> I think it’s more important that we stop collectively pretending that we don’t understand how awful all of this is
Lord forbid if people disagree with you. I know Drew's vibe is always "I'm right because I'm the only one with the correct opinions", but it does get tiring after a while.
Not to say AI isn't having huge drawbacks being introduced, and aren't exactly worry-free, but why not change your frame of mind from "Why don't others understand how awful it is?!" to "People are seeing something I'm not, what am I missing?" so your article could actually contain something else than personal and emotions rants?
> "People are seeing something I'm not, what am I missing?"
I've seen people celebrate horrors beyond my comprehension. Cheer the deaths of innocent people because it may inch some abstract national goal closer to a similarly abstract measurement. Insist that lives in one place are worth less than lives in another.
Should I ask "what am I missing"?
I don't think so, sometimes you draw a line on moral or ethical grounds. Some of those lines should never be given the ability to be fluid. It will always be wrong to bomb a school of children, just like (for Drew and I) it will be wrong to rip the livelyhood from under millions of people's feet for shareholder value. It will be wrong to ignore damaging consequences to the environment. It will be wrong to insist a low quality imitation should ever hold the same value as the original idea.
> rip the livelyhood from under millions of people's feet
I have never gotten this. How is livelihood being "ripped away"? There is enormous capability made available to anyone and everyone who wants to take hold of and do something with it. Just as it's on each individual to go through the process and pains of landing a job (or building a business, etc), it's also on each individual to keep up with changes that may affect their livelihood. If they want to keep it.
I think the point is that regardless of what benefits LLMs are bringing to the table, there are a list of downsides that Drew views as non-negotiables. It doesn't matter what other people are seeing, because he sees a fundamental issue underlying the entire premise.
It does seem like most people completely ignore the obvious harms caused by AI when talking about using LLMs for programming, as though somehow it is disconnected from the other deployments of the technology.
> It does seem like most people completely ignore the obvious harms caused by AI when talking about using LLMs for programming, as though somehow it is disconnected from the other deployments of the technology.
I feel that the people who are completely ignoring the harms are the ones who need and/or benefit from it and do whatever it takes to justify their use of it. The rest are people who understand the harms and minimize interaction followed by the blissfully ignorant.
I was just talking to a content creator who uses AI at work social media platforms to display her personal projects. She talked about how she is fully aware of the harm social media platforms bring while acknowledging they empower her to present her work to the world without gatekeeping. AI allows her to power through boring office tasks but she loathes their use in the art world and replacing people in general.
> It does seem like most people completely ignore the obvious harms caused by AI when talking about using LLMs for programming, as though somehow it is disconnected from the other deployments of the technology.
I would insist that the deployments of a technology should be disconnected from the technology itself - I criticize AI too, and I get a lot of downvotes for it, but I try to separate the science of AI from its economics and politics.
The harms of AI and other technologies come from two sources 1. Capital destroying market bubbles and 2. Deployments motivated and enabled by political and moral corruption.
Both of these are in turn enabled and sustained by legislation. That is, we have to talk politics, not technology and not AI. AI has a great potential - both for improving human life and for making it a lot worse and which way it goes depends entirely on politics.
If we fail to cleanly separate these issues and keep moralizing about technology, we will be chasing red herrings and bumping heads in the dark all the while the tech is being deployed against us.
> there are a list of downsides that Drew views as non-negotiables
Which is all fine and dandy. But why play the "You simply don't understand it as well as I do" rather than something more investigative and curious? Just fuels the whole "holier than thou" vibe Drew been trying to increase seemingly every day.
It's a disagreement of opinion, not some "I'm the only smart person who can realize this", which is why it kind of sours the entire piece.
> But why play the "You simply don't understand it as well as I do" rather than something more investigative and curious?
That's not the tone of the article; he uses the word "pretending". That tells me that he thinks that people do understand, but they don't want to admit that they understand because that would reveal their values.
> Which is all fine and dandy. But why play the "You simply don't understand it as well as I do"
I'll say this from the perspective of a person who publishes content online: because people's revealed preference is for content written this way. You can spend weeks polishing thoughtful, original content that will get few clicks, or you can crank out throwaway op-eds about AI and get thousands of likes and upvotes from people who just wanted to hear their own beliefs explained back to them.
My stuff appeared on HN a couple of times over the years and the less effort I put into it, the better it fared. The temptation to change your writing style and to offer increasingly more provocative and shallow opinions is difficult to resist.
My point is probably this: if you want to see better stuff, I think you gotta stop engaging with articles like this. Patrol /newest and upvote cool in-depth stuff.
In fairness he pretty explicitly states that he thinks people do understand it, but are pretending not to to wash their hands of the consequences. I'm definitely not reading it in the same way you are.
It's a difference in values, not understanding. I understand that AI burns tons of power, and I don't care. Drew understands it the same as I, but he does care. The difference is in what people value, and relative to what.
Right, a difference of opinion, which is fine, OK and even good. But why paint it as "Obviously the rest of you aren't smart enough to understand" instead of "Other's disagree", seems really strange (although in-character).
> I think it’s more important that we stop collectively pretending that we don’t understand how awful all of this is
Would be very different from say:
> I'd like to understand people who don't see it the same way as me, that it's mostly awful and not good.
Or similar, rather than "I'm right, everyone else don't understand it properly". Very HN-esque, but oh so tiring after 100s of articles in the exactly same vein from the same author.
