lol, i hope the author of paint.net doesn't read this. incredibly + gratuitously insulting. one of the problems with llms is their poor or missing "theory of mind" for their users, so it's ironic (morissette sense) to see the blogger here similarly incapable of conceptualising the other & hallucinating a stereotype to replace them.
When people get older, they feel that they deserve upward mobility where eventually they become a manager who tells junior coders to do all the hard work. "AI Agents" allow for that fantasy, they're the tireless junior coders who you get to tell what to do, as if you've been promoted.
Is this a real thing people admit wanting? I am sure there are some people with a power fantasy like this but I can't imagine an LLM agent fulfilling this fantasy for them as it's generally about having authority over people who otherwise have agency unlike LLM agents.
The other two reasons seem right to me, and are enough on their own.
It has been a sad trend of the past years where some people like to compare LLMs to eager junior developers. Whenever I read that I can't help but think that the person has either lack of experience being a mentor/manager, lack of empathy, lack of authority, or all of that together. The reasons for the belittling can be motivated by a power fantasy but in general I think it's motivated by the insecurity that LLMs has brought to our professions.
think it’s impossible to underestimate the temptation that many [especially older] programmers seem to have for wanting a solution that will make their lives “easier” and fill a hole in their heart.
..
Wanting to spend less time at work, more time outside / with loved ones.
I feel like programming makes me feel younger, helps me remember my younger years.
I enjoy opening up a Turbo C++-like color-themed editor (emacs for me) and coding with a blue background/yellow font in a type-safe language, where I track in my head (and with destructors) how resources are used, I feel I understand how things are working between my stuff and down to processes, threads, interrupt handlers -- and with compiler magic dust helping with register allocations, optimized instructions for fetching data, etc.
When coding, I imagine I am in a fairy-tale story of sorts, where various participants have fairy-tale roles. Exceptions are the '666' and '667' and similar-numbered :-) ) in my code. I have recovery knights, defenders (constraints and guard), and so on. That world is something I enjoy spending part of life in.
I do not enjoy YAML configs, sorting out security-forced configurations, CSS, HTML, and type-less stuff (I feel like I'm on thin ice when I have to use those things).
But other than that -- coding is something I enjoy, because it makes me feel good, comfortable, and calm. I feel that I am interfacing with things I understand and that understand me, in a way :-)
Instead I wish we could accept the limitations of our time, our bodies, our minds.
no (with or without LLMs). 99.9999% of the human experience since the dawn of time is to try and go past these limitations.
Does this program NEED to have these extra features? Does this software NEED to work on everything?
yes
On your death bed, nobody is going to care about how many features you pushed on git.
definitely, but in that moment I sure as hell hope I remember all the smiles of the people that have used the features I've added for them to make their artwork / performance / ... and of the audience that saw said shows, art installations, etc.
If there is just one child who only has access to some older machine running Linux that is happy because they made a fun painting with paint.net now available there thanks to the author's work, it validates all the LLM usage in the world for me
Right? In terms of total resource usage, this seems like an incredibly lavishly roundabout way of helping other people. (I wonder why this hypothetical child's family can't afford to upgrade their computer.)
Do you know about all the negative externalities of AI? Does that really balance?
it's up to each to decide ? for me it absolutely does, i wouldn't work in the show industry otherwise, which only exists to convert resources into leisure
I think a lot of people think their private list of benefits or externalities is the whole picture. Rather the picture is unclear and likely to change.
I agree that using an open weight model with an open source harness through a datacenter that uses only green energy would significantly alter the discussion. Where can I sign up? (Honest question; open routers doesn't seem to have a "green energy pls" button :/ )
I don't know about online options, not off the top of my head.
Doing it yourself is still expensive, and likely to remain so until the next RAM price crash.
But basically:
Install solar panels (or convince your government/utility to use renewable power).
Try to secure at least 48GB of fast VRAM, or 196GB of high bandwidth system RAM. This is the expensive part, but a pair of AI Pro R9700s are a decent deal this week, at $3500 for both, giving you 64GB. Or get a Mac Studio.
Install llama-server, VLLM, or some other inference software of your choice.
If using VRAM, run Qwen3.8 27B. If using fast "unified" system RAM, run DeepSeek V4 Flash 0731 or (for non-commercial/internal use) Qwen3.8 Flash Next. If you have 512GB of RAM or more, GLM 5.3 Flash is also an excellent choice.
