[HN Gopher] AI slows down open source developers. Peter Naur can...
       ___________________________________________________________________
        
       AI slows down open source developers. Peter Naur can teach us why
        
       Author : jwhiles
       Score  : 314 points
       Date   : 2025-07-14 14:32 UTC (8 hours ago)
        
 (HTM) web link (johnwhiles.com)
 (TXT) w3m dump (johnwhiles.com)
        
       | mkagenius wrote:
       | AI tends to slow us down because we don't really know what it's
       | good at. _Can it write a proper Nginx config? I don 't know--
       | let's try._ And then we end up wasting 30 minutes on it.
       | 
       | Fully autonomous coding tools like v0, a0, or Aider work well as
       | long as the context is small. But once the context grows--usually
       | due to mistakes made in earlier steps--they just can't keep up.
       | There's no real benefit of "try again" loop yet.
       | 
       | For now, I think simple VSCode extensions are the most useful.
       | You get focused assistance on small files or snippets you're
       | working on, and that's usually all you need.
        
         | ethan_smith wrote:
         | The context switching cost between coding and AI interaction is
         | substantial and rarely measured in these studies. Each
         | prompt/review cycle breaks flow state, which is particularly
         | damaging for complex programming tasks where deep concentration
         | yields the greatest productivity.
        
           | bluefirebrand wrote:
           | This has been my experience too
           | 
           | Ever since my company made switching to Cursor mandatory, I
           | have not been able to hit any kind of flow. I know my own
           | productivity has plummeted and I suspect many others are as
           | well, but no one is saying anything
           | 
           | I have spoken up once or twice and only been smacked down for
           | my troubles, so I am not surprised everyone else is clammed
           | up
        
       | doc_manhat wrote:
       | I directionally disagree with this:
       | 
       | ``` It's common for engineers to end up working on projects which
       | they don't have an accurate mental model of. Projects built by
       | people who have long since left the company for pastures new.
       | It's equally common for developers to work in environments where
       | little value is placed on understanding systems, but a lot of
       | value is placed on quickly delivering changes that mostly work.
       | In this context, I think that AI tools have more of an advantage.
       | They can ingest the unfamiliar codebase faster than any human
       | can, and can often generate changes that will essentially work.
       | ```
       | 
       | Reason: you cannot evaluate the work accurately if you have no
       | mental model. If there's a bug given the systems unwritten
       | assumptions you may not catch it.
       | 
       | Having said that it also depends on how important it is to be
       | writing bug free code in the given domain I guess.
       | 
       | I like AI particularly for green field stuff and one off scripts
       | as it let's you go faster here. Basically you build up the mental
       | model as you're coding with the AI.
       | 
       | Not sure about whether this breaks down at a certain codebase
       | size though.
        
         | horsawlarway wrote:
         | Just anecdotally - I think your reason for disagreeing is a
         | valid statement, but not a valid counterpoint to the argument
         | being made.
         | 
         | So
         | 
         | > Reason: you cannot evaluate the work accurately if you have
         | no mental model. If there's a bug given the systems unwritten
         | assumptions you may not catch it.
         | 
         | This is completely correct. It's a very fair statement. The
         | problem is that a developer coming into a large legacy project
         | is in this spot regardless of the existence of AI.
         | 
         | I've found that asking AI tools to generate a changeset in this
         | case is actually a pretty solid way of starting to learn the
         | mental model.
         | 
         | I want to see where it tries to make changes, what files it
         | wants to touch, what libraries and patterns it uses, etc.
         | 
         | It's a poor man's proxy for having a subject matter expert in
         | the code give you pointers. But it doesn't take anyone else's
         | time, and as long as you're not just trying to dump output into
         | a PR can actually be a pretty good resource.
         | 
         | The key is not letting it dump out a lot of code, in favor of
         | directional signaling.
         | 
         | ex: Prompts like "Which files should I edit to implement a
         | feature which does [detailed description of feature]?" Or
         | "Where is [specific functionality] implemented in this
         | codebase?" Have been real timesavers for me.
         | 
         | The actual code generation has probably been a net time loss.
        
           | Roscius wrote:
           | > I've found that asking AI tools to generate a changeset in
           | this case is actually a pretty solid way of starting to learn
           | the mental model.
           | 
           | This. Leveraging the AI to start to develop the mental model
           | is an advantage. But, using the AI is a non-trivial skill set
           | that needs to be learned. Skepticism of what it's saying is
           | important. AI can be really useful just like a 747 can be
           | useful, but you don't want someone picked off the street at
           | random flying it.
        
             | bluefirebrand wrote:
             | > This. Leveraging the AI to start to develop the mental
             | model is an advantage
             | 
             | Is there any evidence that AI helps you build the mental
             | model of an unfamiliar codebase more quickly?
             | 
             | In my experience trying to use AI for this it often leads
             | me into the weeds
        
           | doc_manhat wrote:
           | Yeah fair points particularly for larger codebases I could
           | see this being a huge time saver.
        
       | gjsman-1000 wrote:
       | What I thought was fascinating, and should be a warning sign to
       | everyone here:
       | 
       | Before beginning the study, the average developer expected about
       | a 20% productivity boost.
       | 
       |  _After ending the study,_ the average developer (potentially:
       | you) believed they actually were 20% more productive.
       | 
       | In reality, they were 0% more productive at best, and 40% less
       | productive at worst.
       | 
       | Think about what it would be like to be that developer; off by
       | 60% about your own output.
       | 
       | If you can't even gauge your own output without being 40% off on
       | average, 60% off at worst; be cautious about strong opinions on
       | anything in life. Especially politically.
       | 
       | Edit 1: Also consider, quite terrifyingly, if said developers
       | were in an online group, together, like... here. The one
       | developer who said she thought it made everyone slower (the truth
       | in this particular case), would be unanimously considered an
       | idiot, downvoted to the full -4, _even with the benefit of
       | hindsight_.
       | 
       | Edit 2: I suppose this goes to show, that even on Hacker News,
       | where there are relatively high-IQ and self-aware individuals
       | present... 95% of the crowd can still possibly be wildly
       | delusional. Stick to your gut, regardless of the crowd, and
       | regardless of who is in it.
        
         | pphysch wrote:
         | Given how deadlines/timelines tend to (not) work in SWE, this
         | is not surprising.
        
           | gjsman-1000 wrote:
           | Perhaps; but this is a developer's own output with an AI
           | tool, compared against _their own historical output_ when
           | they didn 't use it. Apparently, the average developer (read:
           | quite possibly most people here) can't even hit the broadside
           | of a barn in estimating _their own_ productivity.
        
             | sureglymop wrote:
             | That doesn't surprise me at all. Isn't software engineering
             | in essence about being constantly confronted with new
             | problems to solve and having to come up with a sufficient
             | one on the fly? It seems very hard to estimate this, even
             | if you know yourself well.
        
               | lupire wrote:
               | They were 20% underestimating how long it took them to do
               | a 1-8 hr task that they had just completed.
               | 
               | It's like Tog's study that people think Keyboard is
               | faster than the mouse even when they are faster with the
               | mouse. Because they are measuring how they feel, not what
               | is actually happening.
               | 
               | https://www.asktog.com/TOI/toi06KeyboardVMouse1.html
        
               | marcosdumay wrote:
               | That is a very weird set of findings.
               | 
               | This one in particular:
               | 
               | > It takes two seconds to decide upon which special-
               | function key to press.
               | 
               | seems to indicate the study was done on people with no
               | familiarity at all with the software they were testing.
               | 
               | Either way, I don't think there is any evidence out there
               | supporting that either of keyboard-only or mouse-only is
               | faster or equivalent to keyboard+mouse for well known
               | GUIs.
        
             | dragonwriter wrote:
             | That this is generally a problem, and was established as
             | such _before software development existed_ (the big thing
             | people usually point to is a RAND corp from the 1940s) and
             | is the whole motivation for Wideband Delphi estimation
             | methods invented shortly afterwards (of which agile
             | "planning poker" is simply a particular more recent
             | realization) for forward estimation, and why lean methods
             | center on using a plan-do-check-act cycle for process
             | improvements rather than seat of the pants and subjective
             | feel.
             | 
             | But despite the popularity of some of this (planning poker,
             | particularly; PDCA for process improvements is sadly less
             | popular) as ritual, those elements have become part of a
             | cargo cult where almost no one remembers _why we do it_.
        
               | freedomben wrote:
               | But this is still regarding forward estimating of future
               | work, whereas GP is talking about gauging _actual_ , past
               | work done. The problems with forward estimation are
               | indeed widely known, but I doubt most people realize that
               | they are so bad at even knowing how productive they were.
        
         | bluefirebrand wrote:
         | > Also consider, quite terrifyingly, if said developers were in
         | an online group, together, like... here. The one developer who
         | said she thought it made everyone slower (the truth in this
         | particular case), would be unanimously considered an idiot,
         | downvoted to the full -4, even with the benefit of hindsight
         | 
         | Yeah, this is me at my job right now. Every time I express even
         | the mildest skepticism about the value of our Cursor
         | subscription, I'm getting follow up conversations basically
         | telling me to shut up about it
         | 
         | It's been very demoralizing. You're not allowed to question the
         | Emperor's new clothes
        
         | quantumHazer wrote:
         | This should really be top comment. The problem is this tools
         | can really give us some value in certain type of areas, but
         | they are not like they are marketed.
        
       | xyst wrote:
       | Not surprising. Use of LLM has only been helpful in initial
       | exploration of unknown code bases or languages for me.
       | 
       | Using it beyond that is just more work. First parse the broken
       | response, remove any useless junk, have it reprocess with updated
       | query.
       | 
       | It's a nice tool to have (just as search engines gave us easy
       | access to multiple sources/forums), but its limitations are well
       | known. Trying to use it 100% as intended is a massive waste of
       | time and resources (energy use...)
        
       | nico wrote:
       | > They are experienced open source developers, working on their
       | own projects
       | 
       | I just started working on a 3-month old codebase written by
       | someone else, in a framework and architecture I had never used
       | before
       | 
       | Within a couple hours, with the help of Claude Code, I had
       | already created a really nice system to replicate data from
       | staging to local development. Something I had built before in
       | other projects, and I new that manually it would take me a full
       | day or two, especially without experience in the architecture
       | 
       | That immediately sped up my development even more, as now I had
       | better data to test things locally
       | 
       | Then a couple hours later, I had already pushed my first PR. All
       | code following the proper coding style and practices of the
       | existing project and the framework. That PR, would have taken me
       | at least a couple of days and up to 2 weeks to fully manually
       | write out and test
       | 
       | So sure, AI won't speed everyone or everything up. But at least
       | in this one case, it gave me a huge boost
       | 
       | As I keep going, I expect things to slow down a bit, as the
       | complexity of the project grows. However, it's also given me the
       | chance to get an amazing jumpstart
        
         | kevmo314 wrote:
         | You've missed the point of the article, which in fact agrees
         | with your anecdote.
         | 
         | > It's equally common for developers to work in environments
         | where little value is placed on understanding systems, but a
         | lot of value is placed on quickly delivering changes that
         | mostly work. In this context, I think that AI tools have more
         | of an advantage. They can ingest the unfamiliar codebase faster
         | than any human can, and can often generate changes that will
         | essentially work.
        
