this post was submitted on 08 Dec 2025
393 points (96.2% liked)

Technology

4876 readers
484 users here now

Which posts fit here?

Anything that is at least tangentially connected to the technology, social media platforms, informational technologies and tech policy.


Post guidelines

[Opinion] prefixOpinion (op-ed) articles must use [Opinion] prefix before the title.


Rules

1. English onlyTitle and associated content has to be in English.
2. Use original linkPost URL should be the original link to the article (even if paywalled) and archived copies left in the body. It allows avoiding duplicate posts when cross-posting.
3. Respectful communicationAll communication has to be respectful of differing opinions, viewpoints, and experiences.
4. InclusivityEveryone is welcome here regardless of age, body size, visible or invisible disability, ethnicity, sex characteristics, gender identity and expression, education, socio-economic status, nationality, personal appearance, race, caste, color, religion, or sexual identity and orientation.
5. Ad hominem attacksAny kind of personal attacks are expressly forbidden. If you can't argue your position without attacking a person's character, you already lost the argument.
6. Off-topic tangentsStay on topic. Keep it relevant.
7. Instance rules may applyIf something is not covered by community rules, but are against lemmy.zip instance rules, they will be enforced.


Companion communities

!globalnews@lemmy.zip
!interestingshare@lemmy.zip


Icon attribution | Banner attribution


If someone is interested in moderating this community, message @brikox@lemmy.zip.

founded 2 years ago
MODERATORS
 

Article dated June 2025

top 33 comments
sorted by: hot top controversial new old
[–] fargeol@lemmy.world 132 points 1 week ago (3 children)
[–] Prove_your_argument@piefed.social 18 points 1 week ago* (last edited 1 week ago) (1 children)

If this were true(for all of them), AI companies would be way more competent than they are lol

[–] pahlimur@lemmy.world 8 points 1 week ago

Having had a team in India grafted on to help me with a project. AI would've been an equally useless time suck.

[–] Gork@sopuli.xyz 51 points 1 week ago (1 children)

GPT: Gujarati Professional Typist

[–] TheBat@lemmy.world 4 points 1 week ago

How dare you. I'm not Gujarati.😡

[–] Iheartcheese@lemmy.world 46 points 1 week ago (1 children)

They should have noticed something was wrong when the company asked for iTunes gift cards as payment

[–] quick_snail@feddit.nl 6 points 1 week ago

WHY DID YOU REDEEM IT

[–] riskable@programming.dev 42 points 1 week ago (1 children)
[–] henfredemars 23 points 1 week ago (1 children)

Artificial artificial intelligence.

[–] Cabbage_Pout61@lemmy.world 9 points 1 week ago (2 children)

What if the the employed Indian is using builder.ai to help him build said software to builder.ai?

[–] Vex_Detrause@lemmy.ca 9 points 1 week ago

Then there's only one super intelligent coder from the 700 that's actually doing work.

[–] DickFiasco@sh.itjust.works 3 points 1 week ago

That sounds like a perpetual motion machine.

[–] Meron35@lemmy.world 35 points 1 week ago (1 children)

AI = Actually Indians

I feel conflicted. On one hand, the underpaid work often carried out by Indians is getting recognised. On the other, this is also saying that the collective work of 700 Indians is no different to AI slop.

[–] Darohan@lemmy.zip 15 points 1 week ago* (last edited 1 week ago)

Well, look at this way. If I were being paid as little as they probably were to do what they were being asked to do whilst also not getting recognized for my contributions as a human and also being treated like machine by people who don't even speak the same first language as me, I would probably put out pretty sloppy work too.

[–] punrca@piefed.world 24 points 1 week ago (3 children)

It's not true, you can check this article debunking this claim written by Gergely Orosz (Pragmatic Engineer Blog) : https://blog.pragmaticengineer.com/builder-ai-did-not-fake-ai/

The racism against Indians is so normalized, trying to discredit and dehumanize Indians, it's horrendous! Be objective and have some empathy...

