I admittedly flip-flop on this quite a bit. I use LLMs in my work and for a bunch of pet projects I don’t really care much about.
For the pet projects, it’s not really a big deal. I’m the only user, and I as the user don’t care how it works under the hood.
Work is a different story. I have never felt more disconnected from my work than in the last 6-8 months. It takes a huge mental toll to read into our code and figure stuff out, almost like I just got hired and I’m working on new code for the first time. I’m asked to read “specs” written by LLMs that are unintelligible and I find my brain disconnecting immediately. Tickets, also written by LLMs pose the same problem.
I’m asking teammates questions because I want their answer and I’m told to just ask Claude.
None of this feels good, it may be where the industry is heading but as a professional it feels wrong.
when llms write all the code, all we have left are the specs. ive found a lot of success applying software engineering discipline to spec writing and making it a human-only activity. from there hand the spec to LLM with a few guardrails around implemention and writing quality tests that prove the spec.
PRs review more quickly when you have human written spec diffs to orient you and guide (or skip) the code review.
a danger to avoid is writing specs which are basically just really bad code (prose is terrible code). you can work towards patterns to avoid this pitfall.
also, dont coflate spec with plan. spec is durable decisions. plan is execution to turn decisions into code. plan is disposable, dont check it in.
either way, find the human layer and dont use LLMs there. make all the important decisions in the spec, delegate the other decisions to the LLM.
i believe this is the new craft.
that said, i expect a small but significant exodus from software. a lotta folks came for the "programming" that is now being outsourced. were becoming so tired, just absolutely burning out.
i hope we will eventually realize its time to go, instead of giving up our souls for a fucking paycheck.
I'm feeling the same, personal projects have never felt better, but work has slowly felt more miserable.
Code reviews come back with a ream of issues that I just want to scream "I don't give a fuck".
Pre-LLM no-one would have cared that if another class was configured differently ( it isn't ) and called the new class in a different way ( it doesn't ), that you'd in some race-condition occasionally leak a resource, but only on shutdown when the program is about to terminate anyway.
That's the kind of issue that claude loves to find and write up in a code review with the exact same verbiage and write up style as if it had just found a naive SQL injection from string concatenation.
Prior to LLMs I doubt anyone would even have spotted the first issue, let alone bothered to write it up.
But instead it's another round of review, another round of trying to decipher the LLM's comments to discover whether it's a meaningful problem or a nothingburger.
What I hate the most is people vomiting AI slop communicating with others. I have teammates who decided that an agent producing a wall of text summarizing changes in a pull request instead of a simple "LGTM" is a good idea. I have teammates who talk using AI almost exclusively (you can see a small tag "Sent using Claude" or "Sent using Cursor" under their messages)
Posing a well considered, concise and precise question only to get a wall of Claudeisms in response from a colleague is always demoralizing. Then you have to play the guessing game of 'is this detail in the response true, or did Claude make it up for them?'
> I have teammates who talk using Claude almost exclusively.
If somebody told that humans are a form of creature which can forego their most important and powerful capabilities just for convenience, I'd call the person saying that batshit insane, but here we are...
a teammate who exclusively writes slop PRs and copy/paste LLM responses to slack has demonstrated they can be replaced by an LLM integration to slack and github.
I must again point out that we're not even certain yet that we're "losing" any programming capability, and we're not certain the tradeoff isn't worth it.
All of these arguments presuppose we're losing something that's not worth losing.
I want someone to start one square earlier and argue to me why that matters. I've read enough of these "your skills are atrophying!" arguments; I'm also making like 10x as much revenue as I did through a business that delights my clients. Why should I care about my "skills" if I'm still able to make money, pursue hobbies, and enjoy life?
Because you will have an understanding of the complete system. Your domain and the code. This means that you can be more creative with it and also get yourself out of a mess if it occurs (which it will).
From what I have seen AI doesn't understand the code. But even if you argue that it does understand the code it definitely doesn't understand the domain because it's only input channel is a stream of text and you didn't even type in a "full" text representation of the domain.
So you remain the only thing that can understand the whole system. If your competitor has no one (or no team) who can put together a full understanding of the system then you will have an advantage.
What I would love to see is a side by side comparison of two companies (one using AI, and one without) over an extended period of time.
Id wager that they would fare the same. The time saved on some tasks are outweighed by extra time spent on others. Everyone assumes that just because you can pump out more features, you're more successful.
History has a lot to say about the ruling class benefitting from (and ensuring?) the working class cannot organize, because that is exactly when the workers have leverage.
workers that feel powerless and dont try to resist is great news for rulers who only serve to generate revenue.
> it will be sure way of selecting yourself out of the work pool
If, as an engineer, you rely completely on AI then you will be selected out of the work pool. Some engineers will still be needed. If I was interviewing them (I used to do a lot of this) then I will want the ones who understand the code. If they only use AI then it will quickly become clear that they don't understand it. Why would I hire them when another person can understand it?
In this context, a "car" is a vehicle with some sort of engine, as opposed to a horse, mule, carriage or other animal-powered vehicle. Those who rejected cars when they came out would also happily reject buses, cabs and even trains.
Why are we glossing over the fact that it's all trained on stolen code, stolen art, stolen books? It's unethical to use it, it's unethical to take money for it. It should be either free or killed. I really don't understand why we just collectively accepted that.
I gloss over it because it's an argument that makes no sense to me.
Copyright exists to encourage progress in the arts and sciences. Not to enrich artists or scientists.
Ai is progress. Further, the works it creates are fully transformative. It isn't just spitting out Harry Potter verbatim, it's making new things from it.
To me, this argument is like staying that if you read Winnie The Pooh as a child then Milne owns your artistic output for life, because one of your many influences might have been his work. It just seems obvious at face value.
How is it progress if it stops people from sharing. It just reduces everything to the lowest common denominator.
