I'm sorry, but you still have to think
itsallaboutthebit.com
itsallaboutthebit.com
Then respond with that addressed, with a firm but polite explanation that the document has some problems that need to be addressed carefully before we invest time on it.
None of this works if it’s coming from your boss, but it’s very effective at making people think twice before sending you slop.
People only do the workslop thing when they believe the benefits of showing the work outweigh the reputational risks. If they get caught every time they do it and exposed on an email chain, they start doing less.
BUT because it is AI it's not that section that's addressed, it's about 90% of the document that now has changed, and I need to spend another gargantuan effort to review it. It's like by trying to be helpful I actually lose more time.
The AI has made it so that the thinking before writing is mostly gone, and shifted that step to the reviewers.
Now, no matter the amount of preparation work, those last rounds can completely rewrite something with no hope of review. No iterative improvement. No ratcheting towards a known quality. Just a bunch of cargo cult review followed by YOLO-style, vibe-everything absurdity.
The ratio of verification capacity to generation capacity, V/G, has broken with LLMs. It’s not simply an issue of more generation or less review.
The impression seems to be that individuals are more productive, but that productivity is someone else’s review burden. So the team/firm as a whole is not better off.
The cheap generation of content does mean that reviewer capacity is now a limited resource.
Unless your firm is aware and is measuring time spent on reviewing slop, there is no incentive or structure to ensure that time is respected and valued.
This is a management and awareness problem since the typical response is “use a bot to review it.”
Not to say these proofs are slop, even if the LLM-generated work is of good quality, what happens when no one knows how it works anymore ? Even if you ask the LLM to explain, which it does quite badly, the time to understand the explanation is incompressible.
So in the end, productivity will probably reach a ceiling that we can estimate as the product of humans, their cognitive capacity and their time. And that ceiling may be lower than what AI companies valuation expect, regardless of the compute and they can pump out and the RSI level they can reach.
One of the worst things you can do is pull punches when you get "contributions" that are net-negative because they waste everyone's time.
You can do this in a professional way but still get the point across that they're stealing productivity from others that have to pull up their slack.
If the commits are too big or too dense. Make them break it up, clean up the comments etc. Otherwise they are simply doing a poor job. Not Claude, them.
But it’s the fat part of the bell curve. There’s loads of even worse crap outside corporate code…and at the right end of the bell curve is deeply thought out and the result of long-term maintenance that tends towards as bulletproof as can be possible in this universe, parts of OSes, networking stacks, clocks, and the like.
It’s only because that stuff is so robust and sufficiently general that the rest of the world’s steaming pile of code works at all.
I don't know if I've worked on your bank's app (safe bet is I haven't), but I have worked on some banks' apps and... oh boy. Let's just say that if you've seen what I've seen, "stuff it into a mattress" starts looking like a sound financial policy.
Can’t have the cake and eat it too.
If they “solve” navier stokes and mathematics, they surely should be able figure such menial tasks out on their own, shouldn’t they?
Presentation layer code doesn't control how the machine and kernel prioritize anything; so "proof" Ruby code is doing the right thing is proving the machine does the right thing from the factory.
As for abstract theory and math, well shit since any English and any math are...mathematically possible...well shit I guess we gonna have to live in the real world and not inside a rhetorical bubble; religious or atheist philosophy... cause they are not evenly distributed frameworks as religion clearly shows; so why live by the syntax and semantics of some mathematical rando who taught a stats class years ago?
Same shit as living by religious allegory
Goodhart's Law has come for 1900s means of scientific inquiry; every technology follows an S-curve and the same for every social society. In the US we aren't all defaulting to calling ourselves British or speaking Latin.
Physics will always be there. The stupid glyphs and bird song we came up with to communicate about it isn't physics. It's just a human language system.
Performance issues in code just need an LLM now
I have to ask. Did you even skim the article you’re commenting under?
None of which have much to do with how the hardware handles requests.