           | moogleii wrote:
           | That would be an aside, or a comment, not the point of the
           | article.
        
           | antonvs wrote:
           | > You've missed the point of the article
           | 
           | Sadly clickbait headlines like the OP, "AI slows down open
           | source developers," spread this misinformation, ensuring that
           | a majority of people will have the same misapprehension.
        
             | raincole wrote:
             | Which is a good thing for people who are currently
             | benefiting from AI, though. The slower other programmers
             | adopt AI, the more edge those who are proficient with it
             | have.
             | 
             | It took me an embarrassingly long time to realize a simple
             | fact: using AI well is a shallow skill that everyone can
             | learn in days or even hours if they want. And then my small
             | advantage of knowing AI tools will disappear. Since the
             | realization I've been always upvoting articles that claims
             | AI makes you less productive (like the OP).
        
               | rightbyte wrote:
               | So you bother push some sort of self proclaimed false
               | narrative with upvotes but then you try to counteract it
               | by spelling it out?
        
         | markstos wrote:
         | I had a similar experience with AI and open source. AI allowed
         | me to implement features in a language and stack I didn't know
         | well. I had wanted these features for months and no one else
         | was volunteering to implement them. I had tried to study the
         | stack directly myself, but found the total picture to be
         | complex and under-documented for people getting started.
         | 
         | Using Warp terminal (which used Claude) I was get past those
         | barriers and achieve results that weren't happening at all
         | before.
        
         | Vegenoid wrote:
         | I have had similar experiences as you, but this is not the kind
         | of work that the study is talking about:
         | 
         | "When open source developers working in codebases that they are
         | deeply familiar with use AI tools to complete a task, they take
         | longer to complete that task"
         | 
         | I have anecdotally found this to be true as well, that an LLM
         | greatly accelerates my ramp up time in a new codebase, but then
         | actually leads me astray once I am familiar with the project.
        
           | Navarr wrote:
           | > I have anecdotally found this to be true as well, that an
           | LLM greatly accelerates my ramp up time in a new codebase,
           | but then actually leads me astray once I am familiar with the
           | project.
           | 
           | If you are unfamiliar with the project, how do you determine
           | that it wasn't leading you astray in the first place? Do you
           | ever revisit what you had done with AI previously to make
           | sure that, once you know your way around, it was doing it the
           | right way?
        
             | quantumHazer wrote:
             | no, they just claim that AI coding tools are magic and
             | drink their kool-aid
        
             | Vegenoid wrote:
             | In some cases, I have not revisited, as I was happy to
             | simply make a small modification for my use only. In
             | others, I have taken the time to ensure the changes are
             | suitable for upstreaming. In my experience, which I have
             | not methodically recorded in any way, the LLM's changes at
             | this early stage have been pretty good. This is also partly
             | because the changes I am making at the early stage are
             | generally small, usually not requiring adding new
             | functionality but simply hooking up existing functionality
             | to a new input or output.
             | 
             | What's most useful about the LLM in the early stages is not
             | the actual code it writes, but its reasoning that helps me
             | learn about the structure of the project. I don't take the
             | code blind, I am more interested in the reasoning than the
             | code itself. I have found this to be reliably useful.
        
           | pragma_x wrote:
           | Not just new code-bases. I recently used an LLM to accelerate
           | my learning of Rust.
           | 
           | Coming from other programming languages, I had a lot of
           | questions that would be tough to nail down in a Google
           | search, or combing through docs and/or tutorials. In
           | retrospect, it's super fast at finding answers to things that
           | _don't exist_ explicitly, or are implied through the lack of
           | documentation, or exist at the intersection of wildly
           | different resources:
           | 
           | - Can I get compile-time type information of Enum values?
           | 
           | - Can I specialize a generic function/type based on Enum
           | values?
           | 
           | - How can I use macros to reflect on struct fields?
           | 
           | - Can I use an enum without its enclosing namespace, as I can
           | in C++?
           | 
           | - Does rust have a 'with' clause?
           | 
           | - How do I avoid declaring timelines on my types?
           | 
           | - What is an idiomatic way to implement the Strategy pattern?
           | 
           | - What is an idiomatic way to return a closure from a
           | function?
           | 
           | ...and so on. This "conversation" happened here and there
           | over the period of two weeks. Not only was ChatGPT up to the
           | task, but it was able to suggest what technologies would get
           | me close to the mark if Rust wasn't built to do what I had in
           | mind. I'm now much more comfortable and competent in the
           | language, but miles ahead of where I would have been without
           | it.
        
           | Gormo wrote:
           | > I have anecdotally found this to be true as well, that an
           | LLM greatly accelerates my ramp up time in a new codebase,
           | but then actually leads me astray once I am familiar with the
           | project.
           | 
           | How does using AI impact the amount of time it takes you to
           | _become_ sufficiently familiar with the project to recognize
           | when you are being led astray?
           | 
           | One of the worries I have with the fast ramp-up is that a lot
           | of that ramp-up time isn't just grunt work to be optimized a
           | way, it's active learning, and bypassing too much of it can
           | leave you with an incomplete understanding of the problem
           | domain that slows you down perpetually.
           | 
           | Sometimes, there are real efficiencies to be gained; other
           | times those perceived efficiencies are actually incurring
           | heavy technical debt, and I suspect that overuse of AI is
           | usually the latter.
        
         | bko wrote:
         | When anecdote and data don't align, it's usually the data
         | that's wrong.
         | 
         | Not always the case, but whenever I read about these strained
         | studies or arguments about how AI is actually making people
         | less productive, I can't help but wonder why nearly every
         | programmer I know, myself included, finds value in these tools.
         | I wonder if the same thing happened with higher level
         | programming languages where people argued, you may THINK not
         | managing your own garbage collector will lead to more
         | productivity but actually...
         | 
         | Even if we weren't more "productive", millions prefer to use
         | these tools, so it has to count for something. And I don't need
         | a "study" to tell me that
        
           | adrian_b wrote:
           | TFA says clearly that it is likely that AI will make more
           | productive anyone working on an unfamiliar code base, but
           | make less productive those working on a project they
           | understand well, and it gives reasonable arguments for why
           | this is likely to happen.
           | 
           | Moreover, it acknowledges that for programmers working in
           | most companies the first case is much more frequent.
        
             | bko wrote:
             | I have written every line of code in the code base I mostly
             | work in and I still find it incredibly valuable. Millions
             | use these tools and a large percentage of them find them
             | useful in their familiar code base.
             | 
             | Again, overwhelming anecdote and millions of users >
             | "study"
        
               | almatabata wrote:
               | > Interestingly the developers predict that AI will make
               | them faster, and continue to believe that it did make
               | them faster, even after completing the task slower than
               | they otherwise would!
               | 
               | In this case clearly anecdotes are not enough. If that
               | quote from the article is accurate, it shows that you
               | cannot trust the developers time perception.
               | 
               | I agree, its only one study and we should not take it as
               | the final answer. It definitely justifies doing a few
               | follow up evaluations to see if this
        
               | bko wrote:
               | Or I cannot trust a contrived laboratory setting with
               | it's garden of forking paths.
               | 
               | https://mleverything.substack.com/p/garden-of-forking-
               | paths-...
        
               | almatabata wrote:
               | I did not say to trust it. I do not need to trust it.
               | 
               | If I run my own tests on my own codebase I will
               | definitely use some objective time measurement method and
               | a subjective one. I really want to know if there is a big
               | difference.
               | 
               | I really wonder if its just the individuals bias showing.
               | If you are pro-AI you might overestimate one, and if you
               | are against it you might under-estimate it.
        
               | bko wrote:
               | That's fair, I agree.
        
               | overfeed wrote:
               | > If that quote from the article is accurate, it shows
               | that you cannot trust the developers time perception.
               | 
               | The scientific method goes right out the window when it
               | comes to true believers. It reminds me of weed-smokers
               | who insist getting high makes them deep-thinkers: it
               | _feels_ that way in the moment, but if you 've ever been
               | a sober person caught up in a "deep" discussion among
               | people high on THC, _oh boy..._
        
           | rsynnott wrote:
           | > I can't help but wonder why nearly every programmer I know,
           | myself included, finds value in these tools.
           | 
           | One of the more interesting findings of the study mentioned
           | was that the LLM users, even where use of an LLM had
           | apparently degraded their performance, tended to believe it
           | had enhanced it. Anecdote is a _really_ bad argument against
           | data that shows a _perception_ problem.
           | 
           | > Even if we weren't more "productive", millions prefer to
           | use these tools, so it has to count for something.
           | 
           | I mean, on that basis, so does homeopathy.
           | 
           | Like, it's just one study. It's not the last word. But "my
           | anecdotes disprove it" probably isn't a _terribly_ helpful
           | approach.
        
             | ted_bunny wrote:
             | Also, "anecdotes > data" as a general heuristic is a red
             | flag. But like if clowns had a country and their flag were
             | red. That kind.
        
         | quantumHazer wrote:
         | You clearly have not read the study. Problem is developers
         | thought they were 20% faster, but they were actually slower.
         | Anyway from a fast review about your profile you're in conflict
         | of interest about vibe coding, so I will definitely take your
         | opinion with a grain of salt.
        
           | floren wrote:
           | > Anyway from a fast review about your profile you're in
           | conflict of interest about vibe coding
           | 
           | Seems to happen every time, doesn't it?
        
         | samtp wrote:
         | Well that's exactly what it does well at the moment.
         | Boilerplate starter templates, landing pages, throwaway apps,
         | etc. But for projects that need precision like data pipelines,
         | security - it code generated has many subtle flaws that
         | can/will cause giant headaches in your project unless you dig
         | through every line produced
        
         | PaulDavisThe1st wrote:
         | TFA was _specifically_ about people very familiar with the
         | project and codebase that they are working on. Your anecdots is
         | precisely the opposite of the situation is was about, and it
         | acknowledged the sort of process you describe.
        
         | davidclark wrote:
         | > That PR, would have taken me at least a couple of days and up
         | to 2 weeks to fully manually write out and test
         | 
         | What is your accuracy on software development estimates? I
         | always see these productivity claims matched again "It would've
         | taken me" timelines.
         | 
         | But, it's never examined if we're good at estimating. I know I
         | am not good at estimates.
         | 
         | It's also never examined if the quality of the PR is the same
         | as it would've been. Are you skipping steps and system
         | understanding which let you go faster, but with a higher %
         | chance of bugs? You can do that without AI and get the same
         | speed up.
        
         | xoralkindi wrote:
         | How are you confident in the code, coding style and practices
         | simply because the LLM says so. How do you know it is not
         | hallucinating since you don't understand the codebase?
        
