- 7 weeks ago
In this India Today special show from the World Economic Forum in Davos, top artificial intelligence innovators share their perspectives on the global AI race, AI's impact on the Indian economy and much more.
Category
🗞
NewsTranscript
00:00Global AI race heats up
00:03$470 billion invested by US in 2025
00:09$119 billion invested by China in 2025
00:15Tech majors in big battle for survival
00:19US stock markets ride AI wave
00:24Will AI steal your job?
00:27Will robots rule over humans?
00:31Is it a bubble or game changer?
00:34The AI Roundtable
00:39Hello and welcome friends to this India Today Business Today TV special
00:45coming to you from Davos at the World Economic Forum
00:49US President Donald Trump has hogged all the headlines
00:52with his repeated political disruptions
00:55but behind the headlines the real disruptor of our times
01:00could well be artificial intelligence or AI
01:04the other big talking point at Davos
01:07Which is why over the past five days
01:09we spent time talking to some of the best voices in this field
01:13asking them questions on what happens
01:16when will super intelligence come if at all
01:19what will be the impact on India and its economy
01:22can India also stand tall amongst the nations who are ahead in the field of AI
01:28and they had a lot of information and intelligence to share with us
01:33take a look
01:34I wanted to begin by asking you about
01:37according to you the single largest and most fundamental shift
01:41that is happening in human society
01:43on account of the advent of AI as we all know it
01:47Well first of all thank you Siddhartha for having me
01:49it's a pleasure to speak with both of you
01:51and I'm actually looking forward to being in India in a few weeks time by the way
01:54so very much looking forward to that
01:56This is such an exciting moment
01:59I mean I never imagined we would be here
02:02I mean I did my PhD in AI 30 years ago
02:05and we never imagined the kind of progress we're seeing today
02:08I think the reason this is so fundamentally transformational
02:12is because this is one of those rare technologies that has two features to it
02:18it's what economists call a general purpose technology
02:20which means every part of any activity across humanity
02:25will be using this technology
02:26whether it's individuals, companies, businesses, governments
02:29it's a general purpose technology
02:31the other part that's exciting is
02:33which is even more rare
02:35is it's an invention for inventing
02:37which is a different characteristic
02:39so you put all those things together
02:41so when I hear people making comparisons of the industrial revolution
02:44what you should actually do
02:46is to combine the industrial revolution with the enlightenment
02:49and combine the power of the science
02:52and the breakthrough innovations
02:54and the radical shifts that occurred with the enlightenment
02:57we should think about this as both a general purpose technology
03:00and an invention for inventing
03:02that's what's so amazing about this technology
03:05James that sounds like compressing centuries of development, progress, innovation
03:10into one movement
03:11am I correct in sort of...
03:13Oh you're absolutely correct
03:14I mean there's a real example of this
03:16in the sense that look at what AlphaFold did
03:19right because you know up until before AlphaFold
03:22you know thanks my colleagues got a Nobel Prize for this
03:25before that we only understood the structure of about 150,000 proteins
03:30because we're doing them one at a time
03:33in labs with X-ray crystallography one at a time
03:36but with AlphaFold we're able to do all 200 million proteins known to science
03:42now imagine if we'd kept going one protein at a time
03:45it would have taken probably a few hundred years
03:47to get to all of them
03:49with AlphaFold we did all of them in one go
03:51so you start to see the same kinds of things happen in material discovery
03:56one of our AI system genome
03:58was able to discover something like 2.2 million crystals
04:01and of those that were not known to exist before
04:05and of those we've been able to establish something like 300,000 of them are stable enough to be synthesized
04:11so example after example you're seeing this compression
04:15because now you can do scientific discovery at a scale and scope
04:20that's breathtaking and I think that's exciting
04:23and in a way it is democratizing science as well
04:27it's making it perhaps accessible to millions of more people than was the case before AI scaled up
04:36oh absolutely I'll give you a tangible example
04:39I mentioned AlphaFold
04:40so today when that breakthrough happened
04:44we open sourced and open access to the protein structures that we predicted
04:50and up until now so far something like 3.5 million researchers around the world have accessed it
04:56in fact I think about 180,000 of them are in India
04:59and now what that's done is that researchers who have not otherwise been able to work on certain kinds of problems
05:05either neglected diseases and a variety of things
05:08all around the world can now do that work
05:11so yes there's an incredible democratizing effect to enable all of humanity to take advantage of this technology
05:18and drive breakthrough process
05:20I'm very excited about this and I think one of the things that I get excited is when I see
05:24there's so much excitement in innovation in India
05:27and some of the companies and entrepreneurs and researchers we get to see
05:30I think the potential is going to be so exciting
05:33is that number that you said 180,000 if I got that right
05:37does that surprise you from India alone?
05:41no it doesn't
05:43in fact I want it to be even bigger than that
05:45what I meant by the reason why the example is interesting is because
05:48a lot of the research that would have been required to work on protein structures the old way
05:54would have required extraordinary infrastructure and very few researchers were doing that kind of research
06:00which meant a lot of problems and diseases in local communities in the global south
06:04were neglected because the researchers in those countries couldn't do that work
06:09so the fact that now 190
06:11the 3.5 million number that I mentioned of scientists
06:15they come from all 190 countries
06:18I think that's extraordinary
06:19so we now have scientists in all these countries working on problems that they would not have been able to work on before
06:26I think that's incredibly exciting
06:28the two areas that we have been hearing where AI is going to have this hugely transformative effect
06:33one is healthcare
06:34yes
06:35the other is education
06:36both of which in a country like India again become critical given the sheer size of our population
06:41the fact that large numbers of people still are denied quality healthcare and education
06:46do you believe AI will be able to bridge those inequalities in a way that perhaps was not imaginable again pre-AI
06:56or is that a dream that's being sold?