You can use smaller models, or squeeze Qwen3.8 27B into 24-32GB of VRAM. But you'll start making compromises on speed and quality. If you go with specs and models I suggest above, you'll wind up with something about like Opus 4.5-4.6 running at home at tolerable speeds. If you add some additional RAM for context cache, it's quite feasible to share a single server with coworkers, or to set up some kind of co-op.
Environmental costs on these setups are 300-600W when actively running (so less than an electric stove), plus (for training) about a small town's worth of electricity once or twice a year. Which is actually pretty reasonable in the big scheme of industrial civilization.
As for what you'll get? It won't be Fable. It won't allow you to just stop using your brain entirely and to imitate a mollusc while 20 agents grind away on your behalf. But if you want to drive the process, and use the agent to do the boring parts, to debug the strange parts, or to act as a soundjng board, that is entirely possible now.
I mean, it's possible now if you bought hardware before the RAMpocalypse or if you want to spend money on an expensive hobby. It's still far cheaper than boats, horses, motorcycles and project cars. But it won't be something every college kid can afford until after the next price crash.
If you just want to get some cheap tokens while minimizing your energy footprint, try DeepSeek V4 Flash 0731, which is heavily optimized for super cheap serving. Which necessarily means keeping the energy use down, too. And pick a company with data centers in a country with a lot of solar or wind.
I do think it would be neat to set up a non-profit that provided "green energy" tokens to open source authors at cost, using all free software and open weights.
Not 100% what you want, but I use LLMs hosted by Infomaniak in Switzerland for some minor tasks. I like the company because they sponsored a Debian conference, but also, they use 100% green energy, atmospheric-cooled servers, offset 200% of their carbon emissions, send excess heat into the municipal network, etc. It's all quite nice. I don't use it for coding, though.
For what it's worth, I've mostly experienced LLM-assisted software in the form of: uncanny interfaces/interaction models that work, but don't really make sense or feel cohesive at all if you think about it too much; simultaneously verbose and vague UI text; meandering blog posts written in unnecessarily dramatic language and devoid of personality; mountains of complex code to do things which could've been achieved in a much simpler idiomatic way; an explosion of distracting features which have a niche use cases (if any); frequent gratuitous tweaks that don't improve anything to justify the cost of adapting to the change; and piles of project spam posts which have completely drained the enthusiasm I used to have for discovering and hearing about other people's projects; as well as the loss of admiration for more than half of programmers I used to respect for the care they used to put in their projects.
The upsides have been... two or three utilities that are still useful and tolerable despite having all those flaws.
One feeling that keeps popping up is that it feels like the result of someone on a stimulant binge. Which makes sense to me, without the sense of friction for applying effort, it's easy to do a lot without really considering if it's a good idea in the first place.
"an internal, from-scratch, clean-room reverse-engineered rewrite of Direct2D":
Direct2D has always been the biggest hurdle for Paint.NET on WINE, and it's clear that it will never be completed enough for Paint.NET's use. And I can't just "disable" the use of Direct2D. So, instead, Paint.NET now has an internal, from-scratch, clean-room reverse-engineered rewrite of Direct2D that it uses on WINE (triggered by using /wine). It lives in PaintDotNet.Windows.Direct2D1.Managed.dll. This was written by our good friend Claude, without whom this would NOT have been possible and would NEVER have happened.
For it to be truly "internal, from-scratch, clean room", there'd need to be some proof that the original sources (and all pre/successors of it) are not in Claude's training data. Personally, I think that's unlikely.
Same way other illegal stuff ends up in the training data. But that's besides the point; only Anthropic can (theoretically) know. That's not great legally speaking.
Practically though, Microsoft is the copyright holder on the Direct2D codebase, and Microsoft's stated position is that (generally speaking) the code output of AI models is not tainted by the license of the training data, and they will defend this position for you in court.
Attempting legal action against Paint.NET would imply basically destroying the new line of business that they seem to be betting everything on. We can debate here whether it is actually a from-scratch, clean-room, implementation, but the copyright holder's stated position is that it is.
I'm not sure about this diagnosis that "older" programmers are vulnerable to LLM capture. But certainly "senior" in terms of management hierarchy do seem more vulnerable to the siren call.