This is too shallow of a take. Especially when your very next point objects to what he uses as a default reference frame that you disagree with. Lord forbid drew disagree about, I think priorities, and values?
> why not change your frame of mind from "Why don't others understand how awful it is?!" to "People are seeing something I'm not, what am I missing?"
It's the same question. I sympathize with both questions, I constant feel both frustrated, and broken with how few people care about quality, and participating fairly. I try very hard to find the positive aspects "everyone" claims llm codegen provides. I'm looking hard, and can't find them. It's painfully average, often worse so when it gets lost. It doesn't and can not help me, only get in the way, what am I doing wrong? Why is everyone missing something I see as obvious? But again, both could easily be true from both frames you suggest. "Why can't people identify this as trash" could very easily be followed by "what I'm I missing from the equation?" and be the same thought/idea.
> so your article could actually contain something else than personal and emotions rants?
I mean, it's titled, A Eulogy for Vim. That seems to be what it says on the tin, no?
> And at a moment when the climate demands immediate action to reduce our footprint on this planet, the AI boom is driving data centers to consume a full 1.5% of the world’s total energy production in order to eliminate the jobs of the poor and replace them with a robot that lies.
For an environmentmaxxer, eliminating upper-middle class jobs is extremely effective, as this group consumes the lion's share of resources and bears the greatest impact on carbon emissions. Remember that the majority of industry is upstream of consumption.
Not endorsing this world view, just noting that the wealthiest 1% of people in the world (encompasses most US citizens) have an enormously outsized impact on climate.
The "upper middle class" is not strictly defined, but they are pretty clearly the folks below the wealthiest 1%. You can't be in the middle without something on either side.
They certain consume far more than the poor, on account of having resources, but they also consume far less than the wealthiest 1%.
You are almost certainly in the global 1% of the wealthy [1]. Compared to developing nations' residents, we have an order of magnitude greater impact on the environment.
>Remember that the majority of industry is upstream of consumption.
People forget this. Oil companies may have dug up the oil, but they did so because we paid them to, so we could use the energy for good and useful things.
Climate change isn't 'evil billionaire companies are ruining the world', it's 'these things we did to improve our lives turn out to have side effects'.
This is backwards. If it weren't for 'eliminating jobs' we'd both be peasant farmers right now. Automation has improved the standard of living and raised wages for everyone, rich and poor alike.
> The maturity of Vim9 script's modern constructs is now being leveraged by advanced AI development tools. Contributor Yegappan Lakshmanan recently demonstrated the efficacy of these new features through two projects generated using GitHub Copilot
I am not sure I understand the author's concern, is he saying that VIM 9.2 is problematic because it enables AI integration due to the maturity of Vim9 script?
The tl;dr: Drew thinks Vim development has been tainted by LLM contributions, and is thus morally unsuitable to be used, and he will therefore be forking it.
I think I will use vim-classic and possibly contribute to it. Not because of AI, but because I actually want to use Vim over say something like Neovim* and I actually like vimscript, which imo didn't need the development of vim9script to improve it.
Regarding why not Neovim, I think it's because a large section of the community want to create more complex TUI elements or replicate GUI interfaces and make it more like VS Code. I use Vim for the "vim way" not because it's in a terminal or it's not bloated like some other editors.
One of the side effects of AI is definitely that a lot of people have way too much time at their hands which they can now invest in pointless community drama.
This is a weird hill to die on. As much as I resonate with many of the concerns, I don't see refusing to use AI as something that will actually help any of those things. Forking a stable version of vim is something I guess, but I don't really see the sky falling with mainline vim or neovim.
Personally the leverage I have as a bit of a cranky graybeard myself is that I understand how software works and I can distinguish between good and bad uses of AI and think critically about how to influence things towards better software. Just declaring AI as unequivocally bad and evil will do nothing more than make me irrelevant. At some point being right is useless without some measure of also being effective.
Without getting into some of the other things mentioned in the article,
I don't think Vim is going away. Even with all the AI code written, engineers navigate through Claude Code / Codex using Vim (ex: Vim mode in Claude Code).
I really like Vim so much that I've built a gamified way to learn it at https://vimgolf.ai that I am working on completing.
This doesn't seem like a good cost-benefit analysis for AI. Not sure what that would look like, but it seems like making some attempt to quantify the benefits would help.
I love it. Some alternative pathways in code make finding good solutions more likely. I've always liked Neovim, so I'm going to stick with it. I use it in a pretty much vanilla mode. Just deoplete et al.
I've long had great respect for Drew, way since way back when he was sircmpwn writing cool calculator software. Great programmer, and an incredibly based individual. Stays true to himself even in the face of overwhelming pressure.
I completely disagree with his take on this; battleship vibecoder in vimscript is awesome and important, socially, because vibe coding makes computer programming accessible to the masses. I don't expect him to ever agree, but much respect nonetheless
The argument that "vibe coding makes computer programming accessible to the masses" is something I don't understand. With all the free content on the internet, was it not accessible before?
It takes most people years of burying their heads in a computer to become effective programmers of anything more than trivial software. This is rapidly changing.
It does things for them and tells them what to do. Is that really making programming more accessible? I guess in the sense of lowering the barrier to creating stuff. But accessible as in a path to actually start working on things yourself and developing an interest? For most people vibe coding 99% of their lines, I doubt it. And I don't really think that's a problem to be honest, but I don't really buy that it makes programming in and of itself more accessible, more just the result of that programming.
