# Day 49 — YouTube 自动生成字幕（AUTO_GENERATED）

覆盖训练段 000–10 分钟。已去滚动字幕重叠。ASR 生成可能含识别错误。

==================================================

Good evening. I'm Vicramosa and you're watching the big question on NDTV Profit. This week at the India AI impact summit, there is this one message that has been standing out that AI is moving from your phone screen. It's moving to the factory floor. Yes, steel makers, they now run hundreds of AI models to improve quality and safety. There are manufacturers who are scaling predictive maintenance and digital twins and process automation. Logistics players, they're using AI and they are optimizing their supply chains in real time. So

across corporate India, 76% of companies now expect Genai to deliver major business impact and nearly half they already have live use cases [music] in production. production. Deployment speed is now the biggest driver of any AI decision. And yet for millions of Indians, the most visible face of AI is still consumer tech. Social media algorithms, the recommendation engines, the creator tools, the generative [music] AI apps. That's where data pools are deep, users are global, and platforms scale much faster. The government on its part is pushing hard

on the infrastructure side. India's computer capacity is rising as part of the India AI mission. There is this massive policy focus on ethical AI on health AI ecosystems and yes data platforms besides that big R&D push across agriculture education healthcare and we all know the scale at which India's technology sector is growing which is why tonight as investment decisions accelerate and policy priorities shift India faces a strategic fork in the road and what is uh the option for India whether or or not we should go

down the manufacturing route, whether we should build on that productivity index, whether we should build infrastructure on that count or do we go the consumer tech route, build our digital infrastructure, the digital services space which has turned out very well for India. Should we be building on that instead? The platforms are there. Do we make them global? All of these are the questions that we have to ask and we have a debate and for that we have an excellent panel like we do every day. Let

me introduce you to them. Arvin Gupta is joining us. He's head and co-founder of Digital India Foundation. Rajes Singh is also with us. He is of course group CEO at Desh Engineers Limited, Gasonet and Data Express. Magna Krishna is joining us, chief revenue officer at Videow AI. Rohit Bandhar is partner for technology consulting at EI India. Along with them we have Rishi Ahurval the CEO of team le's rect and Rohit Kumar the founder partner of the quantum hub is also joining us on the show. So

let's all begin with this quick round of short responses just a line or two from each of you so our viewers get absolute clarity on where you stand and if I can begin by asking Raj Singh. Mr. Singh do you think Indian manufacturers will genuinely adopt AI at scale in the next 3 years? I want a yes or no answer and you tell me what the biggest barrier to that would be. >> I think the answer is very clear yes and the barrier is basically more

about that the adoption of the policy perspective if this is getting right and right time >> I am certain many of the industries are going to adopt it. Arvin Gupta what do you say in one line should India prioritize AI for manufacturing over consumer tech tell us why too >> um you ask a question to uh uh me uh and I will give you an answer which is in the AI terms the computer science terms the answer is not a boolean zero and one the answer

is and we should do both >> both >> both >> and there is a lot of reasons for that this is not a problem of select ction. This is the problem of diffusion. This is the problem of adoption. And when you do consumer tech and there is a lot of winners that will come out of it that will create the the talent the deep talent required for doing harder problem industrial AI. Of course the question is not u whether we should be adopting AI and manufacturing.

The question is should be adopting it fast enough >> and should we be worried about other nations adopting faster than us. I think you're asking the wrong question. The real question is consumer tech gives us the skills and the returns to the VCs to invest in deeper tech and more patient capital. And I can give go on and on for the reasons but you wanted a very short answer. >> Yes, just to start things off but this is of course an important point. You're saying that

one would feed into the other and that's why you should have both. Mea what's your sense? Should India build a build this global consumer tech platform that is powered by AI or you think that window has closed for us? No, absolutely. I think uh the opportunity lies on both ends. We must adopt in the manufacturing sector as well as well as >> pick one. Pick one m now you need to do that. Where do we have the better chances of scaling? chances of scaling? >> Well,

I don't think that'd be fair because I think for economic strength we must focus on our manufacturing sector but for global dominance the consumer tech will play a large role. Rohit Pandharka should India invest in frontier scale AI models or should we focus on the small and the domain specific models that are tailored to address local problems. address local problems. Certainly the SLMs for India focus like what Sarom is doing with vision models for index scripts etc. But more than that I would say India should

build digital public AI infrastructure. So just like what we did with Aadhaar and UPI at DC, I think we need a GPU scale infrastructure for all the startups and enterprises to use in India. All right, Rishi Agarwal, what do you think is the bigger risk right now? Regulating AI too tightly or regulating it too late. I think regulating it too tightly would be the risk, Vikram. >> And why? We need to allow a lot of headroom for innovation. We're not yet there. Consumer use cases have

seen a lot of traction globally. Industrial use cases are just picking up in in general. India has not invested heavily in tech. If you look at it from a rest of a world perspective and this is our opportunity to let uh this tech power our margins, global competitiveness, cost and quality. Fair enough. Rohit Kumar let me ask you if you had to pick one for India if you had to do it in today's context choose between cheap access to compute or high quality and clean data

