We realised after 1991 that we have to innovate, not wait in Delhi: Murthy

As long as we have corporate leaders who understand the power of AI and ensure that their employees are well trained, I do not see any problem, says Infosys cofounder N R Narayana Murthy

From the challenging pre-1991 business environment that meant waiting patiently for approvals in New Delhi corridors, to the changes that swept the IT sector after liberalisation, Infosys cofounder N R Narayana Murthy has seen it all. In an interview with Avik Das at his Bengaluru residence, he discusses the journey and the road ahead. Edited excerpts:

 

Can you take us through the major pain points that the technology services industry faced before 1991?

 

Friction was extremely high on all fronts for business in general and for start-ups, in particular, before liberalisation. First of all, there were hardly any telephones. It would take about seven years, on an average, to get a connection. But doing business with the US, the most competitive country in the world in one of the most advanced technology areas, without access to a phone was very difficult, if not impossible.

 

Second, there was a marathon of bureaucratic hurdles for businesses. If you wanted a licence to import a computer for software export, you had to provide a guarantee to the government to export 200 per cent of the CIF (Customs, Insurance and Freight) of the import price within five years. This was acceptable, but entailed going  to Delhi 20-30 times to obtain a licence, and the waiting time was 3-4 years. In addition, if you wanted to modify that licence to include a better and cheaper model of the computer system available in the US market, it would take 9-12 months and about 10 visits to Delhi! Those days, the US would release a new model of computers almost once every three months. So, we were eternally behind them by a wide margin. Therefore, our customers, who had installed the latest hardware in the US, found working with Indian software developers a big deterrent as we had access to computers that were three-four generations behind. 

 

Third, there was no current account convertibility. That meant you had to apply to the Reserve Bank of India (RBI) for every foreign trip, and wait for two to three weeks to obtain foreign exchange. That approval came rarely and only if the bureaucrat in the RBI decided that you were lucky.

 

There was an officer in Delhi, called the Controller of Capital Issues, who mostly didn’t know anything about the stock market but decided at what price you could have your IPO (initial public offering). When we applied, this demi-god agreed to have our IPO at ₹11 for a par value of ₹10 after a lot of haggling. We didn’t go ahead because it was not worth selling equity in the firm we built with hard work to investors at just a 10 per cent premium.

 

Lastly, there was no venture capital in the country. Software could not be used as collateral for a term loan since the banks did not understand software. Therefore, access to money was very difficult. Thus, India did everything possible under the sun, unwittingly, to decelerate economic growth. But all of that changed thanks to the architects of 1991 liberalisation: P V Narasimha Rao, Manmohan Singh, P Chidambaram, Montek Singh Ahluwalia and N K Singh. All these hurdles vanished like thin air. 

 

 

What were some of the major mindset changes that took place in the software industry post 1991? 

 

After liberalisation, Indian entrepreneurs realised that it was much more productive to sit in their offices, think of new ideas, study competition, and innovate better to beat competitors rather than waiting in the corridors of Delhi. They realised that the only way to strengthen the future of the company was through innovation, assembling a good team, and creating an open environment that encourages meritocracy, speed, and excellence in execution. Since liberalisation, the focus of entrepreneurs has shifted to satisfying customers through innovation and delivering better value, satisfying employees through market-oriented HR policies and investors through better corporate governance. 

 

Is it easier or more difficult being an entrepreneur today than in the 1980s and 1990s?

 

Personally, I have tremendous respect, admiration and appreciation for today’s entrepreneurs because they are doing a much tougher job in a highly competitive market. During our days, the hurdles were primarily from the government. The market was plentiful. Overcoming these hurdles did not require much intelligence or innovation. It required patience. But today’s entrepreneurs have to compete with the best in the world. Our entrepreneurs have to dream to succeed in the most developed countries in the world. So they have to be smarter, more innovative and daring. I believe that they are much better than I was.

 

But the recent wave of IPOs of new-age firms has turned the focus on founders who seem to have very little skin in the game. How do you see that change?

 

There is a lot of venture capital available today, and many of the VCs are from abroad. They compete very fiercely for a share in the pie of companies they think have a future. So there are many secondaries in the venture market. Today’s entrepreneurs, who are young, bright and highly innovative, have much bigger desires than we had. They would naturally be tempted to sell part of their stake to VCs, create wealth for themselves, obtain as much venture money as possible and spend as much as possible to accelerate their company’s growth, and get to the IPO as early as possible. Therefore, they give away a large part of their equity. 

 

Today’s entrepreneurs compare themselves with their counterparts in the US, where it is considered okay to divest a part of an entrepreneur’s equity to bring in Series A and B money. I don’t blame them because that’s the environment that has been created in the US. Our entrepreneurs too want to lead a life comparable to entrepreneurs abroad. 

 

Further, talent shortages have made acquisition expensive and require significant investment. Compensation levels have gone to astronomical levels. So, our entrepreneurs require lots of venture money. As a consequence, they have to give up a large part of their equity pretty early in the life of their company.

 

Do you think the IT services industry can maintain its previous growth rates in the age of AI?

 

I think there is a lot of confusion about AI. At the end of the day, it is just another technology produced by the human mind. The purpose of any technology is to increase revenue, productivity, quality and profitability, reduce cycle time and cost and make life easier. As long as we have corporate leaders who understand the power of AI and ensure that their employees are well trained, I do not see any problem.

 

It is true that certain parts of the software development lifecycle, like code development, will be taken over by AI. But remember that there is approximately 60 years of legacy code running in thousands of large corporations. Therefore, even maintaining the legacy systems themselves offers a huge revenue opportunity. Add to that the opportunity provided by elevating legacy systems to AI-based systems. This is a huge opportunity if our employees are well trained to do so. 

 

AI will not help you much in a very important aspect of the software development lifecycle: Requirement definition or defining the problem. As long as our youngsters learn how to focus on requirement definition, problem-solving, performance engineering, prompt engineering for generative AI, and trainable and self-learning algorithmic technology, I do not see any problem.

 

Second, AI will also help in creating critical pieces of code that will be reused many times in a large project. For example, validation of customer ID and invoice ID is used thousands of times in large commercial software projects. Such code can be generated by AI, tested thoroughly, and reused many times. 

 

Third, there are two kinds of AI algorithms for corporate use: Trainable and self-learning. The bigger challenge we have is the latter. We have not made as much progress as you and I would like.  

 

Global corporations need the help of Indian software service firms to help them leverage the power of AI. For that to happen, Indian software service companies have to be ahead of prospects and customers. That is where our leaders have to think about what will happen in the next two years, and create suitable training programmes, tools, and technologies.