Or how is it progress if the data centers just eat up all the components skyrocketing the prices for regular people.
>Copyright exists to encourage progress in the arts and sciences. Not to enrich artists or scientists.
Is impoverishing them supposed to encourage progress in the arts and sciences?
Whatever it started out as (and I have my doubts it was ever a tool to help artists), it's now primarily a tool corporations use to milk creativity for profit and squash it where it might interfere with that profit.
> Is impoverishing them supposed to encourage progress in the arts and sciences?
No, but that needs new laws. Not an abuse of the old ones justified under the thinnest of pretenses.
I don't think that folks stop to consider what bending copyright law into these weird shapes to meet their ends will do to other aspects of copyright.
For example, sampling in music. How is that different from AI generated music? If anything AI uses LESS of any individual song. You can't just say "no derivative works." That would fully halt progress on many fronts.
If a song was made purely of samples with no original input I'd argue it's in the same category.
Maybe AI uses less of any individual song, but does AI make something actually original and transformative or is it just bits and pieces of all the things it learnt on (in simplified terms)? I think that's where the discussion lies.
Entirely new works are often created entirely from samples, and can be amazing. See "Frontier Psychiatrist" by The Avalanches for one example done entirely by humans. There are many others.
Amazing or not, that is not the discussion. LLM output can look amazing as well but that's not what I'm arguing about.
Also with LLM's we'll reach a point when everything is made out of samples (metaphorically speaking), where do we fit original content then, or who will make it if it means it's going to be consumed by data centers immediately.
The frequency with which AI boosters leap to "clearly once you've read a book the author owns everything that comes out of your brain" still shocks me a little bit. It's obviously not the same as somebody torrenting books2 or libgen or whatever and training on it for profit.
I see no argument here. Just a note that it seems to be obvious to you. It must not be as obvious to me, or millions of others or we wouldn't be on the opposite side of it.
I'm an open source dev and I appreciate AI. I'm glad it exists and I have no problem with people being rewarded for their work on AI.
That said, I also think everyone who contributed content online deserves to be rewarded for their contribution, especially since OpenAI (which made the largest contribution) was a non-profit in the beginning. So there should be some kind of tax. Otherwise I agree, it's not ethical. It's not 'fair use' of copyrights.
Your heart is in the right place but your conclusion is not.
Because it was trained unethically we have a few moral obligations:
1) USE IT. Do not let only unethical people use a force amplifier on our collective efforts.
2) STOP PUBLISHING HUMAN GENERATED CONTENT. The more original human thought = the more they harvest our thinking.
3) Generate only known things in novel combinations. Do not provide new insights.
I could go on but strategically thinking AI is not going to be stopped in the way you are thinking. So you need to think a bit deeper on methods of optimizing the situation.
IMHO, if they purchased the books etc, it's fair because, for me, it's like a biological neural network (i.e human) reading and watching the contents and memorizing them.
I've never understood the water argument... yes, it takes a lot of water in the initial setup to fill the water cooling loops, but isn't not constantly burning water, right? it gets cooled off again and goes around the loop
It’s not a good argument in my opinion, just something that’s spread because it’s easy to latch onto. The actual numbers on water usage aren’t a big deal.
That’s not to say that AI is all fine and dandy, I genuinely think frontier AI companies have deranged leadership.
Ah ok, then drought aren't a problem I guess? Guess what happens when you pump ground water faster than rain water replenishes it? Guess what climate change is disturbing?
This your your typical HN take, aka "I thought about a random problem I barely understand for 12 seconds and I think I figured it out better than domain experts"
Three percent of two thirds of the entire planet is enormous. We aren't running out, and never will.
Some people just live in places that aren't suitable for human life and survive only through technology. Those places need to have different laws about water.
For example, I shouldn't be forced to use a low flow showerhead when I'm 100 miles from 1.3 quadrillion gallons of fresh water. If you aren't next door to lake Michigan then maybe you should be.
It's about evaporative cooling. Read about how cooling towers work in nuclear and coal plants. The principle is the same for the data centers, but it the wasteheat to get rid of comes from computation.
They obviously shouldn't be doing it like that, but since clean water is cheap and business in absence of reason optimizes for lowest costs that's exactly what they did.
There are various approaches toward AI. Some are radical. I often found extremism to be wrong in most cases.
I have a suspicion that people who reject AI related stuff is just scared under the hood. They fight against AI because of the fear against loosing job, money, or control.
There were many manual jobs. Lets say carpenters. They had manually created things, often pieces of art. Now many things are soulless, created by factory output, and by robots. It is true that we lost some artistry. It is true that every line of code will not be a work of art. There is no going back though. Adapt.
Dude said "Adapt" and you really pulled out Hitler.
Oi. Discussion on this site gets more difficult each day. Instead of pointing out where you disagreed you just pulled out the most egregious appeal to authority and said, "oH yEaH?"
I don't think that's a reasonable read; you can't dismiss a specific criticism on the (unsupported) ground that the person making it would totally have dismissed anything at all.
I don't think the take is bad, and what he says is right. Maybe we shall look at the substance rather than the artistic properties of their words?
I was recently discussing with someone who wanted to put all datacenters into AI datacenter basket, and inadvertently labeled all consolidated computing bad. That person, claiming otherwise was not knowing what they were talking about,
But while having nuances, AI datacenters and heavy AI usage is harmful to the world we live in, as a whole. We should come into terms with this and go from there.
No, what he says is not right. Parts of it are, but broadly (as I pointed out) he conflates a lot of things together that have no business being in the same sentence.
1. Reducing/replacing/automating highly complex work is fundamentally a good thing, it might be the case that society is not ready for the consequences of it, those are two separate things. Concluding that millions of people will die, is simply hyperbole.