sets what would matter more according to you for India India >> uh cheap access to compute is what I would pick u again I do want to say that none of it is a dichotomy all of this has to happen in parallel so perhaps the question needs to change uh but I would think that we will get our act together with the data so it's only a matter of time. time. >> So this is a matter of time. It's about priorities. There's no taking away from

the fact that we need both. Can we have both? What should we place as a priority? Bray Singh, in heavy engineering and industrial services, are you seeing real bottom line impact from AI already? And is that impact more meaningful than what we are seeing in consumer tech? consumer tech? Indeed I think if you look at uh overall contribution of adopting the AI is actually impacting bottom line significantly. There are cost advantage there are margin the efficiencies are improving. So I think in nutshell if you look

at industry particularly heavy industry is getting lot of lot of advantage of adopting AI. Mna from the consumer AI and the media tech side are we in a sense underestimating how powerful consumer platforms are in the way that they shape data in the way they drive demand and the global influence for India as well is that at stake >> I I think we are not underestimating it and I think people do understand the importance of the data and the data flywheel and how important it's going

to become in uh in directing the way our economy grows. Uh so so understanding uh the importance of data and understanding the importance of data flywheel and how user data is going to transform the personalization, the content creation, the social uh interactions is it's it's very important and and if we don't understand it today, yes, it'll be too late. >> So this is a moment where we need to kind of uh ensure that we are firing on all cylinders. AI is that moment pace like one

of you said is going to be the key factor. Now where are you going to apply pace? Do they work at cross purposes at any point? Are we providing enough infrastructure for growth for both bits to kind of rise and scale? Rohit across your clients at EY where is the serious capital actually flowing today because that's going to be an important driver. Is it factory flaws? Are they digital platforms? And what does that signal when you see those capital flows? What is that signaling about the

kind of priorities that India has placed when it comes to AI? What What >> we are seeing is a high ROI use cases for workflow automation is where the capital is flowing either through factory floor automation or through the software automation for back offices. For example, wherever you see people playing a lot of effort in terms of doing something on the floor, analyzing cognitive tasks and data, for example, what should I do next or what is happening? What is likely to happen? How do I improve

my output? Whether this is in the bank or at the manu manufacturing, >> we seem to uh we seem to have a very patchy connection over there and we'll try and rectify that. But just taking away from that uh point that Rohit is making right now rishi from um this entire employment standpoint which AI

# Day 49 — 火山引擎 录音文件识别（volc.seedasr.auc）

豆包大模型 ASR，异步接口。可能含识别错误，仅供听力对照。

==================================================

Good evening.

I'm vikramosa and you're watching the big question on NDTV, profit.

This week at the India AI Impact Summit, there is this one message that has been standing out that AI is moving from your phone screen, it's moving to the factory floor.

Yes.

Steel makers, they now run hundreds of AI models to improve quality and safety.

There are manufacturers who are scaling predictive maintenance and digital twins and process automation.

Logistics players, they're using AI and they're optimizing their supply chains in real time.

So across corporate India, 76% of companies now expect Gen AI to deliver major business impact and nearly half they already have live use cases in production.

Deployment speed is now the biggest driver of any AI decision.

And yet for millions of Indians, the most visible face of AI is still consumer tech.

Social media algorithms, the recommendation engines, the creator tools, the generative AI apps.

That's where data pools are deep, users are global and platforms scale much faster.

The government on its part is pushing hard on the infrastructure side.

India's computer capacity is rising as part of the India AI mission, there is this massive policy focus on ethical AI, on health, AI ecosystems and yes, data platforms.

Besides that big R&D push across agriculture, education, healthcare.

And we all know the scale at which India's technology sector is growing, which is why tonight, as investment decisions accelerate and policy priorities shift, India faces a strategic fork in the road.

And what is the option for India?

Whether or not we should go down the manufacturing route?

Whether we should build on that productivity index, whether we should build infrastructure on that count or do we go the consumer tech route, build our digital infrastructure, the digital services space, which has turned out very well for India?

Should we be building on that instead?

The platforms are there.

Do we make them global?

All of these are the questions that we have to ask and we have a debate.

And for that we have an excellent panel like we do every day.

Let me introduce you to them.

Arvind Gupta is joining us.

He is head and co founder of Digital India Foundation.

Rajesh Singh is also with us.

He is of course Group CEO at Dinish Engineers Limited, Gasonet and Data Express.

Meghna Krishna is joining us, Chief Revenue officer at Videoverse AI.

Rohit Bandarkar is partner for technology consulting at Evi India.

Along with them, we have Rishi Aggarwal, the CEO of Teamlease Regtech and Rohit Kumar, the founder partner of the Quantum Hub is also joining us on the show.

So let's all begin with this quick round of short responses, just a line or two from each of you so our viewers get absolute clarity on where you stand.

And if I can begin by asking Prajesh Singh.

Mr Singh, you think Indian manufacturers will genuinely adopt AI at scale in the next three years?

I want a yes or no answer and you tell me what the biggest barrier to that would be.

I think the answer is very clear, yes.

And the barrier is basically more about that the adoption of the policy perspective.

