Ask HN: Will AI put programmers our of work?

Almost all the work of programmers is already automated in the form of everything you get via `pip`, `apt get` etc.<p>A very large part of what remains is the bit which cannot be automated: modelling real world (business) process <i>in terms of</i> the systems of automation which are available.<p>Programming is a modelling activity which is about phrasing sequences of available actions to represent a process. If AI systems generate code, then programming becomes the sequencing of AI prompts -- which are here then just a more natural language <i>like</i> version of programming.<p>Even in that world a significant amount of technical skill is required to ensure commands are sequenced correctly, the code is correct, etc.<p>For &quot;AI&quot; to replace this process it would not only have to be AGI, but also AGI fully embeded in the human social world of the processes were are modelling.

My immediate guess is that these kinds of tools will greatly enlarge the productivity gap between junior and senior developers.<p>Senior devs will be able (they already are) to generate code, at first boilerplate and gradually more complex code, and effectively work as planners and passive reviewers, in a similar way to how some companies just hire legions of juniors with some architects&#x2F;seniors guiding and reviewing their work.<p>The problem with that flow, I think, is that it completely disrupts the junior to senior pipeline. Senior roles might be valued even more than today, but reaching that stage or simply entering the market might become much more difficult.<p>I feel my career is pretty safe, but I’m not sure about someone joining the industry 5 years from now.

In my experience, non-programmers lack not only the ability to translate concepts to code, but to formulate those concepts in the first place.<p>Ask someone who isn&#x27;t a software engineer to guide you in developing a simple program. You do all the actual coding, they just tell you what they want to happen. You will find that most people are unable to describe even high-level actions in the form of a coherent, procedural sequence.<p>Even the best code-generating AI won&#x27;t solve this problem, since it cannot generate useful code if the operator cannot articulate what they actually need.

Of course not.<p>For the same reason why companies pay $2000 per day for an experienced consultant when an employee could theoretically build the same stuff at minimum wage. Sometimes, mistakes are expensive. And then you need people who can reason about why they are doing what they are doing. AI can maybe churn out CRUD better than other generators, but when you have any significant amount of money depending on the software working, nobody is going to use ChatGPT without a human code review.<p>But ChatGPT code is typically overly lengthy and complicated, just like what a beginner would produce. And that makes for expensive and slow code reviews. That&#x27;s why in the end it&#x27;s cheaper overall to skip all that and just hire a professional.

&gt;Should I worry? Do you think that some form of AI will be able to do the job of an average programmer &gt; any time soon?<p>Yes.<p>&gt; If yes what is your estimate?<p>5 to 10 years will see a noticeable decrease in the number of programming jobs as we know them today.<p>&gt; And how would you try to AI-proof your career<p>Historically tech changes end up leaving a small rump of niche based practitioners, e.g. Blacksmiths servicing riding stables and racing yards, while the majority either exit the industry or take up the skillsets for the new technology. To future proof against AI, it&#x27;s either be about finding the niche in a shrinking market or changing skillsets.<p>In terms of those skillsets.<p>Without Artifical General Intelligence there is going to be a need for someone to translate human requirements into something that the &quot;machine&quot; - however sophisticated - understands _and_ verify the results afterwards. That sounds very much like some form of Behaviour Driven Development.<p>As to niches; there are a lot of complex, ill-defined but essential Cobol, Perl, PHP and Python systems floating around. Verifying a new translations is going to be expensive. QED; specialists keeping those existing systems ticking along is likely to be a thing long enough to make a career from.