2. Saying that AI is not intelligent is at best a non-sequitur but more likely intended misrepresentation and goalpost moving
3. The environmental harm of AI, however significant, is ultimately public knowledge and 100% clear in the public discourse around the current implementation of AI. This says nothing about AI in the general sense, only how we produce it today. This is like preventing cars from becoming a thing because they produce CO2. Do you think the world would be better off if cars were never invented?
None of this is about the 'artistic properties of their words', it is precisely about the core content being objectionable.
> Reducing/replacing/automating highly complex work is fundamentally a good thing, it might be the case that society is not ready for the consequences of it, those are two separate things. Concluding that millions of people will die, is simply hyperbole.
I don't think saying that millions will die is a hyperbole. It might be communicated poorly, but having no ability to work in a world with worsening quality of life (via climate crisis and economic changes) is being pinched between a rock and a hard place. What these people gonna do? Photosynthesis? One of countries which is among the leaders of this technology believes that social governmental systems is communism. So, yeah, perfect.
> The environmental harm of AI, however significant, is ultimately public knowledge and 100% clear in the public discourse around the current implementation of AI.
Yet the people who own and hype this technology is applying the Tobacco/Soda Industry Playbook: Deny and publish studies saying otherwise.
This is not fair, because who has the gold, makes the rules to increase their profit even more.
> This is like preventing cars from becoming a thing because they produce CO2. Do you think the world would be better off without combustion engines?
My current stance is not undoing things to zero, however forcing AI development to be more environmentally conscious and start regulating its power and water use. This is possible by prescribing energy source mix ratios, closed loop water cooling, etc. etc.
Of course this will slow some progress and make it a bit more expensive for a while, but in the end it'll be way more environmentally friendly.
All discourse doesn't have to be on a duotone plane. We can discuss in a 3D space with infinitely more shades.
> but having no ability to work in a world with worsening quality of life
If unemployment leads to death, there are many more things wrong than just AI. Why get mad at AI instead of at capitalism. This is nonsensical.
> Tobacco/Soda Industry Playbook
Except both of those succeeded in destroying health for decades, whereas AI had like a 2 year run before all the datacenter madness came to light.
> My current stance is not undoing things to zero,
This does not at all answer my question, my question is about the past not the present. The author is arguing to resist AI, what if resisting cars had worked? This is my question.
I can buy the idea that AIs are making people a lot lazier intellectually, on average.
But I do think that for the atypical user that is "information-hungry," AIs are a massive productivity booster. I've probably asked ChatGPT 1,000 questions over the last 2 years that I otherwise would have needed to spend 15+ minutes researching. Prior to AI, I would have just... not spent the time looking them up.
The danger is just when you start using AI but otherwise continue doing exactly what you were doing before.
" the idea of using AI as an information-gathering tool."
Programming is mostly exactly that. You gather want information you need to get the job done. What are the names of the API you call, what are good patterns for this kind of situation etc.
But your point still stands as in use of AI can make those skills detoriate. And regarding to coding, that you don't know anymore what is in your codebase, how it actually works. Can happen fast if you let AI run in auto mode.
Count me in. I rejected my promotion (losing 2x gain in salary) from sysadmin to developer at my work exactly because of AI that ate it alive during the last year.
Our devs have been outright complaining about getting dumb, losing touch with the codebase and getting lost in weekly switching between models/agents that have been gushing as if from a firehose, but the management (read investors) knows better and there's a culture of self-censoring the opposition to AI among employees.
"After trying Kagi/Duck Duck Go, I have gone back to Google for search and I freely grant that I find the “AI Overview” to be useful."
I feel compelled to point out that Kagi does offer AI-backed answers if you want it to. Just pose your query with a question mark to trigger the AI response in addition to the usual search results. It is not ever the default or only answer, and I pray Kagi keeps it that way!
I use the AI option more often than I thought I would, usually for things that I don't consider controversial or complicated.
I'm sympathetic to the author's core position. I think LLMs as they're currently deployed are doing net societal harm so far. For example, for every React app whistled up in five hours, there's a high schooler or college student out there sleepwalking through their foundational courses and being set up to catastrophically fail upon trying to enter the job market.
But I think it's important for the quality of public discourse that we forward only the best, strongest arguments for any given position (whether I personally agree with that position or not). A weak argument gives unnecessary purchase to unscrupulous haters who might use that weaker argument as a low-grade "straw man" to dismiss the whole position.
So with that preamble, some constructive criticism (very well-intentioned):
> I am not exaggerating: We need to resist “AI” so that thousands, perhaps millions of people do not die.
On this one you are, in fact, exaggerating (certainly with "millions," likely with "thousands" as well).
The overwhelming majority of people who lose their job for whatever reason (and there are many reasons, not just this one issue) fall back on a support network of their family and friends while they "get back on their feet" by pivoting to another job or form of employment (sometimes branching into entrepreneurship, for example).
The alternative to employment at whatever FAANG or FAANG-adjacent company someone was laid off from is not "death." It is typically "some, sometimes quite significant, discomfort." Granted "AI Layoffs" are starting to ripple out beyond FAANG, but the story is similar in other industries.
If "layoffs kill people" is the moral hill we choose to plant our flag on, then mustn't we also campaign with equal vociferousness against all the other causes of layoffs in general? Since to do otherwise is to sit by while those laid off allegedly die in droves of thousands or perhaps millions?
It also occurs to me that this framing ("AI causes layoffs") also gives incompetent and unaccountable executives a pass by pinning the actual harm done by irresponsible layoffs (which to be clear, is not none!) on "the AI" rather than on the leaders who are in the positions to actually decide what happens in their organizations.
I think the argument is much stronger if we talk about the aggregate cost to humanity being paid in terms of displacement, life disruption, and curtailed opportunity (especially for juniors). We should not make the "toxic escalationism" move of equivocating things we don't like with negligent homicide, even if it makes for a punchy one-liner.