> I guess in the sense of lowering the barrier to creating stuff.
That is the sense which I think most important. There are millions upon millions of very bright people with lots of valuable domain experience in a massive variety of specialities other than computer programming, who will now be able to use their expertise to guide the creation of software which before would have taken them many years of study, or millions of dollars to hire programmers. Empowering people to create their own tools will be a massive boon to humanity.
It doesn't have to be used that way, though. I wouldn't disagree that it mostly is used that way, but it can just as easily be used to teach. My wife has proven that well. AI has been the best development ever for her because it can custom tailor the lesson/task to be hyper relevant to exactly what she is trying to do.
Personally I've always preferred a great book to blogs/tutorials/etc, and even still I'd reach for a book if I had the chance on a new programming language or anything. But not everyone learns well that way, and I accept that.
I disagree, I think the over-reliance in these tools turns AI providers into the final gatekeepers of the profession. And with the raising prices of hardware, I'm afraid AI will make computing as a whole inaccessible to most people.
> because vibe coding makes computer programming accessible to the masses
I've been coding for 24 years and vibe coding has made computer programming accessible to me.
I've burned out on my work several times, to the point that a few years ago I became unable to open my IDE without getting headaches and nausea. This has killed one of the startups where I was fractional CTO and it's debilitating as an engineer to feel this.
Vibe coding has changed this. I'm once again productive. Like, 1000x more productive than I could ever be.
AI is an amplifier. It amplifies shit engineering into shittier code, but I also deeply believe it amplifies people who care about polish and love of their craft into so, so much more.
I've been "as a side project" finishing a bookkeeping app I could never finish (https://financica.app/) and adding so many features that are pure polish, which I always wanted to add but the ROI was just not there.
Like, the other day I wrote (using AI) a PDF parser for a specific type of account statements from the Belgian government, turning those into perfect data for the books. This saves me a ton of time as a user, nobody in the world has this automation for those types of statements, and it would have taken me several months of full time work to write and automate all of this, learning PDF libraries, dealing with the output, figuring out geometry, writing a battery of tests, etc. I would never have done it. But now, in less than an hour the whole feature was built, shipped and announced.
I'm debating using LLMs for my side projects. Does using one remove the "soul" of my project? On the other hand, a friend is actually making progress with his side app _because_ he's able to lean on the LLM after a full day's worth of working the day job. I might be able to actually do some of the things I've dreamed of and never had the capacity for. First world problems, I guess.
I've been doing exactly this now for a little while, and it breathed new life into my projects. It's been amazing, honestly. I was worried about the "soul" as well, especially for some projects where I got intimately deep in bit shifting and things, but realistically that project is now 100x more useful to me because it has a ton of features and even bug fixes that I never would have spent the time on before. I highly recommend it.
I think it depends on what you are doing the side project for.
Are you doing it to learn engineering? The learning potential of a back & forth with LLMs is wasted on people who don't have serious know-how.
Are you doing it to create a product, or learn how to do that? Then no, the LLM is helping you get over the hump of writing slow code.
I think we'll eventually drop the "vibe coding" and retronym coding to "slow coding" or something similar. There's advantages to slow coding in a world of AI coding, just like today there are advantages to dropping other types of abstraction layers (from writing direct code when using a WYSIWYG editor, to dropping into assembly code in a performance-critical branch of a game engine written in C++...).
But spending more time on writing code is not useful if you don't get something out of that additional time.
I disagree. I’ve had almost 20 years of professional programming experience. Spent a decade in FAANG, the rest in startups.
It is unarguable that I am able to program. Vibe coding has absolutely made programming more accessible to me too.
I have two kids and a full time job. Before LLMs I didn’t do side projects; work and parenting plus my other interests took > 100% of my energy.
Now I have many things I’ve worked on or built solely because LLMs lowered the barrier to entry, and I feel that I can fit the remaining human work into the cracks of the time and energy I do have. One can gripe about how I’m less connected to the code, or that I learned fewer substantial technical lessons from the experience; these things are true.
However, I learned more than if I hadn’t done the project at all. It’s like the exercise benefit of an electric bike - you don’t get the aerobic benefit of an unassisted bike, but if it motivates you to ride when you otherwise wouldn’t then the trade off isn’t so clear.
We've had this discussion back when high-level languages started becoming popular. Do the unwashed masses deserve to be programming a computer when they don't have a love and appreciation for assembly, or even the underlying ISA and its instruction encoding? And before that: How dare these whippersnappers just hand in their punched cards when they don't even know how to bit bang the boot sequence of the very computer executing them?
It's not even limited to a given occupation. Many hams were outraged about the FCC handing out amateur radio licenses without ANY demonstrated proficiency in morse!
Fortunately, at least in technology, nobody cares what these gatekeepers say. I guess that's an upside of software engineering never having graduated to be "actual engineering" (i.e. one with certifications and personal liability).
Nobody is preventing anyone from going as deep as they want to, and I expect that going one layer (or ten) deeper in understanding than your peers will still pay off even in a post-AI world. The nice thing is that now, nobody has to to just try something. (And you can ask the same system building these things for you how they work!)
You could use this exact same argument about any discipline and/or tool that has been made to support that discipline. A part of me loathes to make the comparison, but is an "audio engineer" any less of a musician than a traditional pianist? Maybe? It probably depends, but music has been made more accessible by the introduction of digital tools.