No, but tech consolidation probably will. Tech is following in the same footsteps as the American car industry in the 50s.<p>Just as vertical integration in the auto industry killed off the auto startup ecosystem, vertical integration in the tech industry will kill off tech startups. This isnt because there won&#x27;t be demand for innovative new tech or that startups won&#x27;t be able to innovate, but because control of core platforms will allow the bigger players more leeway to crush and swallow smaller companies as well as to siphon their profit margins.<p>Think what aws is doing to elastic on a large scale.<p>Once the tech startup ecosystem dies (which could be soon; high interest rates will suck capital away from startups), the behemoths will probably stop innovating and slash headcounts.<p>Once that happens, I&#x27;m pretty sure that the stewards of capital and captains of industry will <i>scapegoat</i> AI and the Economist and Time magazine and the like will dutifully believe them and so will most of the people who read them.

As I see it, it generates fakes or potentially correct solutions that require review and alteration. So it’s best employed by a professional that can judge the generated result in a wider context.<p>Not too different from interpreting lighthouse scores.<p>It can make anyone with the means to access it a bad coder but the value of a professional has always been in picking the best solution from the possible ones.

Having toyed around with ChatGPT for a week now, I already use it as a sort of &quot;code assistant&quot;, much like how copilot works.<p>How good the results are, depends on how good you are at constructing the queries. It&#x27;s a bit like using google - some people have only the most basic knowledge, while others can find pretty much anything, because they&#x27;re really good at writing queries.<p>Now, imagine how it&#x27;s going to be 10-15-20 years from now? The future models will probably cull a chunk of devs., while the good devs will be even more efficient.<p>But who knows, maybe this will actually help some of the mediocre programmers to focus their energy on other things? Like taking on other roles? Take me for example - I&#x27;m no rock-star programmer. I see programming as a means to an end, it&#x27;s just a tool. I would much prefer to focus on the actual business logic and features of a product, maybe even long-time strategy. If most of the tedious coding was removed, that would make me happy.

We created ChatGPT, Copilot et el. But are they able to generate the code for a more capable version of themselves? Until we&#x27;re at that point I think most jobs are safe, whether we get to that point is in my view still a philosophical question.

I used to reply with a joke answer to questions like these along the lines of<p>&gt; wake me up when it starts to delete bad code instead of writing more of it<p>I too don&#x27;t follow AI space closely, but I have hard time imagining it doing anything of the sort. It probably can&#x27;t even remove a bug in a trivial program.<p>Correct me if I&#x27;m wrong of course, if there&#x27;s a video that demonstrates something important&#x2F;impressive related to this then I would certainly watch it.

Probably the opposite. Technology drives hunger for more complex systems

The current crop of generative AI all seems like it could produce useful tools for various practitioners, programmers included, I don&#x27;t yet see evidence for replacing jobs. But it&#x27;s hard to extrapolate.<p>That said the best way to anything-proof this career that&#x27;s so lucrative is to be frugal and set yourself up to retire early.

I don&#x27;t know how the latest stuff from OpenAI will impract the career of average programmers.<p>But regarding your last question:<p>&gt; how would you try to AI-proof your career?<p>Learn to program from first principles.

In a 5-10 years, generally-available tools resembling ChatGPT will be able to do about 90% of what a programmer does, but the remaining 10% will remain AI-complete.<p>The AI-complete bit is stuff like reading a set of instructions and realising that one of the instructions doesn&#x27;t make sense given the other ones, and pushing back. It&#x27;s looking through an UPDGRADING.md file to see how an API changed, realising that the thing you want isn&#x27;t there, then looking at the actual source-code of a library to understand an undocumented breaking change. It&#x27;s understanding how to write a program such that it&#x27;s easy to parallelise in the future.

Funnily enough I was just asking chatGPT what it thought about it, so here it goes:<p>&gt;It is unlikely that AI will completely replace programmers. While AI and machine learning technology has advanced significantly in recent years, there are still many tasks that require human creativity and intuition, such as coming up with new ideas, solving complex problems, and making decisions that involve subjective judgement. Additionally, as AI and machine learning technology continues to advance, it is likely that new job opportunities will be created in fields related to these technologies, such as developing and managing AI systems.