         | OptionOfT wrote:
         | Now the question is: did you gain the same knowledge and
         | proficiency in the codebase that you would've gained
         | organically?
         | 
         | I find that when working with an LLM the difference in
         | knowledge is the same as learning a new language. Learning to
         | understanding another language is easier than learning to speak
         | another language.
         | 
         | It's like my knowledge of C++. I can read it, and I can make
         | modifications of existing files. But writing something from
         | scratch without a template? That's a lot harder.
        
         | nico wrote:
         | Some additional notes given the comments in the thread
         | 
         | * I wasn't trying to be dismissive of the article or the study,
         | just wanted to present a different context in which AI tools do
         | help a lot
         | 
         | * It's not just code. It also helps with a lot of tasks. For
         | example, Claude Code figured out how to "manually" connect to
         | the AWS cluster that hosted the source db, tested different
         | commands via docker inside the project containers and overall
         | helped immensely with discovery of the overall structure and
         | infrastructure of the project
         | 
         | * My professional experience as a developer, has been that
         | 80-90% of the time, results trump code quality. That's just the
         | projects and companies I've been personally involved with.
         | Mostly saas products in which business goals are usually
         | considered more important than the specifics of the tech stack
         | used. This doesn't mean that 80-90% of code is garbage, it just
         | means that most of the time readability, maintainability and
         | shipping are more important than DRY, clever solutions or
         | optimizations
         | 
         | * I don't know how helpful AI is or could be for things that
         | require super clever algorithms or special data structures, or
         | where code quality is incredibly important
         | 
         | * Having said that, the AI tools I've used can write pretty
         | good quality code, as long as they are provided with good
         | examples and references, and the developer is on top of
         | properly managing the context
         | 
         | * Additionally, these tools are improving almost on a weekly or
         | monthly basis. My experience with them has drastically changed
         | even in the last 3 months
         | 
         | At the end of the day, AI is not magic, it's a tool, and I as
         | the developer, am still accountable for the code and results
         | I'm expected to deliver
        
       | rosspackard wrote:
       | One mediocre paper/study (it should not even be called that with
       | all the bias and sample size issues) and now we have to put up
       | with stories re-hashing and dissecting it. I really hope these
       | don't get upvoted more in the future.
       | 
       | 16 devs. And they weren't allowed to pick which tasks they used
       | the AI on. Ridiculous. Also using it on "old and >1 million line"
       | codebases and then extrapolating that to software engineering in
       | general.
       | 
       | Writers like this then theorize why AI isn't helpful, then those
       | "theories" get repeated until it feels less like a theory and
       | more like a fact and it all proliferates into an echo chamber of
       | AI isn't a useful tool. There have been too many anecdotes and my
       | own personal experience to ignore that it isn't useful.
       | 
       | It is a tool and you have to learn it to be successful with it.
        
         | davidcbc wrote:
         | > And they weren't allowed to pick which tasks they used the AI
         | on.
         | 
         | They were allowed to pick whether or not to use AI on a subset
         | of tasks. They weren't forced to use AI on tasks that don't
         | make sense for AI
        
           | rosspackard wrote:
           | Half the tasks they were not allowed to use AI.
        
             | davidcbc wrote:
             | Yes, and the other half they had the _option_ to use AI.
             | That 's why I said they were allowed to pick whether or not
             | to use AI on a _subset_ of tasks. On the other subset they
             | were not allowed to use AI.
        
           | throwaway284927 wrote:
           | That is not true, usage of AI was decided randomly. From the
           | paper:
           | 
           | "To directly measure the impact of AI tools on developer
           | productivity, we conduct a randomized controlled trial by
           | having 16 developers complete 246 tasks (2.0 hours on
           | average) on well-known open-source repositories (23,000 stars
           | on average) they regularly contribute to. Each task is
           | randomly assigned to allow or disallow AI usage, and we
           | measure how long it takes developers to complete tasks in
           | each condition."
        
             | davidcbc wrote:
             | Directly from the paper:
             | 
             | > If AI is allowed, developers can use any AI tools or
             | models they choose, including no AI tooling if they expect
             | it to not be helpful. If AI is not allowed, no generative
             | AI tooling can be used.
             | 
             | AI is _allowed_ not _required_
        
               | throwaway284927 wrote:
               | True, my bad, I didn't read you correctly. What you said
               | was true.
               | 
               | I do believe however that it's important to emphasize the
               | fact that they didn't got to choose in general, though,
               | which I think your wording (even though it is correct)
               | does not make evident.
        
         | RamblingCTO wrote:
         | It's just the same with all the anecdotal evidence of some hype
         | guys on twitter claiming 10x performance on coding ... Same
         | same but different
        
         | jplusequalt wrote:
         | >One mediocre paper/study (it should not even be called that
         | with all the bias and sample size issues)
         | 
         | Can you bring up any specific issues with the metr study?
         | Alternatively, can you site a journal that critiques it?
        
           | rosspackard wrote:
           | It was just published. Too new for someone to conduct a
           | direct study to critique and journals don't just publish
           | critiques anyway. It would have to be a study that disputes
           | the results.
           | 
           | They used 16 developers. The confidence intervals are wide
           | and a few atypical issues per dev could swing the headline
           | figure.
           | 
           | Veteran maintainers on projects they know inside-out. This is
           | a bias.
           | 
           | Devs supplied the issue list (then randomized) which still
           | leads to subtle self-selection bias. Maintainers may pick
           | tasks they enjoy or that showcase deep repo knowledge--
           | exactly where AI probably has least marginal value.
           | 
           | Time was not independently logged and was self-reported.
           | 
           | No possible direct quality metric is possible. Could the AI
           | code be better?
           | 
           | The Hawthorne effect. Knowing they are observed paid may make
           | devs over-document, over-prompt, or simply take their time.
           | 
           | Many of the devs were new to Cursor
           | 
           | Bias in forecasting.
        
         | steveklabnik wrote:
         | > and then extrapolating that to software engineering in
         | general.
         | 
         | To the credit of the paper authors, they were very clear that
         | they were not making a claim against software engineering in
         | general. But everyone wants to reinforce their biases, so...
        
           | rosspackard wrote:
           | Great for the authors. But everyone else seems to be
           | extrapolating. Authors have a responsibility and should
           | recognize how their work will be used.
           | 
           | Metr may overall have an ok mission, but their motivation is
           | questionable. They published something like this to get
           | attention. Mission accomplished on that but they had to have
           | known how this would be twisted.
        
       | uludag wrote:
       | Great article and I was having very similar thoughts with regards
       | to this productivity study and the "Programming as Theory
       | Building" paper. I'm starting to be convinced that if you are the
       | original author of a program and still have the program's context
       | in the head, _you_ are the asymptote to which any and all AI
       | systems will approach but never surpass: maybe not in terms of
       | raw coding speed, but in terms of understanding the program, its
       | vision of development, its deficiencies and hacks, its context,
       | its users and what they want, the broader culture the program
       | exists in, etc.
       | 
       | I really like how the author then brought up the point that for
       | most daily work we don't have the theory built, even a small
       | fraction of it, and that this may or may not change the equation.
        
         | conartist6 wrote:
         | Thanks, <3
        
       | cratermoon wrote:
       | dissected https://www.fightforthehuman.com/are-developers-slowed-
       | down-...
        
       | omnicognate wrote:
       | All these studies that show "AI makes developers x% more/less
       | productive" are predicated on the idea that developer
       | "productivity" can be usefully captured in a single objectively
       | measurable number.
       | 
       | Just one problem with that...
        
         | narush wrote:
         | Thanks for the feedback! I strongly agree this is not the only
         | measure of developer productivity -- but it's certainly one of
         | them. I think this measure as speaks very directly to how
         | _many_ developers (myself included) understand the impact of AI
         | tools on their own work currently (e.g. just speeding up
         | implementation speed).
         | 
         | (The SPACE [1] framework is a pretty overview of considerations
         | here; I agree with a lot of it, although I'll note that METR
         | [2] has different motivations for studying developer
         | productivity than Microsoft does.)
         | 
         | [1] https://dl.acm.org/doi/10.1145/3454122.3454124
         | 
         | [2] https://metr.org/about
        
         | charcircuit wrote:
         | As long as the true productivity is correlated with that number
         | it should be fine.
        
       | yomismoaqui wrote:
       | Someone on X said that these agentic AI tools (Claude Code, Amp,
       | Gemini Cli) are to programming like the table saw was to hand-
       | made woodworking.
       | 
       | It can make some things faster and better than a human with a
       | saw, but you have to learn how to use them right (or you will
       | loose some fingers).
       | 
       | I personally find that agentic AI tools make me be more ambitious
       | in my projects, I can tackle some things I didn't tthougth about
       | doing before. And I also delegate work that I don't like to them
       | because they are going to do it better and quicker than me. So my
       | mind is free to think on the real problems like architecture, the
       | technical debt balance of my code...
       | 
       | Problem is that there is the temptation of letting the AI agent
       | do everything and just commit the result without understanding
       | YOUR code (yes, it was generated by an AI but if you sign the
       | commit YOU are responsible for that code).
       | 
       | So as with any tool try to take the time to understand how to
       | better use it and see if it works for you.
        
         | bgwalter wrote:
         | "You are using it wrong!"
         | 
         | This is insulting to all pre-2023 open source developers, who
         | produced the entire stack that the "AI" robber barons use in
         | their companies.
         | 
         | It is even more insulting because no actual software of value
         | has been demonstrably produced using "AI".
        
           | yomismoaqui wrote:
           | > It is even more insulting because no actual software of
           | value has been demonstrably produced using "AI".
           | 
           | Claude Code and Amp (equivalent from Sourcegraph) are created
           | by humans using these same tools to add new features and fix
           | bugs.
           | 
           | Having used both tools for some weeks I can tell you that
           | they provide a great value to me, enough that I see paying
           | $100 monthly as a bargain related to that value.
           | 
           | Edit: typo
        
             | jdiff wrote:
             | GP is pointing out the distinct lack of AI driven
             | development in the wild. At this point, agents should be
             | visibly maintaining at least a few popular codebases across
             | this world wide web. The fact that there aren't raises some
             | eyebrows for the claims that are regularly made by
             | proponents. Not just the breathless proponents, either.
             | Even taking claims very conservatively, FOSS maintainer
             | burnout should be a thing of the past, but the only noted
             | interaction with AI seems to be amplifying it.
        
               | yomismoaqui wrote:
               | It's disingenuous to expect that tools that are publicly
               | available for less than a year have a massive adoption in
               | the wild.
               | 
               | Think that these were internal tools that provided value
               | to engineers on Anthropic, OpenAI, Google & others and
               | now are starting to be adopted by the general public.
               | 
               | Some people are overhyped and some seem hurt because I
               | don't know, maybe they define themselves by their ability
               | to write code by hand.
               | 
               | I have no horse in this race and I can only tell you
               | about my experience and I can tell you that the change is
               | coming.
               | 
               | Also if you don't trust a random HN nickname go read
               | about the experiences of people like Armin Ronacher
               | (Flask creator), Steve Yegge or Thomas H. Ptacek.
               | 
               | - https://lucumr.pocoo.org/2025/6/4/changes/ -
               | https://sourcegraph.com/blog/the-brute-squad -
               | https://fly.io/blog/youre-all-nuts/
        
               | asadotzler wrote:
               | >It's disingenuous to expect that tools that are publicly
               | available for less than a year have a massive adoption in
               | the wild.
               | 
               | Github got massive adoption in a year, probably 100K
               | developers and tens of thousands of projects including
               | big names like Ruby on Rails.
               | 
               | I'm sure if I spent more than 2 minutes on this I'd have
               | even more examples but this one is enough to neuter your
               | claims.
        