06:59it's a very ambitious dream
07:01the good news is that it's actually being tackled and addressed
07:03I'll give a few examples
07:05so in the area of healthcare
07:07one of the key issues in many parts of the world is the ability to do diagnosis of any whole range of things
07:15so one of the breakthroughs that we had a few years ago is what's called diabetic retinopathy
07:20where you're able to take a look at retinal scans and observe and diagnose a whole range of things
07:26from cardiovascular conditions and a whole range of things just from retinal scans
07:31and we've already so far in India actually done retinal scans with diabetic retinopathy of something like over 600,000 patients
07:38so the question is how do we scale that
07:40so we're working with partners in India to see can we get to where we can diagnose millions of people using this technology
07:47Tobaculosis is another
07:49so it turns out that something like 40% of the people in the world who have TB go undiagnosed
07:55and most of them live in the global south in India and Africa and Asia and other places
08:00and so we're actually able to show that with AI you can do very accurate TB diagnostics
08:06in fact we actually did some tests in India and in southern Africa to show that it worked
08:12so those are just examples where all of a sudden you're able to bring real solutions to places that otherwise
08:19either because of either skills or capability gaps or infrastructure gaps would not be able to provide healthcare to those places
08:27I think that's exciting and I think that's what that's that's the exciting potential that what we can do with these scientific breakthroughs
08:35and apply them to areas like healthcare I think it's very very exciting
08:39take even agriculture one of the things we're doing for example I think we've partnered with with Wadwani AI
08:46and we're doing work in agriculture where you can try to think about how do you use AI to actually build kind of new kinds of plant structures to be able to assist farmers
08:58so being able to bring these kinds of solutions where they're needed the most I think is one of the most exciting things
09:03and that's scalable all these models are eminently scalable because of the very nature of AI
09:08they're extraordinarily scalable I mean I you know one of the things I love about you know the partnership and collaboration work we're doing in India
09:14is you're able to actually deploy these things at scale we worked last year with the Indian Minister of Agriculture
09:21where we're in and with farmers because we had just developed this breakthrough ways to predict monsoons
09:27and we actually use our latest model neural GCM to predict monsoons actually worked with the with the ministry along with the University of Chicago
09:35and we actually showed that we could actually get monsoon warnings to 38 million farmers in India
09:41that was just the first trial so now we're actually hoping to be able to do that in many more places
09:47so what would your message be to state governments in India or indeed to corporates in India who are excited about AI at one level
09:56but not sure just how far to go to embrace it because familiarly AI also gets associated in our part of the world with potential loss of jobs
10:07it's the other running theme through a lot of our conversations will AI replace blue collar
10:13and indeed in some parts white collar labor what would you tell them to give them a sense of some sense of comfort
10:20that look AI is an enabler it's a disruptor but it's not necessarily inimical to job creation
10:28yeah I think I'd say two different two kinds of things first I think if you're a company one of the things that is exciting about the potential of AI is to enable businesses
10:41whether it's small medium-sized businesses large companies to drive innovation and productivity
10:45so I think it's companies and I think the big challenge there is how do companies actually drive adoption to adopt these tools to innovate their businesses
10:54so I think if you know if nothing else from an innovation growth competitiveness point of view companies should it should embrace the technology
11:04it's going to help their performance their innovation their growth and also ultimately all of that contributes to economic growth and productivity
11:11so you've got that kind of reason to focus on this
11:14I think with regards to the question of jobs I think you know all the research we've done I've done suggest three things will happen all of them at the same time
11:24yes there will be some job declines but there will also be some job growth jobs gained but the biggest effect the biggest effect you know almost all the economic studies show this is jobs that will change
11:38will change meaning people are going to work alongside the technology their jobs will be different so for people to embrace this technology to do their jobs in an assistive way it's going to be the biggest effect
11:51that's what all the studies and work shows
11:53so it's about re-skilling in a way you're going to lead huge amount of re-skilling
11:58I think the re-skilling question is going to be the biggest question because think about if you have all three things right some jobs are declining some are growing
12:06so people are going to need skills as they transition from the jobs that are declined to ones that are growing so that's re-skilling
12:12think about the jobs that are going to be changed whenever there's job changes even if the job still exists but it's different there's going to need re-skilling
12:21so in all of these questions about work and jobs the re-skilling question becomes fundamental and fundamentally important
12:28it's actually one of the reasons why even when we partner with others in India this skilling work is fundamentally important we try to invest in that in re-skilling
12:39and in fact as part of the investment we made and announced last year a big part of that is to do with skilling
12:45how do you enable people to have the skills the essential skills to be able to take advantage of this technology both to help themselves and help them at work but also ultimately contribute to leveraging this technology to
12:57drive and grow the economy James in the time that we have left with you on this conversation and we hope to have more in the future two aspects one your work as the co-chair of the UN
13:10body high-level body on AI and the fact that India is going to host very soon this massive AI summit explain it to our viewers why is this moment significant for that AI summit to be taking place in India
13:25well first of all I'm very very much looking forward to be there I think the reason it's so important for this to take place in India and it's actually one of the things that my colleague and CEO Swindha has always said which is what we don't want is we don't want the digital divide turning into the AI divide
