden of thinking of low-level details. It&#x27;s same the reason I code in Python&#x2F;Javascript rather than C++ or other languages, the goal is to make something of use. That&#x27;s my goal to stay employed, become one with the language du jour, in the same way I&#x27;ve jumped from PHP to Jquery to React to ...
4death4超过 1 年前
I share the sentiment with others here that have become disillusioned with LLMs. Yes, their capabilities are impressive, and their progression is quite striking. But what I have realized, about programming specifically, is that I don&#x27;t think LLMs can do a better job than humans.<p>To many futurists hopping on the hypetrain, this is probably somewhat of a hot take. But let me explain.<p>Ultimately, programming is about instructing a computer on what to do. We use programming languages because they&#x27;re highly efficient representations of what a computer can do. In the case of LLMs, we need to use natural language to specify the program. For simple things, this can work great. But for complex programs, it&#x27;s not easy to articulate exactly what a program should do. As you drill in to the details, you begin to experience an almost exponential growth in the amount of natural language specification you need. And so I think there is something very flawed in believing that you can replace human programmers with LLMs. In my personal opinion, a single programmer will be able to outproduce a single LLM prompt engineer, purely because programming languages are better at expressing ideas compared to natural language. At a certain point, it becomes faster to write code then to write sufficiently detailed prompts.<p>As a thought experiment for the reader: consider a complex system you&#x27;ve worked on. And now try to imagine writing down a specification for the system in natural language that an LLM could use to generate the code that powers the system. I honestly don&#x27;t think I could do it for many of the system I&#x27;ve worked on.<p>So yes, I think LLMs are cool. And I think they&#x27;re cool because they give creative power to people who can&#x27;t program. But for sufficiently complex things, i.e. things people care about, a natural language interface isn&#x27;t the answer.
评论 #38259491 未加载
评论 #38259409 未加载
devit超过 1 年前
I think it&#x27;s more fun to program with LLMs, because rather than having just a single programming phase, you basically have three quite different phases: a first phase where you craft an initial prompt, then a second phase where you review the LLM code and ask it to make changes, and then a third phase where you change the code yourself to exactly what you want it to be.
bigpeopleareold超过 1 年前
Instead of ChatGPTing things in what I work in, I have sought to improve my knowledge a lot in shell scripting, I am reading man pages more diligently (even if I am not dilgent enough) and seek to use documentation first, such as man pages, software documentation, etc. before even going to search for something. I am not perfect, but I feel like most of the problems I have are present in manuals - I am RTFMing myself.<p>I do this because I wasn&#x27;t like that in the past and I hate myself for not being more deliberate in reading and discovering things organically through doing things. This is not about joy, (even though it makes me happy) but my obligation to myself and then to my employer to be good at what I do.<p>Sitting with a chat bot undermines this urge to be a better developer completely. I will probably be argued with, but I have worked with a lot of smarter people, but I realized after many years, (and what the article alludes to), that they have a very strong sense to fight through myriads of almost random problems, but they have the knack at solving those problems intelligently.<p>I remember specific co-worker that used an even plainer Emacs config than I did to write Java code. I was genuinely curious what his thought process was when trying to remember APIs without anything more special than simple modes for jumping around a codebase. He said &quot;I don&#x27;t know, I just read the manual first.&quot;<p>Yes! If he can do that, I can do that. After that, a lot with working with computers is about continually pushing through those random problems to their solution. This is an absolutely precious skill. I don&#x27;t feel the affected yet, but I will be, I expect - lesser quality searches while we have to push our chat bots sub-sub-problems instead of just understanding what we are using. I don&#x27;t want to and I will continue to hold out, like all the other essentially dumb technologies I am now stuck with (I am looking at my two mobile phones now ...)
agentultra超过 1 年前
If your trade as a programmer is in automating tasks with computers... I can see how it might seem like OpenAI is your Squarespace. Most users were never going to learn to code no matter how well we designed scripting languages, macros, etc for them. Just like there was nothing stopping them from learning HTML or using any of the available tools for building their own websites. They still paid people to do it for them. Now we&#x27;re at a moment where they don&#x27;t need to pay such people anymore... just like when Squarespace-et-al killed off the need to hire a web developer to build your website.<p>I don&#x27;t think Djikstra would be <i>appalled</i> in the sense that the author gives it. He wasn&#x27;t saying that C++ or Python were <i>precise</i> languages; they wouldn&#x27;t be precise enough for him. He wouldn&#x27;t be appalled that we don&#x27;t write programs in them. He would be appalled that we don&#x27;t write precise specifications.<p>ChatGPT-4 can generate code. Is it the <i>correct</i> code? We don&#x27;t know unless we specify what <i>correct</i> means in a precise and unambiguous way. Natural language isn&#x27;t precise enough to do that. We have a language for that called, <i>mathematics</i>. ChatGPT-4 is not very good at understanding or speaking this language for the time being.<p>What matters is whether ChatGPT-4-generated-code is good enough for your problem. Generate a script to organize some files or help you with small tasks where correctness isn&#x27;t crucial? Go ahead. I don&#x27;t pull out a whiteboard and fire up a theorem prover when I&#x27;m organizing a photo collection.<p>However, if I&#x27;m working on a system and have a theory that we can parallelize an algorithm over a complex data structure and it&#x27;s critical that this is done correctly or the system cannot be trusted? I wouldn&#x27;t think twice before firing up TLA+ or Lean 4.<p>For now, ChatGPT-4 cannot reason and think for you. The industry isn&#x27;t over, in my opinion. It&#x27;s annoying. It&#x27;s noisier. But not over.