[–] AryaRaiin@lemmy.world 9 points 1 week ago

Thank you. The company failed due to bad upper management and so-so products on unfair deadlines that weren't novel. The workers themselves did just fine, considering the circumstances and bad leadership.

[–] jj4211@lemmy.world 4 points 1 week ago

Yeah, I was thinking it's impossible to fake AI codegen with humans, for however terrible it is, it vomits the code it generates quickly.

I of course take issue with:

LLMs were already more than capable of generating high-quality code.

That claim will remain unjustified so long as LLMs equally generate imagined function calls and such as much as they like 'real' function calls.

But in any event, it's also telling that despite ostensibly being a company focused around AI code gen, when it came to doing stupid in-house stuff that didn't need doing they went with a bunch of human developers. And many of those are knock-offs of projects that should be pretty firmly in the territory of LLM codegen to let you 'vibe code'.

[–] jaybone@lemmy.zip 18 points 1 week ago (2 children)

Outsourced Indian dev code was shit 20 years ago. So now it can be shit again. Progress.

[–] pulsewidth@lemmy.world 10 points 1 week ago* (last edited 1 week ago)

But now instead of blaming a junior dev they can hit you with the, "Our apologies for this issue - we are modifying the prompts for the model and re-running in-unit AI tests, will revert back to you", and charge slightly more.

[–] jj4211@lemmy.world 4 points 1 week ago

Offshored/outsourced code is generally shit regardless of where it comes from.

Question is generally 'how is it that using this company thousands of miles away is so much cheaper? Business folks like to think it's just because the country is poor, but no, the answer is that it is a grift. It is always a grift.

The good developers in that location? They are too busy doing work for real companies paying real money. Maybe less than an American makes, but not as dramatic as management imagines. So companies looking to offshore will never pay enough to get the actual talent in a geography. What happens when someone competent actually lands in one of these arrangements? Gone in 3-4 months after they get a real job after proving themselves.

So what developers are you getting? People who probably had the equivalent of a high school programming course and didn't really get it, but there's a company that will pay money to anyone with some arbitrary 'certiificate' that they can claim to clients means you are a trained software developer. They don't know your use case, and they don't care. They utterly fail at being competent, but who is going to call them on it? Management has no clue, and besides, it was their idea to outsource, so it has to be a good idea. Leftover technical talent that you kept on as a skeleton crew to supervise the outsourced effort? Oh they are just whining and trying to protect their jobs and being overly dismissive of the offshored staff because it suites their self-interest.

Offshoring is a big old grift wherever they set up shop.

[–] oce@jlai.lu 10 points 1 week ago* (last edited 1 week ago) (1 children)

There's also a lot of manual human labor to build LLMs, I was surprised. From exploited African data labelers who have to shift through the most horrific things the web has to offer and remain traumatized for life, to highly paid domain experts who work on improving final results so it ends up looking like a simulation of themselves.

[–] NotMyOldRedditName@lemmy.world 15 points 1 week ago

That sifting through crap isn't just an AI thing.

Pretty much any big site with user generated content and moderators has to have people who deal with seeing that shit.

It's a really shitty underpaid job.

[–] motor_spirit@lemmy.world 9 points 1 week ago
[–] ouRKaoS@lemmy.today 8 points 1 week ago

PAY NO ATTENTION TO THE MAN BEHIND THE CURTAIN!

[–] quick_snail@feddit.nl 8 points 1 week ago (1 children)

Wouldn't the time it takes be a give away?

[–] jacksilver@lemmy.world 4 points 1 week ago

It says they'd been tricking investors for 8 years, we'll before ChatGPT. I assume they overpromised in the early years, leveraging labor as a "stopgap" for the developing AI. Then things eventually caught up and they couldn't deliver, leading to the current situation.

Eight years is a long time, so who was "duped" when can really change the narrative of the headline.

[–] RizzRustbolt@lemmy.world 4 points 1 week ago

Mechanical Turks for everyone!

[–] Rhaedas@fedia.io 3 points 1 week ago

They're all code names "Al".

[–] ryannathans@aussie.zone 2 points 1 week ago

Old article regurgitated yet again