         | candiddevmike wrote:
         | > to programming like the table saw was to hand-made
         | woodworking
         | 
         | This is a ridiculous comparison because the table saw is a
         | precision tool (compared to manual woodworking) when agentic AI
         | is anything but IMO.
        
           | marcellus23 wrote:
           | The nature of the comparison is in the second paragraph. It's
           | nothing to do with how precise it is.
        
       | d00mB0t wrote:
       | Blasphemy! How dare you say our Emperor has no clothes! AI is
       | becoming a cult and I'm not here for it.
        
       | gr8beehive wrote:
       | Mirror neurons got people drinking the same stupid kool aid
       | without realizing it.
        
       | narush wrote:
       | Hey HN -- study author here! (See previous thread on the paper
       | here [1].)
       | 
       | I think this blog post is an interesting take on one specific
       | factor that is likely contributing to slowdown. We discuss this
       | in the paper [2] in the section "Implicit repository context
       | (C.1.5)" -- check it out if you want to see some developer quotes
       | about this factor.
       | 
       | > This is why AI coding tools, as they exist today, will
       | generally slow someone down if they know what they are doing, and
       | are working on a project that they understand.
       | 
       | I made this point in the other thread discussing the study, but
       | in general, these results being surprising makes it easy to read
       | the paper, find one factor that resonates, and conclude "ah, this
       | one factor probably just explains slowdown." My guess: there is
       | no one factor -- there's a bunch of factors that contribute to
       | this result -- at least 5 seem likely, and at least 9 we can't
       | rule out (see the full factors table on page 11).
       | 
       | > If there are no takers then I might try experimenting on
       | myself.
       | 
       | This sounds super cool! I'd be very excited to see how you set
       | this up + how it turns out... please do shoot me an email (in the
       | paper) if you do this!
       | 
       | > AI slows down open source developers. Peter Naur can teach us
       | why
       | 
       | Nit: I appreciate how hard it is to write short titles
       | summarizing the paper (the graph title is the best I was able to
       | do after a lot of trying) -- but I might have written this
       | "Early-2025 AI slows down experienced open-source developers.
       | Peter Naur can give us more context about one specific factor."
       | It's admittedly less of a catchy-title, but I think getting the
       | qualifications right are really important!
       | 
       | Thanks again for the sweet write-up! I'll hang around in the
       | comments today as well.
       | 
       | [1] https://news.ycombinator.com/item?id=44522772
       | 
       | [2] https://metr.org/Early_2025_AI_Experienced_OS_Devs_Study.pdf
        
         | antonvs wrote:
         | > Early-2025 AI slows down experienced open-source developers.
         | 
         | Even that's too general, because it'll depend on what the task
         | is. It's not as if open source developers in general never work
         | on tasks where AI could save time.
        
           | narush wrote:
           | We call this over-generalization out specifically in the "We
           | do not provide evidence that:" table in the blog post and
           | paper - I agree there are tasks these developers are likely
           | sped up on with early-2025 tools.
        
             | 2muchcoffeeman wrote:
             | I think this will be the key. Finding appropriate tasks.
             | Even on code bases I know, I can find tedious things for
             | the AI to do. Sometimes I can find tedious things for it to
             | do that I would never have dreamt of doing in the past.
             | Now, I think "will it do it?".
             | 
             | Once I got a hang of identifying problems, or being more
             | targeted, I was spending less time messing about and got
             | things done quicker.
        
         | seanwilson wrote:
         | If this makes sense, how is the study able to give a reasonable
         | measure of how long an issue/task should have taken, vs how
         | long it took with AI to determine that using AI was slower?
         | 
         | Or it's comparing how long the dev thought it should take with
         | AI vs how long it actually took, which now includes the dev's
         | guess of how AI impacts their productivity?
         | 
         | When it's hard to estimate how difficult an issue should be to
         | complete, how does the study account for this? What percent
         | speed up or slow down would be noise due to estimates being
         | difficult?
         | 
         | I do appreciate that this stuff is very hard to measure.
        
           | krona wrote:
           | An easier way to think about it might be if you timed how
           | long it took each ticket in your backlog. You also recorded
           | whether you were drunk or not when you worked on it, and the
           | ticket was selected at random from your backlog. The
           | assumption (null-hypothesis) is that being drunk has no
           | effect on ticket completion time.
           | 
           | Using the magic of statistics, if you have completed enough
           | tickets, we can determine whether the null-hypothesis holds
           | (for a given level of statistical certainty), and if it
           | doesn't, low large is the difference (with a margin of
           | error).
           | 
           | That's not to say there couldn't be other causes for the
           | difference (if there is one), but that's how science
           | proceeds, generally.
        
             | jiggawatts wrote:
             | The challenge with "controlled experiments" is that saying
             | to developers to "use AI for all of your tickets for a
             | month" forces a specific tool onto problems that may not
             | benefit from that tool.
        
               | msgodel wrote:
               | Most corporate software problems don't need AI at all.
               | They're really coordination/communication/administration
               | problems hiding as technical problems.
        
         | jwhiles wrote:
         | Thanks for the response, and apologies for misrepresenting your
         | results somewhat! I'm probably not going to change the title
         | since I am at heart and polemicist and a sloppy thinker, but
         | I'll update the article to call out this misrepresentation.
         | 
         | That said, I think that what I wrote more or less encompasses
         | three of the factors you call out as being likely to
         | contribute: "High developer familiarity with reposito- ries",
         | "Large and complex repositories", and "Implicit repository
         | context".
         | 
         | I thought more about experimenting on myself, and while I hope
         | to do it - I think it will be very hard to create a controlled
         | enviornment whilst also responding to the demands the job puts
         | on me. I also don't have the luxury of a list of well scoped
         | tasks that could feasibly be completed in a few hours.
        
         | calf wrote:
         | Slowing down isn't necessarily bad, maybe slow programming
         | (literate/Knuth comes to mind as another early argument)
         | encourages better theory formation. Maybe programming today is
         | like fast food, and proper theory and abstraction (and language
         | design) requires a good measure of slow and deliberate work
         | that has not been the norm in industry.
        
         | karmakaze wrote:
         | I would expect any change to an optimized workflow (developing
         | own well understood project) to initially be slower. What I'd
         | like to see is how these same developers do 6 months or a year
         | from now _after using AI has become the natural workflow_ on
         | these same projects. The article mentions that these results
         | don 't extrapolate to other devs, but it's important to note
         | that it may not extrapolate over time to these same devs.
         | 
         | I myself am just getting started and I can see how so many
         | things can be scripted with AI that would be very difficult to
         | (semi-)automate without.
        
       | neuroelectron wrote:
       | Good article and it makes sense. I wish I had sometime in my
       | career worked on a codebase that was possible to be understood
       | without 10 years of experience. Instead most of my development
       | time was spent tracing execution paths through tangles of
       | abstractions in nested objects in 10M LOC legacy codebases. My
       | buddy who introduced me to the job is still doing it today and
       | now uses AI and this has given him the free time to start working
       | on his own side projects. So there's certain types if jobs where
       | AI will certainly speed up your development.
        
       | bunderbunder wrote:
       | > It's a really fabulous study...
       | 
       | Ehhhh... not so much. It had serious design flaws in both the
       | protocol and the analysis. This blog post is a fairly
       | approachable explanation of what's wrong with it:
       | https://www.argmin.net/p/are-developers-finally-out-of-a-job
        
         | narush wrote:
         | Hey, thanks for linking this! I'm a study author, and I greatly
         | appreciate that this author dug into the appendix and provided
         | feedback so that other folks can read it as well.
         | 
         | A few notes if it's helpful:
         | 
         | 1. This post is primarily worried about ordering considerations
         | -- I think this is a valid concern. We explicitly call this out
         | in the paper [1] as a factor we can't rule out -- see "Bias
         | from issue completion order (C.2.4)". We have no evidence this
         | occurred, but we also don't have evidence it didn't.
         | 
         | 2. "I mean, rather than boring us with these robustness checks,
         | METR could just release a CSV with three columns (developer ID,
         | task condition, time)." Seconded :) We're planning on open-
         | sourcing pretty much this data (and some core analysis code)
         | later this week here: https://github.com/METR/Measuring-
         | Early-2025-AI-on-Exp-OSS-D... - star if you want to dig in when
         | it comes out.
         | 
         | 3. As I said in my comment on the post, the takeaway at the end
         | of the post is that "What we can glean from this study is that
         | even expert developers aren't great at predicting how long
         | tasks will take. And despite the new coding tools being
         | incredibly useful, people are certainly far too optimistic
         | about the dramatic gains in productivity they will bring." I
         | think this is a reasonable takeaway from the study overall. As
         | we say in the "We do not provide evidence that:" section of the
         | paper (Page 17), we don't provide evidence across all
         | developers (or even most developers) -- and ofc, this is just a
         | point-in-time measurement that could totally be different by
         | now (from tooling and model improvements in the past month
         | alone).
         | 
         | Thanks again for linking, and to the original author for their
         | detailed review. It's greatly appreciated!
         | 
         | [1]
         | https://metr.org/Early_2025_AI_Experienced_OS_Devs_Study.pdf
        
           | bunderbunder wrote:
           | Thanks for the response, you make some very points. Sorry, I
           | had missed your response on the original post. I don't know
           | if it was there yet, or because for some reason their blog is
           | configured to only show the first two comments by default. :/
           | Either way, my bad.
           | 
           | I think my bias as someone who spends too much time looking
           | at social science papers is that the protocol allows for
           | spillover effects that, to me, imply that the results must be
           | interpreted much more cautiously than a lot of people are
           | doing. (And then on top of that I'm trying to be hyper-
           | cautious and skeptical when I see a paper whose conclusions
           | align with my biases on this topic.)
           | 
           | Granted, that sort of thing is my complaint about basically
           | every study on developer productivity when using LLMs that
           | I've seen so far. So I appreciate how difficult this is to
           | study in practice.
        
       | tomasz_fm wrote:
       | Only one developer in this study had more than 50h of Cursor
       | experience, _including_ time spent using Cursor during the study.
       | That one developer saw a 25% speed improvement.
       | 
       | Everyone else was an absolute Cursor beginner with barely any
       | Cursor experience. I don't find it surprising that using tools
       | they're unfamiliar with slows software engineers down.
       | 
       | I don't think this study can be used to reach any sort of
       | conclusion on use of AI and development speed.
        