> forward only the best, strongest arguments for any given position... A weak argument gives unnecessary purchase to unscrupulous haters who might use that weaker argument as a low-grade "straw man" to dismiss the whole position.
Excellent point. Thank you for writing such a cogent argument.
It is clear that the rich and powerful are making short-term decisions that are causing widespread and long-term harm. Decisions such as, AI-pilled investors demanding unsustainable productivity from companies, and CEOs in turn forcing employees to use agentic coding and laying the rest off.
Such decision-making is not new, of course, but LLM technology is helping grifters and con artists to turbo-charge it. Founders can raise money using cool demos that achieve the impossible. Investment analysts can back up their AI story with huge paper profits.
The result, as you summarized, is life and career displacement and missed opportunities for young people. It's just too abstract and diffuse of an impact to influence people with, and it also seems to be railing against the holy cow of Capitalism.
"Data centers bad!", or "AI is disrupting humanity and the end is nigh!" are much more punchy slogans.
> For example, for every React app whistled up in five hours, there's a high schooler or college student out there sleepwalking through their foundational courses and being set up to catastrophically fail upon trying to enter the job market.
I see a lot of people throw their hands up in the air and say "I have no control over this, might as well go all in on AI".
But we as individual consumers have more power than we think. We don't need to use these AI features and products. We've survived perfectly fine without AI. Made plenty of scientific discoveries and lead enriched lives without it.
The biggest "fuck you" to the pricks who pushed this on us would be if we collectively went "meh".
I understand the appeal of saying fuck you to billionaires and pathological liars CEOs and the annoying usages of AI, but i don't understand the wider "fuck you AI".
AI are here, is it going to demolish and (maybe) reconstruct the world as we know it, trying to avoid it it's just silly. There is no sane world where we can "stop AI", even if every single researcher were to do vanish tomorrow and the hardware and schematics magically gone, it would still only be a stopgap.
Listening to developers complain about AI doing things to their life and the world at large when they have spent the past 30 years automating just about everyone else they could out of a job (Since this has been one of the primary uses of software) is ... something else. It sounds like a a a mix of hubris, arrogance, cognitive dissonance, ignorance, etc.
Only when it started to affect their world did any developers start to care, and even then nobody seems to ever stop and say "hey what were we doing the past 30 years, maybe we should think about this", it's instead "hey look at what these other people are doing with AI, someone should think about this".
It feels like, to paraphrase a movie quote - "Nobody gets to destroy jobs and the the world but us"
Yup. If current-gen AIs were capable of writing articles, but not code, then devs wouldn't care at all. None of these articles would exist, and some new variation of "learn to code" would be the snarky reply to nay-sayers.
In fact, we already pretty much have proof of this: devs didn't care when software destroyed the journalism industry twenty years ago. Turns out that people ultimately only care when their only jobs are on the line.
The difference is that automated tasks are (or should be) deterministic and reproducible, and taking up time devs could be using to do their actual job (writing code, solving technical problems, etc).
I see good ethical reasons to be against AI, but I see no reasons why “resisting” would lead to anything but irrelevance. So I think this is terrible advice
I'm not saying this situation is different from all of the other technological paradigm shifts in history.
But what concerns me is that authors of articles such as this feel no need to differentiate themselves from other anti-technology arguments in history.
I think it is very much different. Previous revolutions allowed us to fall back from manual labour to more mental driven work. Now even this is not going to be an option. Soon we're going to be reduced to be useful for spare parts only.
This is only the case for the Industrial Revolution.
There have been many other thinking technologies and they ALL have been attacked in exactly the same way.
* Socrates attacked the invention of writing, arguing it would make people intellectually lazy.
* Abbot Johannes Trithemius argues in "In Praise of Scribes" in the 15th century that the Gutenberg machine would make people intellectually lazy.
* Lots of people in the 1960s attacked punch card computers, and automation in general (!), on the basis that it would make people intellectually lazy.
* Nicholas Carr attacks Google in a famous article "Is Google Making Us Stupid?", on the basis that it would make people intellectually lazy.
And so on and so on and so on.
"That which has been is what will be, And there is nothing new under the sun."
It is because of this history that any argument such as the author makes, arguing that AI would make people intellectually lazy, must start already from the lamentation of the downfall of human intelligence from the days of Socrates, or must explain why this is different.
It is obviously not different.
There are three mental skills being offloaded onto AI:
1. Domain-specific skills
2. General intelligence (g)
3. Recall
Of these three, the danger with 1 is offloading them too early, before the AI is "good enough" to be better than human experts. This is a genuine concern and we should do things carefully. However, if the AI is good enough, there is no danger here.
Recall is the mental skill that all of these innovations have "hurt", but it can simply be trained explicitly and synthetically. We will certainly have to structure our societies more and more in the future to synthetically exercising recall (like we're now synthetically exercising our muscles), but if we do this, offloading recall at work is not a big issue.
As for general intelligence, this is the most harmless to offload, because, as far as I know from the literature, it cannot really be trained, and it does not really degrade either.
Depends on what you mean by correct. Do you think the world would be better if those anti-technology people succeeded and the technologies did not come to be?
A world without steam engines, cars, computers, printing presses.
Why would they need to differentiate? You're implying those arguments were wrong?
The argument usually isn't "I believe X will not happen," it's usually either "I believe X should not happen" or "I believe we should manage how X happens".
we can all debate the finer points but in any case "resist" is exactly the right message.
AI companies are racing to build empires on products built on the whole of human writing, charging rent to access it, then leveraging their scale to improve the model.
the big data era was a gold rush for greed. the AI era is has upgraded to a warp engine for greed.
the harms and manipulation of social media took ages to enter the social discourse, and are just now turning into legal action.