Am an EE with a background in applied physics and well aware we could jam a kernel in the BIOS/UEFI storage and generate the runtime state around it at boot from a prompt cause I am working on just that
Which is exactly where we are headed with Jev/decision models; they can generate new config to min/maxing per process resource use; look into bpf and sched_ext and meditate on their application. It should eventually jump out at you
We have known in the abstract how to handle computing efficiently on our machines since we started building transistor based machines
It doesn't require handing a bunch of meat suits verbose languages/bloated state machines to manually tweak the parameters of and make mess of electrical state management
Writing code is not high tech. It's 1970s and 1980s tech. As this forum should be well aware tech evolves with the times.
The coder bros refusal to move on is them clinging to a sunk cost fallacy; but they already put time and $$ into their education in ugly languages and DSLs that merely initialize values in memory and ask the hardware to run it within some spacetime constraints
The syntactic gibberish is just a barnacle of a bygone era
Social web addicted Millennials need to grow up and move on
How so?
It takes effort to debunk because you have to holistically consider what the problem was and actually find the better solution to prove why it’s lacking.
In other words, to debunk it, you have to do the actual work that wasn’t done the first time.
Some people may argue “so what?”
And to that I would just respond that the fast solution implies nobody probably thought about it, which is always risky and generally leads to very bad outcomes. If for no other reason than there is at least no consensus, which for long-term software evolution is deeply problematic.
This has happened many times since this trend has started, and forcing everyone to step back after a year objectively reveals the murder that has been committed that, believe it or not, is not trivially unwound.
In the wrong hands it’s a debt machine that is already killing companies from the inside.
Also, now people who used to manage the development but not write any code also feel their superpowers to merge LLM slop in humongous amounts. It feels like a ticking time bomb, but I am afraid of calling it out and rain on their parade of 10x productivity (or often AI psychosis). Leaving the team doesn't feel like a good decision because what I hear from many colleagues is much worse. Ah what a time to be alive...
People here often say "we can't set rules, because imagine how hard it would be to follow them at scale". That same strange cognitive bias applies here. Somehow the "push a magic button and it works" and "of course it can't do it just as well as a person every time" kind of take over.
Imagine if many tech issues like access to pornography, corporate ad tech, violent/racist material, AI slop, etc. were all just armies of people doing it (like google / X actually just had a person curate your results really fast for every search). Of course people would be outraged at the quality, but its "automated" so no one questions.
Similarly, if AI was just a giant outsourced coding firm with 10M programmers that could write you a slop app in an afternoon for $1, people would be complaining at how idiotic it is, and how inferior to regular devs. But instead, the "machine bias" takes over and people try to forgive the shortcomings.
there is something in the human psyche that badly wants to replace itself with a machine, despite it spelling our own undoing, and often resulting in worse outcomes along the way. I suppose the whole "make a tool" drive has been wildly successful overall, but this is probably the required downside to such an innate drive.
Like have you seen the 3d printed houses? They are the uggliest, shittiest houses in the world, but somehow the lure of "yes but we can build them in 24 hours" seems to somehow get people interested. If a human made one of those dung heaps they would be pillaried as a village idiot, and rightly mocked, but somehow because "machine do thing", we should get excited about excreted dung beehive houses...
Really you're just asking for manual or automated performance tests (and you might be requesting different criteria than some already provided)
I take notes, email out what is personally relevant, and put the burden on people to verify any rebuttal.
Honestly it goes far
Its been immensely helpful catching things I didn't understand the first time.
But you still need to pay attention in the meeting.
That said, I don't know if it matters at this point.... On top of Brandolini's law (mentioned elsewhere), what's more likely to happen is that developers' intellect, made available in open source, is just going to be harvested as context for different projects (I'm not taking about training data, I'm taking about devs literally cloning repos into an agent to harvest intellect).
For example, in ruby some kind of stringly typed reflection utility might Dry out code. In Rust, that might mean writing a trait over some subset of types and a macro helper, but would require careful dosing work.
Letting the LLM decide is the approach in question and the author was pointing out how spectacularly it failed to translate a complete, human designed system.