13:41so having having a summit like this in India and being able to actually show that we're all working to make sure this does not become an AI divide both in terms of access infrastructure connectivity people being able to use these tools people being skilled I think it's fundamentally important so I think at this time having this summit in India I think is one of the most exciting thing for quite frankly not just for India but quite frankly for the world because
14:10because there we have a country that's exciting that's growing with the motivation and quite frankly you know one of the things that was actually always fun when I was co-chairing the UN high-level body was that you know there's all there's a wide range of attitudes around AI in the world but I would always enjoy the conversation with Indian colleagues on the panel because they're very optimistic in fact the conversations mostly about okay how do we make how do we fully capitalize on this technology to
14:39transform our economies our education systems and transform our economies our education systems and healthcare and all of that and why do you think that is so let's work on that let's work on those right so the so I think having a country like India hosts a summit in a very optimistic but also responsible way because I think this technology is you know we always say you may have heard Swindha and I and others say that we think it's important to be both bold and responsible
15:09so let's be as unabashedly ambitious about how we apply this technology to transform economies and society improve lives everywhere but at the same time we should be responsible so both things are not in conflict in our minds I think we should work on both I think being able to showcase that in India I think is one of the most exciting opportunities we have okay as we wrap
15:39so I mean what prompted you and more importantly end by telling us what do you think the world will look like with AI 30 years down the line
15:48well so so it's hard to do that in a soundbite but first of all there was nothing original about me doing a PhD 30 years ago remember the field the term came into being in 1956 so this is already an old field it just so happens that in the 90s
16:02it was starting to turn and started to make progress I just happened to be lucky enough to be doing a PhD at that time I think it's the thing to look forward to is that we're seeing such incredible progress and you know what you're seeing of AI today is probably the worst AI you'll ever see
16:21it's only going to get better because you know the the innovations the research the breakthroughs you know we're still so early in this incredibly transformational
16:32technology so I think you know in 30 years time or even less than that in 10 years time we'll be having very different conversations one of the privileges that I have in the work that I do is you know we have a quantum AI team and it's one of my teams that we're you know again that's exciting so we start to see these fundamental breakthroughs in areas like quantum computing tied with AI in fact the team is called quantum AI those are some of the things to look forward to in the future and I can't wait to see where we're going to end up
17:01absolutely James and that it's a very very interesting time and like you say it's only just getting started
17:08as we enter 2026 the debate is is AI a bubble or is it a game changer how do you see it when when critics of AI make an argument that it's just a bubble what would you tell them
17:22so you may be shocked for me to say that this is a transformational technology
17:28you know we're going to be releasing data here over the next couple of days showing that those folks those individuals those companies that are using AI and building it into their systems
17:41are seeing a 7x multiplier on their economic productivity meaning that if you're an individual you're producing seven times more economic value for the company or if you're a company
17:53you're able to translate that into seven times more value for what you are producing through that employee
17:59this is a what's called a general purpose technology and to put that in historical context I know I'm talking to folks all over India and in India history is always really important
18:11like this technology is on par with other general purpose technologies the wheel the combustion engine the digital chip
18:20you know this these are all technologies that fundamentally changed growth changed progress
18:28and so you have to think about it being on that type of a scale
18:32and one of the things that we're really spending a lot of time talking about and I think is you know particularly relevant to India is something called the capability gap
18:40and what we're seeing right now is individuals companies countries that are successfully building this in are getting that economic growth
18:47those that are not potentially are going to be left behind in this and so we really think about this year
18:532026 as the year that we really begin to try to address that capability gap healthcare education what's called enterprise building this in the in the in the companies
19:03and even something that we call open AI for countries which is how can we work with countries
19:07so they're building AI into their socioeconomic infrastructure
19:11and you know India has the AI summit in a couple of weeks like the world I mean the world's here in Davos
19:17but the entire AI world is going to descend on Delhi in about a month's time
19:23and you know India is just at the center of this conversation
19:28you're one of the countries that you're seeing sort of the most adoption at the fastest rate
19:32and so as you think about India building out what I think is going to be the world's biggest strongest most robust middle class out there
19:40like this technology is going to play a transformative role not just globally but particularly in India
19:44Chris before we go further into the India opportunity I just want a broader question because a lot of people especially in the public markets are focused on what is
19:55some people some experts and economists say is an AI bubble very simple question from your point of view is there an AI bubble
20:04this way tomorrow where we have 850 million people using AI on a regular basis we have more than 4 million developers working as more than 1 million companies just working with us
20:17tomorrow those numbers will be higher the day after that they will be higher the real issue here is whether you're actually going to build this in or whether you get left behind
20:26our belief is that you need to be spending this year in particular building this AI in and if you're going to approach this as if it is a bubble
20:34you potentially are going to put yourself on the downside of that credibility that capability gap that we've talked about
20:40India and you spoke about our own efforts and
20:45for the benefit of our viewers where does India stand in terms of
20:50open AI's customer base are we the largest emerging market
20:55what's the data
20:57we look at India and they're just fascinated by it
21:01we've seen
21:03two and a half times growth
