Developers need to be responsible, should focus on trust, inclusivity: Nadella

Microsoft CEO Satya Nadella on Tuesday (February 25) said developers should focus on ethics and building trust when developing solutions to ensure that the benefits of technology are available for everyone.

The India-born top executive also highlighted that developers should work with diverse teams, so that ‘unconscious biases’ do not creep into the models used for developing artificial intelligence (AI).

“Technology is becoming so ubiquitous and pervasive in society in our lives. (It) also comes with responsibility. Every single choice of developer around design ethos, the ethics of the diverse team are going to matter in terms of are we going to create a more inclusive world?” Nadella said at the Future Decoded event.

Related news | AI isn’t unbiased because humans are biased

He further said the question that needs to be kept in mind is that would these solutions be used for prosperity across industries like retail, healthcare and agriculture or “is it just going to be a narrow sliver of society?”

“I think inclusivity around the surplus that gets created by digital technology, is going to be important,” he told an audience comprising developers, partners and customers.

According to Microsoft, there are about 4.2 million developers in India and the country is poised to become the numero uno destination for such talent in the coming years. Nadella, who is on a three-day trip to India, pointed out that platform developers are ‘first responders’ and have to develop trust into technology.

“Every bank that builds an app will have to deal with trust, and trust around AI and the AI models they deploy. The core cybersecurity of their asset, their customers data…if there is a real currency of the next 10 years, (it) is how diverse and inclusive your teams are that are building all these technologies — gender diversity, ethnic diversity – because there is no protection against unconscious bias other than diversity,” he said.

Related news | AI-based systems can be used to determine good, bad credit risk: Nilekani

With AI being used to generate almost human-like responses, ethics is beginning to be a core part of discussions around training of models for AI. For example, if a developer has a negative bias towards a certain thing, this bias can creep into the solution and impact the output of the solution.

Talking about trust, Nadella pointed out that the company is meeting all the data residency laws in different countries.

“We have 57 data centre regions. We have three regions in India (Pune, Chennai and Mumbai)…as we are expanding around the world in all of these regions, that means we are also meeting all the data residency and data sovereignty laws,” he added.

On Monday, Nadella participated in a fireside chat with Reliance Industries chief Mukesh Ambani.