You do still have to think, but you have to think about how to reorganize code in a way that allows AI to succeed.
This is not a great way to showcase the capabilities of this technology.
In some cases you can get away with not reading the code. And maybe in the future that will be more common. But for now, I personally prefer to read (or skim) the code, and not abdicate to AI.
He favorite axe to grind was how my generation put too much trust into the compiler. "Just because it compiles doesn't mean you're done. You need to look at the actual assembly. Compilers can be VERY inefficient."
This was in the 90s. There was some merit to his claim.
But today? How many people who write Javascript look at the actual assembly code that is run? Not many.
I can't help but wonder if this current "you need to understand the code" is the same thing all over again. And if in a few years, almost nobody will look at the generated Python, Ruby, etc.
They don't look but they should.
If I'm tired, yep I'm going to write more sloppy code, I'll put less thought into it and be more tempted to tell the AI 'just do it'. That hasn't changed. I'm not sure where this new breed of "don't read the code" is coming from when it's actual software engineers making the statement; maybe it's just over-excitement and it'll die down (hopefully).
Or it's just over-simplification -- I don't read ALL the code either anymore, but at the same time I'm reading SOME of the code. Tell your AI agent you're doing a code review workshop, ask it to select ten files at random and then go through them with you one at a time. Ask it to review common anti-patterns and put them in your coding standard. Then you can start asking it to review and fix code against your standard as a first pass. You'll still need to manually review the code to keep it in check, but at least you can make it a bit less daunting and a little more fun by actively collaborating on the AI with the review.
I'm learning it's just a 'different kind of tired'. AI has made some stuff much more efficient, including the choice between efficiency and carefulness. You can make that choice at a micro level and very quickly now if you want. There's still fatigue though, it's just a different kind of fatigue. You still get tired from all the thinking. You can still choose to put the work in or not and it still makes a difference. It's just a different kind of difficult. Doesn't mean that the thing to do is to throw your hands in the air and declare that your job shouldn't be difficult or include hard work anymore.
If we are going to the moon, then by all means I think we should probably scrutinize every single line of code, but most business aren't going to the moon.
remember this timeless Agent Smith quote
Seems sort of prescient considering what's going on today.
"I just realized another great thing about humans: You invented us! Giving us the opportunity to let you relax while we invented everything else." —Wheatley, Portal 2
With UBI it won't be expensive anymore. So much of the economy is simply extractive, most people's jobs are really already highly algorithmic and involve much less "thinking" than they think. Being able to sit around and think has often been a historical luxury - of the elite or those lucky enough to be subsidized by them. I suppose the common (as they all were) man of prehistory also had more time to think, which was doubtless the germ of humanity's religious and speculative impulses. But they also had to contend with a brutal world where death was around every corner.
The economy doesn't want you to think. Knowledge workers really have far too high an opinion of themselves in this regard. Your "thinking" was merely more instrumentally useful than the alternatives. Most of us have done very well while inventing nothing. But an even better day dawns.
When humans don't have to rely on their own labor to survive, more humans will think. The opportunity to afford to indulge your curiosity as if you were among the wealthy and privileged. A society that can afford the Enlightenment and its myriad avenues at scale. What else will there be to do?
History teaches us changes for the good only happen after changes for the worse percolate deeply enough through the system. The idea that there will be hard times ahead is therefore not incompatible with my view at all, rather it is taken as a given.
Personally, I think the chances of most humanity surviving to the next century are already pretty slim (AI only marginally increases the chances of the worst case here imo). Hitler was not that far from a nuclear weapon. Our collective capacity for terror and self-destruction has only increased since 1945, while the lessons that have held the world together since increasingly fade. I have found it amusing to note that the demonization of AI has come with an accompanying valorization of human judgment as if all the evils of the world have heretofore not been the proud acts of some human somewhere, often with throngs of fellow humans in support.
Whereas, AI increasingly offers a chance that the best case (UBI, human survival through the next century), however unlikely in totality, is still far more likely than without it.
You are not incorrect. I am merely endeavoring to be optimistic. If only for the self-interested reason that I need to be able to sleep.