21:05over the last year over the last 12 months in India
21:08today it is our second largest consumer market in the world
21:12meaning everyday people using this on a regular basis
21:15the only other country that's ahead of it
21:17is our country the United States
21:19it's also a top five country
21:22in developers and why I think that is really important
21:25I mean you both know this you cover this
21:27all the time your audience knows this
21:29I mean India is producing some of the best
21:31and greatest tech talent
21:32but not just tech not just engineers
21:35although engineers are incredibly important
21:36but actually people who do the research on AI
21:39the AI researchers
21:40they're really the people who are building this
21:42and India is really well positioned to fully participate
21:46to lead in what we think of as the intelligence age
21:50because it has such a talent base
21:52that's domestic
21:54that's in the country
21:55I think sometimes when folks think about this technology
21:58there's a natural tendency to think about it like it's social media
22:02it is not social media
22:04this is a productivity driving technology
22:08the way we think about it is
22:10you know we are in effect trying to make electricity available to the world
22:14but then the world gets to build in that electricity
22:16right people get to build their version of their stoves
22:20their cars
22:21their airplanes
22:22but there's gonna be a whole bunch of new industries built on top of AI
22:25and India is just so well positioned
22:26given the consumer penetration
22:28meaning people are learning to use this on a regular basis
22:32and the fact that you have real talent to build this next generation of companies
22:35you know in November 2025
22:38you officially made ChatGPD OpenAI free for its Indian users
22:43for one year beginning November 4th
22:46now when you decided to take this free offer for users across the country
22:51was the attempt to make India in a sense a testing ground
22:55for large scale AI adoption that you first want to build market access
23:00and then decide how you're going to use it
23:02is was that the goal is that the goal
23:04to see whether India can be the testing ground
23:07to see just how far you can take the AI market
23:10the only place I would maybe be slightly different than the way you put it
23:15is that for us you know India is not necessarily a testing ground
23:19as much as it is one of the most important markets for us
23:23and really for anyone who's going to be in AI in the world
23:27just because the role India is going to play globally
23:29the size of its population
23:31it's a country that when we think about the most important markets for us
23:37where we're going to invest our time
23:39our resources
23:40where we're going to commit to supporting the country
23:42like India is just at the top of that list
23:45I mean if you think about the fact that we have 850 million people using this on a regular basis
23:49and India is the second largest of the countries
23:52that should tell you how important it is
23:54but Chris a lot of people in India
23:57or at least some lose sleep over the fact that we haven't developed a large language model
24:02we lose sleep over the fact that companies from outside of India have access to Indian consumers
24:10and therefore there are also fears that this will have impact on jobs
24:16for India
24:17jobs are very very important given the same size of population that you spoke about
24:21should we be losing sleep over the fact that we haven't yet developed something that is close to chat GPT
24:29or other similar models
24:31great question
24:32I get this question you know not just in India but in other places in the world
24:37and I want to be really respectful in terms of how I address it
24:41I think the reality is is that in the world that we are in today
24:46and again I say this with just the utmost respect
24:49there are two countries in the world that have the capacity to build those frontier models at scale
24:54you know right now it's the US and it's the People's Republic of China
24:58all that said
25:00this technology as I've talked about is akin to electricity
25:05and so at some level it is less whether you're going to build the frontier model
25:11than whether you're going to take that electricity
25:13and then build all the things on top of that electricity
25:18and that's why India is just exceptionally well positioned for that
25:24and so this is not a
25:25this is not social media
25:27it's not an extractive technology
25:29social media is not coming into India
25:31it's not social media that came into India
25:33generated revenue and took that out of India right
25:36this is technology that's going into India
25:38for Indians to build for India
25:40by India
25:41of India
25:43you think some of the biggest companies
25:44they're going to be able to build on top of this technology
25:47you take a small business
25:48and suddenly a small
25:49so I was sharing off
25:51before we came on camera
25:53that I'm traveling with India
25:55with my family
25:56my wife
25:56she is a huge fan of Indian fabrics
25:59think of all the small
26:01what we would call mom and pop shops
26:03that build Indian clothes
26:05in fact
26:05with this technology
26:06those companies can suddenly become international companies
26:09because they're able to use this tech
26:11to project themselves to the world
26:13no but can I push you on that
26:14because you know that's the big debate
26:16that yes you could possibly increase revenues
26:19you mentioned 7x
26:21is what you're suggesting
26:22but it could also take jobs
26:24and in a country like India
26:26just the sheer fear of losing jobs
26:30is now across the board
26:32when it comes to AI
26:33how is
26:34how does one address the critical question
26:36of what AI does
26:38to job markets
26:39to labor markets
26:40in countries like India
26:41I would say is
26:42I don't know if this is two words or three words
26:44AI literacy
26:45you know I mentioned that this is a general purpose technology
26:49I'm going to use one of my favorite historical analogies
26:52this is our generation's version of the printing press
26:55which countries succeeded when the printing press came out
26:58the countries that built in literacy programs
27:01into their social fabric
27:03and so those countries
27:05that are able to actually develop AI literacy
27:08amongst their population
27:09are going to be the countries
27:11that are best positioned to succeed
27:13and you look at what's going on
27:14you look at China
27:15which has a program
27:16to try to make as many of their people literate as possible
27:19I think Japan has a goal of 80% of Japanese people literate in AI by 2030
27:24India already starts off with a big advantage
27:28because of the penetration
27:29because the inherent interest
27:31I do think there's something in the Indian DNA
27:32that leans into technology
27:34that gets excited for this stuff
27:37and so you're just seeing it organically happen