I&#x27;m working at a startup&#x27;s engineering &amp; product team as an engineer. I think it&#x27;s not fundamentally different from how most software engineers work.<p>We work on:<p>- The why: business&#x2F;leadership, and to some extent, product determines key goals (KPIs, OKRs etc.) and markets we are addressing. Us engineers are only informed parties here.<p>- The what: product, and to some extent, engineering determines which features to build or improve to attain these goals in a user-friendly and sustainable way. We create projects and design sketches, assign time constraints to them.<p>- The how: engineering, and to some extent, product ships these features in a maintainable, scalable, performant and supportable way without disrupting existing user experience.<p>A very large part of this work involves creative processes and logical reasoning about business, UX, software engineering problems. (Of course, that&#x27;s why I love it :))<p>Only a tiny amount of this work is &quot;writing boilerplate code&quot; or copying code from Stackoverflow - which ChatGPT is presumed to automate.<p>Of course, more senior engineers are faster at writing boilerplate, but their speed mostly comes from 1) knowing the existing codebase 2) using the right tools &amp; abstractions.<p>Moreover, most of the <i>risk</i> involved in the process is not the time taken to write boilerplate i.e. working on something for too long - but rather working on the <i>wrong thing</i>, or doing an implementation that&#x27;s too slow or hard to maintain (change, test, fix, extend, reason about).<p>All in all, when I think about software engineering from this perspective, I don&#x27;t see AI automating it away anytime soon.<p>I could, though imagine AI being your TDD coder companion. You write some unit or acceptance tests for a service module and the AI generates the code for it. You&#x27;d still need to thoroughly review and test the code though. This would work well for basic CRUD&#x2F;boilerplate modules, but not for anything involving business logic.<p>Nevertheless, this would still remain just a small part of a software engineer&#x27;s work, in my opinion. What do you think?

There was a post on Twitter comparing this question to the effects of calculators on mathematicians, which makes some sense.

I believe that AI will reduce the amount of code software developers write, but writing code is the least important thing we do. The most important is taking nebulous, sometimes contradictory requirements and restating them in such a way that it can be executed by a machine. In my experience, most people don&#x27;t have the wherewithal or desire to do this. So, I&#x27;m guessing coding might go away soon, but software developers will be around for a long while.

Some hunches:<p>- AI will hypercharge us, but we&#x27;ll still need to know what we&#x27;re doing.<p>- Polyglots will be more common (since nailing syntax won&#x27;t take nearly as long)<p>- Increased importance of understanding framework conventions, architecture and &#x27;the system&#x27;<p>- More time debugging crappy AI-written code!<p>- Demand for software development will <i>increase</i> (not decrease) since time to market and cost will go down, making many more projects viable.

Programmer would be the last job to go but it&#x27;s possible that the number of programmer jobs would start decreasing in the mean time.

I wonder if AI can do something like, here is a new programming language that has been written which provides better error handling and concurrency. Anyways, its the hot buzz now and everyone is using it. So do this for me, here is the source code of the language and its documentation.<p>Understand all that and create an app for me that does the work X.

&gt; Will AI put programmers our (sic) of work?<p>Yes. The questions are: &quot;How soon?&quot; and &quot;Which programmers?&quot;<p>&gt; Should I worry?<p>Are you satisfied doing a job that could be done by a simple script?<p>&gt; Do you think that some form of AI will be able to do the job of an average programmer any time soon?<p>This is already the case.<p>Additionally, consider that so many programmers are so bad at their jobs that the average is dismal, and also that computer programming isn&#x27;t that hard.<p>&gt; what is your estimate?<p>I have a comment here on HN predicting it would occur by approximately last year, so we are slightly behind schedule from my POV.<p>&gt; How would you try to AI-proof your career?<p>I wouldn&#x27;t. The very idea is counterproductive. I want to maximize my effectiveness, not lock-in a dead-end &quot;career&quot; of make-work that could be done by a machine.<p>I think what you&#x27;re really asking is, &quot;How can I earn a living once I can no longer do anything that is more economically valuable than what machines can do?&quot;<p>Personally, my answer to that is to change the entire economy to a Star Trek mode or something like that. Or go live in a cave in the woods. Not actually great options, eh?<p>But yeah, broadly speaking, if you as a <i>working</i> programmer could write a script to replace yourself I feel you&#x27;re obliged to do it, eh?<p>Not only is that your job: to write software to solve problems, to economically benefit the company you&#x27;re working for, but it&#x27;s also the way to avoid becoming a &quot;zombie&quot;, working a pointless job just for an excuse to take home a paycheck.<p>Stay home and go on welfare (along with everybody else), we can call it Universal Basic Income so nobody&#x27;s ashamed that machines are better at everything than they are.

