Rapid advances in AI software creation are driving an existential shift in the industry. With this, traditional coding is entering its “handloom” era, according to Sridhar Vembu, Co-founder and Chief Scientist of enterprise software firm Zoho.
Comparing software to a “handloom business,” Vembu said “powerful machine looms” are emerging, referring to automated systems generating complex computer scripts at high speed. Automated systems can already write functional code faster and more efficiently than many human developers, he added.
“AI already can write better code than at least 60-70% of software engineers out there. And in maybe three to six months, when things accelerate, AI can do better software than maybe 90% of software engineers,” Vembu said, speaking at the 17th edition of YourStory’s flagship tech-startup event, TechSparks.
Venture capitalists like Chamath Palihapitiya have suggested that intellectual property, the underlying proprietary ownership of written software code, is effectively dead, he added.
Despite the impending disruption to tech professionals, a large part of India's population will remain largely unaffected, said Vembu, presenting a counter-intuitive macro perspective. He pointed out that roughly 97% of the broader population does not write software and operates as consumers of technology.
Outside major tech hubs such as Bengaluru, districts like Tenkasi in Tamil Nadu or Patna in Bihar will experience minimal direct economic upheaval if coding jobs disappear, said Vembu.
Furthermore, as software creation costs fall, automated efficiency could exert deflationary pressure across the broader economy. Lower production expenses mean non-technology workers, including teachers, nurses, and barbers, effectively gain increased real purchasing power as consumer goods become more affordable.
The technology ownership gap
The true obstacle to national prosperity is not the loss of outsourcing jobs, but a fundamental lack of technology ownership, according to Vembu.
Although Indian engineers write a vast portion of software for Microsoft and Cisco, and design up to 50% of Nvidia graphics processing units, the primary financial returns flow to foreign intellectual property holders, the entrepreneur pointed out.
To highlight this disparity, Vembu compared a Swiss school teacher, who can earn an iPhone in 3-4 days of work, with an Indian teacher who needs to put in months of work for the same item, despite neither being directly involved in technology.
“Ultimately, who has ownership of the technology determines what is the income level of the non-tech people in that economy. Not the income of tech people alone,” Vembu emphasised.
Reflecting on the 1990s, he argued that India was misled by Western commentary, such as Thomas Friedman’s writing celebrating IT services outsourcing, which rewarded labour over true technology ownership.
To achieve lasting economic independence, India must adopt the strategy of East Asian nations such as Japan, South Korea, Taiwan, and China. Instead of rushing into top-level AI applications, China established durable national sovereignty by developing foundational domestic layers first, including search engines like Baidu and messaging ecosystems like WeChat.
Vembu cautioned against prioritising AI over essential industrial hardware and infrastructure, warning that a nation relying entirely on foreign hardware and foreign operating systems risks extreme geopolitical vulnerability.
“One reason I came to AI last is, we obsess about AI. We should not be obsessing about AI first. We should be obsessing about the more foundational layers,” he said.
True national resilience requires mastering core technologies such as lasers, x-rays, robotics, operating systems, and agricultural technology, according to him.
This philosophy guides Zoho's operational strategy amid rising infrastructure demands. The surge in AI adoption has caused server and graphics processing unit costs to explode, with server prices climbing from $8,000 to $45,000. Instead of executing corporate layoffs, Vembu said Zoho is redeploying its engineering talent towards deeptech hardware, including building server infrastructure in-house.
Highlighting this approach, Vembu noted, “As much as our jobs are challenged by the present generation of AI. You know why we don’t lay off anybody? We know we can redeploy them to build all this stuff.”
The emerging tech order, according to the entrepreneur, demands a shift from exporting software services to securing genuine national technological ownership.