The AI Case Against Indian IT Ignores What Indian IT Actually Does (indiadispatch.com)
- Reference: 0180863916
- News link: https://slashdot.org/story/26/02/26/1714241/the-ai-case-against-indian-it-ignores-what-indian-it-actually-does
- Source link: https://indiadispatch.com/p/in-defense-of-indias-it-industry
But the core claim that India's entire $205 billion software export industry [1]rests on cheap labor is roughly 15 years out of date , an analysis argues, custom application maintenance alone accounts for about 35% of a typical Indian IT firm's revenue, per HSBC, and enterprise platforms require deterministic outputs that probabilistic AI systems cannot wholesale replace. HSBC estimates gross AI-led revenue deflation for the sector at 14-16%, a measured headwind rather than an extinction event. The story adds:
> 24 years of software export data that has never posted a decline, $200 billion in annual revenue, partnerships with the very AI labs whose products are supposed to be the instrument of the sector's destruction, possibly a new $1.5 trillion market category emerging at the intersection of services and software, and the largest U.S. corporates in the middle of mapping their entire workforces into process architectures that require technology partners to modernise. I think India's IT is going to be fine.
[1] https://indiadispatch.com/p/in-defense-of-indias-it-industry
15 years out of date? (Score:5, Insightful)
Why the hell do you think American companies continue to contract Indian programmers? It's ALWAYS been about cost. Just that sentence turned me off to the whole article. Nothing has changed in that regard. What we see with our Indian contractors are a bunch of very technically proficient individuals who are utterly disconnected from the actual processes they are supposed to be considering. Oh.... and roosters crowing during meetings. Always with roosters.
Re: (Score:2)
Suppose the counter argument would be that Indian developers are not *solely* providing outsource work to American companies.
Absolutely concur that nearly all offshore, outsourcing outfits are basically there to grift American companies with dubious credentials and work.
Based on experienced, occasionally in that stupid setup, you'll have a really useful and effective person for a while, but ultimately they go 'poof' to get higher pay at a more proper software development shop. So presumably those jobs that
Re:15 years out of date? (Score:4, Interesting)
Having been involved with countless Indian IT outsourcers over the years, There is some truth there but it's poorly worded. One of the things I used to say, and my timeline roughly matches up with theirs, is that the gold mine of skilled labour arbitrage has indeed been gone for about 15 years. Since then, if someone assumes they can go hired a skilled IT resource in Indian for any savings over a North American of similar skill, they are sorely mistaken.
Why? Well, the smart and skilled Indians (many but no doubt a minority in their vast employee counts), are just that - smart and skilled. They aren't going to sit around and accept 50% of the wages they can see the same person in North America makes. All those skilled resources have either immigrated here, or otherwise just don't accept rates below their peers elsewhere in the world now. Actual talented IT people the world over have indeed known their value and haven't accepted less for 15 maybe more years now.
The only thing the Indian outsourcers have left to offer is is just raw body count, and they can deliver huge amounts of bodies on a dime. To be fair, you can't get this in North America, hiring is hard. That said, I truly believed they just rounded up anyone who can fog a mirror off the street and had them listed as an experienced IT resource within hours. They would know one specific thing and literally nothing else in the entire technical realm. I can't imagine there is an actual school you graduate from, and know how to create a very specific type of SQL query, and not a thing else about computers at all. Seriously, it was shocking.
And what they are good at, is convincing executives these are skilled resources and fleecing them dry while still being able to show a cost-saving business case. This is many years ago, but I managed to find out one major Indian firm was charging my major North American firm $85/hr as the "blended rate", which is basically an all-in rate for a resource with their "management fees", etc. An easy sell to sr. management who pays North American professionals ~$150/hr burdened rate (included all overhead). That said, no doubt the actual worker received a fraction of that, and while cheaper, the value was absolutely nowhere near that cheaper cost.
Re:15 years out of date? (Score:5, Interesting)
Oh, I'll also add that for many years we tried in vain to increase the general competency, training, and skill level of the resources they gave us; leveraging every contract requirement, required education and training we could, but it is 100% futile. They can manufacture any credentials instantly, and it is impossible to exceed the lowest common denominator of skills on their teams.