Wanting to spend less time at work, more time outside / with loved ones.
If only. I have a friend who is a writer/editor at a web marketing firm who is forced to use LLMs as part of his job. It used to be that he had a full day to write a 2,000 word article, but now? It's expected for him to write three to four a day because LLMs make the job so much faster! And he hates it, because it's still a ton of work to ensure the LLM doesn't confabulate (most of his clients are lawyers) with the now added stress of having only two hours for a 2,000 word article. I'm sure he would love to spend half a day on a 2,000 word article (half the time as before) and get more time to himself, but his employers are the ones gaining the benefit of the LLMs (fewer writers, more articles written), not my friend.
This post is inspired by the creator of Paint.NET announcing that he has added WINE/Linux support to his art program by using Claude to code it, leading users to have to ditch the program and look for a new art program to use.
No one has to ditch the program. They might choose to, but to me this seems to be more related to personal purity than any actual outcome.
There is a tinkling of an interesting blog post here, but it falls short by not exploring anything introduced further and just stating things as they are. If you are going to write an post like this I do strongly feel that you should spend more time on things like this
From where I stand I feel like this sadly comes from certain desires and insecurities:
Wanting to spend less time at work, more time outside / with loved ones.
Feeling like you should be able to accomplish so much more than your body and mind allows with limited time on Earth.
When people get older, they feel that they deserve upward mobility where eventually they become a manager who tells junior coders to do all the hard work. "AI Agents" allow for that fantasy, they're the tireless junior coders who you get to tell what to do, as if you've been promoted.
I think the author is somewhat in the neighborhood with these points, but calls destination reached before even have reached the right street. To the point that you now have caricatures of reasons that fit what the author wants it to be rather than what is really going on.
The first on is the most reasonable one, which I wouldn't even qualify as a desire or insecurity. In the context of open source development I can understand it if the sole maintainer of something popular considers the aid of AI to balance the time invested a bit more evenly. If that result will be reached is a valid discussion to have, but the reasoning behind it is not something I would actually dispute.
The second point already starts to glide into the ridiculous, mostly be the addition of the last bit. And also here, I don't the desire to be able to do more is unreasonable in many ways. At least not as a standalone bullet point simplified to this point. Again, the valid discussion here is if LLMs can actually make people accomplish more. Extra points for making the discussion about what accomplishing something means if it is largely written by a LLM and if there is a tipping point somewhere.
The third bullet point is just plain ridiculous to me. Sure, I have seen agents described as "tireless juniors who do what I ask". But, that is always in the context the second bullet point and rarely if ever I have seen it in the context of some sort of "I deserve to be management" type of deal.
I've struggled with the decision of using LLMs in my own projects. Just to be upfront about it, I've made the decision of using them, and I use them very extensively for pretty much everything I build lately.
There's no high score, no reward, for endlessly expanding your software as far and fast as possible.
This is true. I agree that the ability to just more or less will some feature into existence makes scope creep and build stuff that nobody cares about extremely tempting.
Does this program NEED to have these extra features? Does this software NEED to work on everything?
For me, my thinking came down to something pretty simple.
I spend a LOT of time on my personal software projects, and I care about them very deeply. This has been the case for over a decade, and I was more than happy to hand-code things. I truly preferred the process of writing code by hand. I got very good at it, the fact that this skill was valuable to others and society in general was huge for my self-worth and livelihood.
Despite all that, I still can't justify not using AI to myself. I just picture the scenario where I decided to hand-code my projects, and then I log into Twitter and see a post from someone who built something very similar (or perhaps even better) with AI instead. Without a doubt, it's going to have more features, it's probably going to be more polished (on the surface at least), and it would have been built in an order of magnitude less time.
Even if that particular project was a shitty pile of vibe-coded slop with no soul or care put into it at all, it would still be crushing to witness. Thinking of seeing some kid with a small fraction of my experience and none of my deep interest and enthusiasm for the tech itself pump out something equivalent in 5% the time makes me sick to my stomach.
So that really only leaves me with a few options. I can despair and quit dumping effort into software projects and pursue my only other goal of getting to Platinum in Valorant. Or I can double down and use AI to accelerate my efforts and try to go far further and do more than I could without it.