Regardless of whether or not AI is generally positive or negative, it's just not a compelling argument on it's face.
> because vibe coding makes computer programming accessible to the masses
This is BY FAR the worst part of LLMs to me. The influx of people i have zero desire to interact with into my normal online spaces has been incredibly painful.
I opened the comment section expecting to find a slew (a slop?) of LLM enthusiasts. I was not disappointed.
Whether you're a fanatical or not, of either side, LLM usage is driving energy and hardware prices to go up, it is an implicit driver of climate change, and it will replace jobs. I don't see what there's to argue.
I've been reading and reading about DeVault for more than a decade now. If I can point to one person on the internet and definitively say that their today's version is better than a decade ago, it would be him. (Yes, I can say that he appears to have improved better in this time than I myself have, which can be interpreted in a more than one way).
In fact, Andrew Kelley, whom I respect fair bit, also chose to stand behind redict, Drew's fork of redis with similar observation.
People change over time, some of them for the better, and I personally like to give them a chance. Some of Drew's opinions and expressions are still a bit much for me, but that is just us both being human.
Interesting he forked Vim 8.2.0148, but I am fine with that. I think I had to update ~/.vimrc to disable some a new default in v9 that annoyed me. I actually forgot what it was :)
I will have to look into his fork because I too do not want to see any form of AI in vim.
I may also look to see what Elvis looks like these days. I really liked the GUI and colors Elvis defaulted to and I stuck with it for a while, but eventually I went to vim in the v5 days for reasons I forgot.
If you're not using any software that might include code that originally came from an LLM, might as well give up on everything now. I'll give up the base if I ever have to remove built-in AI tools, but I don't foresee Vim dev getting that dumb anytime soon.
It doesn't look like they put AI into vim like Microsoft into Notepad. Someone used an outside AI to code something with vimscript, what do you expect? I'll be worried if they mess with even the smallest bit of established muscle memory of any vim user, but a separate language (probably a dead end) and apparently some new diff options don't seem too terrible.
> I won’t speculate on how he would have felt about generative AI, but I can say that GenAI is something I care about. It causes a lot of problems for a lot of people. It drives rising energy prices in poor communities, disrupts wildlife and fresh water supplies, increases pollution, and stresses global supply chains.
This kind of stuff drives me crazy sometimes. There's is little that's unique to AI here. These are the effects of any kind of industrial expansion. They're also the effects of population growth, in general. This stuff is a problem iff AI is a scam or hugely oversold and these resources are being wasted. But that's a different argument and a less clear-cut one.
> It re-enforces the horrible, dangerous working conditions that miners in many African countries are enduring to supply rare metals like Cobalt for the billions of new chips that this boom demands.
This point also deserves special mention. Most green technologies (solar panels, electric cars) also require a bunch of cobalt. Again, the "badness" seems to depend on your a priori evaluation of what the cobalt is being used for and not the cobalt mining itself.
I think there's also a pretty good chance that if a robot that could mine the same cobalt with no human intervention appeared tomorrow, many folks would complain about "hard working cobalt miners in Africa losing their livelihood to automation".
>This point also deserves special mention. Most green technologies (solar panels, electric cars) also require a bunch of cobalt. Again, the "badness" seems to depend on your a priori evaluation of what the cobalt is being used for and not the cobalt mining itself.
Neither solar panels nor Lithium Iron Phosphate batteries require cobalt. Pretty sure all the emphasis on that is mainly meant to cloud things and try to paint these things as just as bad for the environment as eg coal, and apparently it's been very successful based on how frequently I see it repeated, but it's not true currently. It was true with NMC batteries, but I think those have fallen out of favor even in EVs, and grid scale is dominated by LFP. Don't think solar panels have ever needed cobalt, they're glass, aluminum, silicon, and a bit of silver/copper. Thin films have cadmium sometimes, but those aren't the ones in use en masse for solar farms.
There’s a lot of good points in your comment, but fwiw it’s not clear whether they exist to dismiss a complaint or muster focus on the issues.
You’re right to point out that we’re all opted in at multiple levels to tech dependent on mining operations with a terrible human cost. I’d love to see these dangerous mining operations made safer with tech and policy, and you’re quite right that individual opt out is unlikely to have any effect (much less selective opt out from LLMs). Is that the end of the story?
If we’re just here to complain that someone’s marginal harm reduction posture is marginal I’m not sure that’s an effective rebuttal. Collective effort to lay new tracks and untie people off the old ones has more power than complaining someone used their personal trolley switch to shunt to a track with slightly fewer people.
Of course, that goes for people manning their personal switches too. And it’s worthwhile to pause and appreciate the scale and complexity of the problem.
I think my main point is that these particular concerns largely depend on someone already sharing the author's opinion - that AI is bad. They're not convincing otherwise because most other IT buildouts (e.g. "cloud computing", cryptocurrency) have a lot of the same drawbacks. Whether these costs are worth it or not then depends entirely on the nature of the technology they're being used for (which is why I brought up green tech).
> I think there's also a pretty good chance that if a robot that could mine the same cobalt with no human intervention appeared tomorrow, many folks would complain about "hard working cobalt miners in Africa losing their livelihood to automation".
Well, yeah? Just because the current work safety situation is bad, doesn't mean being out of a job couldn't be worse. I'd love a world where more automation meant less, safer, higher paying work for everyone. Our world never worked like that, to my knowledge, and I'm not sure it ever will.