AIs warp-engine speed guarantees greed will fuel warp-engine scale harm before society catches up, and before America's greedy ass politicians are cornered into trying to do something.
resist fucking greed. if you gotta use LLMs, at least don't pay an AI company.
I'm not sure how valuable this advice is. From what I can see, Jeffries hasn't been involved in actual software development for decades (he is ~87 and has a lot of great accomplishments under his belt, but in the year 2026, 0% of his income relies upon him generating solutions for paying clients in a competitive space, and probably hasn't for a long, long time).
For some reason HN readers love this sort of content, largely by periphery players who mostly think it gives them an angle for clicks. And, I mean, as someone who uses lots of AI to delight clients and stay competitive, you be you. It leaves more for me.
I criticize stuff like this a lot. I criticize the doomer “bomb data centers” rationalist cult even more.
At the same time, maybe I shouldn’t, and I sometimes find all the AI skeptic and even doomer discourse heartening.
We really messed up with the Internet. I’m an Xennial and I had a minor role in it but I was there and it feels like the Internet was “our project.” I was literally born with the PC, discovered BBSes as a tween and the pre web Internet as a teen.
Don’t get me wrong. I’m not saying the net is bad. I think it’s more positive than not (still). But my god the addiction and disinformation and brain rot and propaganda and child predators and you get the idea.
The addiction is the big one, the root of all evil. “Maximizing engagement” is destroying civilization. But I digress.
Back in the 90s, nobody talked about this. Well not nobody but maybe this one person pissing into the hurricane. Discourse on the net and computing in general was 99.99% breathlessly positive bordering on delusional utopian.
We didn’t see it coming.
Nobody was talking about how this could go off the rails and do horrible things. Well almost nobody. I guess dystopian cyberpunk nailed some of it but it was too much in the realm of fiction. It wasn’t part of the real world discourse which was all Mondo 2000 and early Wired and open source movement and PC industry optimism.
So ironically all the anti AI discourse makes me cautiously more optimistic about AI. We are talking about the risks, the downsides, the ways it could go majorly bad. Maybe we will notice if it does and correct. If we’d done that with the net, noticed that maybe addictionmaxxing apps and sites was going to work “too well” or that flattening the discourse to 140 characters would harm the collective mind, things might be a little better today.
Even if we go in with one eye a little open and correct a little bit, that could make a big difference. Could have with the net too.
Why? AI is enriching my life and letting me focus on solving hard problems. I don't believe in the sentience crap and am just using it at face value for what it is, a useful productivity tool. Every software developer's main purpose is to automate and reduce the amount of boring manual labour required. By this logic, should people in the past have "resisted the software development department"?
For the pet projects, it’s not really a big deal. I’m the only user, and I as the user don’t care how it works under the hood.
Work is a different story. I have never felt more disconnected from my work than in the last 6-8 months. It takes a huge mental toll to read into our code and figure stuff out, almost like I just got hired and I’m working on new code for the first time. I’m asked to read “specs” written by LLMs that are unintelligible and I find my brain disconnecting immediately. Tickets, also written by LLMs pose the same problem.
I’m asking teammates questions because I want their answer and I’m told to just ask Claude.
None of this feels good, it may be where the industry is heading but as a professional it feels wrong.
PRs review more quickly when you have human written spec diffs to orient you and guide (or skip) the code review.
a danger to avoid is writing specs which are basically just really bad code (prose is terrible code). you can work towards patterns to avoid this pitfall.
also, dont coflate spec with plan. spec is durable decisions. plan is execution to turn decisions into code. plan is disposable, dont check it in.
either way, find the human layer and dont use LLMs there. make all the important decisions in the spec, delegate the other decisions to the LLM.
i believe this is the new craft.
that said, i expect a small but significant exodus from software. a lotta folks came for the "programming" that is now being outsourced. were becoming so tired, just absolutely burning out.
i hope we will eventually realize its time to go, instead of giving up our souls for a fucking paycheck.
Code reviews come back with a ream of issues that I just want to scream "I don't give a fuck".
Pre-LLM no-one would have cared that if another class was configured differently ( it isn't ) and called the new class in a different way ( it doesn't ), that you'd in some race-condition occasionally leak a resource, but only on shutdown when the program is about to terminate anyway.
That's the kind of issue that claude loves to find and write up in a code review with the exact same verbiage and write up style as if it had just found a naive SQL injection from string concatenation.
Prior to LLMs I doubt anyone would even have spotted the first issue, let alone bothered to write it up.
But instead it's another round of review, another round of trying to decipher the LLM's comments to discover whether it's a meaningful problem or a nothingburger.
And it's all so tiring.
If somebody told that humans are a form of creature which can forego their most important and powerful capabilities just for convenience, I'd call the person saying that batshit insane, but here we are...
I have no more words for this.
also, take some fucking pride in your work.
All of these arguments presuppose we're losing something that's not worth losing.
I want someone to start one square earlier and argue to me why that matters. I've read enough of these "your skills are atrophying!" arguments; I'm also making like 10x as much revenue as I did through a business that delights my clients. Why should I care about my "skills" if I'm still able to make money, pursue hobbies, and enjoy life?
I just don't think it's that serious.
From what I have seen AI doesn't understand the code. But even if you argue that it does understand the code it definitely doesn't understand the domain because it's only input channel is a stream of text and you didn't even type in a "full" text representation of the domain.
So you remain the only thing that can understand the whole system. If your competitor has no one (or no team) who can put together a full understanding of the system then you will have an advantage.
Id wager that they would fare the same. The time saved on some tasks are outweighed by extra time spent on others. Everyone assumes that just because you can pump out more features, you're more successful.
workers that feel powerless and dont try to resist is great news for rulers who only serve to generate revenue.
