Rendered at 07:43:05 GMT+0000 (Coordinated Universal Time) with Cloudflare Workers.
socalgal2 6 hours ago [-]
> We are 4 years into this ‘revolution’ and the industry has almost nothing to show for it. Software has not improved in quality,
Apple, Mozillia, and Firefox just released a record number of bug fixes because of AI. So this assertion that Software as not improved in quality is arguably if not provably false. There are 1000s of other examples. I have my own. Personal projects that were stopped because of dependency rot, having 100s of bugs fixed, nearly all dependencies removed, and lots of tests added just by asking. I know tons of others with similar experiences
> I do not feel like I’m falling behind by not using them
Obviously you are correct about "how you feel". But, this argument to me is more like a horse owner claming they don't feel like they're failling behind to trucks that can carry 10x, 100x, 1000x more than their horse. Sure, maybe the horse can go places the truck can't. For most uses the truck is better than the horse.
> The LLM-generated PRs I get are still rubbish.
I can't argue with that. I haven't seed the LLM-generated PRs "you" get. I can say though the LLM-generated bug reports I've seen are 1000x better than human ones. Human bug report "your software breaks, fix it!". LLM bug report "There is a bug in this specfic part of the code for these specific reasons and here's a repo I wrote to demonstrate the bug. Here's instructions on how to run the repo. There's the code to fix the bug. And, here's a test to show the bug is fixed."
Is it perfect? No, sometimes the bug isn't real because the code can only be reached in situtations that can never happen. But still, 7 of 8 times it's a real bug and the bug report are the best I've ever received.
>For most uses the truck is better than the horse.
I think you're talking past the author here, who says:
>By not leaning into LLMs I’ve given myself the room to specialise in a smaller set of skills. So far that conscious decision is working out well, and I remain confident about my job security.
"Most uses" might mean boilerplate and simple frontend work that LLMs can do easily because it's formulaic and doesn't require much creativity. OP is simply not doing that type of work.
I don't think anyone sane is really claiming 100x or 1000x speedups anymore. Most people I talk to who use LLMs are closer to 1.5 or 2. Because their job is not mostly boilerplate, there is simply not that much room to be improved.
bensyverson 5 hours ago [-]
It’s not about speed-ups, it’s the fact that I’ll attempt a project far more ambitious than I would have in the past.
Over the past two months I’ve built out a very complex web app for a client. Without the aid of AI it would be far smaller, simpler and less capable. And honestly I probably wouldn’t have agreed to take on the job in the first place.
5 hours ago [-]
gr_norm 6 hours ago [-]
LLMs are very helpful as a debugging aid, yes, but in large part because the fixes tend to be small and verifiable. That this does not carry over to many other use cases is the crux of the problem.
I myself use them to accelerate programming tasks, so I'm not anywhere near as pessimistic as the author, but the claimed multiples of productivity definitely haven't materialized for me.
qsera 6 hours ago [-]
>Apple, Mozillia, and Firefox just released a record number of bug fixes because of AI.
A hidden metric here is the number of new bugs created by these fixes. If an LLM creates 10x bugs but create 5x more new bugs, then it is not really an improvement. Because that is 5x more bugs that the user has to observe and report (unless it is a security vulnerability, which the LLMs might detect) before it can be fixed.
socalgal2 5 hours ago [-]
This is effectively a claim that Mozillia, Apple, Google's developers are idiots and irresponsible and just accepted the fixes without checking them. That's a pretty strong claim which I think you'd have a hard time proving or even finding any evidence for
qsera 39 minutes ago [-]
I mean no such claim. One does not have to be an idiot or irresponsible to overlook an unintended consequence of some code change in a very large, legacy code base.
skydhash 5 hours ago [-]
Also there's the "let's make time to work on those bugs by using this new tool". Time that wouldn't have been allocated otherwise.
iammrpayments 5 hours ago [-]
Not the 1000x more productive numbers again please. Is this a meme or something? If you said +17% I might have taken you seriously.
socalgal2 5 hours ago [-]
I gave the 1000x example because a single semi-truck can generally carry 240x what a horse can carry. A horse can carry 150-200lbs. A semi-truck can carry 42,000 to 48,000 pounds. The largest truck carries 2.2 million pounds or 11000x the horse.
To see an example of this 1000x, all we have to do is look at the flood of bug reports and fixes to browsers and OSes. Does it do 1000x for every project? No, and neither does a pickup truck do 1000x what a horse does. But, even a small pickup truck does more than a horse for most use cases (carrying people or stuff) and similarly, LLMs do more for most use cases.
wewewedxfgdf 5 hours ago [-]
>> We are 4 years into this ‘revolution’ and the industry has almost nothing to show for it.
If you actually used LLMs instead of preaching against them from the sidelines with no real experience, you'd know such statements are wrong.
This is the real problem with many of the anti LLM folks - they tell us about LLMs and also tell us they don't use them.
"That thing cannot do the stuff you say it can. And there is no way I will ever try it. But I know for sure I am right."
lunar_mycroft 17 minutes ago [-]
> If you actually used LLMs instead of preaching against them from the sidelines with no real experience, you'd know such statements are wrong.
What about the people who do have a lot of LLM experience and agree with OP?
> That thing cannot do the stuff you say it can. And there is no way I will ever try it. But I know for sure I am right.
If someone claims to be able to fly by strapping bird wing shaped pieces of plywood with feathers glued on to their arms, do you need to personally jump off a tower with them to say they don't work, or can you look at the results of others attempting it and draw conclusions based on that? LLM proponents are making claims about their capabilities which can relatively easily be checked without using LLMs yourself. To pick an example where LLM proponents are correct, anyone who says LLMs can't generate syntactically valid code can be proven wrong fairly easily by producing an example of syntactically valid, LLM generated code.
Further, if you read the rest of the paragraph you responded to, it's clear that this claim is about the state of the industry as a whole. "Software has not improved in quality, got faster, become cheaper to produce (when you exclude the mountain of poor-quality demoware that no reputable organisation would touch with a barge pole), or become more capable." Whether this is true or not is something that can be evaluated without ever having prompted yourself, or even arguably without being a developer at all.
m463 2 hours ago [-]
I see this pattern in a lot of things.
- people who preach against PV solar... while people I actually know broke even 10 or more years ago and have been happily getting "free power" for a lot of their needs
- people who preach against electric cars... while people I actually know are driving them with no "range anxiety" or "replaced main batteries", but their non-engineer spouse/kids are driving the cars fine too.
- people who preach that tesla self driving isn't viable/etc... while lots of people are using tesla to drive themselves around for > 90% of their driving
1 hours ago [-]
4 hours ago [-]
quadrifoliate 4 hours ago [-]
> If you actually used LLMs instead of preaching against them from the sidelines with no real experience, you'd know such statements are wrong.
Unlike the author, I'm an LLM optimist; but this take doesn't make sense to me. The thing you show cannot be "hey look at this LLM-based workflow, that's so much better than last year".
Where is the big browser, operating system, or other piece of work that has been able to quickly compete with the existing entries in the market due to LLMs?
The only one that I can think of off the top of my head is the rewrite of Bun in Rust. I think the fact that Claude Code runs on millions of devices on top of that rewrite is the sort of impressive demonstration that would prove that LLMs have lasting, systemic effects -- but it is telling that the only one I can think of was made by the extremely well funded lab that also happens to build the LLM.