> I'd love a world where more automation meant less, safer, higher paying work for everyone. Our world never worked like that, to my knowledge, and I'm not sure it ever will.
I'm not sure what you mean because that's literally what happened. The only remaining caveat is that it's not yet "everyone", but even that part is improving. If I was born in feudal Europe I would have spent my life planting, weeding, and de-pesting potatoes by hand instead of sitting at a computer in a climate-controlled office.
Sorry, I did a quick google search which seemed to indicate that cobalt is present in solar panels and batteries and not a deep dive. The broader point, that whether mining X mineral for Y purpose is bad depends entirely on what you already think of Y, remains.
If emacs can live through Stallman's descent into absurd un-asked-for pedophilliac defense positions, not limited to defense of Jeffrey Epstein himself, Vim can survive the simple passing of its creator.
Please don't do this. "editors outlive their creators' reputations/departures" - is a reasonable point. But to make it land as a zinger, you decided to dig up some
most inflammatory Stallman material possible, that does a lot of collateral damage to the framing.
Emacs the tool and Stallman the person are not nearly as coupled as your comment implies. Stallman created Emacs, yes, but the Emacs community drove him out of the FSF in 2019, pushed back hard when he tried to return in 2021, and has been actively distancing itself from him for years. The community's resilience despite Stallman is kind of the opposite of what you're trying to say - it's not like Emacs users were defending him in solidarity.
Tools transcend their creators - it is actually an interesting point and worth making. You just didn't have to push Stallman shit here.
I don't disagree with the general notion of your sentiment. I just wish there was less dragging Stallman's dick behavior into the mix of Emacs-related discourse. Which doesn't happen a lot, still would be ideal if it didn't happen at all.
Do you have a link for this? What I recall of that whole scenario was that Stallman said something fairly minor regarding Minsky, and the nuance of the words written were lost on the mob and he was accused of saying something worse than that.
I'm not aware of him providing any defense of Epstein himself.
"argued in an email thread last week that Marvin Minskey, the late AI pioneer and longtime MIT professor, was unfairly accused of sexual assault and that one of the underage girls in Epstein’s sex trafficking operation likely presented herself as “entirely” willing to have sex"
Yes, that is the Minsky comment I mentioned that apparently renders people illiterate and incapable of understanding what they read.
What Stallman said is "the people who were trafficked, probably did not tell the people they were trafficked to, that they were being trafficked and were there unwillingly."
I don't see how saying that is a "defense of Jeffrey Epstein".
How is that assigning the blame to Epstein's victims?
The scenario being described was that Epstein was ordering some of his victims, who were ostensibly employed as masseuses at his resort, to go and offer sex to specific people who were at an event taking place there.
You don't keep a sex trafficking operation going as long as he did if you don't coerce victims in that situation to play along with the story that they are masseuses and that the offer of sex is coming from them.
He's saying that Marvin Minsk might not have known that he was on pedo entrapment island and may have assumed a teenage girl was of age. Telling the entrapment targets the deal up front wouldn't be very smart. This is not blaming the girls, it was Epstein's setup.
“We can imagine many scenarios, but the most plausible scenario is that she presented herself to him as entirely willing,” Stallman added. “Assuming she was being coerced by Epstein, he would have had every reason to tell her to conceal that from most of his associates.”
Didn't want to include? That's the quote I'm referring to.
"Sex trafficker probably told the people he trafficked not to tell anyone that they were being trafficked. Trafficked people trapped on an island with their abuser may have done as they were told out of fear." Obviously.
I don't see how that statement is a defense of Epstein, or victim blaming.
It's amusing seeing this brought up in the thread when:
a) Drew is the person who wrote the major "takedown" screed accusing RMS of being a pedo(-defender).
b) Drew was subsequently outed for having a long history on the internet of consuming & sharing lolicon and saying that 14-year olds should be required by law to have IUDs installed.
Didn’t think I’d have to clarify what quotes mean in this context but using an LLM to help with coding is fine and people should get over it as it’s already everywhere anyway.
Lord forbid if people disagree with you. I know Drew's vibe is always "I'm right because I'm the only one with the correct opinions", but it does get tiring after a while.
Not to say AI isn't having huge drawbacks being introduced, and aren't exactly worry-free, but why not change your frame of mind from "Why don't others understand how awful it is?!" to "People are seeing something I'm not, what am I missing?" so your article could actually contain something else than personal and emotions rants?
I've seen people celebrate horrors beyond my comprehension. Cheer the deaths of innocent people because it may inch some abstract national goal closer to a similarly abstract measurement. Insist that lives in one place are worth less than lives in another.
Should I ask "what am I missing"?
I don't think so, sometimes you draw a line on moral or ethical grounds. Some of those lines should never be given the ability to be fluid. It will always be wrong to bomb a school of children, just like (for Drew and I) it will be wrong to rip the livelyhood from under millions of people's feet for shareholder value. It will be wrong to ignore damaging consequences to the environment. It will be wrong to insist a low quality imitation should ever hold the same value as the original idea.
using gpt is like bombing schools?
I have never gotten this. How is livelihood being "ripped away"? There is enormous capability made available to anyone and everyone who wants to take hold of and do something with it. Just as it's on each individual to go through the process and pains of landing a job (or building a business, etc), it's also on each individual to keep up with changes that may affect their livelihood. If they want to keep it.
It does seem like most people completely ignore the obvious harms caused by AI when talking about using LLMs for programming, as though somehow it is disconnected from the other deployments of the technology.