I think in this context, there exists a much more significant uncanny valley than any we know of right now. The usual example from visualization completely pales in comparison. There is no &quot;almost&quot; solution for any non-trivial task. And on this opposite side of the valley only humans can act.

&gt;Do you think that some form of AI will be able to do the job of an average programmer any time soon?<p>No. AI is not a greater threat than stack overflow.<p>The skills of a programmer are not generating 20 lines of code to solve a well known problem. And even a 99% AI is next to useless, since finding errors is exceedingly hard.

It&#x27;s hard to predict the future.<p>One aspect that&#x27;s being slighly neglected is that programming isn&#x27;t that special - tools like ChatGPT can potentially impact any kind of knowledge work. So it&#x27;s not like there&#x27;s a safe white-collar career track that you can easily move to.

As a matter of fact, most software is produced in house tailor made for companies. Since AI is rising, this will make this software less costly, hence the demand for custom made software should increase, since this makes money for the company. So if you are doing something very simple in terms of AI capabilities, you might be in trouble, but otherwise you will simply be asked to do more work in the same time. In any case, the profession will change drastically in the near future, hence you should be alert and willing to learn.

Putting aside AGI types of actual &quot;intelligence&quot;, this is evolution at work, more or less. As soon as someone finds a useful means of automating the &quot;fitness function&quot; testing that the output has desirable qualities (correctness or entertainment or whatever), everything outside the extremes will rapidly leave human hands. The extremes fall soon thereafter.<p>The problems are mostly sociocultural, and I am reminded of &quot;Agent Smith&quot; commenting on humans rejecting a Matrix simulation that didn&#x27;t suck for them.

These tools are incredibly impressive, and they will probably only get better as time goes on.<p>However, most software being developed and maintained out there is not part of greenfield projects. It is new pieces being built upon the pieces that came before it. Years and decades of different styles, formats and convention.<p>Sure, in a few years these tools will speed up how quickly some parts of the UI or some internal logic will be created.<p>But I am not seeing these tools take over the biggest part of my job: connecting vastly different systems together.

I define a programmer as a human who mediates the gateway between the world of people and the world of machines. We are like lawyers hired to talk to genies on behalf of people who have been offered wishes.<p>AI is a dangerous and deceptive tool that requires wise and subtle nudging to make it work.<p>Results I have gotten from OpenAI are all terrible, for instance. I have to learn “prompt engineering” now. It just high level programming.

That&#x27;s too early to know for sure but for now I&#x27;m not that impressed yet. Yes sure those can be great tools for helping developers but replacing jobs? Maybe the most boring jobs possible could be automated with that and I&#x27;m not even sure.<p>Who will review and maintain the code produced apart from a developer anyways?<p>If you think about it it&#x27;s much less revolutionary in terms of reducing coding jobs than WordPress.

When&#x2F;if we reach that point of development, most other jobs will have been automated away by us, the programmers. Only fair, that ours gets automated away too then.<p>I think the true question is, whether we are going to treat human beings like trash, useless, now that their job has been automated away, or we as a society find a good way to deal with this and steer into a happier future.

