Indian here (~15+ years in tech). I've seen this behavior a lot, and unfortunately, I did some of this myself earlier in my career.

Based on my own experience, here are a few reasons (could be a lot more):

1. Unlike most developed countries, in India (and many other develping countries), people in authority are expected to be respected unconditinally(almost). Questioning a manager, teacher, or senior is often seen as disrespect or incompetence. So, instead of asking for clarification, many people just "do something" and hope it is acceptable. You can think of this as a lighter version of Japanese office culture, but not limited to office... it's kind of everywhere in society.

2. Our education system mainly rewards results, not how good or well-thought-out the results are. Sure, better answers get more marks, but the gap between "okay" and "excellent" is usually not emphasized much. This comes from scale problems (huge number of students), very low median income (~$2400/year), and poorly trained teachers, especially outside big cities. Many teachers themselves memorize answers and expect matching output from students. This is slowly improving, but the damage is already there.

3. Pay in India is still severely (serioualy low, with 12-14+ hour work days, even more than 996 culture of China) low for most people, and the job market is extremely competitive. For many students and juniors, having a long list of "projects", PRs, or known names on their resume most often the only way to stand out. Quantity often wins over quality. With LLMs, this problem just got amplified.

Advice: If you want better results from Indian engineers(or designers or anyone else really), especially juniors (speaking as of now, things might change in near future), try to reduce the "authority" gap early on. Make it clear you are approachable and that asking questions is expected. For the first few weeks, work closely with them in the style you want them to follow.. they usually adapt very fast once they feel safe to do so.

This sounds like a real cross-cultural mismatch, but it’s doing too much work with nationality alone. In a lot of Indian (and broader South Asian) work contexts, questioning instructions can be read as challenging authority or admitting incompetence, so people default to executing without asking. That’s often reinforced by education systems and contractor dynamics where producing something quickly feels safer than pausing to clarify.

Add in time zones, language friction, and fear of losing work, and "just run with it" becomes a rational strategy. Meanwhile, many Western workplaces treat clarification and check-ins as professionalism, so the behavior reads as strange or careless.

The key point is that this usually isn’t lack of curiosity or reflection, but risk management under different norms. The pattern often disappears once expectations are explicit: ask questions, check back, iteration is expected.

It is cultural - the whole "not losing face" thing. In a project, I once was squad lead - I was onsite, my squad members were in Bangalore of course. Same experience as you. Once I wanted to talk about a piece of code that we need to improve and refactor, and I was acting in good faith calling the dev that commited that code. When I braught up the code on my screen to start a pair programming, he immediately denied having written the code. Unfortunately for him, being a junior, he did not know about git blame - I entered it in the terminal and his name showed up on that code. Still, he would simply just deny that he wrote it. I then took the git commit hash and looked it up in gitlab, able to bring up the MR he created and the reviewer (wasn't me). Even with that on screen, he still denied being the author - with no arguments or alternative reasoning, he just constantly would repeat "No, I haven't written that". "No no, but I haven't written it". I pulled even the JIRA ticket up, that was about that feature and guess what - he was the assigne and moved it to "In Progress" and "Done". Still with that on screen all I got was a "no, haven't written it".

I had more of those interactions, and we also exchanged some of the indian devs (they were sold to the client by a big consulting group, and immediately replaced by someone else if we wished). I later found out, people that I have had replaced in my sqaud for not being qualified, ended up in different teams in the same corporation, they were basically just moving around inhouse.

After a few month in the project I swore to myself never to work with offshores again. And as a side note, the bank I did the project with, does not exist anymore :)

My guess would be yes, it's cultural. I'm not Indian but spent about 5 months there. Overall my impression was that people act much more on direct feedback.

It would be typical to do the first thing that comes to mind, then see what happens. No negative feedback? Done, move on. Negative feedback? Try the next best thing that makes the negative feed back go away.

