Why India lags behind in AI
Pretty sad imo
MuchosGracias
Stealth
7 months ago
He’s absolutely right. You ask any elite UG student what their northstar is and it’s usually the highest paying job. You have your outliers that go on to start unicorns but even those are nothing more than copies of American companies and models
I struggle to think of an Indian company that has made something truly new and unique
Scarcity mentality at its finest
Indian companies can build something great provided investment is made into it. Just look at UPI, it has put us miles ahead of other countries and it’s pretty robust because the government invested properly into it.
Now, Indian companies and shareholders are very value driven, they don’t invest unless they see immediate ROI in some way or the other. Only when this mindset changes, real research will happen in India.
MuchosGracias
Stealth
7 months ago
UPI was not the product of the elite UG fraternity. It was govt mandated and god bless its soul for what it’s done for this country.
Agree that value driven building will keep us followers only. We need huge influx of research to pivot our innovation compass
See more comments
We’ll be able to move mountains once we stop obsessing with IITians. They work their asses off to get into IITs to get high paying jobs, it’s as simple as that. Train tier-2,3 college students who have nothing to lose. You’ll achieve 100x more optimal results with 100x less investment.
Most IITians who want these high paying job end up leaving India anyways for whatever reasons. So we have a situation where the best and brightest minds end up leaving the country never to return. How then shall we make progress?
Kamlesh
Stealth
7 months ago
Except it's not 100x less investment, it's more than educating someone at a decent starting point compared to basically starting from scratch.
Why this prof isn’t building cutting edge AI in India? We all know the reasons. Start with yourself prof sahib .. 🤦♂️
Kamlesh
Stealth
7 months ago
Not every executor is a good teacher and vice versa
Kamlesh
Stealth
7 months ago
By that logic most people on grapevine should stfu with their Gyan on running startups when they've never been founders themselves
See more comments
TheOatmeal
Stealth
7 months ago
Apna pet paalega ya research karega admi baith ke.
Koi paisa hi nahi de raha research ka. Major fundings are happening in Europe and USA.
Phd schedule in India is horrible. People are pressurised alot in the curriculum.
Kamlesh
Stealth
7 months ago
Aise toh we shouldn't be launching satellites when we still have such widespread poverty.
Need to start some day or another. Better sooner than later.
Hermit
Stealth
7 months ago
Also, because of people like Bhavish Agarwal who believe playing to the tune of jingoism is going to create a popular sentiment within the country.
Yesterday, he tweeted about a reply from Krutrim and ChatGPT proudly and it clearly showcased Krutrim is light years behind as well the training dataset is biased.
AI models need to be fact based and not emotion based and as a country we are so governed and overwhelmed by emotions and virtue signaling that we would lap up any kind of populist idea.
Tweet link for ref :
https://x.com/bhash/status/1753986946827747653?s=46
Hermit
Stealth
7 months ago
I wonder for how long Krutrim would retain its unicorn status with such a LLM training set.
Bhavish thinks that he’s Elon Musk, yet he’s Elon Mast! He’s living in his own delulu.
Pvykey
Stealth
7 months ago
Correct 100%
Completed PG at IIT. Have couple of research papers in hand. Trying to go abroad.
Here, the JD itself contains 3-5 categories of AI job. One person has to do data analytics, ML Engineer, ML testing as well. They want all-in-one as package of very very low as possible.
It's not the case in US, Europe. Pay is very high but still they will take for a single job requirement.
Here, In Simple terms:
Low salary
Very high work requirements
Grub96
Student
7 months ago
Do you think GATE CS is still worth it? I'm not into abroad MS simply because of the high costs. Also how is MTech different from MS by research at IITs.
Pvykey
Stealth
7 months ago
I think it is. MS abroad costs more but 70-80% would get a job earlier. Nowadays it's like more people are giving MS and aren't placed yet. Mtech here costs less and can get same level of recognition here in India.
Mtech is more of like courses and we have to complete the courses and get grades whereas MS research is more like research projects and publications most of the time.
See more comments
As a former researcher let me tell you the problem. The MTech scholarship is 12k pm, the JRF is 31k pm
These are entry level salaries for non-elite students. And for the PhD you need to spend minimum 5 years in that salary bracket and you still won't get funding.
Plus there's a lack of great computers and software in the universities. Noone gets a mac or a powerful laptop to run the numbers at the hostel.