I feel that the people who are completely ignoring the harms are the ones who need and/or benefit from it and do whatever it takes to justify their use of it. The rest are people who understand the harms and minimize interaction followed by the blissfully ignorant.
I was just talking to a content creator who uses AI at work social media platforms to display her personal projects. She talked about how she is fully aware of the harm social media platforms bring while acknowledging they empower her to present her work to the world without gatekeeping. AI allows her to power through boring office tasks but she loathes their use in the art world and replacing people in general.
I would insist that the deployments of a technology should be disconnected from the technology itself - I criticize AI too, and I get a lot of downvotes for it, but I try to separate the science of AI from its economics and politics.
The harms of AI and other technologies come from two sources 1. Capital destroying market bubbles and 2. Deployments motivated and enabled by political and moral corruption.
Both of these are in turn enabled and sustained by legislation. That is, we have to talk politics, not technology and not AI. AI has a great potential - both for improving human life and for making it a lot worse and which way it goes depends entirely on politics.
If we fail to cleanly separate these issues and keep moralizing about technology, we will be chasing red herrings and bumping heads in the dark all the while the tech is being deployed against us.
Which is all fine and dandy. But why play the "You simply don't understand it as well as I do" rather than something more investigative and curious? Just fuels the whole "holier than thou" vibe Drew been trying to increase seemingly every day.
It's a disagreement of opinion, not some "I'm the only smart person who can realize this", which is why it kind of sours the entire piece.
That's not the tone of the article; he uses the word "pretending". That tells me that he thinks that people do understand, but they don't want to admit that they understand because that would reveal their values.
I'll say this from the perspective of a person who publishes content online: because people's revealed preference is for content written this way. You can spend weeks polishing thoughtful, original content that will get few clicks, or you can crank out throwaway op-eds about AI and get thousands of likes and upvotes from people who just wanted to hear their own beliefs explained back to them.
My stuff appeared on HN a couple of times over the years and the less effort I put into it, the better it fared. The temptation to change your writing style and to offer increasingly more provocative and shallow opinions is difficult to resist.
My point is probably this: if you want to see better stuff, I think you gotta stop engaging with articles like this. Patrol /newest and upvote cool in-depth stuff.
Would be very different from say:
> I'd like to understand people who don't see it the same way as me, that it's mostly awful and not good.
Or similar, rather than "I'm right, everyone else don't understand it properly". Very HN-esque, but oh so tiring after 100s of articles in the exactly same vein from the same author.
Ironically, you are not considering that he is seeing something that you are not, but you are not asking "What am I missing?"
See, that sword cuts both ways.
This is too shallow of a take. Especially when your very next point objects to what he uses as a default reference frame that you disagree with. Lord forbid drew disagree about, I think priorities, and values?
> why not change your frame of mind from "Why don't others understand how awful it is?!" to "People are seeing something I'm not, what am I missing?"
It's the same question. I sympathize with both questions, I constant feel both frustrated, and broken with how few people care about quality, and participating fairly. I try very hard to find the positive aspects "everyone" claims llm codegen provides. I'm looking hard, and can't find them. It's painfully average, often worse so when it gets lost. It doesn't and can not help me, only get in the way, what am I doing wrong? Why is everyone missing something I see as obvious? But again, both could easily be true from both frames you suggest. "Why can't people identify this as trash" could very easily be followed by "what I'm I missing from the equation?" and be the same thought/idea.
> so your article could actually contain something else than personal and emotions rants?
I mean, it's titled, A Eulogy for Vim. That seems to be what it says on the tin, no?
That sentence jumped out at me.
note I used "rich" there, not "wealthy"
Not endorsing this world view, just noting that the wealthiest 1% of people in the world (encompasses most US citizens) have an enormously outsized impact on climate.
They certain consume far more than the poor, on account of having resources, but they also consume far less than the wealthiest 1%.
1. https://www.bbc.com/news/magazine-17512040
People forget this. Oil companies may have dug up the oil, but they did so because we paid them to, so we could use the energy for good and useful things.
Climate change isn't 'evil billionaire companies are ruining the world', it's 'these things we did to improve our lives turn out to have side effects'.
OTOH, I can already argue with numbers at hand that Bitcoin made the world poorer and worse off.
> The maturity of Vim9 script's modern constructs is now being leveraged by advanced AI development tools. Contributor Yegappan Lakshmanan recently demonstrated the efficacy of these new features through two projects generated using GitHub Copilot
https://www.vim.org/vim-9.2-released.php#:~:text=The%20matur...
I am not sure I understand the author's concern, is he saying that VIM 9.2 is problematic because it enables AI integration due to the maturity of Vim9 script?
> sadly even Vim now comes under scrutiny in that effort as both Vim and NeoVim are relying on LLMs to develop the software.
...where he links to a comment in a closed issue where someone accuses a contributor of using an LLM to generate patches: https://github.com/vim/vim/issues/18800#issuecomment-3568099...
The tl;dr: Drew thinks Vim development has been tainted by LLM contributions, and is thus morally unsuitable to be used, and he will therefore be forking it.
Regarding why not Neovim, I think it's because a large section of the community want to create more complex TUI elements or replicate GUI interfaces and make it more like VS Code. I use Vim for the "vim way" not because it's in a terminal or it's not bloated like some other editors.
The next release will be the first where the majority of commits will be made by AI, and it has definitely not gone smoothly.