To be clear, I've done plenty of despairing as well along the way. I still feel like the craft that used to define me is quite dead, and when I try to project forward to the logical conclusion of this AI revolution, I find it hard to see anything remotely positive. However, despite all that, I absolutely refuse to just sit here and watch others surpass me because I refuse to make use of tools at my disposal.
I just picture the scenario where I decided to hand-code my projects, and then I log into Twitter and see a post from someone who built something very similar (or perhaps even better) with AI instead. [...] Thinking of seeing some kid with a small fraction of my experience and none of my deep interest and enthusiasm for the tech itself pump out something equivalent in 5% the time makes me sick to my stomach.
We like to think that programming is hard because of the intellectual challenge, but what if that's a convenient lie we tell ourselves? What if the real reason is that, despite all the control computers give us they do nothing to help us control the fundamental un-programable nature of life itself? In other words, the real work of programming is finding a way to fully embrace the "sick to my stomach" feeling you mention. And I mean work, as in it requires actual sincere and sustained effort, which is often even harder than the conventional act of logically thinking through code. We like to think it's all about the logic of code, because at least then the difficulties are at a safe distance from our messy personal lives. But if all you have is the hammer of programming-is-just-about-code, then everything will look like the nail of always-solvable-with-code.
Whether it's a script kiddy, an overbearing boss, spiteful coworkers, an unhappy family life, mental health issues, etc, reality is always sitting next to us when we code. And as much as we'd like it to, it cannot be programmed away. So it's not a matter of finding ways to circumvent this side of making software, but rather a matter of nurturing a practice that predisposes us to noticing and welcoming the inevitable existential discomfort. It's not a sign of failure to have a personal life. Indeed it's the exact opposite. I would even suggest that there is a perspective from which it is impossible to ever waste your time programming. The very fact that we commit to developing something with patience, sincerity and grit, in the face of torrents of slop, is an unmistakable signal. A signal that we have found within ourselves something real and meaningful that speaks to other humans with their own personal and intellectual journeys. Such a signal guides both us as authors and those around us, colleagues, end users and so on.
Don't give up.
This AI thing, like all the other peaks and troughs in the history of computing, is just a changing of the guards. Right now all AI is doing is emulating the old guards that signalled their relevance through lines of code. The new guards express themselves with signals out of reach to LLMs; community, humanity, determination, passion and kindness. What if AI isn't the end of our craft, but inadvertently a return to its roots and a flowering of its original purpose: to truly be of value to humanity?
I feel the same way. I have spent years on some of my projects which at my current pace will take me years to complete. I could probably have AI finish my designs and most of my missing features in a day.
As a master CS student I recently saw a first-year student who has made astonishing software using AI. Entire reverse-engineering debuggers, game engines, application frameworks. Tools tailored exactly to their need, built in the languages they understand (C++).
For myself I think I just need to find a balance. My current way of programming will not hold.
I think I agree by the general sentiment shared on this article, but it's poorly argued and it chose a bad target to argue it against.
Wanting to spend less time at work, more time outside / with loved ones.
Wanting to spend less time working and more time with loved ones is such a slam dunk of a worthwhile pursuit. I can't begrudge anyone for wanting to reduce the amount of time they have to spend working instead of living.
Does this program NEED to have these extra features?
This is the same line of reasoning that is used to discard accessibility features in order to have "simpler" software. Software should be more accessible! If my options are between making unaccessible software and retaining a pure soul; or making accessible software and having to commune with the literal devil, I'm taking the satanic rituals every day of the week.
It's true that you don't have to endlessly expand your software forever, but there is a very real reward to making your software more accessible.
Can someone help me understand why everyone grabs their pitchforks as soon as somebody admits to using AI for code?
The original post by the PAINT.NET author was honest about their AI use. It's clear that without Claude this wine compatibility could've never been written, even if he devoted his life to it. Instead of he spent some time and presumably a lot of money on getting it through AI.
This post suggests that "users have to ditch the program and look for a new art program to use", I don't see why that's the case at all. The original 200.000 lines of code the author worked on for 20 years hasn't changed.
If you don't want to use this vibecoded Direct2D layer then just don't use it? Does it hurt to have it?
Some people hate the idea of LLMs more than the outcomes or externalities. So if someone admits they've ever interacted with a tool that uses an LLM then their humanity and work must be shunned.
