Infosys co-founder and chairman Nandan Nilekani has criticized the trend of creating more complex artificial intelligence (AI) models, saying it’s not leading to better results. Sharing his thoughts on AI at the People+AI Mela in Bengaluru, he pointed out that many of these models are kept secret by companies, making them like “black boxes” to the public. This lack of openness, he said, is slowing down innovation, reports Economic Times.

Nilekani praised China’s approach to building AI models. He noted that China has managed to create effective AI models using fewer resources and has been transparent about its methods. This openness, he believes, helps spread knowledge about AI and supports global development.

He also mentioned that AI models are becoming more like common goods (commodities), with many new models being created. He highlighted China's DeepSeek project, which openly shared its AI-building techniques, making it easier for others to learn and innovate. “They have done the world a service. So that goes back to commoditisation,” he stated.

“I have always believed that models are a commodity. This is proving to be true. Hundreds of models are getting built. China’s DeepSeek demystified model-building. Last week they tweeted all their open-source practices. The knowledge to build models is getting rapidly commoditised,” Nilekani added.

He also stressed the need for high-quality data, especially in Indian languages, to boost AI growth in India. He mentioned AI4Bharat, an initiative by IIT Madras that focuses on open-source AI for Indian languages. Nilekani has supported this project for the past few years and plans to continue for the next three years.

He concluded by saying that while AI models will keep evolving, the real focus should be on using AI in practical ways to make a positive impact on people's lives.