3 hours ago [-]
wewewedxfgdf 4 hours ago [-]
And when you show them evidence of the incredible things that can be done they stick their fingers in their ears and say "I'm not listening. I'm not listening. looping ad infinitum. OR they start saying "Oh yeah, prove it with hard numbers." when nothing about software productivity has ever been able to be proven with hard numbers. It becomes a semantic word game that you cannot win because nothing you say will get them to believe because they don't want to believe. The LLM haters are the most dogmatic of unbelievers.
Which I find weird - the greatest fun for me is making a computer do something awesome. I learned to program because I want that outcome.
I respect if you love the art and craft of hand programming but it's something else entirely to simply deny that LLM's are what they are - which is to say so amazing it is really beyond belief and they are getting more so every year.
If you're a hand programmer forever then fine, just don't make a fool of yourself by saying that everyone is lying about LLMs, to justify your choices.
glimshe 4 hours ago [-]
> which is to say so amazing it is really beyond belief and they are getting more so every year.
We've normalized science fiction. Go back 5-10 years and ask anyone whether we'd soon have something like LLMs.
If you step back and try to look at LLMs without emotions and politics, you will be blown away by what this thing can do. We're talking to a machine, folks.
I just can't believe that LLM haters really like computers or technology, because we haven't seen anything this interesting in decades.
card_zero 1 hours ago [-]
You prompted me to have a little sci-fi fantasy about a world where LLMs are exciting. It goes like this: imagine they were local-only, from the the start, and this was baked into the whole concept because each one has an owner, and it is trained on that owner's mind, in an intimate and slightly trippy process. None of its background information comes from any other source. It's possible to talk to somebody else's LLM, but this is rarely done because it would be awkward and intrusive, and difficult to interpret. Consequently, nobody ever dreams of connecting an LLM with any server. There is no giant ugly industry for training them. They are personal in every way and have nothing to do with the internet or any of its assholes, and do not represent some bland and inscrutable centralized source of truth. Their output always goes to the owner, is hard for other people to understand, and does not provide an opportunity to spam everybody else with vapid boilerplate. Now, I'm excited to own one of these!
rayiner 2 hours ago [-]
> I just can't believe that LLM haters really like computers or technology, because we haven't seen anything this interesting in decades.
Yup. As someone who thinks computing has been in decline since Mac OS X 10.4, LLMs are the first major tech trend that I've been impressed by in 20 years.
quadrifoliate 4 hours ago [-]
> We're talking to a machine, folks.
And? I didn't sign up to talk to a machine.
Has it actually made your daily life better? All I see as a result of LLMs is endless slop everywhere, plateauing in the skill levels of coworkers, and badly informed people in management positions pushing the humans to meet impossible deadlines because "otherwise we'll replace you with a cheaper worker and an LLM".
I am an LLM optimist, but I'll be honest and say it hasn't made my life better. My optimism relies on the models plateauing, and people finally rebelling against the widespread slop to realize that LLMs are just another (very useful!) tool, but don't meaningfully replace people in any capacity.
wewewedxfgdf 1 hours ago [-]
Twice now you have said "I am an LLM optimist" and then spoken as an LLM pessimist.
vidarh 3 hours ago [-]
It has made my daily life better. It has built software for me that I've wanted but previously didn't have time for, for example.
quadrifoliate 2 hours ago [-]
Great! How does this software enhance your daily life? Has it given you more leisure time by leading to less work?
fatata123 2 hours ago [-]
[dead]
ilteris 3 hours ago [-]
What did you sign up for and what would make your life better in this context? Give us a north star
bigstrat2003 4 hours ago [-]
> If you actually used LLMs instead of preaching against them from the sidelines with no real experience, you'd know such statements are wrong.
I have. Which is how I know that such statements are correct.
andai 6 hours ago [-]
>productivity
I recently experienced some cognitive decline due to chronic health issues. This is unfortunate timing because I am shipping a game with online multiplayer for the first time. So "brain works properly" is kind of a basic job requirement! But you work with what you got...
Some day I have the energy to do programming "manually", but on many days, the difference between "tasks get done" and "I am just staring at the screen reading the same code over and over again" is "ask the LLM to do it."
So I have had to shift my identity from "the guy who does the thing" to "the guy who ensures the thing has been done."
That being said, I do check the diffs carefully. (I tried a more hands-off approach a few months ago, and that was a pretty bad idea.) And I've learned to make the diffs as small as possible: more digestible for me, and less room for the AI to go off the rails.
Nice bonus to small, well defined changes is that small, fast, cheap models can handle them. (Having a great time with Luna lately, even on Low reasoning effort, which I wasn't expecting at all.)
microflash 39 minutes ago [-]
My story is a bit similar. Recently, I broke my wrist due to a fall and have also been suffering from eye strain. I couldn’t spend much time “working” on screen as much as I need. And so I ask LLMs to do a major chunk of work, review and test the changes, course correct things to align with architecture that I’ve outlined and get things done as if I were working normally.
vogelke 6 hours ago [-]
By checking the diffs carefully, you're showing how to use an LLM to add value instead of creating a 17000-line PR that nobody with any sense will read.
LLMs are really good at making it harder to ignore ignorant people, and you've demonstrated that you're NOT part of the problem.
hax0ron3 8 hours ago [-]
>The not-so-quiet goal of the LLM push by big tech was to homogenise intellectual labour such that it can be made fungible. When labour is fungible you can swap it out whenever you like for an alternative, decreasing the bargaining power of workers. Software development was one of the last ‘guild fields’ and now we are seeing its industrialisation occur.
I think that the industrialization of software development already happened decades ago. The common professional software development methodologies, both waterfall and sprint-based, ARE the industrialization of software development. Isn't modern "agile", sprint-based software development methodology directly influenced by Japanese manufacturing methods? I don't know when software development actually was a guild field, and my personal experience doesn't reach back this far, but based on what I've read at least, it seems to me that even back in the 1990s it was already not a guild field.
jerf 7 hours ago [-]
A "guild field" is not a well-defined matter, but I think it's still the easiest engineering field to sit at home on your personal gear and learn skills that can get you hired in a real job with a future without attending any professional school for the field.
In fact the four-year degree you'll get from school is getting increasingly distant from the skills I actually want out of a new grad. It's not impossible to bridge the gap or anything but my transition into the commercial realm in the early 2000s was a cakewalk compared to the sheer number of things I'm asking a new grad to learn as soon as they're settled in at their desk... source control, CI/CD, bug trackers, devops, and that's just the beginning of that list not the end.
fwip 7 hours ago [-]
New grads have learnt git and issue trackers as part of standard coursework. Not sure about CI or devops.
skydhash 6 hours ago [-]
> In fact the four-year degree you'll get from school is getting increasingly distant from the skills I actually want out of a new grad
Those skills are easily taught. But in every run of the mill project I’ve been on, I’ve been really happy about all the maths and other science stuff I’ve been taught at college. It’s easier to grasp the web DOM, React, and git when you already know about trees and the related algorithms. Easier to learn SQL after opening some books on relational theory (basically the first chapter on database theory) and learning about projections. And definitely easier to learn the nature of distributed systems.
Learning how to use Bash and Ansible is a walk in the park after that.
