0:00 AI impact summit has got phenomenal
0:02 response.
0:04 This is going to be the fourth and the
0:07 biggest AI summit so far in the entire
0:10 world.
0:12 the response from practically 100
0:15 countries,
0:18 the number of uh heads of state and
0:22 government already 15 confirmations
0:24 received
0:26 and the industry has responded so well
0:29 more than 100 CEOs are joining us in the
0:32 impact summit.
0:35 What's being noted is the systematic
0:38 manner in which India has
0:42 developed its AI ecosystem.
0:46 Very recently honorable prime minister
0:49 interacted with the people who are
0:52 developing the models, the people who
0:54 are developing the applications
0:57 and the people who are setting up the
0:59 infra. As we all know AI architecture
1:04 has five layers. the app layer people
1:08 who interacted with honorable PM
1:10 yesterday. It was very encouraging that
1:13 the pivot that they have taken is such a
1:16 systematic pivot and that the strength
1:20 that our IT industry has had over
1:23 several decades now that strength is
1:25 being put to providing AI solutions
1:32 rather than the old model of having
1:34 software development.
1:37 So it was very heartening all the IT
1:40 majors they
1:42 kind of uh methodical way in which they
1:45 are reskilling their people they are
1:49 rediscovering their customers they are
1:52 re-engaging with their markets that was
1:55 really encouraging when they met
1:57 honorable prime minister they shared
2:01 that they have developed more than 200
2:03 models so far which are very foc focused
2:07 very small and totally oriented towards
2:11 solving the problems of one particular
2:14 sector. I think that is the big learning
2:19 that so many people today express that
2:23 the real value will come from the
2:26 deployment of AI
2:28 getting those productivity gains getting
2:30 those efficiency gains.
2:32 [snorts]
2:32 This was very evident in honorable prime
2:34 minister's interaction with the uh teams
2:38 which are developing on the applications
2:40 and the solutions.
2:42 Few weeks ago, honorable prime minister
2:44 interacted with the teams which are
2:46 developing our the stack of sovereign
2:49 models.
2:50 Very happy to share with you that the
2:52 models are in very good shape and we
2:55 should be launching those models in the
2:57 AI impact summit. So we look forward to
3:01 all of you using them and giving your
3:03 feedback and making them a big success.
3:06 On the infra layer again we had very
3:09 good participation with prime minister's
3:11 round table yesterday.
3:14 So far
3:15 we clearly have about $70 billion worth
3:20 investment
3:22 already getting into the ground getting
3:24 into the way we see the enthusiasm.
3:29 We shouldn't be surprised if this number
3:31 doubles by the time the air impact
3:34 summit is over.
3:36 So that's very significant investment
3:39 which is going into the infra layer.
3:43 on the energy layer. Our robust grid is
3:46 being recognized by
3:49 uh the world today and our ability to
3:52 have almost we have today 50% of our
3:55 power generation capacity is green power
4:00 that is being noted and the robustness
4:02 of the grid the diversity of our grid
4:05 all the investment which has gone in the
4:06 grid in the last 10 years that is being
4:09 noticed by the world. So we are in a
4:13 very uh good shape in that sense and we
4:17 will be focusing a lot on talent
4:20 development.
4:21 I have discussed with the industry. We
4:23 have been discussing for last almost one
4:25 and a half years what is the right model
4:27 for developing talent. Many experiments
4:30 have already gone very well. so many uh
4:33 experiments at skilling the people make
4:36 them AI ready to enable them to use AI
4:39 in their day-to-day lives that has gone
4:42 down very well the industry has done
4:45 their own experiments as part of NASCOM
4:47 and part of mate's future skills program
4:51 they have collaborated with universities
4:54 and developed programs where very high
4:59 level of AI uh skills are being provided
5:02 provided to students. More than almost
5:04 five lakh students have been able to get
5:06 those skills so far. Now we will take it
5:09 to a next level
5:12 and the industry is going to be the main
5:14 participant in that. You recall how we
5:17 did it in telecom where 100 labs were
5:21 set up in universities 5G labs and you
5:24 have seen in case of semiconductors
5:26 where 315 universities now have complete
5:31 stack of EDA tools and students are
5:34 designing chips. Students are taping out
5:37 their chips. We'll have a similar model.
5:40 Now industry is finalizing the course
5:42 curriculum. We'll have a similar model
5:46 where now we'll be more ambitious and
5:48 we'll go to 500 universities
5:51 where we will provide the AI infra
5:54 and the course curriculum where students
5:57 can get very good training on AI
6:02 opportunities
6:04 and then that can become a major talent
6:07 pipeline for the industry.
6:09 So this was also decided in the last few
6:13 reviews that we have been having. Apart
6:16 from that we'll also take up major
6:19 research programs which we will announce
6:21 in the AI summit. So that way our
6:25 program is very comprehensive and one
6:27 more thing which I wanted to share with
6:28 you is the world is taking note of our
6:32 common compute facility in a very big
6:34 way.
6:35 This is a democratization of technology
6:39 approach which our prime minister has
6:42 very clearly forwarded propagated in the
6:45 last 11 years. Whether it is UPI,
6:48 whether it is digilocker, whether it is
6:51 co-in, whether it is 5G rollout, all
6:55 these programs basically showcase the
6:59 philosophy of democratization of
7:01 technology and the common compute is
7:03 part of that. You should be hearing some
7:06 good news, some more additions to the
7:08 common compute stack and uh in the
7:12 coming times as we more or less close
7:17 the first uh part of the AI mission
7:22 another five to 6 months we should be
7:24 going for the AI mission too and I'll be
7:26 seeking your inputs on what else we
7:30 should be doing on a much larger scale
7:32 in the second version of the AI A
7:34 mission.