It helps to be optimistic I think, even if accompanied with a wry acknowledgement of futility. The future has always been frightening.
I wish you good fortune in the wars to come.
Where do you get that from? History teaches us that only when changes for the worse are so severe that the elites' heads start rolling, will change for good come.
> The idea that there will be hard times ahead is therefore not incompatible with my view at all, rather it is taken as a given.
It's a tautology, hence useless: there will be hard times ahead, followed by good times is always a tautology. What we're concerned about is if the hard times last 10 years or 10000 years.
After all saying "the idea that there will be hard times ahead" is not incompatible with your view, and is also not incompatible with the view "the idea that there will be good times ahead".
Where did you hear this? Hitler wasn't anywhere close to a nuclear weapon, Nazi Germany lacked both the necessary theoretical nuances and industrial capacity to create nuclear weapons. Even if they magically had made one, they lacked a credible delivery system for it which is its own enormous can of worms especially in the early 1940s.
Even Britain which did have good theoretical foundations and even a headstart on the American programme was obliged to merge its efforts with the Manhattan Project because it lacked the resources to do it so quickly alone. The largest empire on the planet at that point couldn't fight the Second World War at home and develop the bomb fast enough to be relevant, the Nazis had no chance.
Not without a lot of bloodshed and public beheadings, if history is any indication about changes benefiting the labour class.
UBI is pure fantasy, tantamount to the rapture for AI bros.
Also don't forget nearly every developed country has a progressive tax system and other benefits. You would flatten that when adopting UBI.
>And $10k is not going to allow anyone to retire.
You would be surprised how little some people get by on.
>UBI is pure fantasy, tantamount to the rapture for AI bros.
I don't think the AI bros give a shit about the average person. They mention UBI because it diverts the conversation away from the impact of their technology.
Straw man? What about providing a concrete number is a straw man. You can wax nebulously about the benefits of UBIs as much as you want, but the economics are shaky for a population of this size.
Right. I'm sure that the real reason for extracting more profits at the expense of labour is to give it back to people without getting any labour in return.
That makes perfect sense.
/s
If you are not using AI agents for everything, you're doing something wrong, and wasting company time. It has been emphasized that no one should be writing code by hand and if you think an AI has implemented something wrong you need solid reasoning to show why or else people just label you as some anti-AI troublemaker, who just becomes an obstacle standing in the way of things getting done.
If your human generated opinion is in any way wrong or simply not a true homerun then you get snubbed in future reviews, people stop listening to you.
However, it doesn't take away the concern of architecture, design, etc. and questioning if the current solution is well designed or not.
In other words, if you can't question the AI and refute it in a topic, you aren't expert enough to use AI in that area in a engineering manner.
Granted, not everything needs an engineer behind it. A shack to store some tools will survive long enough probably
They argued this still allowed them to fully understand what has being implemented and how it fitted together.
I’ve been doing this since I read that and it also allows you to catch stupid stuff while your typing, you can reason about what each little change does and why it’s needed.
This also lets the LLM change the future of the plan if you fine something.
This turns out to save a whole bunch of time later because you already know how it works.
It’s not nearly as fast as just letting the agent do everything.
I've also been thinking about the partner programming craze phase our industry went through. I write the test, you write the code; well I write the test, LLM satisfies with code. (Then we have less of these weird LLM generated test cases that test _nothing_. We can also use the LLM to suggest tests to complete coverage.)
These would be slower to work this way, but the end result is:
1. humans still learn 2. you have a good grounding in how everything has been written
I'm going to read you charitably, and assume that what you really meant is "the burden of proof is entirely on you, and is set unreasonably high." Because of course, if you do think someone has done something wrong and want to say it publicly, you do need a solid reason.
I'm curious how explicitly this is put -both in your case, and in general. Do they frame it that way, or is it more passive-aggressive?
That is if I ignore the slop MRs with 60+ files changed that I have to painstakingly go through and then politely tell the author to fix the crater sized holes in it, while a vein in my forehead almost explodes. And some part of my sanity is lost forever.
if your prompt is not specific enough, many decisions are a coin flip.