Anamon超过 1 年前
Another writer mistaking LLMs for AI.<p>As a programmer, I&#x27;m paid to solve novel problems, things that can&#x27;t be solved by just glueing existing things together. But that&#x27;s all LLM-generated code can do. And it&#x27;s a qualitative problem, not a quantitative one. By its very nature, an LLM cannot come up with something novel. Which is why all of the examples in this article are highly formulaic and derivative ideas based on existing stuff, interesting as they may be to the people asking for them.<p>I can get quite well-running scripts from GPT for well-defined, menial tasks that I know would be easy to do, but just can&#x27;t be bothered to fight with the mechanics and syntax in order to get it right. But for solving new and specific problems, that doesn&#x27;t work.<p>Plus, technologies and languages evolve. If LLMs displaced programming, it would mean that progress has stopped. No LLM is going to come up with a nifty new technique or feature. It can only regurgitate what many ceative people first had to come up with.<p>You can have LLMs do your boilerplate, your helper scripts or your clearly defined, self-sufficient algorithms for you. But a real-world application is already impossible because an LLM will always be incapable of covering the big picture, of assembling parts into a whole that doesn&#x27;t only work and make sense, but is also performant, secure, maintainable and extensible. LLMs won&#x27;t kill the craft of programming, it will just lead those who fall for that to producing loads of crap that doesn&#x27;t fulfill any of the above requirements, and then end up having to be trashed and redone from scratch by &quot;real programmers&quot;.<p>The Go analogy doesn&#x27;t work for the same reason. Go has well-defined rules everybody has to follow, and clear goals. Attacking this algorithmically is the extreme opposite of the idea of general artificial intelligence. That doesn&#x27;t mean it&#x27;s easy, but it&#x27;s not even remotely a related problem.<p>There&#x27;s a reason why the people swooning over GPT-based coding are always script kiddies, not professional developers.
评论 #38282820 未加载
JonChesterfield超过 1 年前
There appears to be consensus that these models generate code very quickly, relative to a person typing, and that the code emitted varies from totally wrong to subtly wrong. Thus they&#x27;re going to significantly increase the amount of code in production and that code is going to tend to be of the subtly wrong kind.<p>This is my non-web-dev impression of the javascript world. Loads of code out there, basically all of it wrong in a bunch of places. Reasoning by analogy with that, one should expect the reliability of computer systems to go downhill from here.<p>The interesting question for a professional software dev is where one can add value in a world in which essentially anyone can produce large amounts of code that sometimes works. My hunch is that it&#x27;s going to be really important that the languages and libraries in use are totally solid as they&#x27;re essentially being fuzz tested continually.
rixed超过 1 年前
&gt; At one point, we wanted a command that would print a hundred random lines from a dictionary file. I thought about the problem for a few minutes, and, when thinking failed, tried Googling.<p>Isn&#x27;t the real waning days of the craft the fact that we consider this problem sort of &quot;state of the art&quot; ?
nercury超过 1 年前
The more you use GPT, the more you will understand that it&#x27;s not the replacement for your attention to the work. And without the attention to the work, you can&#x27;t spot bugs, blatant inefficiencies, or better design choices - in other words, if you don&#x27;t take care, you won&#x27;t even know what you are missing.<p>I write code with GPT every day for almost a year now, and it helped me greatly to kickstart code that was hard in the past, namely, audio synthesis, vulkan rendering, and other things that were hard to approach. But it&#x27;s very clear at this point that it shall not be trusted, because it&#x27;s like a coat of a nice paint put on top of all the clever or dumb stuff that exists on the internet. You never know which one you get, but sure it will be worded convincingly.
totallywrong超过 1 年前
I suspect we&#x27;re going to see this article in many different forms over the next few years. That&#x27;s when the hype cycle will settle and people realize that LLMs are amazing search engines on steroid buy will never replace actual developers.
fallingfrog超过 1 年前
You have to consider here that anyone posting on hacker news is a pretty smart cookie. We’re the clever kids in school, the ones who were curious and insightful.<p>Now, chatgpt in its current form cannot think better than me. Faster, for sure, but not better. Its arguments and its explanations are a bit simple and it can’t make really insightful observations. Mostly it just states the obvious.<p>However.<p>There <i>are</i> people in the world that it <i>can</i> out-think. For every one of us there are 10 or 20 people who are pleasant enough but kind of thick, and gpt can already speak more eloquently than them. So where’s the bar for “intelligence?” Does it have to outclass the smartest human alive before we can call it intelligent?
underscoring超过 1 年前
I know we&#x27;re all biased here, as coders, but I don&#x27;t see this as the end, but rather the beginning of something.<p>In the article, the non-coder was able to produce code with an AI&#x27;s help, and therefore, coders are doomed.<p>But in fact what happened is it enabled two &quot;developers&quot; to offload the menial work and focus on the product itself.<p>The development of electric workshop tools didn&#x27;t make woodworkers obsolete, it just enabled them to do the work of 5 people, and do it easier and faster.<p>We are moving from the era of hand-tools in coding.<p>The real risk is that we&#x27;re going to see a lot of cheap shitty furniture made by people with no skills, but at least ChatGPT isn&#x27;t going to chop anyones fingers off.