People will not wonder whether they might bother you. Just start talking. Maybe try to sell you something. That's often annoying. But also just be curious, or offer tea. You react annoyed and tell them to go away? They most likely will and not think anything bad of it. You engage them? They will continue. Most likely won't take "hints" or whatever subtle non-verbal communication a Westerner uses.

I found it quite exhausting in the beginning, it feels like constantly having to defend myself when I want to be left alone. But after I started understanding this mode and becoming more firm in my boundaries, I started to find it quite nice for everyday interactions. Much less guessing involved, just be direct.

Professionally I haven't worked much with Indians, but my expectation would be that it's necessary to be more active in ensuring that things are in track. Ask them to reflect back to you what the stated goal is. Ask them for what you think are obvious implications from the stated goal to ensure they're not just repeating the words. Check work in progress more often.

That is a cultural thing, and one of the first things you learn to handle when working tightly together with Indians as an outsider.

I can't remember all the techniques but a simple trick is to ask them to repeat their understanding back to you before they start working on a thing.

But I don't think it's connected to sending "malicious" reports. That seems rather to be to pad their resume and online presence while studying to get an edge in hiring.

Of course it’s cultural, they have to compete with thousands people just like them in environment where human life is cheap and anyone is replaceable. Any authority have huge weight, which comes from historical system how society is separated. And then any education they receive assumes cheating at exams, then cheat with CV, then cheat with work they do. It’s all about appearances.

I used to work with colleagues from China in contracting and I had the same experience with them. If they don't know something they have hard time saying that they don't know something or don't understand something.

Ficticious Example could be

Q: is this car red? A: it's not green. Q: yeah I know it's not car. But is it red? A: today is Thursday.

One thing I leaned it's not worty pressing forward and causing a scene. Instead Use other ways of finding the information.

When guiding team members I always found it useful to have them explain back to me in their own words what they're tasked to do. It become immediately obvious if they were on the right track or not.

From a half-Indian friend of mine, he described this as "ask vs guess" culture. https://medium.com/redhill-review/navigating-ask-and-guess-c...

Ask culture scales a lot better in a fast changing world full of strangers. Guess culture saves friction, but only in situations where people are mostly guessing correctly because the social structure and expectations are fixed.

> never ever ask for clarifications, would never say they didn't know something, would never say they didn't understand something

I experienced this same thing working with offshore Indian contractors 20 years ago. Interesting to hear someone else echo my observations.

I recently heard from a friend that this is due to something called "izzat". Admitting any sort of wrongdoing would reflect poorly on them and their family, to the point they would rather lie or do the wrong thing than damage their family's reputation.

I had the same exact experience with an Indian contractor. I requested that he used git instead of Shopify CLI for his changes to a store's theme. He acknowledged my request but kept using the CLI. I once again asked him to use git and even offered a detailed, step-by-step guide on how to pull, branch and then push changes. He absolutely ignored everything and simply kept using the CLI. That was actually amazing to witness. The only hypothesis I have is that it's some kind of cultural thing where asking for help is worse than doing the opposite of what's expected from you. I don't know, but your story supports my hypothesis.

I believe it has to do with saving face.

I've worked with mixed nationality teams at a certain 4 letter austinite corporation a couple thousand moons ago. One thing in common with my Asian colleagues back then (many of which i still keep in touch with to this day), is that they would usually refrain from saying things that could rock the boat or disappoint you. If they lacked knowledge for the task at hand, they wouldn't let you know. If they were late on a delivery, they'd insist it would be ready by a certain date. This led to situations where other regional managers would have to plan contingencies to work around the issue.

This is called "saving face"[0] and it's very common in some Asian cultures. Western societies prefer directness, and eastern ones prefer harmony.

0: https://en.wikipedia.org/wiki/Face_(sociological_concept)

There are also a lot of Indian students (there are 1.4bn Indians). There are lots of IT jobs, therefore presumably lots of IT students, and unlike in China Internet access (e.g. to GitHub) is not restricted.

How much are the contractors being paid?

The people having a terrible time with Indian contractors always deal with folks making 3k-10k USD/year. Of course the quality is bad.

