'R&D a National Weakness': Lt Gen Raj Shukla Compares India's Private Sector Funding With US, China

Asianet News English

Watch 'R&D a National Weakness': Lt Gen Raj Shukla Compares India's Private Sector Funding With US, China - Asianet News English on Dailymotion

Transcript

00:00DeepSeq, an open-source model, is basically challenging ChatGPT, those American proprietary models.

00:07And DeepSeq, apart from being at 1 20th of the cost, it is made on older versions and rationed GPUs.

00:19But the most striking thing is that about it is the algorithmic innovation, which has come from talent.

00:25Jo DeepSeq say there are about 211 authors associated with DeepSeq's engineers.

00:3160% of them have not stepped out of China.

00:33So this is the best and the brightest in China.

00:36So this business about talent and algorithmic innovation is something that we have to see to apply in our context.

00:43We also have 23 IITs, a great amount of talent.

00:46Why are we not being able to deliver on the algorithmic innovation?

00:50So innovation talent is something that we need to look at.

00:54Sir, innovation and emerging technologies, when we talk about AI and 5G, they can only happen when we get the R&D sorted.

01:02And what we are seeing is that in India, the government has invested not much in R&D, just a small percent of the GDP.

01:09What do you have to say on that?

01:10R&D is a national weakness.

01:12For the Americans, USA spends $950 billion on R&D every year.

01:18$700 billion comes from the private sector.

01:21China spends in the region of $600 billion.

01:23We are about $30 billion.

01:25And we are about $30 billion.

01:27So $3,600, $975.

01:30That will tell you the story.

01:32So we really need to step up our game.

01:33And not only government, we keep saying government, government, where is the private sector investment in R&D?

01:37So, and it is not that they should spend on R&D.

01:42Why don't they see value in R&D?

01:44If you see value in R&D, you will spend.

01:47So the R&D is also an extremely serious challenge.

01:50So innovation, talent, R&D.

01:54Do you think this is one of the reasons why we are not seeing startups in high tech, deep tech startups?

02:00Exactly.

02:00So this business about startups not pushing deep technologies enough is another dimension of the challenge.

02:09And you look at DeepSeek, it's a very good model.

02:11DeepSeek focuses on R&D.

02:13And the others, Alibaba, Baidu and all, they are doing business and commerce.

02:17So they are complementing each other.

02:19So ideally, we need a DeepSeek, which focuses on R&D.

02:23And the other models, which look at the business and commerce dimensions of AI.

02:27So these two things come up.

02:28So just last question from my side would be, there's a lot of challenges in DeepSeek as well.

02:33For example, there will be a challenge of IP infringement.

02:36There are, those are challenges.

02:38We have the intelligence to attend to these challenges.

02:42But they should not become a reason to create so much regulation that no innovation happens.

Category

Transcript

Be the first to comment

Recommended