tayo42超过 1 年前
Whats the workflow for writing code like the app mentioned with chat? My experience was like the authors, its a massive environment set up, plus even small projects have 10s of files. How does the model keep all that in context plus changes. How do you start a new session? I checked some small cpp project and the code alone is 168,000 words, so thats already larger then the gpt4 context, that doesn&#x27;t even include iteration and chat history. IDK where you would even get started.<p>Also, dont really see the comparison between programming and chess. In chess you can still compete to be the best human chess player. Coding or working in corporate america doesnt really have that.
reactordev超过 1 年前
(as they know it). As someone who also learned to code in the same time era as the author. The craft that we were taught is dying and a new way emerging. Programming has already changed drastically since the 90s. Exceptions around low level languages, everything else is orders of magnitude faster, more efficient, more succinct, and more abundant. Combine this was AI assistance and it’s like you have super powers of productivity.<p>Cry about how you no longer have to code the small and mundane, but solving problems and writing software is more than just coding small mundane things. It’s how those mundane things are architected and called to do a really big thing.
DeepSeaTortoise超过 1 年前
I am quite surprised nobody has called out the elephant in the room yet:<p>A much superior solution than even the next few AI generations has been available for several decades now. You could just outsource your work to a contractor in a developing nation.
mediumsmart超过 1 年前
Why do people think the AI needs to learn how to write programs like a human? It’s going to be a framework with well documented chunksize modules the AI can iterate on using tests it writes itself. People will be out of work because they can’t write those bugfree modules the startup needs for the slots in its version of the AI provided code base framework and if they could the company has no time or money to deal with a human programmer.<p><i>that being said it might still be a non issue if civilization as we know it has less time than currently projected</i>
intrepidsoldier超过 1 年前
* AI is good at hard (to human) tasks like explaining complex code but bad at easy (to human) tasks like counting C files in a codebase. * AI is particularly effective at taking over repetitive work full of toil e.g., reviewing large PRs. As a result, it will allow more developers to focus on fun things like building new features. * AI will increase the number of developers worldwide because it reduces the barrier to entry for programming. * AI will reduce the fear involved in starting large-scale projects like migrating codebases from one framework&#x2F;technology to another.
2devnull超过 1 年前
I’m not going to use gpt, at least not for coding. That’s not something I’m hearing other programmers say, at least not often. They may say it’s not good enough, yet, or that they as a senior dev don’t have any use for it. These strike me as cope. What I say is that I’m just not going to use it. I’m not sure if it’s a political decision I’ve made, but I’m sure it’s the right decision for me. I will simply change careers or go homeless before I try and compete with these great new efficiencies. I won’t take part in the disruption. It’s clearly degenerate.
ern超过 1 年前
I manage a number of development teams, and I don’t get to write much code anymore, but I was asked to urgently build an integration.<p>The team lead I was going to assign the task to was offshore, and not yet online, so I put the URL for the documentation into ChatGPT premium, and it spat out a working solution.<p>My team lead was a bit perturbed and created a more enterprisy solution with factories and DI (more out of pride than necessity), but my original code was production ready.<p>I’m not going to be firing anyone just yet, and certainly not my senior people, but I can see a lot of grunt-level work going away.
评论 #38260687 未加载
MarkusWandel超过 1 年前
It&#x27;s glorified autocomplete. The stuff that is annoying when you&#x27;re at a proper keyboard with proper typing skills, but maybe not so bad when you&#x27;re thumb typing on a smartphone. I&#x27;ve not personally used it, but someone else with very advanced skills, needing to code something in an unfamiliar programming language, asked ChatGPT to do it, got a &quot;mostly right&quot; boilerplate and fixed up the obvious errors, which was much quicker than learning the language enough to start from scratch. Autocomplete!
colinmegill超过 1 年前
Melodramatic, pretentious, they just love these puff pieces outside of tech
brailsafe超过 1 年前
The job market this year, even as someone with a fair bit of experience, is already dogshit for other reasons. I do see the writing on the wall, and am considering re-starting from scratch in something completely different, but I don&#x27;t know what yet. Though I don&#x27;t see dramatic leaps in capability happening, opportunity, or productivity happening as much as labor price reducing even further and competition increasing to the point where I don&#x27;t know if it&#x27;ll be worth trying.<p>Carpentry seems like growing field though.
calibas超过 1 年前
Until there&#x27;s actual general AI, we&#x27;ll still need people that understand how things work on a fundamental level. You can use AI created code, but you still need someone to tie everything together. You&#x27;re also going to want an experienced person to debug when something goes wrong.<p>I think it&#x27;s funny how terrified people are at the prospect of AI and machines doing most of the work for human beings. If used intelligently and equitably, it&#x27;ll mean more free time and prosperity for everyone.
评论 #38260051 未加载
summerisle超过 1 年前
Regardless of the quality of work done by an LLM, all of this AI hype has done an absolutely stellar job of clearly dividing the world of software development into two cohorts: those who care (to literally any degree, about process or product), and those who are only involved in it because it makes money. I have never quite liked the latter, because they are often difficult to work with, tend to lack any motivation to develop an in-depth understanding of certain concepts or problems.