There are shades of gray coming down from that all the way to neutral of course, I don't mean to paint every "anti AI" as this extreme, but it exists for sure and I think it's what were seeing here.
These people really think they can dictate what tools open source maintainers are allowed to use. The sheer amount of thankless work that goes into maintaining project that people put in is incredible. The author's argument carries an air of entitlement that makes it difficult to take seriously.
gulbanana | a day ago
lol, i hope the author of paint.net doesn't read this. incredibly + gratuitously insulting. one of the problems with llms is their poor or missing "theory of mind" for their users, so it's ironic (morissette sense) to see the blogger here similarly incapable of conceptualising the other & hallucinating a stereotype to replace them.
cpurdy | a day ago
Yeah ... that entire blog article was simultaneously creepy and insulting.
pkd | a day ago
Is this a real thing people admit wanting? I am sure there are some people with a power fantasy like this but I can't imagine an LLM agent fulfilling this fantasy for them as it's generally about having authority over people who otherwise have agency unlike LLM agents.
The other two reasons seem right to me, and are enough on their own.
jrgtt | 23 hours ago
It has been a sad trend of the past years where some people like to compare LLMs to eager junior developers. Whenever I read that I can't help but think that the person has either lack of experience being a mentor/manager, lack of empathy, lack of authority, or all of that together. The reasons for the belittling can be motivated by a power fantasy but in general I think it's motivated by the insecurity that LLMs has brought to our professions.
vladislavp | a day ago
..
I feel like programming makes me feel younger, helps me remember my younger years.
I enjoy opening up a Turbo C++-like color-themed editor (emacs for me) and coding with a blue background/yellow font in a type-safe language, where I track in my head (and with destructors) how resources are used, I feel I understand how things are working between my stuff and down to processes, threads, interrupt handlers -- and with compiler magic dust helping with register allocations, optimized instructions for fetching data, etc.
When coding, I imagine I am in a fairy-tale story of sorts, where various participants have fairy-tale roles. Exceptions are the '666' and '667' and similar-numbered :-) ) in my code. I have recovery knights, defenders (constraints and guard), and so on. That world is something I enjoy spending part of life in.
I do not enjoy YAML configs, sorting out security-forced configurations, CSS, HTML, and type-less stuff (I feel like I'm on thin ice when I have to use those things).
But other than that -- coding is something I enjoy, because it makes me feel good, comfortable, and calm. I feel that I am interfacing with things I understand and that understand me, in a way :-)
Would be sad if LLMs take that away from me.
yuriks | 14 hours ago
jcelerier | a day ago
no (with or without LLMs). 99.9999% of the human experience since the dawn of time is to try and go past these limitations.
yes
definitely, but in that moment I sure as hell hope I remember all the smiles of the people that have used the features I've added for them to make their artwork / performance / ... and of the audience that saw said shows, art installations, etc.
If there is just one child who only has access to some older machine running Linux that is happy because they made a fun painting with paint.net now available there thanks to the author's work, it validates all the LLM usage in the world for me
rbr | 23 hours ago
Do you know about all the negative externalities of AI? Does that really balance?
tentacloids | 20 hours ago
Right? In terms of total resource usage, this seems like an incredibly lavishly roundabout way of helping other people. (I wonder why this hypothetical child's family can't afford to upgrade their computer.)
jcelerier | 8 hours ago
it's up to each to decide ? for me it absolutely does, i wouldn't work in the show industry otherwise, which only exists to convert resources into leisure
singpolyma | 16 hours ago
I think a lot of people think their private list of benefits or externalities is the whole picture. Rather the picture is unclear and likely to change.
rbr | 14 hours ago
I agree that using an open weight model with an open source harness through a datacenter that uses only green energy would significantly alter the discussion. Where can I sign up? (Honest question; open routers doesn't seem to have a "green energy pls" button :/ )
emk | 13 hours ago
I don't know about online options, not off the top of my head.
Doing it yourself is still expensive, and likely to remain so until the next RAM price crash.
But basically:
You can use smaller models, or squeeze Qwen3.8 27B into 24-32GB of VRAM. But you'll start making compromises on speed and quality. If you go with specs and models I suggest above, you'll wind up with something about like Opus 4.5-4.6 running at home at tolerable speeds. If you add some additional RAM for context cache, it's quite feasible to share a single server with coworkers, or to set up some kind of co-op.
