Wanna catch the AI bus? Linux Foundation India’s clarion cal…

Indian IT companies have had to hear about having missed the bus – the AI revolution — as they remain engaged in playing catch-up, held back by factors such as low R&D investment, a services-over-products mindset, dependence on headcount, and a general aversion to risk.

Their cautious approach towards innovation, particularly in the AI space, is evident in the fact that most leading IT firms do not disclose AI-related revenue.

However, to be on the bus now would require moving beyond consuming open source software and becoming active contributors in its global ecosystem, suggested Arpit Joshipura, head of the Linux Foundation India (LFI). 

“The only way to get ahead is to participate and contribute to open source. You cannot get ahead by consuming,” he stressed.

Joshipura warned about the pitfalls of merely forking codebases. 

He explained: When companies just fork, consume, and make changes without contributing those back, they face significant challenges. When the mainline code changes, companies must propagate those changes. Before long, there are thousands of different versions and the code base becomes completely fragmented, causing the companies to fall behind and spend time and effort maintaining their divergent code.

“Instead of just catching up, get ahead. Get ahead the open source way,” Joshipura emphasised during an interaction with AIM, on the sidelines of the Linux Foundation Open Source Summit.

Ambitious Roadmap

He highlighted that India has the capacity to host indigenous foundation models publicly. With support through education, training, and developer-centric events, developers are now learning how to architect and manage open foundation models. “There is no shortage of leadership here to do this.”

This perspective underpins the Linux Foundation’s ambitious five-year roadmap focused on growing India’s open source footprint. With the goal of fostering collaboration across India’s tech ecosystem, LFI aims to drive investments in training, education, and infrastructure, ensuring that the Digital Public Infrastructure (DPI) work initiated by the government in the previous decade is backed.

The initiative is designed to prevent India from becoming a mere consumer market for open source, where users download and leverage code without giving back. According to Joshipura, best practices revolve around a cycle: “you use it, you contribute, you lead.”

India Lead Contributor by 2030?

He confidently projected that India will emerge as the number one contributor to open source by around 2030, measured by sheer volume of commits, and aided by efforts towards educational and training events that will instill robust open source practices.

Joshipura anticipated significant convergence globally within three to five years and opined that geopolitics does not affect open source. Global critical infrastructure will increasingly rely on shared open source components, making collaboration vital, he said. Country-specific applications and enterprises will mature using open source as their “plumbing,” while innovation will focus on added layers above it.

From a technical standpoint, Joshipura explained that a large portion of software code, roughly six to seven million lines out of ten million lines (for instance) in typical technology stacks (like mobile phones, UPI, Aadhaar), is “non-differentiated.”

This essential but generic code, handling tasks such as starting and stopping operations, cloud connectivity, logging, database setup, and security, is the core focus of the Linux Foundation’s work.

Companies cannot monetise this non-differentiated code, yet it is critical infrastructure. The aim is for startups and large enterprises to build upon this shared foundation rather than creating proprietary systems, as “the world of proprietary is gone; nobody cares about it anymore.”

Consumption Culture Hinders AI

India currently ranks second in the open source developer community, primarily due to the volume of code committed by enterprise developers from companies like Infosys, IBM, and Red Hat.

While these developers are sophisticated, innovation from Indian developers, especially in AI, is still nascent. A prevalent “consumer mindset” leads many to simply use open source code without contributing enhancements back to the community. 

Cultural factors contribute to the lag in contribution. Many developers, according to Joshipura, work for companies that do not prioritise open source, often continuing the historical outsourcing legacy where client-driven priorities in the US and Europe did not emphasise open source contributions.

However, Joshipura said, “this is changing,” noting that Infosys has become the first among India’s top IT firms to donate a project to the Linux Foundation, signaling growing corporate engagement and leadership.

Global Companies in India Are Participating

Indian AI startups and SaaS companies, according to Joshipura, currently contribute incrementally to open source by focusing on narrow, sector-specific use cases, often without joining open source foundations or driving radical innovation. His call to these startups was to contribute their innovations back to open source foundations and seek global adoption and collaboration.

He reminded that India’s open source ecosystem also includes significant contributions from global R&D centers located within the country. Firms such as Microsoft, IBM, Ericsson, Nokia, Cisco, and Red Hat maintain large engineering operations across Hyderabad, Bangalore, Gurgaon, and elsewhere.

These centres, he observed, are engaged in substantive open source projects, indicating that India’s contribution to open source transcends indigenous companies and includes these global technology hubs.

Open Source a Recent Phenomenon in India

One of the persistent barriers to rapid growth of open source contribution is relatively recent momentum.

“The surge in participation only started about three years ago,” he said.

This trend is distinct from the rise of AI and traces back to government policy initiatives around 2015, notably the Digital Public Infrastructure policies mandating open source for critical infrastructure, he recalled. 

Commenting on compute power inequality, Joshipura said that both globally and within India it is a reality affecting open source AI development. He reiterated that AI demands linear scaling of compute resources, increasing GPU requirements tremendously, which is unsustainable long term. Technological innovations like retrieval-augmented generation (RAG), very large language models (VLLMs), and adaptive training aim to break this linearity, he observed. 

Investment in local data centre infrastructure is essential to address the capital-intensive nature of powering AI at scale. India’s favourable geography offers opportunities for efficient and sustainable data centre deployment, he concluded.

The post Wanna catch the AI bus? Linux Foundation India’s clarion call for Indian IT community appeared first on Analytics India Magazine.

Generated by RSStT. The copyright belongs to the original author.

Source