评论 #38260000 未加载
评论 #38260043 未加载
abhaynayar超过 1 年前
&gt; As he later put it, his own neural network had begun to align with GPT-4’s.<p>&gt; Having found the A.I.’s level, I felt almost instantly that my working life had been transformed. Everywhere I looked I could see GPT-4-size holes; [...]<p>I think this is key for me at least. GPT-4 is really good once you start figuring out what problems NOT to give it. Because there are certain things that it can&#x27;t solve no matter how much you prompt it in the right direction, and it gets really annoying really fast.
teaearlgraycold超过 1 年前
Personally I’m not worried about my employability. Until AGI arrives (which may coincide with an apocalypse for all we know) AI will simply allow me to move faster. If it lets me move 1000x faster then the art will be in what to build. 1000x still does nothing compared to an exponentially exploding field of possibilities. And technical knowledge, experience working as a product focused engineer, and general human intelligence should be enough to beat AI at that art form.
rglover超过 1 年前
<i>Puts hands behind head</i><p>I&#x27;m just going to wait while the bootcamp set phases themselves out of the market and then wait for the &quot;it doesn&#x27;t work any more, please fix it&quot; emails.<p>Assuming they price their skills at a premium, cleaning up the messes that AI will create in the process of it being treated as The Holy Grail are going to make some people <i>absurdly</i> wealthy.
rapind超过 1 年前
One of the concerns I have with AI assisted programming is it might make us more accepting of badly designed APIs and libraries, and eventually APIs designed for AI intuitiveness rather than human intuitiveness (assuming these don&#x27;t always align). If AI takes away the pain of working with badly designed APIs (good), it will also change our perception and acceptance of these badly designed APIs (bad).
sean_the_geek超过 1 年前
My two cents having used AI in programming. It is merely a tool to create. It is not the creator yet, at least in its current iteration.<p>AI is to programming what electric drill were to wood working before they were invented. It made life easy; it sped up the process and may be it did a job a bit finer. But you still need to understand where to drill and how to create the thing. There are a number of analogies like this you can find.
nullptr_deref超过 1 年前
Why are people so unaware of what is happening? The leap in quality from DALL-E 2 to DALL-E 3 was due to improvements in the prompts used for training. With countless people using ChatGPT, it will reach a point where OpenAI will have enough data to train on. There are ideas not yet explored that will be connected to the system. At this point, the pace of development is exponential. Unless it flattens out, we will witness AI being able to engineer, code, craft, design, and build a product on its own.<p>So, what is the basis for my statement? We are just mediators at this point. We don&#x27;t need to be there. Currently GPT-4 translates our requirements to DALL-E 3 during image generation process. What I want to say is, there is no reason to give any form of signal as an &quot;initial seed.&quot; Has anyone tried creating an automatic agent that, when turned on, receives a signal from the external world, whatever it may be, and triggers a chain of thoughts in these LLMs?<p>We, as humans, rely on signals. Currently, we are providing signals to LLMs in the form of text. There will and there must come a point where agents will feed signals to each other. They will perceive signals from the environment on their own and move on from there. This is not far away. It is near because all the components are laid out there.<p>And people argue over things like, &quot;Oh, but I need to prompt it.&quot; No, you don&#x27;t. It will take one engineer to define a way to process signals, then write an objective for the AI and train it. Once trained, it will continue on its own. The projection might sound off to you all, but there is no one stopping you from not doing it. Do you know what the limiting factor is? Capital. Yes, capital. So, as long as the power law holds, we will keep seeing strides in improvement. If you have capital and skills, you can do it right now.<p>I am not saying the signal is &quot;real world physical environment&quot;. Say you have a crawler that scrapes the website. You hook it to a summarizer. The summarizer fires the signals and then it calls AIs to create product. This whole thing once created doesn&#x27;t need a human in the loop. Once something is built, it is done. Also, add to the fact that these AI system only trust other AI system&#x27;s code. Why? Because human are untrustable. Now you will have an ecosystem designed by AI for AI and to be integrated with AI. Why do we need humans at this point? Like why would a company want humans at this point?
_bramses超过 1 年前
How can a field that is less than 80 years old wane? Y’all’s recency bias shows way too strongly. Your definition of junior and senior is purely anecdotal and based around what the brass at your companies “award”. Not to mention the arbitrary conflation of algorithm design and the actual process of writing sustainable code. It’s a show fair of “my way is best”, and science waits for no opinions as such.
resuresu超过 1 年前
I’ve always saw computers as a means to an end. Even in the bladerunner movies no one is sitting around pecking at a phone or a pc keyboard, it’s just holographic AI they talk to and interface with. And that’s the future of technology. Eventually you might do some sort of brain interface and be a cyborg. Who knows but the truth is, we nearing the end of humans needing to sit down and actually write code.
parham超过 1 年前
This will test how egoless each programmer is.<p>If you’re truly egoless and care about the outcome and not the path to the outcome… you’ll see how exciting this is.
评论 #38263974 未加载
devmor超过 1 年前
I’m far more threatened by pennies-on-the-dollar third world contracting firms (and the businesses that abuse and manipulate hiring regulations to make use of them) than the immature novelty tool that is generative AI.<p>That being said, those firms are certainly making use of these tools - relying on their underpaid programming staff to clean up the output into something useable for the client.
BobBagwill超过 1 年前
AFAIK the go and chess models played themselves to improve. They weren&#x27;t trained on games played by junior players. So far, coding LLMs don&#x27;t try to compile and run their code and check the output. They don&#x27;t fuzz the inputs to look for errors. There aren&#x27;t competitive, adversarial LLMs trying to outcode each other or fool each other.