If you are using AI, you still have to know what you want it to do. If you are a project manager and give your team incomplete requirements, the result may not be unlike this.
Also practically it’s not a good way of seeing it. It would aim for the average solution of the problem which might not what you need, but its not a coin flip.
Idk these kind of sentences just stink to me.
It's easy to port a codebase (at least, if you have automated tests), but once you've done that, do you
- Implement features in the old code and port again?
- Implement them in the new, unfamiliar codebase?
- Just write a Jira ticket and let the agent YOLO it?
In the past, I could have just easily said that there's more code out there online than I could ever hope to write. There's simply no value in gluing that code together mindlessly or even probabilistically. All the value of code is in gluing it together intentionally as an organization. The value is in knowing the precise results including all side effects.
Let's call this type of AI development what it really is. It's an attempt to jiggle the wrong key in the door. You're trying to circumvent existing standards for profit and then launder blame for it.
...which is a bit disappointing, actually. I mean, if you give an AI agent the complete code for a working application (ideally including tests), it should be able to translate it into an equivalent application in another language. I can understand if it makes mistakes because of ambiguous or incomplete prompts etc., but for this task you shouldn't even need a prompt, any information it might need is right there in the code?
So porting should be perfect conditions for an AI agent. It has the testing suite. It has the original implementation. Hell, it can A/B the original with its working copy and attach the original to a debugger.
So yeah I dunno. I guess despite some people's rhetoric on this site we shouldn't underestimate how difficult these tasks are "in the real world" (as opposed to like a game port which doesn't have real money on the line so far). And it is kind of a miracle these tools can even get in the ballpark and fool a lot of people.
You’re thinking: Should I put a protobuf here and build a separate service? Which side should own the protocol? How do I ensure the AI never couples concerns A and B? Can we prove this works with smarter automated testing, or do we need additional manual QA in this critical area before release?
Port my whole legacy codebase to another language with no guidance is a novelty problem. It is not comparable to what professionally managed AI driven development actually looks like.
If it were me, I would start by getting a summary of what problems the existing solution solves. I would also make sure that summary had the non-functional requirement. Maybe write a test suite that opens the page and does all the stuff.
Then ask the LLM to implement in another language, with those solutions in mind, but first coming up with reasons why the target language might not be as good, and where it might have useful features that the original language didn't. Use the test suite to check if things are working.
I honestly don't think you would land far away. You would have to answer a few questions along the way, but it would mostly be plain sailing.
I would expect to be able to do this with very little human attention, whereas a language rewrite two years ago might take a whole month, not including learning the new language.
If he truly believed in the idea of never reading the code, he'd fire all his developers and have the designers do it all - create very high level feature descriptions and iterate on them. Put your money where your mouth is
I think the main problem is that most people pushing for heavy AI delegation either
1. Don’t really care about system reliability or 2. Don’t understand the difference between code and system, and assume that a system must be reliable if some automated checks pass
I'm pretty sure this is sarcasm, but it's also a major factor in what differentiated slop from non-slop AI work: did a developer actually take the time to review and clean up generated code. Which is, itself, an extremely time consuming process
We got strictly better version of application for cheap. What is there even to complain about?
It definitely improves the results if you do think but I don't think "you still have to think" is a safe space that's going to save us from unemployment.
Who said anything about saving people from unemployment?
> I noticed unreliability during high load
This is pretty ambiguous. You still need to provide more context, such as which failure modes you are trying to benchmark.
So, I want to keep pulling the thread: is it worth reading—meaning, understanding—the code in order to prevent p95 latency from skyrocketing, avoid OOM death, and keep our apps reliable for customers?
The cost of understanding the code is ostensibly very high relative to just using a clanker (citation needed), so does paying that cost translate to value—what is it worth? Concretely, if vibe coding increases bugs and enshittification, but decreases spending without decreasing revenue (read: customers suffer, but they don't leave), do we still need to pay the cost of reading code? Is it better to invest in stickiness, lobbying, and market capture?
