Indian companies excel at replicating concepts developed in other places.

Infosys co-founder Narayana Murthy recently emphasized the need for Indian youth to work 70 hours a week and highlighted that Indians excel at applying ideas developed elsewhere to improve the nation. He acknowledged that India is still in the early stages of innovation and that creating original inventions will take time. Examples like Ola’s Krutrim model, which replicates OpenAI’s model for Indian use cases, as well as Flipkart, Ola, and Zomato emulating Amazon, Uber, and DoorDash, respectively, illustrate this trend. In contrast, companies like Zepto and Dunzo are introducing unique concepts and thriving.

Kunal Shah, founder of CRED, expressed on X that early-stage startups should be easy to iterate, while late-stage ones should resist distractions, reflecting a mentality among Indian startups that often lack iteration and innovation. Vivian added that startups should be adaptable, indicating a need for more agility in the Indian startup ecosystem.

The notion that India lags in building technology and leveraging AI is becoming more prevalent. Murthy suggested that the first step toward catching up with the West is excelling at applying external ideas locally and gradually innovating. He believes the current generation of young people is significantly smarter than his own was at their age and holds optimism for India’s future innovations.

Some view the application of global ideas locally as a pragmatic and efficient strategy to accelerate innovation by leveraging collective human knowledge. However, India's AI sector faces challenges due to insufficient investment and weak industry-academia partnerships, hindering research and development.

Critics from the Indian tech community, like Nikhil Malhotra from Tech Mahindra and Pratik Desai, argue that most research in India is derivative of Western models. Desai pointed out that India has unique use cases that don't necessitate foundational model research, advocating for more tailored applications of existing models. Gaurav Aggarwal from Jio warned that without developing its own AI infrastructure, India might soon have to import AI technology, akin to its dependence on imported oil. He lamented that India is producing "slightly glorified engineers" rather than true AI experts.

Dharmesh BA, founder of a stealth startup, noted that while many are working on AI, much of it involves merely repurposing existing models like GPT. He emphasized the contemporary challenge of not just building technology but determining what to build.