Dowwie超过 1 年前
Most of the software engineers I&#x27;ve known are problem solvers who use code to accomplish their goals. Problem solvers will use the best tools available for the job. Perhaps engineers will use natural language UI&#x27;s to build software and &quot;drop down&quot; into a languages for debugging or optimization, just as today some drop into assembly.
Pufferbo超过 1 年前
I hate the word “coder”. It’s such an undermining term. Saying that someone “codes” or is a “coder” is like calling a chief a “cooker”.
评论 #38261793 未加载
demondemidi超过 1 年前
Wow what a terrible article. The guy just got tired of programming. Everything his buddy looked up he could have hacked out but chose not to. All the tools exist to code at whatever level he wants to. Just sounds like he lost the spark. The other day on hacker news a 17 year old kid wrote a debugger and was asking for help. That’s the spark.
fungiblecog超过 1 年前
A eulogy for programming as reinventing the wheel over and over again, possibly. But programming as a creative endeavour, no chance.
ubermonkey超过 1 年前
The line that killed me:<p>&gt;the one with the pink-shirted guy on I.B.M. PCs<p>That&#x27;s PETER NORTON, Goddammit. Know your history, New Yorker writer! Damn!
BobBagwill超过 1 年前
AFAIK, the go and chess models played themselves to improve. They weren&#x27;t limited to pieces of recorded games played by junior players, like SO.<p>So far, coding LLMs are not generating code, compiling it, running it, checking the output, timing the program, fuzzing the input, etc. When they can train themselves, they will improve.
maerF0x0超过 1 年前
my understanding of history is the Catholic church wielded great power over the populace by having both literacy, but and also in Latin (a less accessible language). With 99 theses and a German bible people could easily get their religion on without the priestly class. This feels like a metaphor for what has happened in computing.
评论 #38261562 未加载
autonomousErwin超过 1 年前
I think one of the undervalued things is the hack-iness&#x2F;curiosity nature of (good) developers which will be hard to replace. I imagine there&#x27;ll be a very short time where developers are in even more demand than they are now and that will quickly wane as everyone realises that the only job is entrepreneurship.
FrustratedMonky超过 1 年前
Regular expressions are pretty tricky. You can be an expert and it can still take time to tweak. I&#x27;ve been using regular expressions for decades, and now find myself using GPT as a kind of pre-compiler. I&#x27;ll test regular expressions with GPT and it gets really close or the final result.
thaanpaa超过 1 年前
I&#x27;ll be concerned if I see an LLM create a secure, functional SQL schema and API with users, roles, subscriptions, products, MFA, and all the other basic requirements of modern-day apps. We&#x27;re not even close to the required level of complexity yet, and I&#x27;m not holding my breath either.
tangjurine超过 1 年前
I don&#x27;t think you can call this guy a coder.<p>I can solve some math problems&#x2F;write some proofs, but I wouldn&#x27;t call myself a mathematician.<p>It&#x27;s about if you do it for a living, and this guy mostly has some fun side projects.<p>Don&#x27;t think this guy is qualified to comment on ChatGPT taking over, like another comment was saying.
1vuio0pswjnm7超过 1 年前
&quot;From the beginning, I had the sense that there was something wrongheaded in all this.&quot;<p>What was your first clue.
gumballindie超过 1 年前
&gt; For the new crossword project, though, Ben had introduced a third party. He’d signed up for a ChatGPT Plus subscription and was using GPT-4 as a coding assistant.<p>You have to give OpenAI credit. Their marketing campaign is beyond genius. A bit spammy and aggressive but it works.
mastazi超过 1 年前
If you want to become an architect, you don&#x27;t need previous experience as a bricklayer.<p>Many other fields besides construction are similar: most positions don&#x27;t require previous experience as &quot;manual labour&quot;.<p>Software development was an exception to the rule, maybe it no longer is.
cultureswitch超过 1 年前
LLMs are on the way to AGI but they are still laughably bad at logical reasoning.<p>Any remotely interesting coding task is at least somewhat novel and requires some reasoning. So far, LLMs don&#x27;t seem to be very good at handling things they haven&#x27;t seen before.
safaa1993超过 1 年前
To be blunt, i got into programming (nearly 30 years ago) for the results, not the process. In fact if anything i consider the process a sort of burdensome necessity on the way to getting the results i&#x27;m actually after.<p>Anything that shortens the path is good by me.
nickpeterson超过 1 年前
LLMs are going to make a lot of code. The tool will often not be able to fix problems with that code. If anything, this is classic automation expanding a field, we’re going to end up with more programmers and more software, not less.
tasuki超过 1 年前
&gt; At one point, we wanted a command that would print a hundred random lines from a dictionary file. I thought about the problem for a few minutes, and, when thinking failed, tried Googling.<p>Excuse me?
评论 #38282896 未加载
acscott超过 1 年前
Title does not approximate well the article. Ad probandum: &quot;Computing is not yet overcome. GPT-4 is impressive, but a layperson can’t wield it the way a programmer can. I still feel secure in my profession.&quot;
liampulles超过 1 年前
The really big problem with LLMs for general AI usage is the requirement of text input - this is a very limited way to represent ideas and intelligence.<p>Our brain does not think in text, it just spurts out text as a side effect.