For reference:

Good Indian devs out of college make atleast 30k USD. Good senior devs make atleast 50k. The really good ones make much more. Most American companies outsource to bottom of the barrel contracting companies like Infosys.

This is hilarious and reminded me of the two stints I had in India, for about 8 months in total at the turn of the century. I was a hippy traveler and asking directions for almost anything was par for the course. I never had anyone local say they didn't know where something was once asked, even though me following their directions lead to the intended target maybe 10% of the time. It was funny and infuriating at the same time :)

Indian students were the reason that Google's Hacktoberfest was critiziced and ultimately terminated

Indian students have a long history of disrupting free/libre projects, this is nothing new

It's desperational. The desperation of not having to lose any contract. The desperation of being just one bad year away from being on the streets and having to live a terrible life (no food security).

For students, often there is no pathway to actually become good due to lack of resources. So, the only way is to fake it into a job and then become good.

Yes it's "fake till you make it" without the making part

Askers vs Guessers: https://www.theatlantic.com/national/2010/05/askers-vs-guess...

selfishness, laziness, lack of self-awareness, lack of shame, etc are obviously universal traits. But cultures absolutely reinforce them to different degrees. Many cultures around the world are built around the sorts of behaviors we both described.

Whereas other cultures have at least some (if not a lot of) resistance to it - eg publicly ridiculing when people step flagrantly out of line. This is good. My impression is that British culture is like this - "taking the piss", or worse, out of people whose egos start to get too large

Edit: what about this comment could possibly be worth a downvote...? Not that I care about points, but it just seems to be an objective assessment of human nature and cultures, without even singling out any cultures that need improvement.

Possibly as a consequence of this, what I have observed working with Indians is a very hierarchical structure in which you have a "lead" or "architect" who spells out what to do and how to do it in minute details and micromanages, and "devs" who execute as instructed.

I worked at a company where we had a untouchable manager who had some Brahman caste devs report to him and they absolutely HATED this.

I think it's mostly not cultural but just bad engineers lying. IT jobs pays the best in India, and it attracts people who have no skills in IT to just fake their way in.

So for every good developer in India there are probably 20 bad ones who have no idea what they are doing.

>This is generally fine as long as they know what it does.

Thanks to their LLM reliance they'd soon not know what it does, and forget even the little they know about coding

> Resume glorification and LinkedIn / GitHub profile attention do that.

I wondered why people would video themselves going around slapping strangers in public then shouting "its just a prank bro" - turns out it works.

Regard to code maintenance:

I’m actually of the mind it will be easier IF you follow a few rules.

Code maintenance is already a hassle. The solution is to maintain the intent or the original requirements in the code or documentation. With LLMs, that means carrying through any prompts and to ensure there are tests generated which prove that the generated code matches the intent of the tests.

Yes, I get that a million monkeys on typewriters won’t write maintainable code. But the tool they are using makes it remarkably easy to do, if only they learn to use it.

You've been getting PRs? All I've ever seen is "can you assign this issue you me" spam and then disappear. I was nice to them for years but now I just delete the comment and block the users.

This is precisely what we've seen

Those maintainers should be using LLMs to crate their breakup letter with the Issue/PR submitters!

worked well for a bit. but then the program became popular and that’s when it hit the curb. terrible loss, imo. it was a brilliant idea to encourage open source work with a token reward. it relied heavily on good intentions, which quickly disappeared with the popularity.

Everything about this is exactly what is happening in OWASP repos.

It’s not just the Internet. It’s politicians and businesspeople and more generally, shameless citizens.

There’s a lot to dislike about shame as an enforcement mechanism but I’m starting to miss some of the upside it delivered.

Internet reputation became easy to launder, thus meaningless.

> capitalism doesn't give a f8ck anyways

It doesn't until suddenly it does. A glut of junk can eventually trigger a flight to quality.

Sadly, possibly not on a timeline which works for a given individual.