0x696C6961 6 hours ago [-]
There's a big difference between understanding the underlying concepts vs being proficient with a tool.
jgord 6 hours ago [-]
I personally look at react-foo and see something thats the opposite of how I want to work as a software craftsman [ small code, lispy / function flavored style, opinionated, few dependencies ]
For similar aesthetic reasons, have only dabbled with code-gen LLMs. Im happy to vibe code css and html but not, well actual code. I'm aware of my own double standard, Im fine with median-reverting banal css and html.
I do worry about some aspects of LLMs :
- google giving AI code snippets so easily, prevents traffic to sites like stack-overflow, where discussion happens
- vibe coding means less people hit the actual hard parts of coding, which result in learning to think as a developer. Thence where the next generation of developers who understand code ?
- code will become a magical, inscrutable black box where no-one understands how it works. DNNs and LLMs have this feature already. At least with LLM discovered math, the proof in lean exists, which is independent at how it was arrived at / searched for.
- if I write a novel, and the DNN AI detects it as AI slop with 68% probability, then what recourse do I have to prove its human written ?
- circular training : subsequent rounds of LLMs trained not on human input, but on LLM slop of the previous generation [ the %ge of which increases ]
- you will own nothing : fortunately it looks like self-hosted open weight LLMS will remain a thing, but the massive spend by hyperscaler companies on Datacenters/GPUs/RAM has driven down supply and driven up prices of CPU, GPU, RAM for enthusiasts / gamers / home AI nerds / garage startups .. which might stifle innovation.
Grombobulous 8 hours ago [-]
The last bullet point is the most important. The industry is most hyped about AI so that they can pay software developers $70k instead of paying them $200k.
I don’t really agree that AI can’t make development faster, though. What the author describes as the negative AI outcome is blind vibecoding. There are many other ways to use it, and even the basic “enhanced autocomplete” is a net benefit, especially since that functionality is dirt cheap or free.
7 hours ago [-]
pydry 7 hours ago [-]
Not really. The other bullet points are the reality. The last one is the MBA wet dream.
We have been programmed to believe that MBA wet dreams are inevitable because "$x hundreds of billions invested can't be wrong" but they very often are.
classictraffic 10 hours ago [-]
To the author’s point about there not being enough evidence to support pro-AI claims — Additionally, there are studies and articles coming out showing those people are wrong and AI only makes people barely more efficient. Not the 10x tech CEOs are claiming while slashing thousands of jobs
The problem is that LLMs improve so fast, your article may very well be out of date by now.
jerf 7 hours ago [-]
I'm definitely able to spew out more code than I used to. Good code. Maybe not great code, but good code.
But I spent a lot of time in the second half of this week dealing with friction with a team that is very annoyed that I'm moving fast and using an agile methodology so I can't tell them the exact, precise REST calls that I'm going to have for them in six months designed to a tee and signed off in triplicate before they start development against it. Manifesting that increase in code production as real value to the business is going to take more from me than just spewing the code out more quickly.
AI isn't creating this problem. I would have had this problem anyhow even if I were writing all the code by hand again. I know, because I've been there before. But the increased velocity is manifesting in increased organizational stress and not just increased velocity.
AI is perhaps even helping solve it to some degree, though far from totally. I have written before about how people eventually learned not to play the "oh well we can't do this until we have documentation" card on me [1]. This week they played the "well, I see you have docs but they aren't in our precise format". Guess what AI can do in about 15 minutes really well? You may recall the term "style transfer" getting tossed about a lot 3-4 years ago, and it is still something AI is extremely good at, and "take these docs in this format and convert them to that format" is just a style-transfer problem. AI really does chew at the "oh but we need docs" old-school card... and they can't even complain about the quality of the AI docs because in order to do that, they'd have to actually read them, and that is not the point of the "but we need docs" card, you see....
Yeah, as you said, this isn't an AI problem. The interface between an agile team and the non-agile part of the company was a problem over 20 years ago. (At least, that was when it bit me.)
You can think of it as an impedance mismatch, or as a translation problem. But no matter how you think about it, it's real, and it's a problem - especially if upper management lives on the non-agile side of the fence.
E-Reverance 9 hours ago [-]
(March 2026)
spl757 6 hours ago [-]
All AI hallucinates, and therefore is unpredictable and unreliable. 1.5 trillion dollars to prove the obvious. It has no place in production use. IMHO at least. They want so badly to get rid of all of us, and replace us with something that doesn't complain, doesn't have kids, or outside lives or want ick, vacations. Most of you are paycheck to paycheck, statistically speaking, so you have a vested interest in how this all turns out.
edit: I get it, it's not a popular opinion.
But am I wrong?
If the problem of hallucination in AI has been solved, maybe I missed it?
reverius42 4 hours ago [-]
The "problem" of hallucination has not been solved, and never will be fully solved, because that's just how LLMs work. They're "hallucinating" 100% of the time but we only call it a hallucination when they're wrong.
But they're right much more often than they used to be. If you haven't tried a near-Frontier model lately (say, Claude Opus or ChatGPT Sol) you probably need to update your priors for just how often they can be right.
spl757 4 hours ago [-]
I only run models I can run locally. Perhaps I'm framing my question incorrectly.
AI is being used in contexts that involve life and death. Being right more often is not good enough in that context. If we are going to put AI in drones and have them autonomously acquire and fire on enemies, perhaps the fact that all AI hallucinates at all makes it morally and ethically unusable for purposes in which the stakes are very high, such as life and death.
This is what scares me about it. And also what I was able to accomplish air-gapped with a Ryzen 7 and RTX 3060 12GB scared me.
This is not the kind of thing we should all have to speculate about, someone should have solid answers. Perhaps the fact that some, and maybe most, AI models are black boxes even to those that trained them, should have been the first red flag and that was years ago.
TiredOfLife 37 minutes ago [-]
Cars are generally useless. I am using only the ones i make at my house from paper, but it can't be that different.
spl757 4 hours ago [-]
One of the problems with this is companies and governments have invested 1.5 Trillion dollars on the bet that AI would revolution everything and reshape the world. AI companies lose money. All of the money invested is being recirculated among the major players in the game. It's not the people paying money for AI as a product, it's the companies wanting so badly for that to happen that they have created a spending bubble not too dissimlar to the sub-prime mortgage disaster that resulted in the housing bubble to burst. This is no different. AI companies have proven that they cannot make a profit at this point. There's 1.5 trillion at stake, and who do you think is going to be left holding the bag? My guess is the people.
mrdomino- 6 hours ago [-]
> The not-so-quiet goal of the LLM push by big tech was to homogenise intellectual labour such that it can be made fungible. When labour is fungible you can swap it out whenever you like for an alternative, decreasing the bargaining power of workers. Software development was one of the last ‘guild fields’ and now we are seeing its industrialisation occur.
This is really well said. I had not put my finger on it before. Maybe this is a next move in the saga to avoid labor unions in engineering.
thataccount 15 hours ago [-]
>The not-so-quiet goal of the LLM push by big tech was to homogenise intellectual labour such that it can be made fungible. When labour is fungible you can swap it out whenever you like for an alternative, decreasing the bargaining power of workers.
