Break-through productivity from Generative AI a potential threat for IT Services: Fractal Analytics

Generative AI has shown breakthrough in productivity gains across segments, says Fractal Analytics co-founder and vice chairman Srikanth Vellamakani.

Generative AI has shown breakthrough in productivity gains across segments. These gains and challenges associated with AI can be seen especially in the technology sector.

In coding, Generative AI has shown more than 50 percent in productivity gains, and this could prove to be a threat to the Indian IT sector as dependency on humans for jobs involving coding will be reduced, said Fractal Analytics co-founder and vice chairman Srikanth Vellamakani. He was speaking as a part of Motilal Oswal’s Global Investor conference.

Fractal Analytics is a global analytics firm.

Watch: 'Humans who don’t use AI will be replaced'

According to projections, over the next few decades, labour contribution to overall growth will continue to reduce and that gap will be filled by AI. “The only skill you may soon need will be English. If you can input the information into AI, you will no longer need to know languages like Java, C++ etc.,” Vellamakani said. What this means is that while the overall technology spending from corporates will only increase, given the focus on the space, it may shift spending from IT services to other areas, which are better suited for AI-led corporate environment, the Motilal Oswal report added.

Some of the top Indian companies in the AI space include Tata Elxsi, Kelton Tech, Bosch, Happiest Minds Technology and Zensar Technology.

Also Read: Coforge launches generative AI platform for enterprises

The impact of AI has been seen not only in India but also globally. A recent report stated that Italy banned Chat GPT for a few months and coding productivity dropped by 50 percent as a direct result of it.

Currently, three types of AI are being deployed globally – Supervised, Unsupervised, and Reinforced learning. Open AI, the report explains, was created by combining supervised and unsupervised AI to create a “self-supervised” model.