27:39and so if India from a government perspective
27:42from a business perspective
27:43from a community perspective
27:44can really think about that literacy piece
27:47then you're going to grow all sorts of economic opportunities
27:50look I think some of this technology
27:53is going to reward those
27:54who are willing and interested
27:56in using it to empower themselves
27:58it does bring down barriers
28:00I mean you have a device in your hand right now
28:03think of the wealthiest person in the world
28:06and maybe just a regular person
28:07if they have that device in their hand with AI
28:10they're on equal ground
28:12and equal footing
28:13and so for a country like India
28:15which is looking to make the great leap forward
28:17right that big step
28:18into becoming the world's largest middle class economy
28:21this technology unlocks that
28:23so what happens to those who say
28:24that projections of blue collar
28:26and white collar workforce
28:27could get affected
28:28if companies decide to use AI
28:31in a way to cut costs
28:32so I think what we've been seeing so far
28:34is what you're seeing
28:36is that people are using the technology
28:38to increase the productivity
28:40of their current workforce
28:41now I do think as we go forward
28:43this technology is not static
28:45it's going to continue to accelerate
28:47continue to get better
28:48continue to get stronger
28:50we think this year
28:51you're going to see breakthroughs in health
28:52in science
28:53along with education
28:54and companies building it in
28:55but to your specific question
28:57I do think countries need to think about this
28:59from a transitional perspective
29:01that comes back to the literacy piece
29:02do you start to build in AI training
29:05job certification
29:06those types of programs
29:08not just for kids going through
29:10you know public school
29:11but also for that mid-career person
29:13for that person who's maybe doing
29:14a job right now
29:16that either
29:16if they learn this technology
29:18they're going to be better at that job
29:19and they're going to be more valued
29:21so not just young people
29:22but also people mid-career
29:23I think what Chris is saying is
29:2510-20 years ago
29:27in similar conversations
29:29it was said that
29:30oh you know
29:31coding should become part of school curriculum
29:33and now it's AI
29:35because AI can do coding for you
29:37but there's a very interesting piece of data
29:39and I wonder if you have a comment about this
29:42Indian IT
29:43and IT enabled services
29:45are a massive job creator in India
29:47have been so for the past three decades
29:49very significant export earnings and revenues
29:52just 17 jobs added Chris
29:56across Indian IT companies
29:58net additions
29:59in the past one year
30:00now we don't have a breakup
30:02of how many
30:02of those jobs
30:04which used to normally be
30:05in the tens of thousands
30:06did not happen on account of AI
30:09but Indian IT companies
30:11are responding to AI
30:12by buying AI companies
30:14by winning AI projects
30:16what are the prospects
30:18according to you
30:19for the Indian IT
30:20and IT enabled services companies
30:22because it's one of the largest
30:23job creators
30:24revenue generators
30:25in India
30:26you're really hitting on this concept
30:28that I've mentioned a couple times here
30:30but this capability gap
30:33I do think for the Indian IT sector
30:37there is going to be a real imperative
30:40for them to build the AI
30:42into their systems right now
30:45so that they're going to be well positioned
30:47as this economy goes into the future
30:49it's not just the new companies
30:52it is existing companies
30:53I think one of the variables
30:55that I'm not sure
30:57anyone really knows
30:58the answer to
30:58but it's going to be
31:00are legacy companies
31:02able to build this in
31:03benefit from it
31:04or is this going to help
31:05new companies that emerge
31:07and I think in India
31:08those IT companies
31:09which have generally
31:10been pretty innovative
31:11if they are able
31:14to have the foresight
31:16and ability
31:16to build this into their systems
31:18then I think they're going to be
31:19well positioned
31:20as we go forward
31:21so would you agree with a question
31:23that will AI kill the back office
31:25I think what it will do
31:27is make the back office more powerful
31:28it will make the back office
31:30able to do even more
31:31I do think some of this stuff
31:33will transition
31:33like I think we do need
31:35to be honest about that
31:36right
31:37every time you have gone
31:39through technology shifts
31:40in global history
31:42right
31:42you've had some jobs
31:44that have been displaced
31:45typically a lot more new jobs
31:47get created
31:48on the other hand
31:49I mean in the US right now
31:51there are more jobs
31:52that exist now
31:53than existed
31:53you know new jobs
31:54than existed prior
31:55to the second
31:56give me an example
31:57give me an example
31:57of an industry
31:58where new jobs
31:59will be created through AI
32:00well I think that
32:01what you're going to see
32:02is in the healthcare space
32:03you're going to be able
32:05to have people
32:06who right now
32:07may not be licensed
32:09to perform
32:10direct medical
32:11primary care
32:12who are going to be able
32:14to use this technology
32:15in the US
32:16think of someone
32:16like a domestic healthcare worker
32:18they're typically people
32:19who go into a home
32:20make sure that an elderly person
32:22or a young person
32:23is physically okay
32:24but they're not really able
32:26to provide direct healthcare
32:28with this technology
32:28there's going to be
32:29the potential
32:30for those types of folks
32:31to actually directly
32:32provide that healthcare
32:33that's going to be
32:33a version of what you saw
32:35when electricity came
32:36when lamp lighters
32:38the people who used
32:38to light the lamps
32:39with the oil
32:40became electricians
32:42when the people
32:43who drove the horses
32:44became truck drivers
32:45and each time
32:46you ended up having more
32:48now I do think
32:49that what's really important
32:50with this particular technology
32:52is the pace
32:54that it's going to move at
32:55and if jobs get displaced
32:57while new jobs
32:58are getting created
32:59what is the time lag
33:00between those two things
33:01and so what I think
33:02is really really important
33:04for both for businesses
33:06but also for countries
33:07and I know the prime minister
33:09is thinking about this
33:10is how are you able
33:12to take the real time data
33:13that you're seeing from AI
33:14and translate that
33:16into government policies
33:17so that you're making it
33:19as easy and straightforward
33:20as possible
33:21for this generation's
33:22lamplighter to become
33:24an electrician very fast
33:25As a final question
33:27what AI has done
33:30is totally disrupt
33:30it disrupted the world
33:32as we knew it
33:33I come from the typewriter age
33:35I started as a journalist
33:36working on a typewriter