Especially if you replace people with robots that don't need days off and have the people train their replacements.
smokedetector1 10 hours ago [-]
idk. I get more stuff done, faster, that I wouldn’t have been able to. You have to use it judiciously and with supervision - I can see how letting it get out of control would negate the benefits. But with good use, its undeniable.
ChrisMarshallNY 7 hours ago [-]
That’s been my experience, too.
I’m in the final prep for ship phase of a project that started February. It’s a bottom-to-top rewrite of a project that’s been shipping for two years, and took two years to write, initially.
I’m deliberately doing it all with the $20/month ChatGPT sub. After it ships, I’ll move to the $100/month sub; but I want to be able to say that the entire project was done on a low-tier LLM subscription.
twelve40 9 hours ago [-]
the decision i have to make dozens if not hundreds times a day: do i actually look at this crap or trust it blindly and let it slide? I'm honestly reading maybe 20% tops of what "I" produce (and what other people post for my review) I think you do lose skills and understanding that way. Sure, you can always ask the robot to do the tldr for you, but.... yeah, something is still lost, not in a good way.
smokedetector1 4 hours ago [-]
my way of working is make sure I understand every bit of it. Something is still lost but I feel confident my ability to understand systems is not atrophied at all. And its actually improving because I can crunch through new systems rapidly and have a private tutor. Of course its best supplemented by human team members who actually know whats going on. Its not a panacea. But its incredibly useful
wat10000 6 hours ago [-]
It takes some time to learn how to get these systems to work well, and learn what they’re good for. I think people who don’t see the benefits either haven’t dug in enough or haven’t used them for things they’re good at.
Do coding agents help me with highly technical work where it’s some deep thought and discussion with coworkers and eventually I change five lines of code? Not really.
But I can debug a weird kernel panic or optimizer bug from a standing start in an hour when it would have taken me all day without these tools.
smokedetector1 4 hours ago [-]
Yeah theyre amazing at debugging.
As long as you make sure you genuinely understand everything that comes out of it, and your domain has enough coverage in the LLM training data/context, its an incredible tool.
mistercheph 6 hours ago [-]
your perception is that you get more stuff done faster, and that seems to be the conclusion that most using LLM's come to (including myself): but there are no productivity gains that are externally observable to show for it (yet).
That might be because of some lag, but another explanation might be that the tools are really effective at increasing someone's perception of their productivity while net productivity gains measured over long time windows are low or zero. Another explanation might be that the incompetent/negligent users are causing enough harm to nearly neutralize the gains brought by proficient / effective users. I don't know what the explanation is, but there is a mystery in the divergence between users' perceptions of their own productivity, and the observed productivity gains in the economy.
smokedetector1 4 hours ago [-]
I think its that theres a way to use them really well and a way to use them not well and it comes out as a wash in the study if you dont stratify it. Im sorry, call it anecdotal if you want, I dont really care - I know what its doing for me on a day to day basis.
wat10000 5 hours ago [-]
It would be a weird coincidence if a pretty different way of working turned out to be exactly as productive as the old way, no more and no less. I’m skeptical of these claims. Measuring programmer productivity is notoriously difficult.
mistercheph 5 hours ago [-]
It might be weird, but it might be the result of forces in equilibrium:
e.g. one explanation might be that time spent prompting is time spent not improving, or learning, so it would not be surprising then that the productivity gains might be similar in scale to the productivity lost from learning / improvement that never happens.
another explanation might be that prompting just delays work that will have to be done later when fixing errors. When encountering an error that the LLM doesn't fix on the first attempt the human-in-the-loop doesn't understand the problem very well: they will either pull the slot machine for a while until they hit lucky, or grit their teeth and sit down to understand what the code is doing in the excruciating and boring detail required to find and solve bugs, which is exactly the work that they had offloaded onto the LLM earlier.
wat10000 3 hours ago [-]
Sure, it’s possible. But I don’t see anything that could constrain those things to be of similar magnitude. If they just happen to cancel each other out by luck, that would be pretty strange.
err4nt 3 hours ago [-]
Anecdotally, I've had an iPhone for 15 years, and that whole time I've dreamed of apps - some silly, some useful, that I would tell people "Imagine if there was an app that did x…". In those 15 years I learned to program and built websites and web apps, and built very basic web apps to use on my phone but shipped zero iOS apps.
Thanks to Claude, as a hobby, I've built and shipped 4 iOS apps in the last month. None of them are world-changing, but that's 4 in the last month thanks to Claude and zero in the 15 years before it. I'm clearly getting my subscription's worth out of Claude and squeezing more juice out of my own devices. It's a fun hobby! (And I'm sure if I used Android I could do more and customize my own device even more!)
lookACamel 1 hours ago [-]
> It's a fun hobby!
This is what skeptics are missing. The strongest answer to what AI has delivered to society so far is the new, fulfilling hobby of vibecoding. Their disparagement is like being skeptical of the routine, regular investment into iterative improvements in smartphone cameras purely on the basis that you can't pinpoint a specific lump of business process value. The vibecoding is the point. That's what people want and what they are willing to pay for.
People like OP have frivolously wasted large chunks of their life writing and maintaining open source code that no valuable businesses really care about, just because they have this dumb hobby of typing out code and thinking about logic. (I'm saying this tongue in cheek.)
They, of all people, should be to understand the unlock that AI is for a large swath of humanity.
andy99 8 hours ago [-]
> Software development was one of the last ‘guild fields’ and now we are seeing its industrialisation occur.
Doesn’t really support the rest of the argument. If you’re a skeptic and think it’s all hype, there’s nothing to worry about.
Barrin92 5 hours ago [-]
that'd only be true if you have a lot of trust that people care about quality. It's wrong to think of industrialization as only affecting producers. Fast fashion is a good example of this. There's studies on the quality of for example jeans and they've declined over decades. Mid 20th century cotton denim was a lot better than what you can buy now. Animation is another one as many people have noticed, you have movies that have somehow managed to decline in quality but nobody cares because you can churn out ten times as many. The nutritional quality of food has declined for 60 years as ultra-processed and industrially produced foods have taken over.
The worst scenario isn't actually that machines are intelligent and you lose your job, it's that they're stupid and you lose your job because in our quantitative glutton culture qualitative judgement is entirely gone to begin with
hingler36 7 hours ago [-]
I have started using AI a bit, but I'm still trying to be careful in that if I ever lose access I can still perform my job at pre-AI ability. My understanding is that tokens are still heavily subsidized, and when that well runs dry I don't want it to be some nightmare to go back to the way I was doing my work in like 2022.
elgertam 6 hours ago [-]
From my reading on the topic, the tokens are subsidized when considering average cost, but are profitable at marginal cost. Basically, they're super expensive when considering the training cost, but aren't super expensive when doing inference. Since AGI is quite unlikely now, my guess is that we'll see consolidation and will see frontier models released at a slower clip, such that they can pay for training for from the profits from prediction tokens themselves. There may be slight increases in token prices, but there's sufficient competition between Google, Anthropic, OpenAI and X alongside the open weights providers that I don't see huge price increases happening. Even if OpenAI is absorbed into one of the others (which I think is most likely), that would still make token price collusion difficult.
lokar 5 hours ago [-]
People keep saying that, but there are no real confirmed numbers. And an incredibly high incentive to lie.
vivzkestrel 4 hours ago [-]
- you forgot the biggest and most important point of all!