It was quite common in the steady stream of bodies they were throwing at your projects and contracts to come across someone new who clearly did have talent and skill. No matter how hard you tried to latch on to that individual, again they are the few smart ones - it took no time at all for them to become aware of their level of ability compared to the others and they'd leave in no time for better pay and treatment, as they should clearly do. They'd quickly be replaced by a handful of complete morons.
Not to disparage ... (Score:4, Interesting)
There are many great, hard working, smart and motivated Indian IT workers. More power to them. Live long and prosper.
What Indian IT firms will do it employ the most brain dead, cheapest, uneducated and unmotivated individuals and sell them as experts. And somehow, they get away with it. Distance also makes it harder to hold them to account.
Says as much about the quality of the management that I have been inflicted with, as it does about the ethics of the Indian IT firms.
Free clues: In any so-called low wage country, the good ones have already left the country. What you're left with is "cheap", in all senses of the word. You'll also have to deal with major cultural clashes. You'll have to spend a lot of time micro-managing and evaluating personnel, more-so than if you'd had outsourced locally, or had direct employees. Caveat emptor.
Re: Not to disparage ... (Score:3)
So absolutely true. Some folks who immigrated from India to the US that I have worked with were some of the best developers and tech people I've known. The numerous Indian based outsourcing firms, however, we're never above mediocre. Maybe one person is fairly good but it's always diluted with a sea of clueless newbies who say yes, I understand and have no idea what is going on. The most comical has been seeing non-technical startup founders secure funding, pay an Indian outsourcing company to make the
Re: (Score:2)
> Some folks who immigrated from India to the US
Apologies for grammar policing, but in case it is educational: you emigrate from a place, and immigrate to a place.
Nonsense (Score:2)
It's not just "cheap labor", it's also quality, which is absolutely still a problem with "India IT" (both abroad and domestically).
You absolutely can replace H1B-quality work with an AI - at less cost, time, and effort.
The common theme is: they have one competent developer who's a complete slave driver and he instructs an army of incompetents what to do. The incompetents are paid laborer wages and the competent guy (who may be a $300k/year prior SV employee who wants to live in India) is making a comparable
Custom application support (Score:2)
You mean IT support? That's actually the most replaceable part of the IT market. AI can handle turning things off and on again with superior success rate.
Buh-bye, Tata (Score:2)
Fuck those guys.
Guess you can pay someone to say anything... (Score:2)
Pointing at 35% of your current work to explain why a likely significantly faster and lower cost replacement, that shows no sign of topping out in capabilities or interest to use it, isn't going to decimate your "not objectively better or faster, but hopefully cheaper" based industry... ballsy at least. Likely deluded too.
"Deterministic" (Score:1)
"enterprise platforms require deterministic outputs that probabilistic AI systems cannot wholesale replace". So, humans are deterministic now?
Re: (Score:3)
Some people do believe that to be the case. If you ever meet one make sure to slap them and tell them that it was their fate to be slapped and nothing could have changed the outcome. If the protest, slap them again. This exercise can be repeated for some, possibly determined, number of times or at least until your hand gets sore. Remember that you can quit anytime you want, but they apparently lack such luxury.
More seriously, even if humans are entirely deterministic I don't think we could function under
Re: (Score:2)
I think it is more alluding to the need that using AI to wholesale replace existing/legacy systems doesn't work because the outputs aren't consistent. AI can't do payroll with probabilistic checks and withholding.
Re: (Score:2)
Though I think AI proponents would claim that, *largely* that's not the idea, that they would hang their hat on the codegen, where a probabilistic engine is used to direct code changes that when done are determenistic.
That companies that outsource to offshore coding outfits, no *matter* the rationalization (despite what they say) will see the new hype is insourcing to random local dev guys that can LLM themselves to '10x engineers'. A few low end local devs will be cheaper than a lot of the remote devs and
Re: (Score:3)
I dunno, I bet you could ask a large language model to do the needful, and it would.
Re: "Deterministic" (Score:2)
Can we please stop saying that LLMs are 'nondeterministic'... It's misinformed. And it matters (because some people associate nondeterministism with some sort of magic, and others associate the word with inaccuracy -- neither of which are true). These models are clearly, architecturally, mathematically deterministic at a fundamental level, although this determinism is hidden from users. Let me explain. If you take any recent LL model (GPT, Llama, Gemini, etc...) and run it with 'temperature' set to zer