Environmental costs on these setups are 300-600W when actively running (so less than an electric stove), plus (for training) about a small town's worth of electricity once or twice a year. Which is actually pretty reasonable in the big scheme of industrial civilization.
As for what you'll get? It won't be Fable. It won't allow you to just stop using your brain entirely and to imitate a mollusc while 20 agents grind away on your behalf. But if you want to drive the process, and use the agent to do the boring parts, to debug the strange parts, or to act as a soundjng board, that is entirely possible now.
I mean, it's possible now if you bought hardware before the RAMpocalypse or if you want to spend money on an expensive hobby. It's still far cheaper than boats, horses, motorcycles and project cars. But it won't be something every college kid can afford until after the next price crash.
If you just want to get some cheap tokens while minimizing your energy footprint, try DeepSeek V4 Flash 0731, which is heavily optimized for super cheap serving. Which necessarily means keeping the energy use down, too. And pick a company with data centers in a country with a lot of solar or wind.
I do think it would be neat to set up a non-profit that provided "green energy" tokens to open source authors at cost, using all free software and open weights.
rkaw92 | 11 hours ago
Not 100% what you want, but I use LLMs hosted by Infomaniak in Switzerland for some minor tasks. I like the company because they sponsored a Debian conference, but also, they use 100% green energy, atmospheric-cooled servers, offset 200% of their carbon emissions, send excess heat into the municipal network, etc. It's all quite nice. I don't use it for coding, though.
singpolyma | 14 hours ago
You can put solar panels on your roof and do it in your basement :) there's no reason most models need to be run in a data centre at all.
I'd for sure like to see what you're saying as well though, for any cloud computing not just "AI".
matthewcroughan | a day ago
“It is not the man who has too little, but the man who craves more, that is poor. ”
yuriks | 14 hours ago
For what it's worth, I've mostly experienced LLM-assisted software in the form of: uncanny interfaces/interaction models that work, but don't really make sense or feel cohesive at all if you think about it too much; simultaneously verbose and vague UI text; meandering blog posts written in unnecessarily dramatic language and devoid of personality; mountains of complex code to do things which could've been achieved in a much simpler idiomatic way; an explosion of distracting features which have a niche use cases (if any); frequent gratuitous tweaks that don't improve anything to justify the cost of adapting to the change; and piles of project spam posts which have completely drained the enthusiasm I used to have for discovering and hearing about other people's projects; as well as the loss of admiration for more than half of programmers I used to respect for the care they used to put in their projects.
The upsides have been... two or three utilities that are still useful and tolerable despite having all those flaws.
One feeling that keeps popping up is that it feels like the result of someone on a stimulant binge. Which makes sense to me, without the sense of friction for applying effort, it's easy to do a lot without really considering if it's a good idea in the first place.
rbr | 23 hours ago
Post of Paint.NET author announcing the change' https://forums.paint.net/topic/134563-%F0%9F%8D%B7-extremely-experimental-winelinux-support-how-to-get-started/. He has some words to say about it.
simonw | 23 hours ago
"an internal, from-scratch, clean-room reverse-engineered rewrite of Direct2D":
rbr | 22 hours ago
For it to be truly "internal, from-scratch, clean room", there'd need to be some proof that the original sources (and all pre/successors of it) are not in Claude's training data. Personally, I think that's unlikely.
simonw | 15 hours ago
Direct2D is closed source though, right? How could that end up in the training data?
rbr | 14 hours ago
Same way other illegal stuff ends up in the training data. But that's besides the point; only Anthropic can (theoretically) know. That's not great legally speaking.
strugee | 8 hours ago
Practically though, Microsoft is the copyright holder on the Direct2D codebase, and Microsoft's stated position is that (generally speaking) the code output of AI models is not tainted by the license of the training data, and they will defend this position for you in court.
Attempting legal action against Paint.NET would imply basically destroying the new line of business that they seem to be betting everything on. We can debate here whether it is actually a from-scratch, clean-room, implementation, but the copyright holder's stated position is that it is.