         | Art9681 wrote:
         | This is exactly my same take. Any tool an engineer is
         | inexperienced with will slow them down. AI is no different.
        
           | bluefirebrand wrote:
           | This runs counter to the starry eyed promises of AI letting
           | people with no experience accomplish things
        
             | TeMPOraL wrote:
             | That promise is true, though, and the two claims are not
             | opposite. The devil is in details, specifically in what you
             | mean by "people" and "accomplish things".
             | 
             | If by "people" you mean "general public", and by
             | "accomplish things" you mean solving some immediate
             | problems, that may or may not involve authoring a script or
             | even a small app - then yes, this is already happening, and
             | is a big reason behind the AI hype as it is.
             | 
             | If by "people" you mean "experienced software engineers",
             | and by "accomplish things" you mean meaningful
             | contributions to a large software product, measured by high
             | internal code and process quality standards, then no - AI
             | tools may not help with that directly, though chances are
             | greater when you have enough experience with those tools to
             | reliably give them right context and steer away from
             | failure modes.
             | 
             | Still, solving one-off problems != incremental improvements
             | to a large system.
        
               | bluefirebrand wrote:
               | > If by "people" you mean "experienced software
               | engineers",
               | 
               | My post is a single sentence and I literally wrote
               | "people with no experience"
        
               | helloplanets wrote:
               | He addressed your point in the paragraph before that. The
               | paragraph from which you quoted was meant to show the
               | difference between your point and the fact that the
               | original research was indeed measuring software
               | engineers.
        
               | bluefirebrand wrote:
               | My point is that I was very clear about what people I was
               | referring to.
               | 
               | No need for all the "if by people you mean" rigamarole
        
               | ben_w wrote:
               | Then your previous point is false, because "X helps Y"
               | doesn't run counter to any promise that "X helps Z".
               | 
               | You said the second. You responded to the first.
               | 
               | Y = [experts]
               | 
               | Z = [noobs]
               | 
               | {Y, Z} [?] [all humans]
        
             | jonfw wrote:
             | AI let's people with no experience accomplish things.
             | People who have experience can create those things without
             | AI. Those experienced folks will likely outperform novices,
             | even when novices leverage AI.
             | 
             | None of these statements are controversial. What we have to
             | establish is- Does the experienced AI builder outperform
             | the experienced manual coder?
        
         | narush wrote:
         | Hey, thanks for digging into the details here! Copying a
         | relevant comment
         | (https://news.ycombinator.com/item?id=44523638) from the other
         | thread on the paper, in case it's help on this point.
         | 
         | 1. Some prior studies that find speedup do so with developers
         | that have similar (or less!) experience with the tools they
         | use. In other words, the "steep learning curve" theory doesn't
         | differentially explain our results vs. other results.
         | 
         | 2. Prior to the study, 90+% of developers had reasonable
         | experience prompting LLMs. Before we found slowdown, this was
         | the only concern that most external reviewers had about
         | experience was about prompting -- as prompting was considered
         | the primary skill. In general, the standard wisdom was/is
         | Cursor is very easy to pick up if you're used to VSCode, which
         | most developers used prior to the study.
         | 
         | 3. Imagine all these developers had a TON of AI experience. One
         | thing this might do is make them worse programmers when not
         | using AI (relatable, at least for me), which in turn would
         | raise the speedup we find (but not because AI was better, but
         | just because with AI is much worse). In other words, we're
         | sorta in between a rock and a hard place here -- it's just
         | plain hard to figure out what the right baseline should be!
         | 
         | 4. We shared information on developer prior experience with
         | expert forecasters. Even with this information, forecasters
         | were still dramatically over-optimistic about speedup.
         | 
         | 5. As you say, it's totally possible that there is a long-tail
         | of skills to using these tools -- things you only pick up and
         | realize after hundreds of hours of usage. Our study doesn't
         | really speak to this. I'd be excited for future literature to
         | explore this more.
         | 
         | In general, these results being surprising makes it easy to
         | read the paper, find one factor that resonates, and conclude
         | "ah, this one factor probably just explains slowdown." My
         | guess: there is no one factor -- there's a bunch of factors
         | that contribute to this result -- at least 5 seem likely, and
         | at least 9 we can't rule out (see the factors table on page
         | 11).
         | 
         | I'll also note that one really important takeaway -- that
         | developer self-reports after using AI are overoptimistic to the
         | point of being on the wrong side of speedup/slowdown -- isn't a
         | function of which tool they use. The need for robust, on-the-
         | ground measurements to accurately judge productivity gains is a
         | key takeaway here for me!
         | 
         | (You can see a lot more detail in section C.2.7 of the paper
         | ("Below-average use of AI tools") -- where we explore the
         | points here in more detail.)
        
           | brulard wrote:
           | 1. That does not support these results in any way 2. Having
           | experience prompting is quite a little part of being able to
           | use agentic IDE tools. It's like relating cutting onion to
           | being a good cook
           | 
           | I think we should all focus on how the effectivity is going
           | to change in the long-term. We all know AI tooling is not
           | going to disappear but to become better and better. I
           | wouldn't be afraid to lose some productivity for months if I
           | would acquire new skills for the future.
        
         | WhyNotHugo wrote:
         | An interesting little detail. Any seasoned developer is likely
         | going to take _substantially_ longer if they have to use any
         | IDE except their everyday one.
         | 
         | I've been using Vim/Neovim for over a decade. I'm sure if I
         | wanted to use something like Cursor, it would take me at least
         | a month before I can productive even a fraction of my usual.
        
       | whatever1 wrote:
       | They didn't use the latest model that was released yesterday
       | night. Follow my paid course to learn how to vibe code/s
        
       | methuselah_in wrote:
       | Those of current generation students who have access to ai might
       | become slow over time. Because when things are not readily
       | available then they have to struggle and work harder in that
       | process, at that time I thing human a lot of secondary things !
       | Now when everything is easily available especially knowledge
       | without knowing how to struggle with basics. It will eventually
       | make kids dumb. But can be opposite also. Eventually even I
       | become slow even I keep on using chat gpt or gemini.
        
       | piker wrote:
       | My main two attempts at using an "agentic" coding workflow were
       | trying to incorporate an Outlook COM interface into my rust code
       | base and to streamline an existing abstract windows API
       | interaction to avoid copying memory a couple of times. Both
       | wasted tremendous amounts of time and were ultimately abandoned
       | leaving me only slightly more educated about windows development.
       | They make great autocompletion engines but I just cannot see them
       | being useful in my project otherwise.
        
         | crinkly wrote:
         | This is typically what I see when I've seen it applied. And as
         | always trying to hammer nails in with a banana.
         | 
         | Rather than fit two generally disparate things together it's
         | probably better to just use VSTO and C# (hammer and nails)
         | rather than some unholy combination no one else has tried or
         | suffered through. When it goes wrong there's more info to get
         | you unstuck.
        
           | piker wrote:
           | To be fair though, unsafe rust (where the COM lives) is
           | basically just C, so I totally expected it to be tractable in
           | the same way it has been tractable for the last 20ish years?
           | But it isn't.
           | 
           | Why is interacting with the OS' API in a compiled language
           | the wrong approach in 2025? Why must I use this managed
           | Frankenstein's monster of dotnet? I didn't want to ship or
           | expect a whole runtime for what should be a tiny convenience
           | DLL. Insane
        
         | charcircuit wrote:
         | I had the opposite experience. Gemini was able to work with COM
         | and accomplish what I needed despite me never using COM before.
        
           | tonyedgecombe wrote:
           | I've done a lot of work with COM over the years and that is
           | the last technology I would trust to an AI. It's very easy to
           | write COM code that appears to work but contains subtle bugs.
        
             | piker wrote:
             | That was my issue. Integration works, Outlook itself not so
             | much, afterwards. (I.e. memory error.)
        
           | piker wrote:
           | Actually hadn't tried Gemini with it yet. Perhaps worth
           | taking a look.
        
         | jdiff wrote:
         | They make great autocompletion engines, most of the time. It's
         | nice when it can recognize that I'm replicating a specific math
         | formula and expands out the next dozen lines for me. It's less
         | nice when it predicts code that's not even syntactically valid
         | for the language or the correct API for the library I'm using.
         | Those times, for whatever reason, seem to be popping up a lot
         | in the last few weeks so I find myself disabling those
         | suggestions more often than not.
        
       | joshmarlow wrote:
       | I've gotten some pretty cool things working with LLMs doing most
       | of the heavy lifting using the following approaches:
       | 
       | * spec out project goals and relevant context in a README and
       | spec out all components; have the AI build out each component and
       | compose them. I understand the high-level but don't necessarily
       | know all of the low-level details. This is particularly helpful
       | when I'm not deeply familiar with some of the underlying
       | technologies/libraries. * having an AI write tests for code that
       | I've verified is working. As we all know, testing is tedious - so
       | of course I want to automate it. And we written tests (for well
       | written code) can be pretty easy to review.
        
       | antimora wrote:
       | I'm one of the regular code reviewers for Burn (a deep learning
       | framework in Rust). I recently had to close a PR because the
       | submitter's bug fix was clearly written entirely by an AI agent.
       | The "fix" simply muted an error instead of addressing the root
       | cause. This is exactly what AI tends to do when it can't identify
       | the actual problem. The code was unnecessarily verbose and even
       | included tests for muting the error. Based on the person's
       | profile, I suspect their motivation was just to get a commit on
       | their record. This is becoming a troubling trend with AI tools.
        
         | meindnoch wrote:
         | >a deep learning framework in Rust [...] This is becoming a
         | troubling trend with AI tools.
         | 
         |  _The serpent is devouring its own tail._
        
           | LoganDark wrote:
           | Deep learning can be incredibly cool and not just used for AI
           | slop.
        
           | TeMPOraL wrote:
           | OTOH when they'll start getting _good_ AI contributions,
           | then... it 'll be too late for us all.
        
         | lvl155 wrote:
         | This is a real problem that's only going to get worse. With the
         | major model providers basically keeping all the data
         | themselves, I frankly don't like this trend long term.
        
         | tomrod wrote:
         | As a side question: I work in AI, but mostly python and theory
         | work. How can I best jump into Burn? Rust has been intriguing
         | to me for a long time
        
         | dawnerd wrote:
         | That's what I love about LLMs. You can spot it doesn't know the
         | answer, tell it that it's wrong and it'll go, "You're
         | absolutely right. Let me actually fix it"
         | 
         | It scares me how much code is being produced by people without
         | enough experience to spot issues or people that just gave up
         | caring. We're going to be in for wild ride when all the
         | exploits start flowing.
        
           | candiddevmike wrote:
           | > tell it that it's wrong and it'll go, "You're absolutely
           | right. Let me actually fix it"
           | 
           | ...and then it still doesn't actually fix it
        
             | colechristensen wrote:
             | Sometimes it does... sometimes.
             | 
             | I recently had a nice conversation looking for some reading
             | suggestions from an LLM. The first round of suggestions
             | were superb, some of them I'd already read, some were
             | entirely new and turned out great. Maybe a dozen or so
             | great suggestions. Then it was like squeezing blood from a
             | stone but I did get a few more. After that it was like
             | talking to a babbling idiot. Repeating the same suggestions
             | over and over, failing to listen to instructions, and
             | generally just being useless.
             | 
             | LLMs are great on the first pass but the further you get
             | away from that they degrade into uselessness.
        