33:38Was that it?
33:39Okay great
33:40Okay so I was on a typewriter
33:42we've seen the world
33:43change enormously
33:44A year from now
33:45what's that one big thing
33:47you think the world
33:48should be prepared
33:49from open AI
33:50something that you think
33:51they've never seen before
33:52is there something out there
33:54that's going to be
33:55the next big thing in AI?
33:57Okay I'm going to give you
33:58two things
33:59since you asked
33:59The first is
34:01I do think you will see
34:03and whether it's from us
34:04or others
34:04but I think you'll see it
34:05off of our technology
34:06some incredible advances
34:08in science
34:10Last year was
34:11agentic AI
34:12where you literally
34:13could be a non-coder
34:14go into chat
34:16and actually do coding
34:18I am someone
34:19who does not know
34:20how to do coding
34:20I'm actually able
34:21to build programs now
34:22using this
34:23I mean I think all of us
34:24those of us
34:24from the typewriter age
34:25are able to do coding now
34:26I think 2026
34:28is going to be
34:29the year of science
34:31where you see
34:32this technology
34:33leading to acceleration
34:35and breakthrough
34:36in science
34:38and in ways
34:38that are going to
34:38impact folks
34:39every day
34:40I think a second
34:41and I'm not going to say
34:42too too much about this
34:43because I don't want
34:44to get too far ahead
34:45but I think
34:46maybe towards
34:47the latter part
34:48of the year
34:49you'll see AI
34:50in devices
34:52right
34:52you know
34:54it's well known
34:54that we're working
34:55with Johnny I
34:56the maker of the iPhone
34:57on a device
34:59and there's a lot more
35:00that will be coming
35:01on that
35:02and so I would put
35:03people in that direction
35:04if you were to be called
35:06one of the godfathers
35:08of AI
35:08does that make you feel
35:09comfortable
35:10in the age
35:11in which we live now
35:12I'm not sure
35:15there's a lot of
35:16unknown unknowns
35:17and risks
35:18associated with AI
35:19I think
35:20most people
35:22don't realize
35:23the magnitude
35:24of the transformations
35:25that are likely
35:26to come
35:27most people
35:29don't realize
35:30how much better
35:31the intelligence
35:33the capabilities
35:33of AI
35:34are likely to be
35:35in a few years
35:36as the data shows
35:38it's been going
35:39up and up
35:39exponentially
35:41over the last
35:41five years
35:42so
35:43we need to
35:45consider
35:46the possible scenarios
35:47and we need to
35:48plan accordingly
35:49give us those
35:51possible scenarios
35:52when you started
35:52up all those years ago
35:54and now you see
35:55AI
35:55moving in every
35:58dimension of society
35:59give us
36:00what are the
36:00possible scenarios
36:01that you see
36:02of what AI
36:03will be doing
36:04in our lives
36:04over the next decade
36:06there's a spectrum
36:07of scenarios
36:08if things continue
36:10as they have been
36:11in the last few years
36:11we'll have systems
36:13that can
36:14do most of the
36:15tasks that humans
36:16can do
36:17within
36:18you know
36:18in the coming years
36:19depending on the tasks
36:20and reaching
36:21human level
36:22in many ways
36:23within maybe
36:24five years
36:25and it might
36:28come at different
36:28speeds for different
36:29kinds of abilities
36:30it could also
36:31even go faster
36:32some people think
36:33that AI
36:34will be used
36:35to do AI research
36:35and accelerate
36:36the advances
36:37the big question
36:39is it a bubble
36:40or is it a game changer
36:41I think
36:42even if
36:44the progress
36:46was to slow down
36:47in terms of the science
36:48it would still be transformative
36:50just to take
36:51what we already have
36:52and you know
36:53bring it
36:54to the economy
36:55in many different ways
36:56do you think
36:56market hype
36:57and if the market
36:59valuation cycle
37:00were to break
37:01all cycles move
37:02eventually
37:03would that impede
37:04the progress
37:05of the development
37:05of the science
37:06of AI
37:07so it is possible
37:10that there will be
37:10some financial ups
37:11and downs
37:12because there may be hype
37:13there may be overselling
37:14and things like that
37:15but the science
37:17is going to continue
37:18and yes
37:19it could slow down
37:20if there is less money
37:21put into it
37:22for sure
37:22Professor
37:23from the India perspective
37:24I ask you this question
37:25there are two global
37:27superpowers in AI
37:28one is the United States
37:29the other is China
37:31both have their own
37:32sort of specific
37:33areas of strength
37:34but within that
37:36where does that leave
37:37a country like AI
37:38and more particularly
37:39this race between the two
37:41in some ways
37:42are there concerns
37:44around safety
37:45ethics
37:46standards
37:47and the lack
37:48of governance
37:49absolutely
37:50I'm very concerned
37:51about the race
37:53both at the level
37:54of nations
37:55mostly the US
37:56and China
37:57and between corporations
38:00as a result
38:01of that race
38:02the incentive
38:02for each player
38:03is to go
38:05as quickly as possible
38:06and cut corners
38:06on safety
38:08the public goods
38:09you know
38:09risks to society
38:10to democracy
38:11risks to individuals
38:14like you know
38:14psychological effects
38:15children
38:16suicides
38:18there are many
38:19like
38:19unknown unknowns
38:21that
38:23when you are in the race
38:24you're just
38:24trying to win
38:25and that's dangerous
38:27so
38:29so we need to
38:30understand that race
38:31and countries like India
38:33I think can be part of
38:34a larger movement
38:35around the world
38:36what is this AI safety
38:38report all about
38:39if you could give
38:39a concrete example
38:41as to how
38:42you can put
38:43some kind of
38:44guardrails
38:45to prevent AI
38:46from spiraling
38:47out of control
38:48that's right
38:48that's exactly why
38:49this report
38:52was created
38:52it was created
38:54at the Bletchley
38:55AI summit
38:56which is the
38:57ancestor of this
38:58Indian summit
38:58that took place
38:59two years ago
39:00and it is about
39:02understanding
39:03the scientific
39:05data
39:06about
39:07the capabilities
39:08of AI
39:09the risks
39:10there's a
39:11you know
39:12many different
39:13kinds of risks
39:14and the mitigation
39:15methods
39:16that exist
39:17or that researchers
39:18are working on
39:19and it's taking
39:19the scientific
39:20literature
39:21and creating
39:22a synthesis
39:23for policy makers
39:25so it is written
39:26in the language
39:27of policy makers
39:27what
39:28for a country
39:29of the size
39:30and the nature
39:30of our economy
39:31is the best case
39:32for AI
39:33in your view
39:33so first
39:36I think
39:36that
39:37the part about
39:38the foundation
39:39models
39:39and
39:40not just
39:42copying
39:43what the others
39:44are doing
39:44or using
39:45open source models
39:46and doing
39:47something with them
39:47but actually
39:48pushing the frontier
39:50is going to be
39:52key
39:52in the future
39:53because
39:54if the AI
39:55capabilities
39:55continue to grow