Additionally, when it comes to the software field, it will be sure way of selecting yourself out of the work pool.
https://www.youtube.com/watch?v=--bH6O2_Vvk
If, as an engineer, you rely completely on AI then you will be selected out of the work pool. Some engineers will still be needed. If I was interviewing them (I used to do a lot of this) then I will want the ones who understand the code. If they only use AI then it will quickly become clear that they don't understand it. Why would I hire them when another person can understand it?
i'm ok with that
Copyright exists to encourage progress in the arts and sciences. Not to enrich artists or scientists.
Ai is progress. Further, the works it creates are fully transformative. It isn't just spitting out Harry Potter verbatim, it's making new things from it.
To me, this argument is like staying that if you read Winnie The Pooh as a child then Milne owns your artistic output for life, because one of your many influences might have been his work. It just seems obvious at face value.
Is impoverishing them supposed to encourage progress in the arts and sciences?
Whatever it started out as (and I have my doubts it was ever a tool to help artists), it's now primarily a tool corporations use to milk creativity for profit and squash it where it might interfere with that profit.
No, but that needs new laws. Not an abuse of the old ones justified under the thinnest of pretenses.
I don't think that folks stop to consider what bending copyright law into these weird shapes to meet their ends will do to other aspects of copyright.
For example, sampling in music. How is that different from AI generated music? If anything AI uses LESS of any individual song. You can't just say "no derivative works." That would fully halt progress on many fronts.
That said, I also think everyone who contributed content online deserves to be rewarded for their contribution, especially since OpenAI (which made the largest contribution) was a non-profit in the beginning. So there should be some kind of tax. Otherwise I agree, it's not ethical. It's not 'fair use' of copyrights.
Because it was trained unethically we have a few moral obligations:
1) USE IT. Do not let only unethical people use a force amplifier on our collective efforts.
2) STOP PUBLISHING HUMAN GENERATED CONTENT. The more original human thought = the more they harvest our thinking.
3) Generate only known things in novel combinations. Do not provide new insights.
I could go on but strategically thinking AI is not going to be stopped in the way you are thinking. So you need to think a bit deeper on methods of optimizing the situation.
CMIIW
Copyright violation is not theft
legally i have no fucking clue, but morally, this is stealing humanity to produce a profit.
realistically the toothpaste is outta the tube, so the second best option is accept LLMs, but do not enrich AI companies.
That’s not to say that AI is all fine and dandy, I genuinely think frontier AI companies have deranged leadership.
Earth is two thirds water, and it rains. The water I drank this morning is million of years old. It will be around for millions more.
So, they pump some water out of the earth and it evaporates... It will come back down, and we aren't running out anytime soon.
Why do we care? Are they building these things in deserts? If so that's the problem. They should use closed loops if they want to build in the desert.
This your your typical HN take, aka "I thought about a random problem I barely understand for 12 seconds and I think I figured it out better than domain experts"
The 1/3rd that choose to live in arid places where droughts are a concern should probably just pass laws that require closed loop cooling.
Some people just live in places that aren't suitable for human life and survive only through technology. Those places need to have different laws about water.
For example, I shouldn't be forced to use a low flow showerhead when I'm 100 miles from 1.3 quadrillion gallons of fresh water. If you aren't next door to lake Michigan then maybe you should be.
They obviously shouldn't be doing it like that, but since clean water is cheap and business in absence of reason optimizes for lowest costs that's exactly what they did.
I have a suspicion that people who reject AI related stuff is just scared under the hood. They fight against AI because of the fear against loosing job, money, or control.
There were many manual jobs. Lets say carpenters. They had manually created things, often pieces of art. Now many things are soulless, created by factory output, and by robots. It is true that we lost some artistry. It is true that every line of code will not be a work of art. There is no going back though. Adapt.
Oh yes, buddy, those stupid people should better act out of love, not fear, right?
> There is no going back though. Adapt.
“He who would live must fight, and he who does not want to fight in this world, where permanent struggle is the law of life, has no right to exist.”
- Adolf Hitler, Mein Kampf, 1925.
Oi. Discussion on this site gets more difficult each day. Instead of pointing out where you disagreed you just pulled out the most egregious appeal to authority and said, "oH yEaH?"
It's as if all man-made change was placed in the same ballpark as tectonic movements and extinction events - as a blind force of nature.
Humans are given intelligence not only to construct things, but also not to construct things that better be not, FFS.
If Mao had said it, it would've been "absolutely communist", if Washington had said it, it would've been "absolutely Patriotic".
> I do not come to you with clean hands. > I am not here to dictate morality
These two things contradict eachother.
> My primary focus here will be software development > Environmental Harm > Stanley Tucci
These three things also contradict eachother.
> If we lived in a post-scarcity
Maybe this is your real problem, and you're just confusing cause and effect.
> We need to resist “AI” so that thousands, perhaps millions of people do not die
Hyperbole.
> Look up the “Eliza effect”. Even the weakest conceivable program like this was seductive. These new ones are masters of seduction.
Sounds like software development alright.
This person has a familiar name, I'm sure he is amazing at something. This blog is not the peak of his work.
EDIT:
> I quote “AI” to emphasize that these things are not intelligent.
https://en.wikipedia.org/wiki/AI_effect
I was recently discussing with someone who wanted to put all datacenters into AI datacenter basket, and inadvertently labeled all consolidated computing bad. That person, claiming otherwise was not knowing what they were talking about,
But while having nuances, AI datacenters and heavy AI usage is harmful to the world we live in, as a whole. We should come into terms with this and go from there.
1. Reducing/replacing/automating highly complex work is fundamentally a good thing, it might be the case that society is not ready for the consequences of it, those are two separate things. Concluding that millions of people will die, is simply hyperbole.