               | aquariusDue wrote:
               | Yeah, when I first heard about "one-shot"ing it felt more
               | like a trick instead of a useful heuristic but with time
               | my experience mimics yours, nowadays I try to one-shot
               | small-ish changes instead of going back and forth.
        
               | daxfohl wrote:
               | I've had some luck in these cases prompting "your context
               | seems to be getting too bloated. summarize this
               | conversation into a prompt that I can feed into a new
               | chat with a fresh context. make sure to include <...>".
               | 
               | Sometimes it works well the first time, and sometimes it
               | spits out a summary where you can see what it is confused
               | about, and you can guide it to create a better summary.
               | Sometimes just having that summary in its context gets it
               | over the hump and you can just say "actually I'm going to
               | continue with you; please reference this summary going
               | forward", and sometimes you actually do have to restart
               | the LLM with the new context. And of course sometimes
               | there's nothing that works at all.
        
               | dawnerd wrote:
               | I've had really good luck with having gpt generate a todo
               | list that's very, very detailed. Then having Claude use
               | it to check items off. Still far from perfect but since
               | doing that haven't run into context issues since I can
               | just start a new chat and feed it the todo (the todo also
               | contains project info).
        
             | mlyle wrote:
             | So, I recently have done my first couple heavily AI
             | augmented tasks for hobby projects.
             | 
             | I wrote a TON of LVGL code. The result wasn't perfect for
             | placement, but when I iterated a couple of times, it fixed
             | almost all of the issues. The result is a little hacked
             | together but a bit better than my typical first pass
             | writing UI code. I think this saved me a factor of 10 in
             | time. Next I am going to see how much of the cleanup and
             | factoring of the pile of code it can do.
             | 
             | Next I had it write a bunch of low level code to init
             | hardware. It saved me a little time compared to reading the
             | reference manual, and was more pleasant, but it wasn't
             | perfectly correct. If I did not have domain expertise I
             | would not have been able to complete the task with the LLM.
        
               | stavros wrote:
               | > If I did not have domain expertise I would not have
               | been able to complete the task with the LLM.
               | 
               | This kind of sums up my experience with LLMs too. They
               | save me a lot of time reading documentation, but I need
               | to review a lot of what they write, or it will just
               | become too brittle and verbose.
        
               | la_fayette wrote:
               | When you argued that it saved you time by a factor of 10,
               | have you even measured that properly? I initially also
               | had the feeling that LLMs save me time, but in the end it
               | didn't. I roughly compared my performance to past
               | performance by the amount of stories done and LLMs made
               | me slower even if I thought I am saving time...
               | 
               | From several month of deep work with LLMs I think they
               | are amazing pattern matchers, but not problem solvers.
               | They suggest a solution pattern based on their trained
               | weights. This even could result in real solutions, e.g.,
               | when programming Tetris or so, but not when working on
               | somewhat unique problems...
        
               | mlyle wrote:
               | I am pretty confident. Last similar LVGL thing I did took
               | me 10-12 hours, and I had a quicker iteration time
               | (running locally instead of the test hardware). Here I
               | spent a little more than an hour, testing on real
               | hardware, and the last 20 minutes was nitpicking.
               | 
               | Writing front-end display code and instantiating
               | components to look right is very much playing to the
               | model's strength, though. A carefully written sentence
               | plus context would become 40 lines of detail-dense but
               | formulaic code.
               | 
               | (I have also had a lot of luck asking it to make a first
               | pass at typesetting things in Tex, too, for similar
               | reasons)
        
               | delusional wrote:
               | There was a recent study that found that LLM users in
               | general tend to feel like they were more productive with
               | AI while actually being less productive.
        
               | asadotzler wrote:
               | presumably the study this very HN discussion responds to.
        
               | delusional wrote:
               | Heh, yep. Guess I sometimes forget to read the content
               | before commenting too.
        
             | seunosewa wrote:
             | Some prompts can help:
             | 
             | "Find the root cause of this problem and explain it"
             | 
             | "Explain why the previous fix didn't work."
             | 
             | Often, it's best to undo the action and provide more
             | context/tips.
             | 
             | Often, switching to Gemini 2.5 Pro when Claude is stumped
             | helps a lot.
        
             | brazzy wrote:
             | My favourite recent experience was switching multiple times
             | between using a library function and rolling its own
             | implementation, each time claiming that it's "simplifying"
             | the code and making it "more reliable".
        
             | Retr0id wrote:
             | I was trying out Copilot recently for something trivial. It
             | made the change as requested, but also added a comment that
             | stated something obvious.
             | 
             | I asked it to remove the comment, which it enthusiastically
             | agreed to, and then... didn't. I couldn't tell if it was
             | the LLM being dense or just a bug in Copilot's
             | implementation.
        
           | colechristensen wrote:
           | I also get things like this from very experienced engineers
           | working outside their area of expertise. It's obviously less
           | of the completely boneheaded suggestion but still doing
           | exactly the wrong thing suggested by AI that required a
           | person to step in and correct.
        
           | cogman10 wrote:
           | My favorite LLM moment. I wrote some code, asked the LLM
           | "Find any bugs or problems with this code" and of course what
           | it did was hyperfocus on an out of date comment (that I
           | didn't write). Since the problem no longer existed identified
           | in the comment, the LLM just spat out like 100 lines of
           | garbage to refactor the code.
        
           | rectang wrote:
           | > _" You're absolutely right."_
           | 
           | I admit a tendency to anthropomorphize the LLM and get
           | irritated by this quirk of language, although it's not bad
           | enough to prevent me from leveraging the LLM to its fullest.
           | 
           | The key when acknowledging fault is to _show_ your sincerity
           | through actual effort. For technical problems, that means
           | demonstrating that you have worked to analyze the issue, take
           | corrective action, and verify the solution.
           | 
           | But of course current LLMs are weak at understanding, so they
           | can't pull that off. I wish that the LLM could say, "I don't
           | know", but apparently the current tech can't know that that
           | it doesn't know.
           | 
           | And so, as the LLM flails over and over, it shamelessly
           | kisses ass and bullshits you about the work its doing.
           | 
           | I figure that this quirk of LLMs will be minimized in the
           | near future by tweaking the language to be slightly less
           | obsequious. Improved modeling and acknowledging uncertainty
           | will be a heavier lift.
        
           | daxfohl wrote:
           | It'd be nice if github had a feature that updated the issue
           | with this context automatically too, so that if this agent
           | gives up and closes the PR, the next agent doesn't go and do
           | the exact same thing.
        
         | pennomi wrote:
         | This is the most frustrating thing LLMs do. They put wide
         | try:catch structures around the code making it impossible to
         | actually track down the source of a problem. I want my code to
         | fail fast and HARD during development so I can solve every
         | problem immediately.
        
           | daxfohl wrote:
           | Seems like there's a need for github to create a separate
           | flow for AI-cretaed PRs. Project maintainers should be able
           | to stipulate rules like this in English, and an AI "pre-
           | reviewer" would check that the AI has followed all these
           | rules before the PR is created, and chat with the AI
           | submitter to resolve any violations. For exceptional cases, a
           | human submitter is required.
           | 
           | Granted, the compute required is probably more expensive than
           | github would offer for free, and IDK whether it'd be within
           | budget for many open-source projects.
           | 
           | Also granted, something like this may be useful for human-
           | sourced PRs as well, though perhaps post-submission so that
           | maintainers can see and provide some manual assistance if
           | desired. (And also granted, in some cases maybe maintainers
           | would want to provide manual assistance to AI submissions,
           | but I expect the initial triaging based on whether it's a
           | human or AI would be what makes sense in most cases).
        
           | kfajdsl wrote:
           | This is my number one complaint with LLM produced code too.
           | The worst thing is when it swallows an error to print its own
           | error message with far less info and no traceback.
           | 
           | In my rules I tell it that try catches are completely banned
           | unless I explicitly ask for one (an okay tradeoff, since
           | usually my error boundaries are pretty wide and I know where
           | I want them). I know the context length is getting too long
           | when it starts ignore that.
        
         | jampa wrote:
         | > I suspect their motivation was just to get a commit on their
         | record. This is becoming a troubling trend with AI tools.
         | 
         | It has been for a while, AI just makes SPAM more effective:
         | 
         | https://news.ycombinator.com/item?id=24643894
        
         | doug_durham wrote:
         | You should be rejecting the PR because the fix was
         | insufficient, not because it was AI agent written. Bad code is
         | bad code regardless of the source. I think the fixation on how
         | the code was generated is not productive.
        
           | RobinL wrote:
           | The problem is that code often takes as long to review as to
           | write, and AI potentially reduces the quality bar to pull
           | requests. So maintainers have a problem of lots of low
           | quality PRs that take time to reject
        
           | glitchc wrote:
           | No, that's not how code review works. Getting inside the mind
           | of the developer, understanding how they thought about the
           | fix, is critical to the review process.
           | 
           | If an actual developer wrote this code and submitted it
           | willingly, it would either constitute malice, an attempt to
           | sabotage the codebase or inject a trojan, or stupidity, for
           | failing to understand the purpose of the error message. With
           | an LLM we mostly have stupidity. Flagging it as such reveals
           | the source of the stupidity, as LLMs do not actually
           | understand anything.
        
           | rustyminnow wrote:
           | > You should be rejecting the PR because the fix was
           | insufficient
           | 
           | I mean they probly could've articulated it your way, but I
           | think that's basically what they did... they point out the
           | insufficient "fix" later, but the root cause of the "fix" was
           | blind trust in AI output, so that's the part of the story
           | they lead with.
        
         | Macha wrote:
         | I recently reviewed a MR from a coworker. There was a test that
         | was clearly written by AI, except I guess however he prompted
         | it, it gave some rather poor variable names like "thing1",
         | "thing2", etc. in test cases. Basically, these were multiple
         | permutations of data that all needed to be represented in the
         | result set. So I asked for them to be named distinctively,
         | maybe by what makes them special.
         | 
         | It's clear he just took that feedback and asked the AI to make
         | the change, and it came up with a change that gave them all
         | very long, very unique names, that just listed all the unique
         | properties in the test case. But to the extent that they sort
         | of became noise.
         | 
         | It's clear writing the PR was very fast for that developer, I'm
         | sure they felt they were X times faster than writing it
         | themselves. But this isn't a good outcome for the tool either.
         | And I'm sure if they'd reviewed it to the extent I did, a lot
         | of that gained time would have dissipated.
        