0:02 response.
0:04 This is going to be the fourth and the
0:07 biggest AI summit so far in the entire
0:10 world.
0:12 the response from practically 100
0:15 countries,
0:18 the number of uh heads of state and
0:22 government already 15 confirmations
0:24 received
0:26 and the industry has responded so well
0:29 more than 100 CEOs are joining us in the
0:32 impact summit.
0:35 What's being noted is the systematic
0:38 manner in which India has
0:42 developed its AI ecosystem.
0:46 Very recently honorable prime minister
0:49 interacted with the people who are
0:52 developing the models, the people who
0:54 are developing the applications
0:57 and the people who are setting up the
0:59 infra. As we all know AI architecture
1:04 has five layers. the app layer people
1:08 who interacted with honorable PM
1:10 yesterday. It was very encouraging that
1:13 the pivot that they have taken is such a
1:16 systematic pivot and that the strength
1:20 that our IT industry has had over
1:23 several decades now that strength is
1:25 being put to providing AI solutions
1:32 rather than the old model of having
1:34 software development.
1:37 So it was very heartening all the IT
1:40 majors they
1:42 kind of uh methodical way in which they
1:45 are reskilling their people they are
1:49 rediscovering their customers they are
1:52 re-engaging with their markets that was
1:55 really encouraging when they met
1:57 honorable prime minister they shared
2:01 that they have developed more than 200
2:03 models so far which are very foc focused
2:07 very small and totally oriented towards
2:11 solving the problems of one particular
2:14 sector. I think that is the big learning
2:19 that so many people today express that
2:23 the real value will come from the
2:26 deployment of AI
2:28 getting those productivity gains getting
2:30 those efficiency gains.
2:32 [snorts]
2:32 This was very evident in honorable prime
2:34 minister's interaction with the uh teams
2:38 which are developing on the applications
2:40 and the solutions.
2:42 Few weeks ago, honorable prime minister
2:44 interacted with the teams which are
2:46 developing our the stack of sovereign
2:49 models.
2:50 Very happy to share with you that the
2:52 models are in very good shape and we
2:55 should be launching those models in the
2:57 AI impact summit. So we look forward to
3:01 all of you using them and giving your
3:03 feedback and making them a big success.
3:06 On the infra layer again we had very
3:09 good participation with prime minister's
3:11 round table yesterday.
3:14 So far
3:15 we clearly have about $70 billion worth
3:20 investment
3:22 already getting into the ground getting
3:24 into the way we see the enthusiasm.
3:29 We shouldn't be surprised if this number
3:31 doubles by the time the air impact
3:34 summit is over.
3:36 So that's very significant investment
3:39 which is going into the infra layer.
3:43 on the energy layer. Our robust grid is
3:46 being recognized by
3:49 uh the world today and our ability to
3:52 have almost we have today 50% of our
3:55 power generation capacity is green power
4:00 that is being noted and the robustness
4:02 of the grid the diversity of our grid
4:05 all the investment which has gone in the
4:06 grid in the last 10 years that is being
4:09 noticed by the world. So we are in a
4:13 very uh good shape in that sense and we
4:17 will be focusing a lot on talent
4:20 development.
4:21 I have discussed with the industry. We
4:23 have been discussing for last almost one
4:25 and a half years what is the right model
4:27 for developing talent. Many experiments
4:30 have already gone very well. so many uh
4:33 experiments at skilling the people make
4:36 them AI ready to enable them to use AI
4:39 in their day-to-day lives that has gone
4:42 down very well the industry has done
4:45 their own experiments as part of NASCOM
4:47 and part of mate's future skills program
4:51 they have collaborated with universities
4:54 and developed programs where very high
4:59 level of AI uh skills are being provided
5:02 provided to students. More than almost
5:04 five lakh students have been able to get
5:06 those skills so far. Now we will take it
5:09 to a next level
5:12 and the industry is going to be the main
5:14 participant in that. You recall how we
5:17 did it in telecom where 100 labs were
5:21 set up in universities 5G labs and you
5:24 have seen in case of semiconductors
5:26 where 315 universities now have complete
5:31 stack of EDA tools and students are
5:34 designing chips. Students are taping out
5:37 their chips. We'll have a similar model.
5:40 Now industry is finalizing the course
5:42 curriculum. We'll have a similar model
5:46 where now we'll be more ambitious and
5:48 we'll go to 500 universities
5:51 where we will provide the AI infra
5:54 and the course curriculum where students
5:57 can get very good training on AI
6:02 opportunities
6:04 and then that can become a major talent
6:07 pipeline for the industry.
6:09 So this was also decided in the last few
6:13 reviews that we have been having. Apart
6:16 from that we'll also take up major
6:19 research programs which we will announce
6:21 in the AI summit. So that way our
6:25 program is very comprehensive and one
6:27 more thing which I wanted to share with
6:28 you is the world is taking note of our
6:32 common compute facility in a very big
6:34 way.
6:35 This is a democratization of technology
6:39 approach which our prime minister has
6:42 very clearly forwarded propagated in the
6:45 last 11 years. Whether it is UPI,
6:48 whether it is digilocker, whether it is
6:51 co-in, whether it is 5G rollout, all
6:55 these programs basically showcase the
6:59 philosophy of democratization of
7:01 technology and the common compute is
7:03 part of that. You should be hearing some
7:06 good news, some more additions to the
7:08 common compute stack and uh in the
7:12 coming times as we more or less close
7:17 the first uh part of the AI mission
7:22 another five to 6 months we should be
7:24 going for the AI mission too and I'll be
7:26 seeking your inputs on what else we
7:30 should be doing on a much larger scale
7:32 in the second version of the AI A
7:34 mission.