Because their usernames are Indian and profiles have links to Indian universities, and sometimes descriptions of the 101 classes they're currently taking. That doesn't stop them from saying things like "I see this sort of vulnerability all the time"

Ever been on Stack Overflow before LLMs became a thing?

Sir, I agree with moralestapia. Not a singular one of the 20 lakh lines in the PR were written by ChaiGPT.

"Students" sounds very speculative. "Indian" likely based on usernames, which are often a South Asian first name followed by a random integer.

Subjecting every real contributor to the "AI guardian" would be unfair, and shadow banning is ineffective when you're dealing with a large number of drive-by nuisances rather than a small number of dedicated trolls. Public humiliation is actually a great solution here.

the only way to stop a bad guy with a llm is with a good guy with a llm

I intensely dislike the idea that we need more AI in order to deal with AI.

If I ever need to start using an AI to summarize text that someone else has generated with AI from a short summary, I'm gonna be so fucking done.

> GitHub is under Microsoft’s AI division.

Finally an explanation to why GitHub suddenly have way more bugs than usual for the last months (year even?), and seemingly whole UX flows that no longer work.

I don't understand how it happens, do developers not at least load the pages their changes presumable affects? Or is the developers doing 100% vibe-coding for production code? Don't get me wrong, I use LLMs for development too, but not so I can sacrifice quality, that wouldn't make much sense.

So much for the curation argument of the price justification of professional journals.

Medical? What's the point? I'm happy with 98% of doctors being able to handle known conditions and only the few percent that are really interested to do research.

Not necessarily. Reviewing an issue report is already enough time. Reviewing a patch is even more developer time.

The problem they had before was a financial incentive to sending reports, leading to crap reports that wasted time to review. Incentivizing sending reports + patches has the same failure mode, but they now have to waste even more time to review the larger quantity of input.

Anyway, for most cases I'd bet that Daniel can produce and get reviewed a correct patch for a given security bug quicker than the curl team can review a third-party patch for the same, especially if it's "correct, but ai-written".

> Even if a machine generates them.

Why? If it is a purely machine generated report there is no need to have dozens of third parties that throw them around blindly. A project could run it internally without having to deal with the kind of complications third parties introduce, like duplicates, copy paste errors or nonsensical assertions that they deserve money for unrelated bugfixes.

A purely machine generated report without any meaningfull contribution by the submitter seems to be the first thing you would want to exclude from a bug bounty program.

I've read this idea that we could make people pay for security reports a few times here on HN (and you get back the money if the report is deemed good). That feels very wrong.

If I find a security issue, I'm willing to responsibly disclose it, but if you make me pay, I don't think I will bother.

Punishing bad behavior to disincentivize it seems more sensible.

> Even if a machine generates them.

That sounds wonderfully meritocratic, but in the real world, a machine generating it is a very strong signal that it's bullshit, and the people are flooding maintainers using the machines. Maintainers don't have infinite time.

To be clear, no, it is not, because of the opportunity cost of all the other slop. That's what this is all about.

The discussion requirement is often to prevent disappointment, waste of time, and anger, when maintainers simple close PRs, because it's not the direction they want the project to go. A lot of people will take this very personally, so it's much better to have a conversation about it beforehand.

Plagiarism is much different than collaborating on open source projects but I'm glad that calling them out worked.

That's sort of the whole point of this thought exercise, no? If shame worked in an environment with anonymous/pseudonymous users, then we wouldn't be here. The only people you stand to harm are the ones who attach their real identities to their profile (and they're more likely to be good faith IMO)

Besides, I've seen plenty of profiles here on HN who advertise their real name and espouse (in my view) awful takes that would most likely not fly in real life. I'd recommend reading this article[0] for an example of when people, with their real names exposed, can still cause a shitstorm of misunderstanding.

0: https://lwn.net/Articles/973782/

Then they should exclude specific groups from their CoC.

"You can't be a contributor if you're an Indian using AI".

I don't think this is the way ..

Heh.

I understand they people hate to to waste time. They should just be polite about it.

Or you know .. update or delete the CoC.