- Everyone and their mother using an LLM is slowly and steadily destroying their brain cells and abilities in the process while creating a learned dependence
- Give it about 5 yrs, mark my words. The studies ll come out saying "95% programmers cannot write a hello world without an LLM now"
- Now someone smart ll hop in on my comment and bring calculators into the argument and I have answer waiting for them that those do (not able to put a direct link here for some reason)
E-Reverance 9 hours ago [-]
I think the fundamental rebuttal to this is merely that, we’re still so early, the $1.5 trillion number is irrelevant for developing technologies in the same way two women can’t birth a baby in 4.5 months
CoolestBeans 9 hours ago [-]
The $1.5 trillion number is extremely relevant because it makes the revenue numbers the industry has to hit higher and there's an expectation to see progress sooner before funding starts getting pulled back.
worik 7 hours ago [-]
>...it makes the revenue numbers the industry has to hit higher...
Yes. But.
Industry can miss its revenue targets, Oracle can go bankrupt (looking more and more likely) OpenAI, Anthropic and SpaceX might find "true value" at about 10% of current valuations (that feels right to me), all that can happen with the associated economic chaos... and still out of it comes world shaping new technology
Both things can be true
CoolestBeans 6 hours ago [-]
Certainly, but a flop of $1.5 trillion would create a deep and long trough. I think in that situation OpenAI and Anthropic probably don't survive as independent companies (where is the revenue? who would want to put more money in at that point?). In that situation there would definitely be a fire sale in the prices of licensing models (open weights are already pushing this way the hell down), semiconductors, data centers, and compute in general. And like in that situation, who knows what products come out of that dividend of cheap stranded assets? It might not even be AI related. You can do a lot of things with a lot of cheap GPU power. This boom-bust cycle is nothing new, especially to the tech industry. This deep crash outcome is the risk that a bet of $1.5 trillion makes. And it makes the current paradigm of coding agents and chatbots less likely because it is so volatile.
AnimalMuppet 5 hours ago [-]
Oracle can go bankrupt? Oracle was a license to print money for decades. If they go bankrupt, that is one insanely bad bet they made.
CoolestBeans 5 hours ago [-]
Oracle has bet the house on AI. Free cash flow is deeply negative with tons more capex guidance on the books. They are currently the largest non-financial corporate issuer of new debt in the corporate bond market. They are literally cutting their headcount to the bone to help fund this build out. If their compute deals do not pan out because their customers have liquidity issues or revenue stalls, they are in dire straits.
fragmede 8 hours ago [-]
Given the advances since 2022 why would funding get pulled back? The endgame is replacing all office workers globally and so you get 10% of that market that's that's easily $1 trillion. A competent AI employee replaces a human the same way the car replaced the horse. Businesses will pay more for an AI employee that doesn't get sick or come in hung over over a human one and never goes to sleep. If OpenAI manages to crack that, and capture even 10% of the market globally, that's trillions.
CoolestBeans 6 hours ago [-]
Total addressable market doesn't pay interest expenses. Even if this technology really did have the potential to be as revolutionary as the move from the horse to the car, if you can't get there with what the world economy's liquidity can provide then you can't get there. Free cash flow is drying up and data centers depreciate. This creates a limit and if that limit is hit before the profit starts coming in it doesn't matter what the potential could be. Paying back interest is financial gravity and it exists today. And this isn't an appeal to Wall Street's short term profit motive, there just is an economic limit to what the capital markets can spend on a bet this big and this risky.
twelve40 9 hours ago [-]
Yes, that's going on optimism. It's also possible that we're 80% there and the rest will take forever, while the loans and investments were taken out now.
gitaarik 4 hours ago [-]
I wonder how much AI was used by the human submission PRs. Maybe they also just use AI, but just in the way you should: as a tool and not a replacement. And the people that use it in the wrong way are of course much more, making it harder to see the people who are actually benefitting from it.
andai 6 hours ago [-]
>We should all know by now that code is an input to the software development process and not an output: yet even the most forward-looking organisations still tactitly treat lines of code as a productivity metric
What exactly is this referring to? My thinking here is "code is a hypothesis", and you don't know if it's solid or not until it contacts the real world. (And ideally tests, and maybe a proof or two.) And what we're seeing lately is that the hypotheses often get disproven decades later (longstanding kernel bugs etc.)
skydhash 5 hours ago [-]
Maybe because the actual work is done by the process, which may not even be the code itself, just its compiled form. In this code the code is both a communication tool (between humans) and the input that leads to the process. It's like a group of musicians collaborating on a music sheet. The final valuable form is a concert. The audience won't care about the music sheet.
The comments here gush about how they can write more code, but they don't say anything about the quality of the software being built and if they have improved on some points with it (cheaper maintenance, faster reaction to feedback, less bugs overall). Very much like a musician clamoring how he can write a lot more music sheets, but there's no concert in sight to listen to his music.
iammrpayments 5 hours ago [-]
I have a similar experience that I expected several competitors to popup in my app niche, and so far only one has come up, and after some digging, they already had 8+ apps in another niche that had some overlap, so they would be able to complete the competing one without AI.
My apps are also not super complicated.
rushabh 7 hours ago [-]
I agreed right upto the last quote. That made the post political, similar to the ones being written by the AI evangelists.
The thing about complex systems is that they are not predictable. Right now the answer is whatever you think is right, and a lot of scenarios seem plausible.
It is a fun intellectual moment trying to make sense, but maybe only time will tell how the dynamics work out.
whateveracct 7 hours ago [-]
it is a political issue though
qarl2 6 hours ago [-]
The author claims that software has become no more secure - despite the avalanche of security fixes appearing for all software everywhere.
Never mind the more subtle issues.
wolttam 7 hours ago [-]
I think an argument can be made that some software has gotten more secure.
I don’t think it can be denied that the models do show an ability to find security vulnerabilities that may have otherwise been missed
youre-wrong3 4 hours ago [-]
Seems like a lot of people who aren’t very good remain skeptical.
wbkang 6 hours ago [-]
This is equally as biased as those who claim AI can just code all night autonomously for you. You don't have to use it to generate "fully LLM-generated PRs". It's true GitHub is flooded with poor-quality LLM slop, but that is not the only way to use it.
It's ironic the author links to Naur's paper because I use LLMs to dig through code faster, question my assumptions, review my drafts, or generate drafts for my review. It's not because I am bad at it, I used to be very proud of being able to do this better than others. Anyway, this does not lead to more "lines of code" as the author suggests and instead it can lead to higher quality code.
FWIW, I also don't believe you can let LLMs make critical judgement by somehow stringing together more LLMs. And I am also not making a claim on whether the $1.5T investment is proportional to its benefits. But to say "industry has nothing to show for it", or to bring up strawmen like "LLM-generated PRs" and "lines of code", makes me question whether the author has tried studying the applications of LLMs.
8note 9 hours ago [-]
are open ai, anthropic, zai, deepseek, cursor etc not new players? that benchmark seems like one defined to never be good enough
jansport123 7 hours ago [-]
the way i see, LLM's in their current shape are neither good enough to replace software engineers but they have become a necessary part of everyone tool kit. ergo, actual companies themselves aren't getting much benefit because their competitors are using the same tools and they can't fire their human labor either just it's just an extra cost. At my org, i see good engineers use LLM's in wise ways and bad engineers just let it write slop for them. What gets to me is the verbosity and noise though, i consider brevity, and curation to be hallmarks of intelligence but most people at work just spit out so many words in their design docs or code comment these days.
j45 6 hours ago [-]
The reality is new technology that works, will get used for what it best works for.
