I see generative AI as a helper, regulation is needed: Salesforce India CEO

When veteran banker ARUNDHATI BHATTACHARYA was appointed chairperson and chief executive officer of Salesforce India in 2020, one of her first tasks at the tech giant was to set up company systems on her home computer, all by herself. This showcases her passion to learn new things. These days, Bhattacharya and her husband are testing various features of ChatGPT, an AI chatbot that uses natural language processing to create humanlike conversational dialogue. “My husband gives it a topic and asks it to write a poem. Sometimes they are good, but there are a number of times where I end up laughing my guts off, because it is either not correct or uses words that you wouldn’t use,” said Bhattacharya. At an enterprise level, she expects generative AI and ChatGPT to change the future of work. Bhattacharya told Peerzada Abrar in a video interview that she sees the technology as an assistant and a helper to workers, rather than taking away their jobs. Edited excerpts:

Q. What impact do you see generative AI and ChatGPT having on businesses, including the information technology (IT) industry?

I expect that the productivity and efficiency of the workforce will go up. Much of the grunge work can be done by generative AI, which can then be checked. More people will be called upon to do things that need creativity, problem-solving and maybe emotional intelligence. There are certain areas where we need the physical presence of people to either convince us or to carry conviction. When you’re doing sales, unless it’s absolutely impossible to meet, we always try to meet the client. I think the client needs to see us and we need to understand the client better. I don’t think that is going to change due to ChatGPT.

Generative AI allows you to pose the query, not in any particular computer language, but in your own language. It uses a large language model to gather trillions of data points and therefore provide you with material with which you will then innovate and create.

Q. Does it mean change in the future of work?

I do think there will be a change in the future of work. There was a time when we did not believe that people could work remotely and still be productive, or that we could hold them accountable. Covid taught us that it was not really true. So, generative AI is another trigger, which will cause some amount of change in the rules that we see around us today.

Q. How do you see the banking industry changing due to generative AI and ChatGPT?

One of the things that we used to constantly battle with is the quality of service that we render to our customers. The quality of service used to differ from place to place. I think we can bring about far better outcomes there and make service standards far more uniform across people. Generative AI will tell you what script you need to follow, the answer that you need to give, and the outreach that you need to do. Most enterprises are dependent on IT, and generative AI will have the ability to generate at least the basic code. I see it as an assistant and a helper to people rather than taking away people’s jobs.

Q. What kind of innovations are you building in the generative AI space?

Our Einstein platform now has a GPT layer. It will allow you to use large language models and ask what is required to be done of the system in plain language. While you use your data for generating your answers, it also ensures that we are building a trust layer where your data remains your own. Nothing of this is going to cause any security or privacy concerns. All of our cloud products, whether it be sales, service, marketing, or commerce, are now getting infused with the GPT layer.

The company’s investment arm, Salesforce Ventures, is also expanding its ‘Generative AI Fund’, doubling the $250 million fund to $500 million as part of its continuing commitment to bolster the AI start-up ecosystem and spark the development of responsible generative AI.

Q. The government recently said that India will regulate AI to protect citizens from harm. How do you see concerns related to security, and the ethical implications of generative AI?

I think AI is not capable enough yet, that it can learn things totally on its own and develop beyond a point where humans wouldn’t be able to control it. However, one doesn’t know the future. That is why it is important to be open to being regulated. It is important to be open to understanding the threats that may emanate. Even when nuclear weapons were first tried out, people quickly realised the threat and they came together for some regulatory oversight.

If AI really poses that kind of threat, though it has not been established, we still need to get together and make sure that we don’t allow anything of that sort. Will there be rogue nations that (misuse) it? Yes, there would be. Some kind of regulation has to be there. For instance, when we deploy any of these large language models or generative AI, we make very sure that there is a trust layer, so that our customers are reassured that their data is not going anywhere, and their data is not being used to train anybody else without their consent.

Q. The government has a vision of ‘make AI in India’ and ‘make AI work for India’. How big is this opportunity and do you plan any partnerships with the government?

At this point in time, we are not very deep into the government because of various issues, on procurement and stuff like that. We are beginning to slowly get into the PSUs (public sector units). Regarding building such AI for India, OpenAI CEO Sam Altman recently was trying to say that one would need a huge amount of investment as well as a large number of people dedicated to it. But there is absolutely no reason why it should not be tried. Just because the model has been built in a particular manner, it doesn’t mean that the next person who tries it will also need to build it in the same manner.

What will be more useful for our people to get better at it is acquiring the right skill-sets in order to handle generative AI as it exists today. I definitely see it as a huge opportunity for India, mainly on account of the India stack. The country has a lot of data on which it can train its models, keeping privacy concerns in mind. We also have a large amount of young talent in the country.