If this is getting right and right time, I'm certain many of the industries are going to adopt it.

Arvind Gupta, what do you say in one line, should India prioritize AI for manufacturing over consumer tech?

Tell us why, too.

You ask a question to me and I will give you an answer, which is, in the AI terms, the computer science terms, the answer is not a boolean, 0 and 1.

The answer is, and you should do both, both.

And there is a lot of reasons for that.

This is not a problem of selection.

This is a problem of diffusion.

This is a problem of adoption.

And when you do consumer tech and there is a lot of winners that will come out of it, that will create the, the talent, the deep talent required for doing harder problem industrial AI.

Of course, the question is not whether we should be adopting AI and manufacturing.

The question is, should be adopting it fast enough and should we be worried about other nations adopting faster than us?

So, so the pace is asking the wrong question.

The real question is, uh, consumer tech gives us the skills and the returns to the vcs to invest in deeper tech and more patient capital, right?

I can give go on and on for the reasons, but you wanted a very short answer.

Yes, just to start things off.

But this is, of course, an important point.

You're saying that one would feed into the other and that's why you should have both.

Magna, what's your sense?

Should India build us build this global consumer tech platform that is powered by AI or you think that window has closed for us?

No, absolutely.

I think the opportunity lies on both ends.

We must adopt in the manufacturing sector as well.

But pick one, pick one, Magna.

You need to do that.

Where do we have the better chances of scaling?

Well, I don't think that'd be fair because I think for economic strength, we must focus on our manufacturing sector.

But for global dominance, the consumer, tech will play a large role.

Rohit Pandharkar, should India invest in frontier scale AI models or should we focus on the small and the domain specific models that are tailored to address local problems?

Certainly the SLMs for India focus like what Sarom is doing with vision models for index scripts, etc.

But more than that, I would say India should build digital public AI infrastructure.

So just like what we did with Adhar and UPI and ONDC, I think we need a GPU scale infrastructure for all the startups and enterprises to use in India.

All right, Rishi Aggarwal, what do you think is the bigger risk right now?

Regulating AI too tightly or regulating it too late?

I think regulating it too tightly would be the risk, Vikram, and why we need to allow a lot of headroom for innovation, right?

Not yet there.

Consumer use cases have seen a lot of traction globally.

Industrial use cases are just picking up in, in.

In general, India has not invested heavily in tech.

If you look at it from a rest of a world perspective, and this is our opportunity to let this tech power our margins, global competitiveness, cost and quality.

Fair enough.

Rohit Kumar, let me ask you, if you had to pick one for India, if you had to do it in today's context, choose between cheap access to compute or high quality and clean data sets.

What would matter more according to you, for India?

Cheap access to computer is what I would pick.

Again, I do want to say that none of it is a dichotomy.

All of this has to happen in parallel.

So perhaps the question needs to change.

But I would think that we will get our act together with the data, so it's only a matter of time.

So this is a matter of time.

It's about priorities.

There's no taking away from the fact that we need both.

Can we have both?

What should we place as a priority?

Rajasingh in heavy engineering and industrial services?

Are you seeing real bottom line impact from AI already?

And is that impact more meaningful than what we are seeing in consumer tech?

Indeed, I think if you look at overall contribution of adopting the AI is actually impacting bottom line significantly.

There are cost advantage, there are margin.

The efficiencies are improving.

So I think in nutshell, if you look at industry, a particularly heavy industry is getting a lot of, lot of advantage of adopting AI Magna from the consumer AI and the media tech side.

Are we in a sense underestimating how powerful consumer platforms are in the way that they shape data, in the way they drive demand and the global influence for India as well.

Is that at stake?

I think we are not underestimating it and I think people do understand the importance of the data and the data flywheel and how important it's going to become in, uh, in directing the way our economy grows.

Uh, so, so understanding, uh, how the importance of data and understanding the importance of data flywheel and how user data is going to transform the personalization, the content creation, the social, uh, interactions is, it's, it's very important.

And, and if we don't understand it today, yes, it will be too late.

So this is a moment where we need to kind of ensure that we are firing on all cylinders.

AI is that moment pace, like one of you said is going to be the key factor.

Now where are you going to apply pace?

Do they work at cross purposes at any point?

Are we providing enough infrastructure for growth for both bits to kind of rise and scale.

Rohit, across your clients at Evi, where is the serious capital actually flowing today?

Because that's going to be an important driver.

Is it factory floors?

Are they digital platforms?

And what does that signal?

When you see those capital flows, what is that signaling about the kind of priorities that India has placed when it comes to AI?

We are seeing is a high ROI use cases for workflow automation is where the capital is flowing either through factory floor automation or through the software automation for back offices, for example, wherever you see people playing.

A lot of effort in terms of doing something on the floor, analyzing cognitive tasks and data, for example, what should I do next?

Or what is happening, what is likely to happen, how do I improve my output, whether this is in the bank or in the manufacturing.

Manufacturing, we seem to, uh, we seem to have a very patchy connection over there and we'll try and rectify that.

But just taking away from that point that Rohit is making right now, Rishi, from this entire.