Plus publishing has a huge bias against Indians
We need to overhaul the research infra before we ask students to participate.
Also as a professor you're probably making 1.5l at the highest level. You get a lot of other money but that's the cap for you after 30ish years in the profession.
A good software engineer makes that while starting
Kamlesh
Stealth
7 months ago
Thanks for sharing, appreciate your insight. I guess we can start with demanding more R&D funding and ancillary programs from regulators for universities.
But funding crunch is a problem everywhere from what I've seen. Even doctors get barely anything for the amount of work they put in while studying and doing residencies.
Is it not possible for people to do independent research on the side?
Yep funding crunch is a worldwide problem but, people mitigate it with grants. Our grants are pretty low too.
I agree some independent research is possible but, I worked with AI and it requires compute. I was training a specific computer vision AI and it used to take around 3 days to train and re-train alone.
Now, if there were centres with databases from the government that would allow us to use them and research, maybe, but that will cost an insane amount of money, something the government probably won't give for a pipe dream
See more comments
Is there any field where India isn't lagging? Heck,we are a billion people and play one sport,and haven't won any intl trophies in that in the last decade. Is there any field where we are the best?
Sapthinker
Stealth
7 months ago
Kabaddi😎
If Mr. Amit Sethi wants me to work my ass off for 2-3 years to get into one of the world’s most premier colleges and end up making a fraction of what the likes of Elvish Yadav and Sourav Joshi makes then I have a bad news for him, I’m either moving abroad or killing myself.
I think we’re a generation or two behind in terms of qualified kids being able to take 0 to 1 startup risks or building something foundational in tech in India.
Everyone’s in a survival mode and once a generation passes through when they get wealthy, their kids can start taking risks or think on solving problems with first principles thinking.
It’ll take time cuz there’s no two ways about it. America has decades of headstart in terms of wealth and risk taking.
BinaryBillionaire
Stealth
7 months ago
In indian every other company is busy in just using buzz words like ‘AI’, ‘ML’ in their product advertisement, but in reality they don’t spend 0.01% of their revenue into R&D of it
Ofcourse there are few outliners
Dihaadi
Stealth
7 months ago
It’s about the incentives and how our people are psychologically conditioned towards maximising money. In the western society people are attuned towards building something bigger than an individual. Here we’re taught that innovation, science and technology are only means to an end not the path itself. Ask any Indian VC to fund 10 mn on just experimenting and innovating, they won’t. In the west, this culture starts from college endowments to angels to VCs . Haldiram investors can’t create a culture which incentivises risk taking in the name of innovation. For morons quoting shit like Krutim, you don’t understand the basics of venture capital in India
Kamlesh
Stealth
7 months ago
I do believe India has been able to break out of that cycle of money and survival in many parts. Otherwise we never would have gotten nukes or a space program.
It is the outliers who do it yes , but we also have more outliers due to sheer numbers. Just need to make sure our national ecosystem develops well for them to reduce brain drain.
Dihaadi
Stealth
7 months ago
Nukes and Space programs should have happened 15 years earlier than they happened and it was due to misaligned priorities
See more comments
Tbh, Majority of the Core Research is happening in the west as of now and on top of that the Chinese are having their own scholars working in such field for a long time now (they heavily invested in the development of computer vision models to be deployed inside every inanimate object having a camera and a microphone out there)
All we need is investment in innovation instead of mediocrity. We never support doing extra work. Not all UGs want max paying job, 2nd part of point 3 can be applied for UGs as well. Most of my friends are going for PhD in US are because they have been just writing and debugging some stupid SQL since last 2-3 years.
Kamlesh
Stealth
7 months ago
I have a friend doing PhD in US too, working on foundational research. He's pretty happy with the work and atmosphere there and said he has people willing to fund him if he chooses to startup too. But he's more into research, not really an entrepreneur.
ihateWFO
Stealth
7 months ago
The real number of people who can work with AI in India is probably less than 500, while the US has thousands
Demon
Stealth
7 months ago
Why isn't this professor producing great research work if he is the one not running behind money ?
And even the startup bro are running behind copy cat ideas and vc valuation
AI is one of the highest paid jobs, But we don't have that big investment for now because of various reasons. So people end up joining in US companies who has better capital.
totally correct.
couldn't agree more.
however I'd differ with EU as whole ... i have been seeing some research group in France coming up as a strong contributors. other countries...not so much though.
Discover More
Curated from across