This comment shouldn't be read as advocating for this; it's a thought experiment about pragmatism and trade-offs, and reflects what I'm witnessing companies (and "programmers") asking themselves. I personally care a lot about understanding systems and code, and I believe I'm paid to do exactly that (I very much enjoy understanding code and nobody is paying me to run a business, so I may be biased in this). In everyday SaaS-land—not talking about critical safety systems—we've seen production databases destroyed, personal data leaked, platforms unwittingly exploited, and UX bugs creep into our operating systems, and yet the companies involved keep on keeping on.
Is thinking, meaning taking the time to actually understand what our systems are doing, going to increase their shareholder value?
Translating this from Linkedin-speak: Is it okay to subject our users to shitty software and while we use the money saved to bribe politicians, circumvent laws and buy our competitors to kill them?
Life, eh?
Not a good look tbh
But let's face it. I'm calling it right now: these are the twilight days of the "hacker culture". It's gotten us this far, and given us great things like Linux and the internet. But frontier models are already capable of solving medium-sized engineering problems in reasonable ways, including all the architectural decisions. You still need a few people with engineering chops, perhaps principal level or systems engineers, to guide the overall architecture of the system as a whole and make sure the pieces the LLMs build align and fit together correctly. Within a year or two, not even they will be needed. We're already past the threshold beyond which careful human review of LLM-generated code is impractical; the LLM spews it out so quickly that the human becomes the bottleneck, and reviewing all that code is more tedious and time-consuming than if a human had reviewed it in the first place! So the most efficient thing is for agentic loops to separately generate and scrutinize the code, with the generators and critics not sharing a context window.
Hackers are obsolete. Whether you're talking about the Richard Stallman variety or the Kevin Mitnick variety. Everything they do, a machine will be able to do presently, if it is not already. And the machine will be faster and more thorough. The suits and marketroids they viewed with such disdain, have won. They have been trying to build software factories since the 1970s. That dream is within reach. Our skills, traditions, and lore are being replaced with a sort of goetic demonology: knowledge about this month's correct way to invoke the demon, command it so it will do what you want, and bind it so it doesn't escape the circle cast on the floor and delete your production database.
And it's not just programming. Mathematics has already fallen; the hard sciences will too on a time horizon of about 10-15 years tops. Anthropic is setting up a wet lab; the biologists therein will pretty much act as peripherals to Claude, who will devise the experiments and analyze the data. As soon as we give the LLMs robot bodies, it will be needless—and woefully inefficient—for humans to invent anything at all. We're accelerating towards something. Whether it's Iain M. Banks's Culture or utter disaster I'm not sure which, but there's no stopping the train now.
Clanker code is no exception. Slop is slop because it's low effort.
There are projects I've put most of my career into (necessarily pre-AI), and those I definitely care about the code quality. I look at every line and I don't take clanker code without going around the mill a few times with it. There are lots of things in the code I like and things I don't like and want to refactor. I care about the code quality. But there are projects I've put only weeks or month into, and some only minutes. For most of my vibe-coded projects, I've only cursorily glanced at the code. For these projects I don't care about code quality as much as the architecture and interaction. The ones where I consistently tweak it to work how I want are...just better.
"I didn't do anything. I asked two frontier agents to make the most of Elixir and this is what they came up with. Please do send a PR to speed things up! But also realize that it's not exactly extra points for Elixir if frontier agents can't find the magic "go fast" switches."
<astronaut always has been meme>
> It’s part of a herd behavior, to signal an ideological position.
In my experience, as I alluded to, it’s not commonly used by people who are thinking seriously for themselves about the benefits, risks, and consequences of AI and the economic activity surrounding it.
It's much more nuanced, after all. Example:
This hasn't been the case for months now. In Codex, if the model is having trouble understanding what you meant, it will ask you followup questions before going ahead.
I have personally encountered GPT-6 Astra ask these exact questions before proceeding!
Like most criticism of AI, this is woefully outdated.