There are things that LLMs remain uniquely suited for, while the echo chambers go for everything.
There's no requirement to use new technology. Maybe what you do, or how you do it is just fine, and that's OK.
Skepticism today can remind me of someone who went into a cave for 3-4 years with a dated take of LLMs on day 1 and not remember software improves, and LLMs are software.
There are no shortage of uniquely working solutions with LLMs if sone seeks to find them, and also the self-validating echo chamber that says that it doesn't if that's desired. Which parts of software development, or how software development may better happen differently with the involvement of AI remains to be seen.
hexasquid 8 hours ago [-]
"Well, actually, vim doesn't make you faster"
honeybadger1 6 hours ago [-]
it's not as good as the hype, but it's definitely revolutionary. i am still blown away at what I can achieve in 20 minutes with a couple of agents
rhelz 8 hours ago [-]
AI skeptics hate him. Here's how he doubled his productivity using AI with one simple trick.
You have to stop thinking like an employee or a worker bee, and start thinking like a boss.
Guess what, the work you give your boss isn't 100% gold either. And sometimes you are preoccupied, or tired, so what you deliver isn't the best of all possible deliverables. Sometimes you hallucinate, i.e. you are dead wrong when you think you are dead right.
This is the job of a boss: take fallible employees and turn what they produce into something actually valuable. The boss lets the grunts do the grunt work, and then they combine the output together, and do the due diligence, etc etc.
It's an entirely new skillset. But learning the new skillset isn't the real blocker--the real blocker is that you've got to stop thinking like a grunt and start thinking like an executive.
WHAT is it that you really want to do? Frame that question as clearly and as concretely as you can. Then farm out thee work to human or AI grunts, and be prepared to do your due diligence on the result.
jgord 6 hours ago [-]
... theres a lot of anecdotal evidence executives aren't good thinkers.
Why not replace all mid-execs with LLMs trained on every MBA textbook and linked-in blog post ? why stop at grunt coders, graphic designers and phone marketeers ?
Can we reduce the role of CEO to the following prompt : "make money, by any legal means"
Apple, Mozillia, and Firefox just released a record number of bug fixes because of AI. So this assertion that Software as not improved in quality is arguably if not provably false. There are 1000s of other examples. I have my own. Personal projects that were stopped because of dependency rot, having 100s of bugs fixed, nearly all dependencies removed, and lots of tests added just by asking. I know tons of others with similar experiences
> I do not feel like I’m falling behind by not using them
Obviously you are correct about "how you feel". But, this argument to me is more like a horse owner claming they don't feel like they're failling behind to trucks that can carry 10x, 100x, 1000x more than their horse. Sure, maybe the horse can go places the truck can't. For most uses the truck is better than the horse.
> The LLM-generated PRs I get are still rubbish.
I can't argue with that. I haven't seed the LLM-generated PRs "you" get. I can say though the LLM-generated bug reports I've seen are 1000x better than human ones. Human bug report "your software breaks, fix it!". LLM bug report "There is a bug in this specfic part of the code for these specific reasons and here's a repo I wrote to demonstrate the bug. Here's instructions on how to run the repo. There's the code to fix the bug. And, here's a test to show the bug is fixed."
Is it perfect? No, sometimes the bug isn't real because the code can only be reached in situtations that can never happen. But still, 7 of 8 times it's a real bug and the bug report are the best I've ever received.
you can go through this list: https://www.firefox.com/en-US/firefox/150.0/releasenotes/ and for those very few issues not still blocked you can read the examples
I think you're talking past the author here, who says:
>By not leaning into LLMs I’ve given myself the room to specialise in a smaller set of skills. So far that conscious decision is working out well, and I remain confident about my job security.
"Most uses" might mean boilerplate and simple frontend work that LLMs can do easily because it's formulaic and doesn't require much creativity. OP is simply not doing that type of work.
I don't think anyone sane is really claiming 100x or 1000x speedups anymore. Most people I talk to who use LLMs are closer to 1.5 or 2. Because their job is not mostly boilerplate, there is simply not that much room to be improved.
Over the past two months I’ve built out a very complex web app for a client. Without the aid of AI it would be far smaller, simpler and less capable. And honestly I probably wouldn’t have agreed to take on the job in the first place.
I myself use them to accelerate programming tasks, so I'm not anywhere near as pessimistic as the author, but the claimed multiples of productivity definitely haven't materialized for me.
A hidden metric here is the number of new bugs created by these fixes. If an LLM creates 10x bugs but create 5x more new bugs, then it is not really an improvement. Because that is 5x more bugs that the user has to observe and report (unless it is a security vulnerability, which the LLMs might detect) before it can be fixed.
To see an example of this 1000x, all we have to do is look at the flood of bug reports and fixes to browsers and OSes. Does it do 1000x for every project? No, and neither does a pickup truck do 1000x what a horse does. But, even a small pickup truck does more than a horse for most use cases (carrying people or stuff) and similarly, LLMs do more for most use cases.
If you actually used LLMs instead of preaching against them from the sidelines with no real experience, you'd know such statements are wrong.
This is the real problem with many of the anti LLM folks - they tell us about LLMs and also tell us they don't use them.
"That thing cannot do the stuff you say it can. And there is no way I will ever try it. But I know for sure I am right."
What about the people who do have a lot of LLM experience and agree with OP?
> That thing cannot do the stuff you say it can. And there is no way I will ever try it. But I know for sure I am right.
If someone claims to be able to fly by strapping bird wing shaped pieces of plywood with feathers glued on to their arms, do you need to personally jump off a tower with them to say they don't work, or can you look at the results of others attempting it and draw conclusions based on that? LLM proponents are making claims about their capabilities which can relatively easily be checked without using LLMs yourself. To pick an example where LLM proponents are correct, anyone who says LLMs can't generate syntactically valid code can be proven wrong fairly easily by producing an example of syntactically valid, LLM generated code.
Further, if you read the rest of the paragraph you responded to, it's clear that this claim is about the state of the industry as a whole. "Software has not improved in quality, got faster, become cheaper to produce (when you exclude the mountain of poor-quality demoware that no reputable organisation would touch with a barge pole), or become more capable." Whether this is true or not is something that can be evaluated without ever having prompted yourself, or even arguably without being a developer at all.
- people who preach against PV solar... while people I actually know broke even 10 or more years ago and have been happily getting "free power" for a lot of their needs
- people who preach against electric cars... while people I actually know are driving them with no "range anxiety" or "replaced main batteries", but their non-engineer spouse/kids are driving the cars fine too.
- people who preach that tesla self driving isn't viable/etc... while lots of people are using tesla to drive themselves around for > 90% of their driving
Unlike the author, I'm an LLM optimist; but this take doesn't make sense to me. The thing you show cannot be "hey look at this LLM-based workflow, that's so much better than last year".
Where is the big browser, operating system, or other piece of work that has been able to quickly compete with the existing entries in the market due to LLMs?