         | nerdjon wrote:
         | I will never forget being in a code review for a upcoming
         | release, there was a method that was... different. Like
         | massively different with no good reason why it was changed as
         | much as it was for such a small addition.
         | 
         | We asked the person why they made the change, and "silence".
         | They had no reason. It became painfully clear that all they did
         | was copy and paste the method into an LLM and say "add this
         | thing" and it spit out a completely redone method.
         | 
         | So now we had a change that no one in the company actually knew
         | just because the developer took a shortcut. (this change was
         | rejected and reverted).
         | 
         | The scariest thing to me is no one actually knowing what code
         | is running anymore with these models having a tendency to make
         | change for the sake of making change (and likely not actually
         | addressing the root thing but a shortcut like you mentioned)
        
         | 0xbadcafebee wrote:
         | > The "fix" simply muted an error instead of addressing the
         | root cause.
         | 
         | FWIW, I have seen human developers do this countless times. In
         | fact there are many people in engineering that will argue for
         | these kinds of "fixes" by default. Usually it's in closed-
         | source projects where the shittiness is hidden from the world,
         | but trust me, it's common.
         | 
         | > I suspect their motivation was just to get a commit on their
         | record. This is becoming a troubling trend with AI tools.
         | 
         | There was already a problem (pre-AI) with shitty PRs on GitHub
         | made to try to game a system. Regardless of how they made the
         | change, the underlying problem is a policy one: how to deal
         | with people making shitty changes for ulterior motives. I
         | expect the solution is actually _more_ AI to detect shitty
         | changes from suspicious submitters.
         | 
         | Another solution (that I know nobody's going to go for): stop
         | using GitHub. Back in the "olden times", we just had CVS,
         | mailing lists and patches. You had to perform some effort in
         | order to get to the point of getting the change done and
         | merged, and it was not necessarily obvious afterward that you
         | had contributed. This would probably stop 99% of people who are
         | hoping for a quick change to boost their profile.
        
       | hartator wrote:
       | I am not super sure how to quickly writing benchmark scripts that
       | are one-shot used slows anyone down, but okay.
        
       | lsy wrote:
       | Typically debugging, e.g., a tricky race condition in an
       | unfamiliar code base would require adding logging, refactoring
       | library calls, inspecting existing logs, and even rewriting parts
       | of your program to be more modular or understandable. This is
       | part of the theory-building.
       | 
       | When you have an AI that says "here is the race condition and
       | here is the code change to make to fix it", that might be
       | "faster" in the immediate sense, but it means you aren't
       | understanding the program better or making it easier for anyone
       | else to understand. There is also the question of whether this
       | process is sustainable: does an AI-edited program eventually fall
       | so far outside what is "normal" for a program that the AI becomes
       | unable to model correct responses?
        
         | sodapopcan wrote:
         | This is always my thought whenever I hear the "AI let me build
         | a feature in a codebase I didn't know in a language I didn't
         | know" (which is often, there is at one in these comments).
         | Great, but what have you learned? This is fine for small
         | contributions, I guess, but I don't hear a lot of stories of
         | long-term maintenance. Unpopular opinion, though, I know.
        
           | threetonesun wrote:
           | I guess it's a question of how anyone learns. There's some
           | value in typing code, I suppose, but with tab complete that's
           | been gone for a long time. Letting AI write something and
           | then reading it seems as good as copying and pasting from
           | some other source.
        
             | sodapopcan wrote:
             | I'm not super qualified to answer as I haven't gone deep
             | into AI at all. But from my limited observations I'd say
             | yes and no. You generally aren't copy/pasting entire
             | features, just snippets that you yourself have to string
             | together in a sensible way. Of course there are lots of
             | people who still do this and what's why I find most people
             | in this industry infuriating to work with. It's all good
             | when it's boilerplate, and that's actually my primary use
             | of "AI"--it's essentially been a snippets replacement (and
             | is quite good at that).
        
       | andix wrote:
       | What I noticed: AI development constantly breaks my flow. It
       | makes me more tired, and I work for shorter time periods on
       | coding.
       | 
       | It's a myth that you can code a whole day long. I usually do
       | intervals of 1-3 hours for coding, with some breaks in between.
       | Procrastination can even happen on work related things, like
       | reading other project members code/changes for an hour. It has a
       | benefit to some extent, but during this time I don't get my work
       | done.
       | 
       | Agentic AI works the best for me. Small refactoring tasks on a
       | selected code snippet can be helpful, but isn't a huge time
       | saver. The worst are AI code completions (first version Copilot
       | style), they are much more noise then help.
        
         | rightbyte wrote:
         | It would be interesting to record what one do in a day at the
         | desk. Probably quite depressing to watch.
         | 
         | Like, I think 1h would be streaching it for mature codebases.
        
           | andix wrote:
           | The 1h I'm talking about is not all the time I might spend
           | reading on code. It's the time I might procrastinate on my
           | tasks with reading unrelated code.
           | 
           | Like doom scrolling on social media: Let's see what the fancy
           | new guy got done this week. I need to feel better, I'm just
           | going to look at the commits of the guy in the other team
           | that always breaks production. Let's see how close he got to
           | that recently, ...
        
       | afro88 wrote:
       | I said this when the linked paper was shared and got downvotes:
       | it's based on early 2025 data. My point isn't that it should be
       | completely up to date, but that how we need to consider it in
       | that context. This is pre Claude 4, Claude Code. Pre Gemini 2.5
       | even. These models are such a big step up from what came
       | previously.
       | 
       | Just like we put a (2023) on articles here so they are considered
       | in the right context, so too this paper should be. Blanket "AI
       | tools slow sown development" statements with a "look this
       | rigorous paper says so!" is ignoring a key variable: the rate of
       | effectiveness improvement. If said paper evaluated with the
       | current models, the picture would be different. Also in 3 months
       | time. AI tools aren't a static thing that either works or don't
       | indefinitely.
        
         | tonyedgecombe wrote:
         | >This is pre Claude 4, Claude Code. Pre Gemini 2.5 even.
         | 
         | The most interesting point from the article wasn't about how
         | well the AI's worked, rather it was the gap between peoples
         | perception and their actual results.
        
       | blake1 wrote:
       | I think a reasonable summary of the study referenced is that: "AI
       | creates the perception of productivity enhancements far beyond
       | the reality."
       | 
       | Even within the study, there were some participants who saw mild
       | improvements to productivity, but most had a significant drop in
       | productivity. This thread is now full of people telling their
       | story about huge productivity gains they made with AI, but none
       | of the comments contend with the central insight of this study:
       | that these productivity gains are illusions. AI is a product
       | designed to make you value the product.
       | 
       | In matters of personal value, perception is reality, no question.
       | Anyone relying heavily on AI should really be worried that it is
       | mostly a tool for warping their self-perception, one that creates
       | dependency and a false sense of accomplishment. After all, it
       | speaks a highly optimized stream of tokens at you, and you really
       | have to wonder what the optimization goal was.
        
         | BriggyDwiggs42 wrote:
         | I've noticed that you can definitely use them to help you learn
         | something, but that your understanding tends to be more
         | abstract and LLM-like that way. You definitely want to mix it
         | up when learning too.
        
           | daxfohl wrote:
           | I've also had bad results with hallucinations there. I was
           | trying to learn more about multi-dimensional qubit
           | algorithms, and spent a whole day learning a bunch of stuff
           | that was fascinating but plain wrong. I only figured out it
           | was wrong at the end of the day when I tried to do a
           | simulation and the results weren't consistent.
           | 
           | Early in the chat it substituted a `-1` for an `i`, and
           | everything that followed was garbage. There were also some
           | errors that I spotted real-time and got it to correct itself.
           | 
           | But yeah, IDK, it presents itself so confidently and "knows"
           | so much and is so easy to use, that it's hard not to try to
           | use as a reference / teacher. But it's also quite dangerous
           | if you're not confirming things; it can send you down
           | incorrect paths and waste a ton of time. I haven't decided
           | whether the cost is worth the benefit or not.
           | 
           | Presumably they'll get better at this over time, so in the
           | long run (probably no more than a year) it'll likely easily
           | exceed the ROI breakeven point, but for now, you do have to
           | remain vigilant.
        
           | tonyedgecombe wrote:
           | I keep wondering whether the best way to use these tools is
           | to do the work yourself then ask the AI to critique it, to
           | find the bugs, optimisations or missing features.
        
         | thinkingemote wrote:
         | It's like the difference between being fast and quick. AI tools
         | make the developer feel quick but they may not be fast. It's
         | less cognitive effort in some ways. It's an interesting
         | illusion, one that is based on changing emotions from different
         | feedback loops and the effects of how memory forms.
        
           | asadotzler wrote:
           | Quickness is a burst; speed is a flow.
           | 
           | Or, "slow is smooth, and smooth is fast"
        
       | Kim_Bruning wrote:
       | I think different people use these tools differently. I've got
       | mine set up to start in "rubber duck" mode, where I do rubber
       | duck programming, before asking the AI to help me with certain
       | tasks (if at all). Low impact utility scripts? The AI gets let
       | off the leash. Critical core logic? I might do most of the work
       | myself (though having a rubber duck can still be good!)
        
       | ringeryless wrote:
       | not to mention the annoyance of AI assisted issues being opened,
       | many times incorrectly due to hallucinations. these tickets
       | hammer human teams with nonsense and suck resources away from
       | real issues.
        
       | alganet wrote:
       | This idea that some developers have some "mental model" and
       | others not is an extraordinary claim, and I don't see
       | extraordinary evidence.
       | 
       | It sounds like a good thing, right? "Wow, mental model. I want
       | that, I want to be good and have big brain", which encourages you
       | to believe the bullshit.
       | 
       | The truth is, this paper is irrelevant and a waste of time. It
       | only serves the purpose of creating discussion around the
       | subject. It's not science, it's a cupholder for marketing.
        
         | imiric wrote:
         | You couldn't be more wrong. If you've ever programmed, or
         | worked with programmers, that is not an extraordinary claim at
         | all, but a widely accepted fact.
         | 
         | A mental model of the software is what allows a programmer to
         | intuitively know why the software is behaving a certain way, or
         | what the most optimal design for a feature would be. In the
         | vast majority of cases these intuitions are correct, and other
         | programmers should pay attention to them. This ability is what
         | separates those with a mental model and those without.
         | 
         | On the other hand, LLMs are unable to do this, and are usually
         | not used in ways that help build a mental model. At best, they
         | can summarize the design of a system or answer questions about
         | its behavior, which can be helpful, but a mental model is an
         | abstract model of the software, not a textual summary of its
         | design or behavior. Those neural pathways can only be activated
         | by natural learning and manual programming.
        