IANAL.
marginalia | 15 hours ago
As much as I dislike vibecoding and AI slop, this is just incredibly entitled Eric Cartman-ass behavior.
draxil | 18 hours ago
I'm not sure about this diagnosis that "older" programmers are vulnerable to LLM capture. But certainly "senior" in terms of management hierarchy do seem more vulnerable to the siren call.
spc476 | 10 hours ago
If only. I have a friend who is a writer/editor at a web marketing firm who is forced to use LLMs as part of his job. It used to be that he had a full day to write a 2,000 word article, but now? It's expected for him to write three to four a day because LLMs make the job so much faster! And he hates it, because it's still a ton of work to ensure the LLM doesn't confabulate (most of his clients are lawyers) with the now added stress of having only two hours for a 2,000 word article. I'm sure he would love to spend half a day on a 2,000 word article (half the time as before) and get more time to himself, but his employers are the ones gaining the benefit of the LLMs (fewer writers, more articles written), not my friend.
tao_oat | 16 hours ago
No one has to ditch the program. They might choose to, but to me this seems to be more related to personal purity than any actual outcome.
creesch | 22 hours ago
There is a tinkling of an interesting blog post here, but it falls short by not exploring anything introduced further and just stating things as they are. If you are going to write an post like this I do strongly feel that you should spend more time on things like this
I think the author is somewhat in the neighborhood with these points, but calls destination reached before even have reached the right street. To the point that you now have caricatures of reasons that fit what the author wants it to be rather than what is really going on.
The first on is the most reasonable one, which I wouldn't even qualify as a desire or insecurity. In the context of open source development I can understand it if the sole maintainer of something popular considers the aid of AI to balance the time invested a bit more evenly. If that result will be reached is a valid discussion to have, but the reasoning behind it is not something I would actually dispute.
The second point already starts to glide into the ridiculous, mostly be the addition of the last bit. And also here, I don't the desire to be able to do more is unreasonable in many ways. At least not as a standalone bullet point simplified to this point. Again, the valid discussion here is if LLMs can actually make people accomplish more. Extra points for making the discussion about what accomplishing something means if it is largely written by a LLM and if there is a tipping point somewhere.
The third bullet point is just plain ridiculous to me. Sure, I have seen agents described as "tireless juniors who do what I ask". But, that is always in the context the second bullet point and rarely if ever I have seen it in the context of some sort of "I deserve to be management" type of deal.
Ameo | 22 hours ago
I've struggled with the decision of using LLMs in my own projects. Just to be upfront about it, I've made the decision of using them, and I use them very extensively for pretty much everything I build lately.
This is true. I agree that the ability to just more or less will some feature into existence makes scope creep and build stuff that nobody cares about extremely tempting.
For me, my thinking came down to something pretty simple.
I spend a LOT of time on my personal software projects, and I care about them very deeply. This has been the case for over a decade, and I was more than happy to hand-code things. I truly preferred the process of writing code by hand. I got very good at it, the fact that this skill was valuable to others and society in general was huge for my self-worth and livelihood.
Despite all that, I still can't justify not using AI to myself. I just picture the scenario where I decided to hand-code my projects, and then I log into Twitter and see a post from someone who built something very similar (or perhaps even better) with AI instead. Without a doubt, it's going to have more features, it's probably going to be more polished (on the surface at least), and it would have been built in an order of magnitude less time.
Even if that particular project was a shitty pile of vibe-coded slop with no soul or care put into it at all, it would still be crushing to witness. Thinking of seeing some kid with a small fraction of my experience and none of my deep interest and enthusiasm for the tech itself pump out something equivalent in 5% the time makes me sick to my stomach.
So that really only leaves me with a few options. I can despair and quit dumping effort into software projects and pursue my only other goal of getting to Platinum in Valorant. Or I can double down and use AI to accelerate my efforts and try to go far further and do more than I could without it.
To be clear, I've done plenty of despairing as well along the way. I still feel like the craft that used to define me is quite dead, and when I try to project forward to the logical conclusion of this AI revolution, I find it hard to see anything remotely positive. However, despite all that, I absolutely refuse to just sit here and watch others surpass me because I refuse to make use of tools at my disposal.
tombh | 21 hours ago
We like to think that programming is hard because of the intellectual challenge, but what if that's a convenient lie we tell ourselves? What if the real reason is that, despite all the control computers give us they do nothing to help us control the fundamental un-programable nature of life itself? In other words, the real work of programming is finding a way to fully embrace the "sick to my stomach" feeling you mention. And I mean work, as in it requires actual sincere and sustained effort, which is often even harder than the conventional act of logically thinking through code. We like to think it's all about the logic of code, because at least then the difficulties are at a safe distance from our messy personal lives. But if all you have is the hammer of programming-is-just-about-code, then everything will look like the nail of always-solvable-with-code.