The only one that I can think of off the top of my head is the rewrite of Bun in Rust. I think the fact that Claude Code runs on millions of devices on top of that rewrite is the sort of impressive demonstration that would prove that LLMs have lasting, systemic effects -- but it is telling that the only one I can think of was made by the extremely well funded lab that also happens to build the LLM.
Which I find weird - the greatest fun for me is making a computer do something awesome. I learned to program because I want that outcome.
I respect if you love the art and craft of hand programming but it's something else entirely to simply deny that LLM's are what they are - which is to say so amazing it is really beyond belief and they are getting more so every year.
If you're a hand programmer forever then fine, just don't make a fool of yourself by saying that everyone is lying about LLMs, to justify your choices.
We've normalized science fiction. Go back 5-10 years and ask anyone whether we'd soon have something like LLMs.
If you step back and try to look at LLMs without emotions and politics, you will be blown away by what this thing can do. We're talking to a machine, folks.
I just can't believe that LLM haters really like computers or technology, because we haven't seen anything this interesting in decades.
Yup. As someone who thinks computing has been in decline since Mac OS X 10.4, LLMs are the first major tech trend that I've been impressed by in 20 years.
And? I didn't sign up to talk to a machine.
Has it actually made your daily life better? All I see as a result of LLMs is endless slop everywhere, plateauing in the skill levels of coworkers, and badly informed people in management positions pushing the humans to meet impossible deadlines because "otherwise we'll replace you with a cheaper worker and an LLM".
I am an LLM optimist, but I'll be honest and say it hasn't made my life better. My optimism relies on the models plateauing, and people finally rebelling against the widespread slop to realize that LLMs are just another (very useful!) tool, but don't meaningfully replace people in any capacity.
I have. Which is how I know that such statements are correct.
I recently experienced some cognitive decline due to chronic health issues. This is unfortunate timing because I am shipping a game with online multiplayer for the first time. So "brain works properly" is kind of a basic job requirement! But you work with what you got...
Some day I have the energy to do programming "manually", but on many days, the difference between "tasks get done" and "I am just staring at the screen reading the same code over and over again" is "ask the LLM to do it."
So I have had to shift my identity from "the guy who does the thing" to "the guy who ensures the thing has been done."
That being said, I do check the diffs carefully. (I tried a more hands-off approach a few months ago, and that was a pretty bad idea.) And I've learned to make the diffs as small as possible: more digestible for me, and less room for the AI to go off the rails.
Nice bonus to small, well defined changes is that small, fast, cheap models can handle them. (Having a great time with Luna lately, even on Low reasoning effort, which I wasn't expecting at all.)
LLMs are really good at making it harder to ignore ignorant people, and you've demonstrated that you're NOT part of the problem.
I think that the industrialization of software development already happened decades ago. The common professional software development methodologies, both waterfall and sprint-based, ARE the industrialization of software development. Isn't modern "agile", sprint-based software development methodology directly influenced by Japanese manufacturing methods? I don't know when software development actually was a guild field, and my personal experience doesn't reach back this far, but based on what I've read at least, it seems to me that even back in the 1990s it was already not a guild field.
In fact the four-year degree you'll get from school is getting increasingly distant from the skills I actually want out of a new grad. It's not impossible to bridge the gap or anything but my transition into the commercial realm in the early 2000s was a cakewalk compared to the sheer number of things I'm asking a new grad to learn as soon as they're settled in at their desk... source control, CI/CD, bug trackers, devops, and that's just the beginning of that list not the end.
Those skills are easily taught. But in every run of the mill project I’ve been on, I’ve been really happy about all the maths and other science stuff I’ve been taught at college. It’s easier to grasp the web DOM, React, and git when you already know about trees and the related algorithms. Easier to learn SQL after opening some books on relational theory (basically the first chapter on database theory) and learning about projections. And definitely easier to learn the nature of distributed systems.
Learning how to use Bash and Ansible is a walk in the park after that.
For similar aesthetic reasons, have only dabbled with code-gen LLMs. Im happy to vibe code css and html but not, well actual code. I'm aware of my own double standard, Im fine with median-reverting banal css and html.
I do worry about some aspects of LLMs :
- google giving AI code snippets so easily, prevents traffic to sites like stack-overflow, where discussion happens
- vibe coding means less people hit the actual hard parts of coding, which result in learning to think as a developer. Thence where the next generation of developers who understand code ?
- code will become a magical, inscrutable black box where no-one understands how it works. DNNs and LLMs have this feature already. At least with LLM discovered math, the proof in lean exists, which is independent at how it was arrived at / searched for.
- if I write a novel, and the DNN AI detects it as AI slop with 68% probability, then what recourse do I have to prove its human written ?
- circular training : subsequent rounds of LLMs trained not on human input, but on LLM slop of the previous generation [ the %ge of which increases ]
- you will own nothing : fortunately it looks like self-hosted open weight LLMS will remain a thing, but the massive spend by hyperscaler companies on Datacenters/GPUs/RAM has driven down supply and driven up prices of CPU, GPU, RAM for enthusiasts / gamers / home AI nerds / garage startups .. which might stifle innovation.
I don’t really agree that AI can’t make development faster, though. What the author describes as the negative AI outcome is blind vibecoding. There are many other ways to use it, and even the basic “enhanced autocomplete” is a net benefit, especially since that functionality is dirt cheap or free.
We have been programmed to believe that MBA wet dreams are inevitable because "$x hundreds of billions invested can't be wrong" but they very often are.
https://getdx.com/blog/ai-productivity-gains-are-10-percent-...
But I spent a lot of time in the second half of this week dealing with friction with a team that is very annoyed that I'm moving fast and using an agile methodology so I can't tell them the exact, precise REST calls that I'm going to have for them in six months designed to a tee and signed off in triplicate before they start development against it. Manifesting that increase in code production as real value to the business is going to take more from me than just spewing the code out more quickly.
AI isn't creating this problem. I would have had this problem anyhow even if I were writing all the code by hand again. I know, because I've been there before. But the increased velocity is manifesting in increased organizational stress and not just increased velocity.
AI is perhaps even helping solve it to some degree, though far from totally. I have written before about how people eventually learned not to play the "oh well we can't do this until we have documentation" card on me [1]. This week they played the "well, I see you have docs but they aren't in our precise format". Guess what AI can do in about 15 minutes really well? You may recall the term "style transfer" getting tossed about a lot 3-4 years ago, and it is still something AI is extremely good at, and "take these docs in this format and convert them to that format" is just a style-transfer problem. AI really does chew at the "oh but we need docs" old-school card... and they can't even complain about the quality of the AI docs because in order to do that, they'd have to actually read them, and that is not the point of the "but we need docs" card, you see....
[1]: https://news.ycombinator.com/item?id=48412349
You can think of it as an impedance mismatch, or as a translation problem. But no matter how you think about it, it's real, and it's a problem - especially if upper management lives on the non-agile side of the fence.
edit: I get it, it's not a popular opinion.
But am I wrong?
If the problem of hallucination in AI has been solved, maybe I missed it?
But they're right much more often than they used to be. If you haven't tried a near-Frontier model lately (say, Claude Opus or ChatGPT Sol) you probably need to update your priors for just how often they can be right.
AI is being used in contexts that involve life and death. Being right more often is not good enough in that context. If we are going to put AI in drones and have them autonomously acquire and fire on enemies, perhaps the fact that all AI hallucinates at all makes it morally and ethically unusable for purposes in which the stakes are very high, such as life and death.