           | alganet wrote:
           | > You couldn't be more wrong.
           | 
           | Explanation missing.
           | 
           | > If you've ever programmed, or worked with programmers, that
           | is not an extraordinary claim at all.
           | 
           | One step ahead of you. I already say this is engineered to
           | encourage belief "I want to be good, big brain, and open
           | source is good, I want to be good big brain".
           | 
           | It's marketing.
           | 
           | > A mental model of the software is what allows a programmer
           | [yadda yadda]
           | 
           | I'm not saying it doesn't exist, I'm saying the paper doesn't
           | provide any relevant information regarding the phenomena.
           | 
           | > Those neural pathways can only be activated by natural
           | learning and manual programming.
           | 
           | Again, probably true. But the paper doesn't provide any
           | relevant information regarding this phenomena.
           | 
           | ---
           | 
           | Your answer seems to disagree with me, but displays a
           | disjointed understanding of what I'm really addressing.
           | 
           | ---
           | 
           | As a lighthearted fun analogy, I present:
           | 
           | https://isotropic.org/papers/chicken.pdf
           | 
           | The paper does not prove the existence of chickens. It says
           | chicken a lot, but never addresses the phenomena of chickens
           | existing.
        
             | imiric wrote:
             | I'm confused by what your point is, then. You want evidence
             | of an abstraction that exists in the minds of experienced
             | developers? That's like asking for evidence of humor or
             | love. We accept these things as real because of shared
             | experiences, not because of concrete evidence.
        
               | alganet wrote:
               | My point is that the paper has no point, the article on
               | the paper is a stretch, and none of this is relevant in
               | any way except creating chatter.
               | 
               | It's useless from the research perspective. But it is a
               | cup-holder for marketing something.
               | 
               | I already laid this out very clearly in my first comment.
        
       | wellpast wrote:
       | The fact that the devs thought the AI saved them time is no
       | surprise to me... at least at this point in my career.
       | 
       | Developers (people?) in general for some reason just simply
       | cannot see time. It's why so many people don't believe in
       | estimation.
       | 
       | What I don't understand is why. Is this like a general human
       | brain limitation (like not being able to visualize four
       | dimensions, or how some folks don't have an internal monologue)?
       | 
       | Or is this more psychodynamic or emotional?
       | 
       | It's been super clear and interesting to me how developers I work
       | with _want_ to believe AI (code generation) is saving them time
       | when it's clearly obviously not.
       | 
       | Is it just the hope that one day it will? Is it fetishization of
       | AI?
       | 
       | Why in an industry that so requires clarity of thinking and
       | expression (computer processors don't like ambiguity), can we be
       | so bad at talking about, thinking about... time?
       | 
       | Don't get me started on the static type enthusiasts who think
       | their strong type system (another seeming fetish) is saving them
       | time.
        
       | munificent wrote:
       | _> The inability of developers to tell if a tool sped them up or
       | slowed them down is fascinating in itself, probably applies to
       | many other forms of human endeavour, and explains things as
       | varied as why so many people think that AI has made them 10 times
       | more productive, why I continue to use Vim, why people drive in
       | London etc._
       | 
       | In boating, there's a notion of a "set and drift" which describes
       | how wind and current pushes a boat off course. If a mariner isn't
       | careful, they'll end up far from their destination because of it.
       | 
       | This is because when you're sitting in a boat, your perception of
       | motion is relative and local. You feel the breeze on your face,
       | and you see how the boat cuts through the surrounding water. You
       | _interpret_ that as motion _towards your destination_ , but it
       | can equally consist of wind and current where the medium itself
       | is moving.
       | 
       | I think a similar effect explains all of these. Our perception of
       | "making progress" is mostly a sense of motion and "stuff
       | happening" in our immediate vicinity. It's not based on a
       | perception of the goal getting closer, which is much harder to
       | measure and develop an intuition for.
       | 
       | So people tend to choose strategies that make them feel like
       | they're making progress even if it's not the most effective
       | strategy. I think this is why people often take "shortcuts" when
       | driving that are actually longer. All of the twists and turns
       | keep them busy and make them feel like they're making more
       | progress than zoning out on a boring interstate does.
        
         | thinkingemote wrote:
         | Exactly! Waze the navigation app tends to route users on longer
         | routes but which feels more fast. When driving we perceive our
         | journey as fast or slow not by the actual length but by our
         | memories of what happened. Waze knows human drivers are happier
         | with driving a route that may be longer in time and distance of
         | they feel like they are making progress with the twists and
         | turns.
         | 
         | Ai tools makes programming feel easier. That it might be
         | actually less productive is interesting but we humans prefer
         | the easier shortcuts. Our memories of coding with AI tells us
         | that we didn't struggle and therefore we made progress.
        
           | tjr wrote:
           | That sounds like a navigation tool that I absolutely do not
           | want! Occasionally I do enjoy meandering around, but usually
           | fastest / shortest path would be preferred.
           | 
           | And I'm not sure about the other either. In my 20+ year
           | career in aerospace software, the most memorable times were
           | solving interesting problems, not days with no struggle just
           | churning out code.
        
             | thinkingemote wrote:
             | Indeed it is removing the memorable events of achievement!
             | 
             | Generally memorable things are different than unmemorable
             | things. Work is unmemorable. Driving is unmemorable except
             | when something negative happens. Waze tries to give some
             | positive feelings to the driving route. Waze knows that
             | people want positive experiences sometimes more than
             | efficiency.
             | 
             | Being stuck in a traffic jam is more memorable than not
             | being so. Or we remember the negative feeling more than the
             | fact that our drive actually wasn't inefficient.
             | 
             | AI tools makes us have a less negative day of work. so we
             | feel like we have no traffic jams. "I got so much done"
             | really means "I didn't get stuck". But it's also removing
             | the positive feelings too!
             | 
             | It's an illusion of progress through our feelings and
             | memories.
             | 
             | Or programming with AI brings different feedback mechanisms
             | and systems and different emotional engagements and
             | different memory behaviours. It's very interesting!
        
         | wrsh07 wrote:
         | Something I noticed early on when using AI tools was that it
         | was great because I didn't get blocked. Somehow, I always
         | wanted to keep going and always felt like I could keep going.
         | 
         | The problem, of course, is that one might thoughtlessly invoke
         | the ai tool when it would be faster to make the one line change
         | directly
         | 
         | Edit
         | 
         | This could make sense with the driving analogy. If the road I
         | was planning to take is closed, gps will happily tell me to try
         | something else. But if that fails too, it might go back to the
         | original suggestion.
        
         | Alex_L_Wood wrote:
         | We all as humans are hardwired to prefer greedy algorithms,
         | basically.
        
         | PicassoCTs wrote:
         | I also think that AI written code- is just not read. People
         | hate code-reviews, and actively refuse to read code- because
         | that is hard work, reading into other peoples thoughts and
         | ideas.
         | 
         | This is why pushing for new code, rewrites, new frameworks is
         | so popular. https://www.joelonsoftware.com/2000/04/06/things-
         | you-should-...
         | 
         | So a ton of ai-generated code- is just that, never read. Its
         | generated, tested against test-functions - and thats it. I
         | wouldn't wonder, if some of these devs themselves have only
         | marginal ideas whats in there codebases and why.
        
           | tjr wrote:
           | I have mostly worked in aerospace software, and find this
           | rather horrifying. I suppose, if your tests are in fact good
           | and comprehensive enough, there could be a logical argument
           | for not needing to understand the code, but if we're talking
           | people's safety in the hands of your software, I don't know
           | if there is any number of tests I would accept in exchange
           | for willingly giving up understanding of the code.
        
             | asadotzler wrote:
             | You're transferring the need to be really good at coding
             | and understanding code to the need to be really good at
             | testing and understanding tests, which 9/10 times requires
             | being good at coding and understanding code. There are no
             | free lunches.
        
         | jiggawatts wrote:
         | _> The inability of developers to tell if a tool sped them up
         | or slowed them down is fascinating in itself_
         | 
         | Linux/UNIX users are convinced of the superiority of keyboard
         | control and CLI tools, but studies have shown that the mouse is
         | faster for almost all common tasks.
         | 
         | Keyboard input _feels_ faster because there are more actions
         | per second.
        
           | mhuffman wrote:
           | >but studies have shown that the mouse is faster for almost
           | all common tasks.
           | 
           | Do you think that daily CLI Linux/UNIX users might have a
           | different list of what they consider "common tasks"?
        
       | i_love_retros wrote:
       | The AI hype will die off just like block chain and web3. LLMs are
       | a solution in search of a problem.
       | 
       | All the VCs are gonna lose a ton of money! OpenAI will be
       | NopenAI, relegated to the dustbin of history.
       | 
       | We never asked for this, nobody wants it.
       | 
       | Companies using AI and promoting it in their products will be
       | seen as tacky and cheap. Just like developers and artists that
       | use it.
        
       | diamond559 wrote:
       | Measure twice cut once. Not cut 100 times and hope it does it
       | right once.
        
       | trey-jones wrote:
       | Doing my own post-mortem of a recent project (the first that I've
       | leaned on "AI" tools to any extent), my feeling was the
       | following:
       | 
       | 1. It did not make me faster. I don't know that I expected it to.
       | 
       | 2. It's very possible that it made me slower.
       | 
       | 3. The quality of my work was better.
       | 
       | Slower and better are related here, because I used these tools
       | more to either check ideas that I had for soundness, or to get
       | some fresh ideas if I didn't have a good one. In many cases the
       | workflow would be: "I don't like that idea, what else do you have
       | for me?"
       | 
       | There were also instances of being led by my tools into a rabbit
       | hole that I eventually just abandoned, so that also contributes
       | to the slowness. This might happen in instances where I'm using
       | "AI" to help cover areas that I'm less of an expert in (and these
       | were great learning experiences). In my areas of expertise, it
       | was much more likely that I would refine my ideas, or the "AI"
       | tool's ideas into something that I was ultimately very pleased
       | with, hence the improved quality.
       | 
       | Now, some people might think that speed is the only metric that
       | matters, and certainly it's harder to quantify quality - but it
       | definitely felt worth it to me.
        
         | jpc0 wrote:
         | I do this a lot and absolutely think it might even improve it,
         | and this is why I like the current crop of AIs that are more
         | likely to be argumentative and not just capitulate.
         | 
         | I will ask the AI for an idea and then start blowing holes in
         | its idea, or will ask it to do the same for my idea.
         | 
         | And I might end up not going with it's idea regardless but it
         | got me thinking about things I wouldn't have thought about.
         | 
         | Effectively its like chatting to a coworker that has a
         | reasonable idea about the domain and can bounce ideas around.
        
           | trey-jones wrote:
           | I'm on record saying it's "like the smartest coworker I've
           | ever had" (no offense).
        
       | remorses wrote:
       | Using AI agents productively requires setting up a repository for
       | collaboration, it means writing docs and making the build process
       | easy and fast.
       | 
       | As any other tool AI is slow to adopt but has huge gains later on
        
       | stevekrouse wrote:
       | Such a great essay! Peter Naur's thesis is also the central point
       | in my talk about vibe coding from last month:
       | https://www.youtube.com/watch?v=1WC8dxMC4Xw
       | 
       | I'm spending an inordinate amount of time turning that video into
       | an essay, but I feel like I'm being scooped already, so here's my
       | current draft in case anyone wants to get a sneak preview:
       | https://valdottown--89ed76076a6544019f981f7d4397d736.web.val...
       | 
       | Feedback appreciated :)
        
       ___________________________________________________________________
       (page generated 2025-07-14 23:01 UTC)