After a dozen or so bug reports, it's mostly in a working state, but I worry the output is no longer reliable in subtle ways.
Personally the leverage I have as a bit of a cranky graybeard myself is that I understand how software works and I can distinguish between good and bad uses of AI and think critically about how to influence things towards better software. Just declaring AI as unequivocally bad and evil will do nothing more than make me irrelevant. At some point being right is useless without some measure of also being effective.
I don't think Vim is going away. Even with all the AI code written, engineers navigate through Claude Code / Codex using Vim (ex: Vim mode in Claude Code).
I really like Vim so much that I've built a gamified way to learn it at https://vimgolf.ai that I am working on completing.
I completely disagree with his take on this; battleship vibecoder in vimscript is awesome and important, socially, because vibe coding makes computer programming accessible to the masses. I don't expect him to ever agree, but much respect nonetheless
That is the sense which I think most important. There are millions upon millions of very bright people with lots of valuable domain experience in a massive variety of specialities other than computer programming, who will now be able to use their expertise to guide the creation of software which before would have taken them many years of study, or millions of dollars to hire programmers. Empowering people to create their own tools will be a massive boon to humanity.
Personally I've always preferred a great book to blogs/tutorials/etc, and even still I'd reach for a book if I had the chance on a new programming language or anything. But not everyone learns well that way, and I accept that.
I've been coding for 24 years and vibe coding has made computer programming accessible to me.
I've burned out on my work several times, to the point that a few years ago I became unable to open my IDE without getting headaches and nausea. This has killed one of the startups where I was fractional CTO and it's debilitating as an engineer to feel this.
Vibe coding has changed this. I'm once again productive. Like, 1000x more productive than I could ever be.
AI is an amplifier. It amplifies shit engineering into shittier code, but I also deeply believe it amplifies people who care about polish and love of their craft into so, so much more.
I've been "as a side project" finishing a bookkeeping app I could never finish (https://financica.app/) and adding so many features that are pure polish, which I always wanted to add but the ROI was just not there.
Like, the other day I wrote (using AI) a PDF parser for a specific type of account statements from the Belgian government, turning those into perfect data for the books. This saves me a ton of time as a user, nobody in the world has this automation for those types of statements, and it would have taken me several months of full time work to write and automate all of this, learning PDF libraries, dealing with the output, figuring out geometry, writing a battery of tests, etc. I would never have done it. But now, in less than an hour the whole feature was built, shipped and announced.
It's awesome.
Are you doing it to learn engineering? The learning potential of a back & forth with LLMs is wasted on people who don't have serious know-how.
Are you doing it to create a product, or learn how to do that? Then no, the LLM is helping you get over the hump of writing slow code.
I think we'll eventually drop the "vibe coding" and retronym coding to "slow coding" or something similar. There's advantages to slow coding in a world of AI coding, just like today there are advantages to dropping other types of abstraction layers (from writing direct code when using a WYSIWYG editor, to dropping into assembly code in a performance-critical branch of a game engine written in C++...).
But spending more time on writing code is not useful if you don't get something out of that additional time.
This was never the case.
The only people vibe coding has made programming accessible to is people who don't have such motivation.
It is unarguable that I am able to program. Vibe coding has absolutely made programming more accessible to me too.
I have two kids and a full time job. Before LLMs I didn’t do side projects; work and parenting plus my other interests took > 100% of my energy.
Now I have many things I’ve worked on or built solely because LLMs lowered the barrier to entry, and I feel that I can fit the remaining human work into the cracks of the time and energy I do have. One can gripe about how I’m less connected to the code, or that I learned fewer substantial technical lessons from the experience; these things are true.
However, I learned more than if I hadn’t done the project at all. It’s like the exercise benefit of an electric bike - you don’t get the aerobic benefit of an unassisted bike, but if it motivates you to ride when you otherwise wouldn’t then the trade off isn’t so clear.
It's not even limited to a given occupation. Many hams were outraged about the FCC handing out amateur radio licenses without ANY demonstrated proficiency in morse!
Fortunately, at least in technology, nobody cares what these gatekeepers say. I guess that's an upside of software engineering never having graduated to be "actual engineering" (i.e. one with certifications and personal liability).
Nobody is preventing anyone from going as deep as they want to, and I expect that going one layer (or ten) deeper in understanding than your peers will still pay off even in a post-AI world. The nice thing is that now, nobody has to to just try something. (And you can ask the same system building these things for you how they work!)
Regardless of whether or not AI is generally positive or negative, it's just not a compelling argument on it's face.
This is BY FAR the worst part of LLMs to me. The influx of people i have zero desire to interact with into my normal online spaces has been incredibly painful.
Whether you're a fanatical or not, of either side, LLM usage is driving energy and hardware prices to go up, it is an implicit driver of climate change, and it will replace jobs. I don't see what there's to argue.
Great article through and through.
In fact, Andrew Kelley, whom I respect fair bit, also chose to stand behind redict, Drew's fork of redis with similar observation.
People change over time, some of them for the better, and I personally like to give them a chance. Some of Drew's opinions and expressions are still a bit much for me, but that is just us both being human.
I will have to look into his fork because I too do not want to see any form of AI in vim.
I may also look to see what Elvis looks like these days. I really liked the GUI and colors Elvis defaulted to and I stuck with it for a while, but eventually I went to vim in the v5 days for reasons I forgot.