This is what scares me about it. And also what I was able to accomplish air-gapped with a Ryzen 7 and RTX 3060 12GB scared me.
This is not the kind of thing we should all have to speculate about, someone should have solid answers. Perhaps the fact that some, and maybe most, AI models are black boxes even to those that trained them, should have been the first red flag and that was years ago.
This is really well said. I had not put my finger on it before. Maybe this is a next move in the saga to avoid labor unions in engineering.
Especially if you replace people with robots that don't need days off and have the people train their replacements.
I’m in the final prep for ship phase of a project that started February. It’s a bottom-to-top rewrite of a project that’s been shipping for two years, and took two years to write, initially.
I’m deliberately doing it all with the $20/month ChatGPT sub. After it ships, I’ll move to the $100/month sub; but I want to be able to say that the entire project was done on a low-tier LLM subscription.
Do coding agents help me with highly technical work where it’s some deep thought and discussion with coworkers and eventually I change five lines of code? Not really.
But I can debug a weird kernel panic or optimizer bug from a standing start in an hour when it would have taken me all day without these tools.
As long as you make sure you genuinely understand everything that comes out of it, and your domain has enough coverage in the LLM training data/context, its an incredible tool.
That might be because of some lag, but another explanation might be that the tools are really effective at increasing someone's perception of their productivity while net productivity gains measured over long time windows are low or zero. Another explanation might be that the incompetent/negligent users are causing enough harm to nearly neutralize the gains brought by proficient / effective users. I don't know what the explanation is, but there is a mystery in the divergence between users' perceptions of their own productivity, and the observed productivity gains in the economy.
e.g. one explanation might be that time spent prompting is time spent not improving, or learning, so it would not be surprising then that the productivity gains might be similar in scale to the productivity lost from learning / improvement that never happens.
another explanation might be that prompting just delays work that will have to be done later when fixing errors. When encountering an error that the LLM doesn't fix on the first attempt the human-in-the-loop doesn't understand the problem very well: they will either pull the slot machine for a while until they hit lucky, or grit their teeth and sit down to understand what the code is doing in the excruciating and boring detail required to find and solve bugs, which is exactly the work that they had offloaded onto the LLM earlier.
Thanks to Claude, as a hobby, I've built and shipped 4 iOS apps in the last month. None of them are world-changing, but that's 4 in the last month thanks to Claude and zero in the 15 years before it. I'm clearly getting my subscription's worth out of Claude and squeezing more juice out of my own devices. It's a fun hobby! (And I'm sure if I used Android I could do more and customize my own device even more!)
This is what skeptics are missing. The strongest answer to what AI has delivered to society so far is the new, fulfilling hobby of vibecoding. Their disparagement is like being skeptical of the routine, regular investment into iterative improvements in smartphone cameras purely on the basis that you can't pinpoint a specific lump of business process value. The vibecoding is the point. That's what people want and what they are willing to pay for.
People like OP have frivolously wasted large chunks of their life writing and maintaining open source code that no valuable businesses really care about, just because they have this dumb hobby of typing out code and thinking about logic. (I'm saying this tongue in cheek.) They, of all people, should be to understand the unlock that AI is for a large swath of humanity.
Doesn’t really support the rest of the argument. If you’re a skeptic and think it’s all hype, there’s nothing to worry about.
The worst scenario isn't actually that machines are intelligent and you lose your job, it's that they're stupid and you lose your job because in our quantitative glutton culture qualitative judgement is entirely gone to begin with
- Everyone and their mother using an LLM is slowly and steadily destroying their brain cells and abilities in the process while creating a learned dependence
- Give it about 5 yrs, mark my words. The studies ll come out saying "95% programmers cannot write a hello world without an LLM now"
- Now someone smart ll hop in on my comment and bring calculators into the argument and I have answer waiting for them that those do (not able to put a direct link here for some reason)
Yes. But.
Industry can miss its revenue targets, Oracle can go bankrupt (looking more and more likely) OpenAI, Anthropic and SpaceX might find "true value" at about 10% of current valuations (that feels right to me), all that can happen with the associated economic chaos... and still out of it comes world shaping new technology
Both things can be true
What exactly is this referring to? My thinking here is "code is a hypothesis", and you don't know if it's solid or not until it contacts the real world. (And ideally tests, and maybe a proof or two.) And what we're seeing lately is that the hypotheses often get disproven decades later (longstanding kernel bugs etc.)
The comments here gush about how they can write more code, but they don't say anything about the quality of the software being built and if they have improved on some points with it (cheaper maintenance, faster reaction to feedback, less bugs overall). Very much like a musician clamoring how he can write a lot more music sheets, but there's no concert in sight to listen to his music.
My apps are also not super complicated.
The thing about complex systems is that they are not predictable. Right now the answer is whatever you think is right, and a lot of scenarios seem plausible.
It is a fun intellectual moment trying to make sense, but maybe only time will tell how the dynamics work out.
Never mind the more subtle issues.
I don’t think it can be denied that the models do show an ability to find security vulnerabilities that may have otherwise been missed
It's ironic the author links to Naur's paper because I use LLMs to dig through code faster, question my assumptions, review my drafts, or generate drafts for my review. It's not because I am bad at it, I used to be very proud of being able to do this better than others. Anyway, this does not lead to more "lines of code" as the author suggests and instead it can lead to higher quality code.
FWIW, I also don't believe you can let LLMs make critical judgement by somehow stringing together more LLMs. And I am also not making a claim on whether the $1.5T investment is proportional to its benefits. But to say "industry has nothing to show for it", or to bring up strawmen like "LLM-generated PRs" and "lines of code", makes me question whether the author has tried studying the applications of LLMs.
There are things that LLMs remain uniquely suited for, while the echo chambers go for everything.
There's no requirement to use new technology. Maybe what you do, or how you do it is just fine, and that's OK.
Skepticism today can remind me of someone who went into a cave for 3-4 years with a dated take of LLMs on day 1 and not remember software improves, and LLMs are software.
There are no shortage of uniquely working solutions with LLMs if sone seeks to find them, and also the self-validating echo chamber that says that it doesn't if that's desired. Which parts of software development, or how software development may better happen differently with the involvement of AI remains to be seen.
You have to stop thinking like an employee or a worker bee, and start thinking like a boss.
Guess what, the work you give your boss isn't 100% gold either. And sometimes you are preoccupied, or tired, so what you deliver isn't the best of all possible deliverables. Sometimes you hallucinate, i.e. you are dead wrong when you think you are dead right.
This is the job of a boss: take fallible employees and turn what they produce into something actually valuable. The boss lets the grunts do the grunt work, and then they combine the output together, and do the due diligence, etc etc.
It's an entirely new skillset. But learning the new skillset isn't the real blocker--the real blocker is that you've got to stop thinking like a grunt and start thinking like an executive.
WHAT is it that you really want to do? Frame that question as clearly and as concretely as you can. Then farm out thee work to human or AI grunts, and be prepared to do your due diligence on the result.
Why not replace all mid-execs with LLMs trained on every MBA textbook and linked-in blog post ? why stop at grunt coders, graphic designers and phone marketeers ?
Can we reduce the role of CEO to the following prompt : "make money, by any legal means"