评论 #38261196 未加载
评论 #38260768 未加载
29athrowaway超过 1 年前
If you train LLMs on code from GitHub it will output what the average GitHub repo has in it. That is, &quot;competent level&quot;, not expert or virtuoso.<p>And it will also carry over bugs found at the &quot;competent&quot; level.
评论 #38258663 未加载
pjmlp超过 1 年前
Definitly, this is what many seem to be missing, today LLM might help write the code, tomorrow they will compile the whole application themselves, they won&#x27;t need the babysitter developer any longer.
assimpleaspossi超过 1 年前
Whenever someone calls a programmer a &quot;coder&quot;, I feel like it&#x27;s calling Stephen King a &quot;typist&quot; and that&#x27;s insulting. Programming is much more than just the coding part.
xbmcuser超过 1 年前
There is a lot of office work that could be automated as chat gpt etc get better more and more efficiency experts will crop up selling efficiency ie less workers&#x2F;cost&#x2F;salary to do the same jobs.
tromp超过 1 年前
&gt; Perhaps what pushed Lee Sedol to retire from the game of Go<p>But he didn&#x27;t; he merely retired from professional play, and remains intensely interested in the game and the new directions it&#x27;s taking.
stratigos超过 1 年前
The article sounds like it was written by someone with minimal career experience. Its overwhelmingly naive, and Im sure the intended readership isnt any more or less naive than the author.
chasing超过 1 年前
Every five-to-ten years some technology comes around that&#x27;s finally going to put software engineers out of work.<p>Turns out the only thing truly successful at that task is economic mismanagement.
wiradikusuma超过 1 年前
I&#x27;ve been trying to get AI to write a component to embed in my Flutter app. The component is a zoomable whiteboard.<p>These people who said AI built programs, are they using different AI?
aspyct超过 1 年前
Genuine question, as I didn&#x27;t use LLMs much:<p>Sure, they&#x27;re good for quick prototypes or even launching production grade services.<p>But what about maintenance? What about fixing issues in legacy software?
评论 #38259717 未加载
评论 #38259726 未加载
samuelec超过 1 年前
&quot;I worried that it would rob me of both the joy of working on puzzles and the satisfaction of being the one who solved them.&quot;<p>I think every programmer feel it
aurbano超过 1 年前
I see a lot of comments wondering how AI could be useful as a software engineer so I&#x27;ll give my take on it:<p>I envision it being able to fully replace a junior engineer, and in some use-cases senior engineers as well.<p>In the case of junior engineers: the AI should have access to an internal knowledge base (i.e. Confluence) and the task&#x2F;ticketing system (Jira), and ideally the internal chat (Slack). I would assign tickets to it, and I&#x27;d expect the AI to reply with comments asking for clarification when there is something unclear, or proposing an implementation before starting if it&#x27;s not a very simple task (this could even be defined in the task, using story points for instance.<p>Once cleared, the AI submits a PR linked to the task - so far just like any engineer in the team would. The PR gets reviewed as usual, with suggestions&#x2F;requests for changes made by the (human) team, which then get addressed by the AI. With the big difference that all this process may happen in less than 1h from ticket creation to PR review.<p>I wouldn&#x27;t expect it to be able to implement complex features, onboard new libraries, or rearchitect the system in major ways to accommodate future features - just like I wouldn&#x27;t expect that from junior team members.<p>It would obviously be amazing if it could incorporate previous PR comments into it&#x27;s &quot;context&quot; for future work, so it could learn and improve.<p>Separately I mentioned it could also do part of the job of senior team members - in the form of PR reviews. If it has access to every previous PR review and learns from them it might be able to give insightful suggestions, and for very large codebases it could have an advantage over humans as it could find patterns or existing code that may be overlooked (i.e. &quot;It looks like this util function you added is the same as this one that was already there&quot;, or &quot;This code looks different to similar areas, but follows a different pattern, you might want to rewrite it this way &#x2F;&#x2F;...&quot;<p>Is GPT-4 there? Definitely not, perhaps an LLM is not even the way to achieve this, but I absolutely see this becoming an incredible asset to a tech team.
rip_netrunner超过 1 年前
Reading the responses on this page reminds of the quote<p>“it is difficult to get a man to understand something, when his salary depends on his not understanding it.” - Upton Sinclair.
pipeline_peak超过 1 年前
I think people like the author are starting realize coding is a very generic thing. And that it’s not itself “beautiful” but can be used to apply “beautiful” concepts.
mnd999超过 1 年前
Even if AI were actually good at greenfield dev, that’s still only a tiny part of the job, and not the hard part. Try fitting a huge legacy codebase in your prompt.
cultureswitch超过 1 年前
LLMs are pretty good at going from complex code&#x2F;legalese to prose which is easier to understand. Not so much the other way around.
totallywrong超过 1 年前
&gt; What I learned was that programming is not really about knowledge or skill but simply about patience, or maybe obsession.<p>Damn, the secret is out.
karmasimida超过 1 年前
Is driving a craft? Is plumbing a craft?<p>Coding to me is the same to those &#x27;crafts&#x27;, with GPT it will be done faster and more efficient, that is it.