Whether it's a script kiddy, an overbearing boss, spiteful coworkers, an unhappy family life, mental health issues, etc, reality is always sitting next to us when we code. And as much as we'd like it to, it cannot be programmed away. So it's not a matter of finding ways to circumvent this side of making software, but rather a matter of nurturing a practice that predisposes us to noticing and welcoming the inevitable existential discomfort. It's not a sign of failure to have a personal life. Indeed it's the exact opposite. I would even suggest that there is a perspective from which it is impossible to ever waste your time programming. The very fact that we commit to developing something with patience, sincerity and grit, in the face of torrents of slop, is an unmistakable signal. A signal that we have found within ourselves something real and meaningful that speaks to other humans with their own personal and intellectual journeys. Such a signal guides both us as authors and those around us, colleagues, end users and so on.
Don't give up.
This AI thing, like all the other peaks and troughs in the history of computing, is just a changing of the guards. Right now all AI is doing is emulating the old guards that signalled their relevance through lines of code. The new guards express themselves with signals out of reach to LLMs; community, humanity, determination, passion and kindness. What if AI isn't the end of our craft, but inadvertently a return to its roots and a flowering of its original purpose: to truly be of value to humanity?
Sanity | 9 hours ago
Thank you, that was beautiful. The cynic in me isn't quite convinced of the last part, but this helps :)
Levitating | 18 hours ago
I feel the same way. I have spent years on some of my projects which at my current pace will take me years to complete. I could probably have AI finish my designs and most of my missing features in a day.
As a master CS student I recently saw a first-year student who has made astonishing software using AI. Entire reverse-engineering debuggers, game engines, application frameworks. Tools tailored exactly to their need, built in the languages they understand (C++).
For myself I think I just need to find a balance. My current way of programming will not hold.
annieversary | 15 hours ago
I think I agree by the general sentiment shared on this article, but it's poorly argued and it chose a bad target to argue it against.
Wanting to spend less time working and more time with loved ones is such a slam dunk of a worthwhile pursuit. I can't begrudge anyone for wanting to reduce the amount of time they have to spend working instead of living.
This is the same line of reasoning that is used to discard accessibility features in order to have "simpler" software. Software should be more accessible! If my options are between making unaccessible software and retaining a pure soul; or making accessible software and having to commune with the literal devil, I'm taking the satanic rituals every day of the week.
It's true that you don't have to endlessly expand your software forever, but there is a very real reward to making your software more accessible.
Levitating | 18 hours ago
Can someone help me understand why everyone grabs their pitchforks as soon as somebody admits to using AI for code?
The original post by the PAINT.NET author was honest about their AI use. It's clear that without Claude this wine compatibility could've never been written, even if he devoted his life to it. Instead of he spent some time and presumably a lot of money on getting it through AI.
This post suggests that "users have to ditch the program and look for a new art program to use", I don't see why that's the case at all. The original 200.000 lines of code the author worked on for 20 years hasn't changed.
If you don't want to use this vibecoded Direct2D layer then just don't use it? Does it hurt to have it?
emk | 14 hours ago
singpolyma | 15 hours ago
Some people hate the idea of LLMs more than the outcomes or externalities. So if someone admits they've ever interacted with a tool that uses an LLM then their humanity and work must be shunned.
There are shades of gray coming down from that all the way to neutral of course, I don't mean to paint every "anti AI" as this extreme, but it exists for sure and I think it's what were seeing here.
Yogthos | 15 hours ago
These people really think they can dictate what tools open source maintainers are allowed to use. The sheer amount of thankless work that goes into maintaining project that people put in is incredible. The author's argument carries an air of entitlement that makes it difficult to take seriously.
ngrilly | 13 hours ago
Counter-argument: I'm seeing a lot of young people who are all-in on AI and very much willing to spend a lot of time at work.