2. Saying that AI is not intelligent is at best a non-sequitur but more likely intended misrepresentation and goalpost moving
3. The environmental harm of AI, however significant, is ultimately public knowledge and 100% clear in the public discourse around the current implementation of AI. This says nothing about AI in the general sense, only how we produce it today. This is like preventing cars from becoming a thing because they produce CO2. Do you think the world would be better off if cars were never invented?
None of this is about the 'artistic properties of their words', it is precisely about the core content being objectionable.
I don't think saying that millions will die is a hyperbole. It might be communicated poorly, but having no ability to work in a world with worsening quality of life (via climate crisis and economic changes) is being pinched between a rock and a hard place. What these people gonna do? Photosynthesis? One of countries which is among the leaders of this technology believes that social governmental systems is communism. So, yeah, perfect.
> The environmental harm of AI, however significant, is ultimately public knowledge and 100% clear in the public discourse around the current implementation of AI.
Yet the people who own and hype this technology is applying the Tobacco/Soda Industry Playbook: Deny and publish studies saying otherwise.
This is not fair, because who has the gold, makes the rules to increase their profit even more.
> This is like preventing cars from becoming a thing because they produce CO2. Do you think the world would be better off without combustion engines?
My current stance is not undoing things to zero, however forcing AI development to be more environmentally conscious and start regulating its power and water use. This is possible by prescribing energy source mix ratios, closed loop water cooling, etc. etc.
Of course this will slow some progress and make it a bit more expensive for a while, but in the end it'll be way more environmentally friendly.
All discourse doesn't have to be on a duotone plane. We can discuss in a 3D space with infinitely more shades.
If unemployment leads to death, there are many more things wrong than just AI. Why get mad at AI instead of at capitalism. This is nonsensical.
> Tobacco/Soda Industry Playbook
Except both of those succeeded in destroying health for decades, whereas AI had like a 2 year run before all the datacenter madness came to light.
> My current stance is not undoing things to zero,
This does not at all answer my question, my question is about the past not the present. The author is arguing to resist AI, what if resisting cars had worked? This is my question.
But I do think that for the atypical user that is "information-hungry," AIs are a massive productivity booster. I've probably asked ChatGPT 1,000 questions over the last 2 years that I otherwise would have needed to spend 15+ minutes researching. Prior to AI, I would have just... not spent the time looking them up.
The danger is just when you start using AI but otherwise continue doing exactly what you were doing before.
You mean letting AI write the code, but still trying to write the same code by hand?
Joke aside, this is still what I am struggling with. I don't really code anymore - not worth it, when I want something done, but I miss it.
I guess the analogy could be: with AI you should be doing more complex and more (quantity) programming than you did before. But don't quote me.
Programming is mostly exactly that. You gather want information you need to get the job done. What are the names of the API you call, what are good patterns for this kind of situation etc.
But your point still stands as in use of AI can make those skills detoriate. And regarding to coding, that you don't know anymore what is in your codebase, how it actually works. Can happen fast if you let AI run in auto mode.
Our devs have been outright complaining about getting dumb, losing touch with the codebase and getting lost in weekly switching between models/agents that have been gushing as if from a firehose, but the management (read investors) knows better and there's a culture of self-censoring the opposition to AI among employees.
I feel compelled to point out that Kagi does offer AI-backed answers if you want it to. Just pose your query with a question mark to trigger the AI response in addition to the usual search results. It is not ever the default or only answer, and I pray Kagi keeps it that way!
I use the AI option more often than I thought I would, usually for things that I don't consider controversial or complicated.
But I think it's important for the quality of public discourse that we forward only the best, strongest arguments for any given position (whether I personally agree with that position or not). A weak argument gives unnecessary purchase to unscrupulous haters who might use that weaker argument as a low-grade "straw man" to dismiss the whole position.
So with that preamble, some constructive criticism (very well-intentioned):
> I am not exaggerating: We need to resist “AI” so that thousands, perhaps millions of people do not die.
On this one you are, in fact, exaggerating (certainly with "millions," likely with "thousands" as well).
The overwhelming majority of people who lose their job for whatever reason (and there are many reasons, not just this one issue) fall back on a support network of their family and friends while they "get back on their feet" by pivoting to another job or form of employment (sometimes branching into entrepreneurship, for example).
The alternative to employment at whatever FAANG or FAANG-adjacent company someone was laid off from is not "death." It is typically "some, sometimes quite significant, discomfort." Granted "AI Layoffs" are starting to ripple out beyond FAANG, but the story is similar in other industries.
If "layoffs kill people" is the moral hill we choose to plant our flag on, then mustn't we also campaign with equal vociferousness against all the other causes of layoffs in general? Since to do otherwise is to sit by while those laid off allegedly die in droves of thousands or perhaps millions?
It also occurs to me that this framing ("AI causes layoffs") also gives incompetent and unaccountable executives a pass by pinning the actual harm done by irresponsible layoffs (which to be clear, is not none!) on "the AI" rather than on the leaders who are in the positions to actually decide what happens in their organizations.
I think the argument is much stronger if we talk about the aggregate cost to humanity being paid in terms of displacement, life disruption, and curtailed opportunity (especially for juniors). We should not make the "toxic escalationism" move of equivocating things we don't like with negligent homicide, even if it makes for a punchy one-liner.
Excellent point. Thank you for writing such a cogent argument.
It is clear that the rich and powerful are making short-term decisions that are causing widespread and long-term harm. Decisions such as, AI-pilled investors demanding unsustainable productivity from companies, and CEOs in turn forcing employees to use agentic coding and laying the rest off.
Such decision-making is not new, of course, but LLM technology is helping grifters and con artists to turbo-charge it. Founders can raise money using cool demos that achieve the impossible. Investment analysts can back up their AI story with huge paper profits.
The result, as you summarized, is life and career displacement and missed opportunities for young people. It's just too abstract and diffuse of an impact to influence people with, and it also seems to be railing against the holy cow of Capitalism.