39:57you have to
39:58understand
39:58that
39:59intelligence
39:59gives power
40:01so
40:01the people
40:03the corporations
40:04and the countries
40:05which have
40:07control over
40:08the most advanced
40:09AI
40:09will have a lot
40:10of power
40:10in the future
40:11now
40:12I believe
40:13that the best bet
40:14for a country
40:14like India
40:15is
40:16not to try
40:17to like
40:18completely
40:18do it alone
40:19but to
40:20start thinking
40:21about partnerships
40:21with other countries
40:23there are also
40:23concerns
40:24around it
40:25about India
40:25just becoming
40:26a data mining
40:27market
40:28in some ways
40:28where do you
40:29stand on this
40:29well
40:30I mean
40:31it's great
40:31to see
40:32more deployment
40:33and access
40:35of course
40:36but you
40:37have to think
40:38of the dependence
40:42on AI systems
40:43like
40:44you know
40:45the geopolitics
40:46especially in the past
40:48of thinking about
40:49the dependence
40:49on oil
40:50so
40:51you don't want
40:52to be
40:53prisoner
40:53of a single
40:54vendor
40:55or single
40:56corporation
40:57or even a single
40:58country
40:58you'd like to
41:00establish your
41:01deployment of AI
41:02in a way that
41:03you can easily
41:04you know
41:05switch
41:06or even
41:06participate
41:07in the
41:08development
41:08of AI
41:09that will be
41:10coming with
41:12stronger
41:12governance
41:13guarantees
41:13both in terms
41:14of safety
41:14and in terms
41:15of access
41:17you know
41:17but
41:18somewhere
41:18the real
41:19concern
41:19a lot
41:20of the AI
41:21debate
41:21especially
41:21in India
41:22is about
41:23to what
41:24extent
41:24will AI
41:25take
41:25the role
41:26of jobs
41:27that are
41:27today
41:28done by
41:28both
41:29blue
41:29and white
41:29collar
41:30forces
41:30should we
41:32be very
41:32very wanted
41:32I think
41:35on the
41:35economic
41:35side
41:36it's going
41:37to be
41:37difficult
41:37to say
41:38stop
41:38in some
41:39aspects
41:40there could
41:40be
41:40international
41:41agreements
41:42when it
41:43comes to
41:44catastrophic
41:45risks
41:45like
41:46the use
41:46of AI
41:47for
41:47launching
41:48all kinds
41:49of attacks
41:49or the
41:50possibility
41:51of even
41:51losing
41:52control
41:53of a
41:53rogue AI
41:54this is
41:54also
41:54something
41:55that the
41:55scientists
41:55are studying
41:56and we
41:57are seeing
41:57signs
41:58over the
41:58last year
41:59in both
41:59the major
42:00American
42:01labs
42:02and outside
42:03labs
42:05that the
42:06most advanced
42:07AI's
42:08don't want
42:09to be
42:09shut down
42:10they are
42:12trying to
42:12preserve
42:12themselves
42:13they are
42:13trying to
42:14hack other
42:14computers
42:14to save
42:15themselves
42:16somewhere else
42:17or they are
42:17willing to
42:17even be
42:18strategic
42:19and start
42:20blackmailing
42:21the engineers
42:21to make sure
42:22they are not
42:22going to be
42:22replaced by
42:23a new
42:23version
42:23super
42:24intelligence
42:25now this
42:25is a
42:26controversy
42:27that was
42:27sparked
42:27some time
42:28ago
42:28but as
42:29an academic
42:30as a
42:30scientist
42:31as a
42:31researcher
42:32is AI
42:34going to
42:35be
42:35sentient
42:35and is
42:38that going
42:38to happen
42:38in the
42:38future
42:39I don't
42:41know but
42:41we are
42:41on track
42:42for that
42:42we are
42:43well yeah
42:44like the
42:45capabilities of
42:45AI system
42:46is just
42:46going up
42:47and in
42:47fact it's
42:48already
42:48stronger than
42:49humans in
42:49some aspects
42:50and it's
42:50going up
42:51on other
42:52aspects where
42:52it's lagging
42:53quite a lot
42:54with respect
42:54to humans
42:55so there's
42:55no reason
42:56to think that
42:57human intelligence
42:58is the apex
42:59of intelligence
42:59and effectively
43:02you're pitching
43:03to take up
43:03from what
43:04Siddharth said
43:04a kind of
43:05terminator
43:05scenario
43:06in which a
43:08machine endowed
43:09with human
43:10emotions
43:10turns on its
43:12creator
43:12is that a
43:12possibility
43:13let's not say
43:14that is a
43:14possibility
43:15even if we
43:16don't know
43:17what is the
43:18probability
43:19for this
43:19the severity
43:22of the risk
43:22is so great
43:23that we should
43:24try to prevent
43:25it
43:25which we try
43:26to make sure
43:26it doesn't
43:27happen
43:27what is your
43:28piece of
43:28advice you'd
43:29give Prime
43:29Minister Modi
43:30let me turn
43:31it around
43:31to the person
43:32who the world
43:33is looking at
43:33most carefully
43:34at the moment
43:35President Trump
43:35if you had to
43:37give a piece
43:37of advice
43:38to President
43:38Trump
43:39on the entire
43:39AI dimension
43:40what would
43:41that be
43:41a lot of
43:44the same
43:44things
43:44so
43:45we see
43:48the scientists
43:49don't all
43:49agree
43:50but they can
43:51see a range
43:52of scenarios
43:52where
43:53very likely
43:55AIs will
43:55become much
43:56smarter than
43:56they are now
43:57and if we
43:58continue on
43:58that track
43:59they will be
44:00smarter than
44:00us
44:01and right
44:02now we don't
44:03know how to
44:03control them
44:04so
44:05the United
44:06States
44:06who's
44:07leading
44:07in AI
44:08right now
44:08should also
44:09lead on
44:10how we
44:10make sure
44:11AI will
44:12be beneficial
44:12to humans
44:14and
44:15not just
44:16press on
44:17the accelerator
44:17for capability
44:18but there's
44:19a special
44:19capability
44:20which is
44:21safety
44:21which is
44:23do no harm
44:24now one of
44:26the striking
44:26aspects
44:27of
44:28what we are
44:29seeing around
44:29us at the
44:30World Economic
44:31Forum
44:31is how
44:32Indian minds
44:33are shaping
44:35the world
44:36AI economy
44:37and many of
44:38them are
44:39products of
44:40the Indian
44:40education
44:41system
44:41and I'm
44:42joined by
44:43one such
44:43very special
44:44individual
44:44joined by
44:45Deepak
44:45Prathak
44:46CEO and
44:47co-founder
44:47of Skilled
44:48AI
44:49Deepak is a
44:50Bareli boy
44:51went to
44:51IIT
44:52Kanpur
44:52and is now
44:53leading a
44:54company that's
44:55building software
44:56that can act
44:57as a general
44:57purpose
44:58brain for
45:00robots
45:00and you've
45:02just done a
45:03major round
45:03of funding
45:04first of all
45:05let's talk
45:06about the
45:07company itself
45:08building
45:09brains for
45:11robots
45:12allowing
45:13different
45:14machines to
45:14perform
45:15different tasks
45:16break it down
45:17for us
45:17what makes
45:18skilled AI
45:19so different
45:20so when you
45:21think of
45:22models like
45:23chat GPT
45:23for instance
45:24right
45:24you think of
45:25intelligence
45:26you think of
45:27okay it's like a
45:27human like brain
45:28behind the
45:29computer
45:29when you think
45:30of machines
45:31or robots
45:31what do you
45:32think of
45:32you think of
45:33hardware
45:34you almost
45:35never think of
45:36the intelligence
45:36behind it
45:37and if you
45:38have been
45:39seeing like
45:39demos around
45:41the world
45:41you can find
45:42demos last
45:42thing like
45:4350 years
45:4360 years
45:4470 years
