Technology regulation should be in the national interest: Satya Nadella

The company’s $1 billion investment in Open AI LP, the for-profit arm of Open AI Inc.—the popular nonprofit behind Chat GPT that’s currently trending in the tech world, is being considered to give Microsoft an edge in incorporating AI into search to open up new ways in which we search for information online.

During a 4-day visit to India, Hyderabad-born Nadella sat down with mint for a wide-ranging conversation about AI, tech regulation, startups, modern work, India’s public digital infrastructure, and how he sees his work at Microsoft. Edited excerpts:

Are you watching Bing launch with ChatGPT or another language model built in?

We’re excited about the work we’re doing with OpenAI, and you can expect us to do a lot with platform models. Like what we did with the GitHub co-pilot, it was an early experience of what I think could be a fundamental change in what will happen when major platform products and models like GitHub coming together to empower people. A software developer today can be a lot more productive, a lot more creative, and can really flatten the learning curve. Quite frankly, we want to do more with AI across every Microsoft product. So I would say wait and see because we’re just getting started.

In terms of accuracy in search results, are these language models really there?

I think big language and search models will come together in very novel ways. There are fundamental challenges around any model — in terms of how you build the foundation of the model and how you maintain its freshness — those are all great technical challenges that I think will answer. Good technical answer. And in due time you will see the output of all that.

Since the English language is the language in which such models are primarily trained, what will happen in countries like India, where there are countless languages?

That is a great question. I think we shouldn’t think of it just as this one shot. It’s the continuous improvement of the model, with human feedback and AI-generated content. So let’s say I want to read Lorca translated into Hindi or English. I can do it today. Is it a Hindi poem? Is there an English transliteration? Is it a spanish poem?

What is considered knowledge in one language can be used in another and can be modified in another, after which it will come back. So we should think about the fact that more and more people around the world consume these big models in all languages ​​and add to that; that’s how we really democratize these differences and also these biases. Ultimately, collective human judgment is the only test against bias.

What do you think people don’t fully understand about where AI is today?

I would say that what perhaps needs to be better understood are some of the scaling effects and emerging behavior of large patterns. I’m not saying this is the be-all and end-all or the ultimate evolution of model architectures.

But GPT3 or 3.5 are not linear steps. So the question is, where does the non-linear progress come from? There are studies where if you train a large model on a data set of, say, all mathematical equations. It was good there. But if you train it on all the math equations plus all the web text and all the documents in the world, it will be better at math than just the corpus it was trained on.

And you might say, why is that? That’s less intuitive because you and I don’t go to school and just do math. We learned history, we learned languages, and we learned a lot of other things, which, by the way, created common circuits (in our brains) that help us learn math better. So that’s the key, I think.

In the evolution of the Indian tech industry, we have legacy large technology services companies. We’ve seen a wave of consumer startups lately. Are we about to see another wave of startups mining AI and the cloud?

Honestly, one of the things that gives me great optimism for India is all the different ways technology is being used. I spent time with a small business called Senco Gold that was expanding its retail stores with technology. I met a blind man at State Bank of India and was building an app using Power Apps to automate a group of workflows.

India already has the second largest number of developers and is the number one in terms of AI repositories which means that Indian developers are scouring every open source AI repository and becoming fiduciaries. core. So the simple answer to your question is that in the long run, AI will be used more. One of the things that you should talk about, even in India, is the strong use of new technology, whether it is a gentleman at SBI using AI or the latest unicorn, because both have super value for the development of India.

As the super boss of both Microsoft Teams and LinkedIn, what can you tell us about how work has changed in the post-pandemic world?

There are three data points about the post-pandemic world that we are all learning about. One is the paranoia about productivity. Some leaders say we’re productive enough, while employees say we’re exhausted because I’m working from home or sleeping at work? I think the way to connect that is through data. There are matching results. Every organization has to compete, and ultimately, work needs to be done. Use that instead of dogma about any particular way a pre-pandemic workflow could be helpful.

The second data point is that people come for others; they don’t come for policy. I think all of us as leaders have to learn some soft skills about how to bring people together. I think, learning better soft skills on how to make connections is very important.

The bottom line is, just because you hire them once doesn’t mean you shouldn’t re-hire them every day. None of us can take anything lightly, especially in a market like India.

What role do you see India playing in the overall regulation of technology and how do our current regulations affect your business?

I think every country will have regulations so that technology that is both introduced into the country and created in that country ultimately serves the social purpose of that country. We welcome and also do not expect regulation to have fundamental trust in technology and technology business models. That’s how I approach it.

It is the continuous evolution of management modes and technology, and everything has to keep up. And just because regulation can’t keep up with the technology that’s coming, frankly, people like us who are working in this business, have to say we have a fundamental belief How core is into what we build, by design.

Some of the things I’m seeing in India, even in terms of data privacy and data flow, are very illuminating. Finally, India has three things. One is that we’re making money in India by building our data centers and human resources so others can make more money in India—whether it’s Senco Gold, SBI or InMobi .

But they don’t just manufacture in India for India; they are doing in India for the world. So then you have to have rules explaining all of that. Data doesn’t come from one place. It could have been made somewhere else, and it should have gone to India. Then in India people will be able to use it, add value to it, then get it out of India and let others use it. That’s how regulators think about it. And by the way, all should be done in the national interest. It is not about benefiting anyone other than the people of this country.

When you take on the CEO position, you have a lot of work to do. Now you are being mentioned as one of the great CEOs of the world. How do you look back on your tenure at Microsoft?

I feel that if I have to live the culture I follow, I have to focus more on the mistakes I make every day so I can learn from them and become better tomorrow. And if anything, we are all temporary custodians of organizations that hopefully, if we do it right, will outlast us.

If there’s one thing I’ve learned from my Indian civil servant father, it’s that getting organized after you leave is more important than anything you’ve achieved during your tenure. .

I wanted to make sure that I personally and everyone at Microsoft had the confidence to admit my imperfections.