"Data centers bad!", or "AI is disrupting humanity and the end is nigh!" are much more punchy slogans.
Oh man how did we ever survive calculators.
But we as individual consumers have more power than we think. We don't need to use these AI features and products. We've survived perfectly fine without AI. Made plenty of scientific discoveries and lead enriched lives without it.
The biggest "fuck you" to the pricks who pushed this on us would be if we collectively went "meh".
AI are here, is it going to demolish and (maybe) reconstruct the world as we know it, trying to avoid it it's just silly. There is no sane world where we can "stop AI", even if every single researcher were to do vanish tomorrow and the hardware and schematics magically gone, it would still only be a stopgap.
So...What are you trying to achieve?
Sure we can't stop the technology, but we _can_ stop the importance and money heaped into it.
Listening to developers complain about AI doing things to their life and the world at large when they have spent the past 30 years automating just about everyone else they could out of a job (Since this has been one of the primary uses of software) is ... something else. It sounds like a a a mix of hubris, arrogance, cognitive dissonance, ignorance, etc.
Only when it started to affect their world did any developers start to care, and even then nobody seems to ever stop and say "hey what were we doing the past 30 years, maybe we should think about this", it's instead "hey look at what these other people are doing with AI, someone should think about this".
It feels like, to paraphrase a movie quote - "Nobody gets to destroy jobs and the the world but us"
In fact, we already pretty much have proof of this: devs didn't care when software destroyed the journalism industry twenty years ago. Turns out that people ultimately only care when their only jobs are on the line.
But what concerns me is that authors of articles such as this feel no need to differentiate themselves from other anti-technology arguments in history.
There have been many other thinking technologies and they ALL have been attacked in exactly the same way.
* Socrates attacked the invention of writing, arguing it would make people intellectually lazy.
* Abbot Johannes Trithemius argues in "In Praise of Scribes" in the 15th century that the Gutenberg machine would make people intellectually lazy.
* Lots of people in the 1960s attacked punch card computers, and automation in general (!), on the basis that it would make people intellectually lazy.
* Nicholas Carr attacks Google in a famous article "Is Google Making Us Stupid?", on the basis that it would make people intellectually lazy.
And so on and so on and so on.
"That which has been is what will be, And there is nothing new under the sun."
It is because of this history that any argument such as the author makes, arguing that AI would make people intellectually lazy, must start already from the lamentation of the downfall of human intelligence from the days of Socrates, or must explain why this is different.
It is obviously not different.
There are three mental skills being offloaded onto AI:
1. Domain-specific skills
2. General intelligence (g)
3. Recall
Of these three, the danger with 1 is offloading them too early, before the AI is "good enough" to be better than human experts. This is a genuine concern and we should do things carefully. However, if the AI is good enough, there is no danger here.
Recall is the mental skill that all of these innovations have "hurt", but it can simply be trained explicitly and synthetically. We will certainly have to structure our societies more and more in the future to synthetically exercising recall (like we're now synthetically exercising our muscles), but if we do this, offloading recall at work is not a big issue.
As for general intelligence, this is the most harmless to offload, because, as far as I know from the literature, it cannot really be trained, and it does not really degrade either.
A world without steam engines, cars, computers, printing presses.
The argument usually isn't "I believe X will not happen," it's usually either "I believe X should not happen" or "I believe we should manage how X happens".
Why should they? Other "anti-tech" arguments include
- Concern about global warming/pollution
- Resistance to nuclear/chemical warfare
It's not just, or even primarily, that anti-tech arguments are the equivalent of 'cameras will steal souls'.
AI companies are racing to build empires on products built on the whole of human writing, charging rent to access it, then leveraging their scale to improve the model.
the big data era was a gold rush for greed. the AI era is has upgraded to a warp engine for greed.
the harms and manipulation of social media took ages to enter the social discourse, and are just now turning into legal action.
AIs warp-engine speed guarantees greed will fuel warp-engine scale harm before society catches up, and before America's greedy ass politicians are cornered into trying to do something.
resist fucking greed. if you gotta use LLMs, at least don't pay an AI company.
For some reason HN readers love this sort of content, largely by periphery players who mostly think it gives them an angle for clicks. And, I mean, as someone who uses lots of AI to delight clients and stay competitive, you be you. It leaves more for me.
At the same time, maybe I shouldn’t, and I sometimes find all the AI skeptic and even doomer discourse heartening.
We really messed up with the Internet. I’m an Xennial and I had a minor role in it but I was there and it feels like the Internet was “our project.” I was literally born with the PC, discovered BBSes as a tween and the pre web Internet as a teen.
Don’t get me wrong. I’m not saying the net is bad. I think it’s more positive than not (still). But my god the addiction and disinformation and brain rot and propaganda and child predators and you get the idea.
The addiction is the big one, the root of all evil. “Maximizing engagement” is destroying civilization. But I digress.
Back in the 90s, nobody talked about this. Well not nobody but maybe this one person pissing into the hurricane. Discourse on the net and computing in general was 99.99% breathlessly positive bordering on delusional utopian.
We didn’t see it coming.
Nobody was talking about how this could go off the rails and do horrible things. Well almost nobody. I guess dystopian cyberpunk nailed some of it but it was too much in the realm of fiction. It wasn’t part of the real world discourse which was all Mondo 2000 and early Wired and open source movement and PC industry optimism.
So ironically all the anti AI discourse makes me cautiously more optimistic about AI. We are talking about the risks, the downsides, the ways it could go majorly bad. Maybe we will notice if it does and correct. If we’d done that with the net, noticed that maybe addictionmaxxing apps and sites was going to work “too well” or that flattening the discourse to 140 characters would harm the collective mind, things might be a little better today.
Even if we go in with one eye a little open and correct a little bit, that could make a big difference. Could have with the net too.