Generative AI is having a massive impact on many sectors. But to none, it poses as much, and as quick, a threat of disruption as the IT Sector itself
By Prof. Vidhu Shekhar, Faculty, Bhavan’s SPJIMR, Mumbai
Generative AI, with models like ChatGPT, is as disruptive as the internet and smartphones once were. In the near term, as tools like ChatGPT permeate enterprises, we could see a massive surge in white-collar productivity, potentially reshaping the trajectories of numerous sectors.
However, it is not all good news. J P Morgan predicts that, over the next 4-8 years, there will be significant job shifts across various roles – customer service, graphic design, legal work, healthcare, content creation, and more. Entire professions risk being disrupted. Including Tech. Already, among many others, a giant like Google is reportedly likely to lay off 30,000 employees post new AI innovation.
Interestingly, no sector faces as immediate and significant a threat as IT Services.
IT services: First in the firing line
While almost every industry faces generative AI-induced churn, perhaps no domain will witness as swift and intense tectonic shifts as IT services.
IT services giants have thrived on outsourcing arbitrage by leveraging wage gaps and abundant engineering talent. However, this entire offshoring construct may now be threatened. From AI tools surpassing human coders in capability to the rebalancing of outsourcing, the very foundations of this sector may get disrupted.
AI coders well outpace humans in capability and cost
ChatGPT launch last year showed to all how capable AI has become in human conversations.
As it turns out, AI is even more better at speaking the language of a “dumb machine” (which is programming/coding) than it is in human languages.
Coding plays to Generative AI’s strengths – it is essentially language and communication for machines. Coding languages have more rigid grammar & syntax than complex, context-heavy human languages. Acting as coding co-pilots, they can create functional code in various programming languages from prompts. And they can even debug itself as well if the errors are pointed out!
Their effectiveness is such that at major coding hackathons around the world, we are now seeing a new kind of winner – solo product managers paired with AI Copilots for coding and without any coder. Afterall Product Managers are already experienced in giving instructions for coding – earlier to coders, now to AI.
Redefining and shattering coding timelines
With product managers using these tools for quick prototyping and even shipping production apps without dedicated developer teams, the need for human coders drops significantly. The few coders required for it gain better support/efficiency and the Product managers gain more precise insight into the effort level of coding tasks – setting new meanings of Agile Development.
And we are seeing this start to unfold. Reportedly, Vijay Shekhar Sharma at Paytm is pushing his 10,000-member technology, product, and engineering teams to adopt these AI tools. According to him, this move has dramatically sped up product development – from weeks to just days in many cases!
Paytm is already seeing a 10-15% reduction in human costs, with further savings expected as AI utilization increases. We are likely to see a significant decrease in the demand for coding manpower across sectors, not restricted to IT itself. But in IT, it poses the most threat.
Cheap, fast, and disrupting IT outsourcing models
Not only can Generative AI write programs and codes faster and cheaper, but the cost of implementing such AI engines is also in rapid decline. For e.g. Bloomberg invested over a million dollars in developing a finance-domain-focused Large Language Model (LLM) named BloombergGPT. Just recently, within just six months of BloombergGPT’s release, an open source model (AdaptLLM-7B) costing merely $100 is surpassing BloombergGPT in performance.
Closer home, we recently had a significant Indian IT services firm lose a $1.5 billion deal on providing AI digital transformation support services cancelled by a client – most likely to bring more work in-house using AI tools.
Irony abounds as AI disrupts its developers’ first
With solutions able to be built 10x faster – from weeks to days – at a fraction of the cost, outsourcing models face significant disruption. Additionally, firms are increasingly seeing new custom IT solutions not as expendable commodities but as strategic assets. Take conversational AI – a bank may not want to outsource a cutting-edge Conversational AI Financial advisor that interfaces with customers and has deep access to their data. We could see more IT functions being retained rather than outsourced, even by non-tech firms.
Already reeling under the impact of an uncertain global economy, this disruption can cause significant troubles to IT services firms. The impact already seems to be reflected in Engineering Schools and B-Schools placements this year, which saw a drastic fall in IT recruitment this year, perhaps because of this increasingly lower need for human talent in the whole IT services process.
Navigating the AI storm – What Indian IT must do
The $15 billion Indian IT services industry, is likely entering a significant transformation phase due to Generative AI. IT outsourcing in non-core, non-strategic areas may be able to endure this, albeit with slimmer margins. Combined with clients insourcing via AI, this might significantly reduce both growth and margins.
As a first response, IT services are beginning to adopt generative AIs and Co-Pilots internally to lower their own human costs. This though means that the sector could experience significant job losses, a trend that is already visible and may accelerate. There is no escaping that.
But even though this internal transformation could make the industry leaner and more prepared to tackle future challenges, it may not be enough.
The most critical step for IT services would be to pivot towards developing their own large-scale AI solutions. We need systems like Infosys GPT, TCS Co-Pilots, and Jio LLMs. New-age services firms will need to have their home-grown smartest AI coding agents. Smart firms like Jio are already taking steps in this direction.
In the famous 2011 series Margin Call on the 2008 Global Crisis, CEO John Tuld, talking about the financial industry, says, “There are three ways to make a living in this business: be first, be smarter, or cheat.”. For IT services in today’s world, AI is the smart. Indian Firms need to be really fast, if not the first.
