India doing ‘very, very well’ in AI: Karp

Chapel Hill (North Carolina), Sep 3 (IANS) Palantir Chief Executive Alex Karp said that India could emerge as a major artificial intelligence (AI) power, citing its technical talent, energy initiatives and capacity to develop advanced AI models.

“I could see it doing very, very well,” Karp told IANS when asked where he saw India in the field of AI.

“I mean, first of all, model development, energy source. I mean, you guys are doing a lot of scrappy things to get energy,” he told IANS during a media interaction on the sidelines of the G20 Innovation Ministerial here.

Karp described energy and technical expertise as vital to the development of AI.

“One of the bottlenecks is energy, technical talent,” he said, adding that every country possessed distinctive technical capabilities.

The Palantir CEO called AI possibly the most consequential technological advance in a century.

He also cautioned that it could be deployed for constructive as well as harmful purposes.

“The most important technological development. It, it may be in the last hundred years,” Karp said.

“And because of that, it can be used for good and ill. So you really do need to have experts looking at it.”

Karp said nations would have to find ways to build and retain control over their own sovereign AI systems.

“I think what’s particularly interesting about the conference is every country has to figure out a way to have its own sovereign AI,” he said.

He pointed to several questions governments must address while building such systems.

“How do you control the weights? How do you control the data? How do you control where it goes?” Karp said.

“How do you make sure it’s indigenous to your own country?”

He identified the United States, China and Israel as leaders in the field, but said other countries must also develop domestic capabilities and determine how best to use them.

“America, really America, China, Israel are in the forefront here, but it’s really important that other countries have their own AI developments and figure out ways to use them,” he said.

Karp added that organisations using AI needed computing capacity and an application layer.

They also needed to decide where to deploy closed and open models to protect valuable information, he said.

“You need compute, you need an application layer, you need to be able to control where you put closed and open models so that you’re not losing your telemetry, your alpha, your government’s secrets maybe,” he added.

Palantir’s platform, he said, could run models from different developers.

The company preferred American products but supported competition in commercial markets.

“We have to win this battle, and we have to have American models be the best,” Karp said.

“But you can’t just say our models are gonna be the worst and people should use them even though they’re more expensive and equally not as safe.”

Karp also warned governments to approach AI regulation carefully, saying poorly designed rules could hurt the countries they were intended to protect.

“One of the biggest problems is it’s gotta be done by people who understand the issues,” he said.

“That doesn’t mean there should be no regulation. So I, I’m not an anti-regulation person.”

Palantir Technologies is a US software company co-founded by Karp in 2003. Its platforms help government agencies and commercial organisations integrate and analyse large volumes of data.

The company is known for its work with US defence and intelligence agencies. It has also expanded its commercial business, providing software for sectors, including manufacturing, healthcare, energy and financial services.

–IANS

int/lkj/khz