评论 #38258523 未加载
TDiblik超过 1 年前
as a high school student, I&#x27;m more hesitant to go into CS degree not because of chatgpt but because it is constantly marketed as a high income job and A LOT of people are choosing&#x2F;learning it. I wonder if it&#x27;s gonna be the same in 10 years or so. Chatgpt&#x2F;copilot make stuff easier and there are only so much CRUD apps to be made :&#x2F;
androtheos超过 1 年前
&quot;Programmers are people who can endure an endless parade of tedious obstacles.&quot;, maybe my new favorite quote.
awinter-py超过 1 年前
this is by james somers! long arc from &#x27;there&#x27;s a gold rush and I&#x27;m the shovel&#x27; back in the day
评论 #38259218 未加载
LAC-Tech超过 1 年前
LLMs have not significantly changed the way I code at all. It&#x27;s only good at very basic mechanical tasks.
hcks超过 1 年前
Coders had decades to unionise and setup legal barriers to protect their livelihood.<p>(Doctors won’t ever be out of a job because of AI because they’ll always get a fat paycheck even just to look at the output spat out by the machine.)<p>They didn’t out of misplaced hubris and elitism (&quot;just be the best 10x engineer ever and you’ll be safe&quot;).<p>Now they have less than 10 years before their wages start crashing hard.
ingen0s超过 1 年前
AI is doing nothing but raising expectations, and so it should. Stop crying and adapt!
epcoa超过 1 年前
This article is inadvertently more about bullshit jobs and post clown world low interest rate free money.<p>&gt; He dialled up an FTP server (where files were stored) in an MS-DOS terminal and typed obscure commands.<p>This is some code a UI in Visual Basic to trace a IP address level exposition right here (this is from a self proclaimed subject matter expert, so no excuse), they’re not deeply knowledgeable and just because they got paid 6 figures at some dumb post recession startup doesn’t change anything.<p>(Others have already addressed the other howlers in this article)
AndrewKemendo超过 1 年前
Yeah but they’ll never master system design&#x2F;architecture so I’m safe!<p>&#x2F;s<p>We all knew it was coming
hm-nah超过 1 年前
This is a badass article.
dmingod666超过 1 年前
most impacted are technical consultants ( outside experts that merely provide high level consulting ) -- their recommendations will be scrutinized more closely
fuzztester超过 1 年前
Nostradamus wannabes are everywhere, in every field.
lawlessone超过 1 年前
&gt;For instance, it was considered foolish to estimate how long a coding task might take, since at any moment the programmer might turn over a rock and discover a tangle of bugs.<p>Still true tbf
Loxicon超过 1 年前
This sounds like a Chat GPT sales letter.
gatinsama超过 1 年前
I&#x27;m amazed by the fact that this was written by a human when AI would have done a good job.
StrangeSmells01超过 1 年前
It&#x27;s interesting and kind of a bummer that all this learning is made obsolete.
评论 #38257547 未加载
评论 #38257602 未加载
评论 #38257708 未加载
评论 #38300779 未加载
评论 #38257515 未加载
spacecadet超过 1 年前
New Yorker lol.
itsokimbatman超过 1 年前
Saw;
joebergeron超过 1 年前
I wonder if there are parallels to be drawn here against the industrialization of other, physical, goods, and how their production evolved over time. If you look at something like, say, printing (I&#x27;ll use the example of woodblock printing here, since I know a decent amount about it), it originated very much as a craft and discipline, as a means to fulfill a particular function. As printing technology improved and became better industrialized, the craft of printing was gradually replaced with other means that fulfilled the same function, but possessed a different form. i.e., lacking those characteristics of products of craft that we find desirable; artless.<p>We&#x27;ve been able to fulfill the function of printing very cheaply for what seems like ages now, and we&#x27;ve reached a point where some niche and particularly attuned segment of the population wants a bit more out of the actual form of the printed product, the depth of form that was once common. There is a growing community of people that deeply care about woodblock prints now, favoring their physical characteristics, despite such prints falling out of fashion for a period of time during the heights of industrialization. This group of people understands the value of such craft, and is willing to spend more for it, since the difference in the form of the end product from mass-produced stuff is so stark.<p>The key thing here, and with other categories of physical goods (e.g. pottery, glassware, furniture, etc.), is that there&#x27;s an obvious and tangible difference in the form of products produced via traditional means, and those mass manufactured, despite them serving fundamentally the same function.<p>With software however, I worry that this isn&#x27;t the case, and the sort of resurgence of interest we see now in products produced by traditional means won&#x27;t ever translate, assuming that we do move in the direction of more and more software engineering being &quot;automated&quot; by AI assistance. To an end-user of a piece of software, I imagine that there will be very little visible difference in the observable characteristics between fully hand-written and AI-produced software. Indeed, given the same requirements, there ought not be a difference between these two things. It&#x27;s exactly this delta, however, which drives the passionate and less cost-sensitive enthusiasts to prefer handmade physical goods over manufactured ones. If both the form and the function of AI-produced software is identical to those of traditional software, but the AI-produced software is cheaper, why would anyone go with the traditional stuff? I understand that there are other factors at play here as well (e.g., particularly principled consumers etc.) but really, some combination of form, function, and cost seem like the biggest levers to me, and they seem on the face of it to be pulled toward the direction of AI, for better or worse.
b20000超过 1 年前
who is james somers?
hiddencost超过 1 年前
&gt; At one point, we wanted a command that would print a hundred random lines from a dictionary file. I thought about the problem for a few minutes, and, when thinking failed, tried Googling. I made some false starts using what I could gather, and while I did my thing—programming—Ben told GPT-4 what he wanted and got code that ran perfectly.<p>I mean ... IDK, if you can&#x27;t write that on your own without research, I don&#x27;t think you should be writing a eulogy for programming in the New Yorker.<p>(I just wrote two version, one using python&#x27;s random.sample and the other more verbosely, to double check myself.)