45:44back from
45:451970s
45:46anything you
45:48saw in
45:481970s
45:49in terms of
45:50technology
45:50it has come
45:52to the real
45:52world today
45:52but where are
45:53robots
45:53and the reason
45:55is it is not
45:56the hardware
45:56it's the brain
45:57which has been
45:58missing for
45:58all these
45:59years
45:59for these
46:00systems to
46:00work very
46:00very in the
46:01real world
46:01even when as
46:02a human
46:03when you go
46:03do a simple
46:04task like
46:04opening a
46:05door
46:05you may not
46:07give it much
46:07value because
46:08every almost
46:09every human
46:09can do it
46:10but that is
46:11the reason
46:11we have
46:12intelligence
46:12and language
46:14as we speak
46:14as we think
46:15math is all
46:16byproduct of
46:16that
46:17so
46:18you're saying
46:20you're creating
46:20robots with
46:21brains
46:22am I correct
46:22brain for
46:23robots
46:23like it's a
46:24single model
46:25you're creating
46:25brains for
46:26robots which
46:27will enable
46:28them to
46:28do
46:29what humans
46:30can do
46:31so effectively
46:31you'll replace
46:32humans and
46:33there are those
46:33who are listening
46:34who will fear
46:35you know the
46:36tasks which
46:37today are done
46:38by humans
46:39if skilled
46:40AI scales up
46:41robots will
46:43take over
46:43give me an
46:44example of
46:45what your
46:45robots will
46:46do which
46:47humans are
46:47doing at
46:48the moment
46:48well if you
46:50put it that
46:51way it seems
46:52to be sensational
46:53but we are
46:54very far from
46:55the general
46:56capability
46:56but if you
46:58think about
46:58in terms of
46:59what humans
47:00can do
47:00today like
47:01in US for
47:02instance there
47:03are already
47:04over a million
47:05jobs unfilled
47:06which means
47:07there are more
47:07jobs available
47:08than people
47:09available to
47:09fill them
47:10and that gap
47:11is continuously
47:11increasing
47:12it's supposed
47:13to go 30
47:13million by 2030
47:14so at the
47:15same time
47:16there is also
47:17unemployment
47:17so which means
47:19there are jobs
47:20which are unfilled
47:21and people
47:22unemployed
47:22which means
47:22people do not
47:23want to do
47:23certain kind
47:24of jobs
47:24and this is
47:25becoming more
47:26and more
47:26clear over
47:26time
47:26you are now
47:30deploying
47:31robots in
47:32enterprises
47:33such as
47:33manufacturing
47:34inspections
47:35data centers
47:36so you are
47:37already
47:37half way
47:38down there
47:39so what's
47:43the future
47:43Deepak
47:44you see a
47:45world where
47:45robots say
47:4610 years
47:47from now
47:47do a lot
47:48of those
47:49tasks that
47:49you say
47:49remain unfilled
47:51at the moment
47:52or will they
47:53end up
47:53taking away
47:54jobs which
47:55is the big
47:55fear
47:55in countries
47:56like India
47:57the fear
47:57is you'll
47:58take away
47:59jobs
47:59see in India
48:00right now
48:01there is
48:01labor surplus
48:02right and
48:03the cost of
48:04wages do not
48:05justify full
48:06automation yet
48:06so India is
48:07in a good
48:07place but if
48:08you look at
48:08many other
48:09countries for
48:09instance Japan
48:11Korea and
48:12other countries
48:13you'll find
48:14there's one
48:14noticing thing
48:15the birth rate
48:16is declining
48:16and the
48:18average age
48:19is increasing
48:19and it is
48:21expected in
48:21like 10 to
48:2215 years
48:23there will be
48:24way more
48:24elderly people
48:25than we had
48:26even like 30
48:27years back
48:27and at the
48:28cost of less
48:29young people
48:29so while
48:31this is a good
48:31question about
48:31India but
48:32rest of the
48:32world really
48:33needs this
48:33right now
48:34not even 10
48:35years into the
48:35future
48:36so what do I
48:37see as the
48:38future
48:38well I do not
48:39know the long
48:40term it's like
48:40very far away
48:41but in the
48:42near future
48:43we will see
48:43enterprise like
48:44jobs getting
48:45filled by robots
48:46more and more
48:47than we see
48:48today
48:48I must ask you
48:49this where I
48:50started off
48:50the journey of
48:51a Bareilly boy
48:52going to IIT
48:54Kanpur and now
48:54in California
48:55is this the
48:56are Indian
48:58shaping the
48:58world will
48:59you all come
48:59back at some
49:00stage to
49:01innovate in
49:02India because
49:02one of the
49:03challenges for
49:03AI is going
49:04to be the
49:05pace of
49:05innovation
49:06China is way
49:07ahead of many
49:09parts of the
49:09world certainly
49:10India when it
49:11comes to
49:11innovation in
49:11AI will
49:12many of you
49:13want to come
49:14back will you
49:14partner with
49:15companies in
49:16India do you
49:16see that as
49:17your vision
49:18office in India
49:19in Bangalore and
49:20we are already
49:20actively hiring in
49:21this space I
49:23think one thing I
49:23would say is
49:24usually if you
49:25look at something
49:25already is
49:25happening outside
49:26we tend to
49:27build the same
49:28thing in house
49:29but in case of
49:30India we have to
49:31think from first
49:32principles like for
49:33instance if you
49:34look at the
49:34current manufacturing
49:35the hardware side
49:36of things China
49:37has major lead
49:38compared to any
49:39other country
49:39supply chain
49:41rare earth
49:41metals with the
49:42global trade
49:43war and all
49:44the sanctions
49:44going on
49:45everybody is
49:47looking for
49:48next player
49:49beyond China
49:50where manufacturing
49:51can increase
49:51and India is a
49:52great contender
49:53in that field
49:54why?
49:55because India
49:56already has big
49:56automotive market
49:57manufacturing
49:58you have
49:59electronics
50:00for robotics
50:01it is not yet
50:03there in any
50:03country in a
50:03major way
50:04but India can
50:05have a chance
50:06like they can
50:06surpass everybody
50:07because of
50:07automotive
50:08manufacturing
50:09piggyback on
50:10the supply
50:10chain
50:10electronic supply
50:12chain
50:12and we are
50:13already trying
50:13to encourage
50:14that in India
50:15do you have a
50:16picture of one
50:17of your new age
50:18robots that you
50:18can show our
50:19viewers
50:19well I don't have
50:20right now
50:21I don't carry the
50:22photo but I have
50:22given the videos
50:23so they can
50:24yeah so we are
50:25going to use
50:25your videos
50:26over the
50:27conversation
50:27because it is
50:29transformative
50:30and as I said
50:31it's great to
50:32see young Indians
50:33how old
50:33when I first
50:34saw Deepak
50:35entering the
50:36studio I
50:37wasn't sure
50:37whether some
50:38intern was
50:38just walking
50:39in then I
50:40realized he is
50:41a CEO and
50:42a co-founder of
50:43one of the
50:44cutting edge
50:44robotic companies
50:46how old are you
50:47now?
50:4733
50:4833
50:48well that's
50:494 maybe
50:51I don't know
50:51plus minus 1
50:53okay but it's
50:54great to have
50:54you on the
50:55show Deepak
50:55and giving us
50:56a sense of
50:56what this new
50:57age of robotics
50:58is going to do
50:59thanks very much
51:01for joining me
51:02at the World
51:02Economic Forum
51:03thank you
Comments