How to apply the power of AI for the public good

There is no longer a debate about whether artificial intelligence (AI) is a disrupter or enhancer, but rather about how to use its powers responsibly and for the public good.

That remained a tough challenge, Aptagrim chief executive and founder Sreegopal Mandapaka said in a virtual address at the Nelson Mandela University Faculty of Business and Economic Sciences this week.

Acting director of the university’s Business School, Prof Paul Poisat, and Prof Grant Freedman, a research associate and part-time lecturer, interviewed Mandapaka via Zoom from his base in India.

They had invited the technology entrepreneur to speak as a guest of Breakfast Club @Business School, with a view to understanding the implications of AI for people and business.

“Business will have to build in policies, and it will have to be standing on the shoulders of large multinational companies such as Microsoft and Google, which have the power of access to data, to define these strategies,” Mandapaka said.

Most countries lacked policies, he said, and were unable to control or channel the now enormous flow of data.

Mandapaka created AI technology firm Aptagrim in 2012 to address real-world problems and transform business operations.

Based in Hyderabad and with branches in other countries, Aptagrim is a deep-tech AI consulting company that builds products for clients, and helps organisations to incubate AI innovations.

It has worked with private entities and national governments, including those of India, Singapore and Ireland.

Mandapaka said his own background — he has a master’s degree in biochemistry from the University of Pune, India — had shown him the power of data in bioinformatics.

“It is all about being able to make some substantial progress in leveraging these technologies to have an impact on human life.

“I saw how you could use data to make a difference to the quality of people’s lives.”

Mandapaka cited two of Aptagrim’s major innovations: smart glasses to assist visually impaired individuals and a video-based psychometric assessment tool.

The smart glasses help the wearer to navigate physical spaces, such as walking into a shop, and advise of obstacles ahead.

The glasses also can “read” text, whether it is on a menu or in a book.

The psychometric tool helps businesses to speed up the traditionally time-consuming task of job interviews.

It does this by asking candidates a set of questions that people naturally respond to, over a few minutes.

It then uses AI-powered subject profiling to look at factors such as facial expression, body language and speech patterns.

Noting the tremendous amount of progress Aptagrim has been making on the psychometric tool in particular, Mandapaka sees it as a game changer.

Poisat highlighted how a discussion on the topic of artificial or augmented reality could become heated.

“The adoption of this technology is developing and growing at an unprecedented rate never witnessed before,” he said.

“It sparks fears of loss of jobs and displacement, and yet AI innovation can — and does — enhance business and has many more applications.”

It was therefore imperative that the university, and the faculty of business and economic sciences in particular, provided education for the numerous new careers that would emerge in the future world of work.

Mandapaka agreed that AI was changing the landscape of work, saying it was also giving developing countries an opportunity to scale innovation.

AI had also advanced far beyond relying on a single language of input, such as English.

“Whenever there is new technology, there are new areas of responsibility and if it is not controlled, or co-ordinated, it can lead to a lot of challenges,” he said.

When ChatGPT “blew up” towards the end of last year, for example, he said “we were all caught off guard; nobody saw it coming”.

“We are already seeing a race between the big multinationals and it all boils down to data security and how optimally these AI modules are utilised.

“Unfortunately, there is no road map or clear, immediate answer.

“One of the problems we have is fabricated data that gets spread across social media and the internet.

“These are challenging times as well as great, exciting times. It will take some time for these policies to fall in place and for us to have a much more secure world adopting AI.”

However, he said AI was “definitely not as complex as people sometimes think it is”.

“What we want to do with that knowledge, and what challenges we would like to address, that is really the key factor,” Mandapaka said.