评论 #38258881 未加载
评论 #38258932 未加载
评论 #38259105 未加载
评论 #38259041 未加载
评论 #38258833 未加载
评论 #38259133 未加载
评论 #38259113 未加载
评论 #38258960 未加载
评论 #38259216 未加载
评论 #38258918 未加载
评论 #38259130 未加载
评论 #38259117 未加载
评论 #38259024 未加载
评论 #38259124 未加载
评论 #38259369 未加载
bawana超过 1 年前
AI will surpass humanity at abstraction and the manipulation of abstraction in all its forms - laws, finance, literature, poetry, mathematics... Humans will finally be coralled into the activity their evolution optimized them for - innovation, adaptation, exploration, creation. All the billionaires of today will fade into obscurity as AI strips them of their wealth and leaves only the real humans - unfortunately this may only intensify the rat race if humans continue to compete in a capitalist market.
RagnarD超过 1 年前
Not to be unkind but simply factual: the author is clearly not a great developer. e.g.:<p>&quot;At one point, we wanted a command that would print a hundred random lines from a dictionary file. I thought about the problem for a few minutes, and, when thinking failed, tried Googling.&quot; Really?<p>&quot;But he had no idea what a pain it is to make an iPhone app. I’d tried a few times and never got beyond something that half worked. I found Apple’s programming environment forbidding.&quot; Perhaps. But compared to what? Yeah, it&#x27;s not using VI to write Javascript or whatever he considers a programming environment.<p>Power tools didn&#x27;t stop people from making things. Autocad didn&#x27;t put architects out of business. Finite element modelling didn&#x27;t destroy civil engineering. Quickbooks hasn&#x27;t destroyed the need for accountants. Word processors didn&#x27;t destroy writing. Google translate didn&#x27;t remove the need for translators.<p>LLMs will be a power tool for good developers. I think many underestimate what it&#x27;ll take to fully replace good or excellent software developers and what they do in totality.
评论 #38263984 未加载
评论 #38267161 未加载
评论 #38267138 未加载
n_ary超过 1 年前
Here is the controversial take from my perspective:<p>All these AI replacing coders and creatives is just a broad marketing campaign to put downward pressure on dev and creative salary.<p>We could&#x27;ve said the same thing back when Code generation was popular. Or we couod say the same thing about stackoverflow, because a non-tech person can bungle together enough solutions from there to build something. Heck NoCode(TM) was putting coders out of job in last decade and here we are in 2023.<p>Compared to GPT&#x27;s code-fu, most image generation models are light years ahead(imagine how those models can combine best of different artists into one painting), yet I still have plenty trouble describing in excruciatingly details about the painting I want with all the prompt-engineering-sauce and the end result is usually far away from desired state unless it is just normal sticker stuff.<p>Explaining requirements is hard, adapting to changes is hard, we still write tests and relatively every MR needs some fixes&#x2F;updates and we have a whole army of management people trying to do the prompt engineering thing and we still have to interpret in between lines to get a piece of software(even basic CRUDs that makes millions), so yeah, it is all hype driven marketing catering to enterprise about saving all those money but in the end still spend more on the token count and prompt engineering and another army of people figuring out the right keywords to describe the product and whatnot.<p>All this things are mostly the same, it is like claiming that we&#x27;ll habe quantum computers just like next year and any Joe rando can train their own GPT or crack open the SSL layer.<p>Don&#x27;t feed the hype cycle of these marketers with ypur fear because more you engage on this, more their boat is lifted.<p>Show me a real product that was built&#x2F;written by GPT and is making even few hundred in monthly revenue and I will be happy to change my mind.
评论 #38260430 未加载
评论 #38260219 未加载
评论 #38260127 未加载
评论 #38260190 未加载
评论 #38260168 未加载
fumeux_fume超过 1 年前
A lot of this article and many of the comments here don&#x27;t really comport with my experience programming and designing programs. Many or all of the examples in the article seem like toy programs meant to do something relatively straightforward&#x2F;simple with loose constraints, but in a domain or using tools unfamiliar to the programmer. However, many programmers find themselves in the opposite situation: using tools they&#x27;re familiar with to solve complex problems with hard constraints. It&#x27;s in that kind of situation where I&#x27;ve found that GPTs aren&#x27;t much of a game changer.<p>Many of the top comments voice a hated of programming--that programming gets in the way of creating. For a certain set of problems I completely agree (like why is making a website so damn complex), but for many of the problems I work on there are a wealth of good, helpful tools and APIs such that the writing of the code melts away and I&#x27;m left with the intellectual challenge of organizing my thoughts and thinking creatively. To me, the article and some of the top comments tell me more about the problems they choose to tackle than the act of programming itself.
评论 #38259831 未加载
评论 #38259739 未加载
评论 #38259826 未加载
评论 #38259854 未加载
Uptrenda超过 1 年前
&gt;At one point, we wanted a command that would print a hundred random lines from a dictionary file. I thought about the problem for a few minutes, and, when thinking failed<p>Wow, this guys working on rocket science, everyone. Watch out! We might get replaced!
评论 #38258698 未加载
评论 #38257645 未加载
评论 #38257722 未加载
评论 #38257541 未加载