I submitted an example[1] of a programmer trying to use ChatGPT to generate code for two different tasks: download and extract text from a PDF file; and then parse the PDF using regex etc.<p>It doesn&#x27;t always work, and the reasons are interesting.<p>[1]:<a href="https:&#x2F;&#x2F;news.ycombinator.com&#x2F;item?id=33940515" rel="nofollow">https:&#x2F;&#x2F;news.ycombinator.com&#x2F;item?id=33940515</a>

There are many other things you could choose to worry about:<p><pre><code> - nuclear conflict erasing 95% of humanity - deadly pandemic erasing 67% of humanity - devastating climate crisis erasing 80% of humanity - impact event erasing 100% of humanity </code></pre> IMHO any of these events have higher probability when some AI taking over our jobs

As with all new technology it will change the jobs we do and how we do them. AI seems likely to have a large impact on jobs heavily involving communicating with computers. However is it likely that less people will be communicating with computers as a result; I would say no. Did the shovel make it less likely you would employ someone to dig?

My take is that it will decrease the &quot;value&quot; of developers by reducing the demand for them. If a company can be 10x more productive it will probably need less developers. if you can do your job 10x faster it will mean your team would less team members.<p>There&#x27;s no shortage that is not followed by a glut..

No way. Just look at problems like translation. Its a nice tool but you cant rely on it to do a propper translation, it&#x27;s actually very bad. How long have they been working on the translation problem?<p>Same with self driving. We get some nice tools to assist drivers but replacement won&#x27;t come in my lifetime.

No..our work might morph into refactoring AI&#x27;s spaghetti but a lot of a programmers job is comforting the person that hired them (whether in house or contract), knowing they have someone on hand to solve whatever problem may arise. This simply won&#x27;t be the case with AI.

I don&#x27;t think that&#x27;s the case, at least for the near future.<p>One, it gives small blocks of code, that too, for the most common use cases. Two, the code often contains a few errors (doesn&#x27;t compile) or has a few security vulnerabilities.

Nope. But one must learn to use these new tools to improve. Competition will not come from AI, it will come from other humans who leverage AI to increase output.

If anything, there have to be people in the loop to check balances (write good prompts to generate code and fix them) and take responsibilities for downtime.

Even if AI automates 90% of my work, I could just move full-time into products &#x2F; business. Programming is just dealing with systems and abstractions.

Things for now just get more high level and less tedious.

Some programmers? Maybe. All programmers? Definitely not. In fact, AI has only created more job opportunities so far...

Most programming is just plumbing. AI will probably replace that.<p>That leaves algorithm research, and so we can then spend 100% of our time on hard CS.

It will replace the programmer in exactly the same way as Google and Stack Overflow did. Or like calculators replaced mathematics.

I found solutions like ChatGPT, Copilot, and CodeWhisperer too basic to replace a programmer.<p>But they can help with around 20% of the coding part.

(i) AI does not exist<p>(ii) already you can automate low hanging fruit with python + excel but its not done<p>It will increase their demand, if anything.

No, absolutely not. I think it will even boost the need for devops and frotend work.

If you ask most workers if their jobs are vulnerable to automation, they will say no

It will put everyone else producing ppts and excel sheets out of work first.

Kind of ironic that we are putting ourselves out of work.

If AI technology continues to advance and is able to take over most jobs, it will fundamentally change the way society functions.<p>In a capitalist economy, people are expected to work hard in order to produce goods and services, which drives economic growth. But if AI can do all of the work, there may be no need for people to work at all.<p>This could lead to a new economic model, one where prosperity is not tied to the labor of individuals. Instead, the focus could shift to ensuring that everyone has access to the resources and opportunities they need to thrive, regardless of whether they are working or not.

Did compilers put programmers out of work?

No, neither compilers and interpreters did.

No, as soon as it&#x27;s not open source.

the glue work will remain. Maybe we will become &quot;AI whisperer&quot;