This kind of stuff drives me crazy sometimes. There's is little that's unique to AI here. These are the effects of any kind of industrial expansion. They're also the effects of population growth, in general. This stuff is a problem iff AI is a scam or hugely oversold and these resources are being wasted. But that's a different argument and a less clear-cut one.
> It re-enforces the horrible, dangerous working conditions that miners in many African countries are enduring to supply rare metals like Cobalt for the billions of new chips that this boom demands.
This point also deserves special mention. Most green technologies (solar panels, electric cars) also require a bunch of cobalt. Again, the "badness" seems to depend on your a priori evaluation of what the cobalt is being used for and not the cobalt mining itself.
I think there's also a pretty good chance that if a robot that could mine the same cobalt with no human intervention appeared tomorrow, many folks would complain about "hard working cobalt miners in Africa losing their livelihood to automation".
Neither solar panels nor Lithium Iron Phosphate batteries require cobalt. Pretty sure all the emphasis on that is mainly meant to cloud things and try to paint these things as just as bad for the environment as eg coal, and apparently it's been very successful based on how frequently I see it repeated, but it's not true currently. It was true with NMC batteries, but I think those have fallen out of favor even in EVs, and grid scale is dominated by LFP. Don't think solar panels have ever needed cobalt, they're glass, aluminum, silicon, and a bit of silver/copper. Thin films have cadmium sometimes, but those aren't the ones in use en masse for solar farms.
You’re right to point out that we’re all opted in at multiple levels to tech dependent on mining operations with a terrible human cost. I’d love to see these dangerous mining operations made safer with tech and policy, and you’re quite right that individual opt out is unlikely to have any effect (much less selective opt out from LLMs). Is that the end of the story?
If we’re just here to complain that someone’s marginal harm reduction posture is marginal I’m not sure that’s an effective rebuttal. Collective effort to lay new tracks and untie people off the old ones has more power than complaining someone used their personal trolley switch to shunt to a track with slightly fewer people.
Of course, that goes for people manning their personal switches too. And it’s worthwhile to pause and appreciate the scale and complexity of the problem.
Well, yeah? Just because the current work safety situation is bad, doesn't mean being out of a job couldn't be worse. I'd love a world where more automation meant less, safer, higher paying work for everyone. Our world never worked like that, to my knowledge, and I'm not sure it ever will.
I'm not sure what you mean because that's literally what happened. The only remaining caveat is that it's not yet "everyone", but even that part is improving. If I was born in feudal Europe I would have spent my life planting, weeding, and de-pesting potatoes by hand instead of sitting at a computer in a climate-controlled office.
Technology Connections did a great video that goes into this: https://www.youtube.com/watch?v=KtQ9nt2ZeGM
Emacs the tool and Stallman the person are not nearly as coupled as your comment implies. Stallman created Emacs, yes, but the Emacs community drove him out of the FSF in 2019, pushed back hard when he tried to return in 2021, and has been actively distancing itself from him for years. The community's resilience despite Stallman is kind of the opposite of what you're trying to say - it's not like Emacs users were defending him in solidarity.
Tools transcend their creators - it is actually an interesting point and worth making. You just didn't have to push Stallman shit here.
I'm literally describing the resilience of the emacs community exactly as you described.
And the emacs community deserves the right to call him out to distance ourselves from them.
Do you have a link for this? What I recall of that whole scenario was that Stallman said something fairly minor regarding Minsky, and the nuance of the words written were lost on the mob and he was accused of saying something worse than that.
I'm not aware of him providing any defense of Epstein himself.
MIT scientist Richard Stallman resigns in the wake of his Jeffrey Epstein remarks https://share.google/L9w5zAnDjbvnrWhex
What Stallman said is "the people who were trafficked, probably did not tell the people they were trafficked to, that they were being trafficked and were there unwillingly."
I don't see how saying that is a "defense of Jeffrey Epstein".
He is assigning the blame to Epstein's victims.
The scenario being described was that Epstein was ordering some of his victims, who were ostensibly employed as masseuses at his resort, to go and offer sex to specific people who were at an event taking place there.
You don't keep a sex trafficking operation going as long as he did if you don't coerce victims in that situation to play along with the story that they are masseuses and that the offer of sex is coming from them.
if you dont understand what that is saying, i can't help you.
You can't see the difference between "she presented herself to him as entirely willing" and "she was entirely willing"?
Stallman may be a dick, but at least he's precise with his speech - this means exactly what it says, and in no way means what you want it to mean.
"The trafficked person did not reveal they were being trafficked, because they were trapped on an island with their abuser and were afraid".
This is not blaming the victim, nor a defense of the abuser.
“We can imagine many scenarios, but the most plausible scenario is that she presented herself to him as entirely willing,” Stallman added. “Assuming she was being coerced by Epstein, he would have had every reason to tell her to conceal that from most of his associates.”
"Sex trafficker probably told the people he trafficked not to tell anyone that they were being trafficked. Trafficked people trapped on an island with their abuser may have done as they were told out of fear." Obviously.
I don't see how that statement is a defense of Epstein, or victim blaming.
a) Drew is the person who wrote the major "takedown" screed accusing RMS of being a pedo(-defender). b) Drew was subsequently outed for having a long history on the internet of consuming & sharing lolicon and saying that 14-year olds should be required by law to have IUDs installed.
I know that my point of view is considered .+(cist|phobic) (based on the post). I'm sorry for that.
That's great you can spend more time with your family, but the code you're writing this way is, by and large, probably crap.