Nandan Nilekani, co-founder of Infosys and the man behind India’s Aadhaar project, joined Satya Nadella, the third ever chief executive of technology giant Microsoft in front of a tightly packed audience in Bengaluru recently. Nadella is in India right now visiting his colleagues, developers, entrepreneurs and government officials. On stage, Nilekani and Nadella mostly spoke about the Microsoft platforms and Aadhaar and India Stack, a set of applications being built on the Aadhaar platform. If there was one area that their conversation veered into again and again it was artificial intelligence (AI).
AI is very much in news across the world, not least because it threatens to replace jobs in the short run and impact us in unpredictable ways in the long run. There is debate over how societies will deal with its growing power. Some are advocating universal basic income to all. Bill Gates recently said governments should tax the robots that take jobs away. Tesla Motors’ Elon Musk suggested that humans must merge with machines to avoid becoming irrelevant.
So, what do the two tech titans make of AI? Their views matter. Microsoft’s market cap shot up after Nadella took over as CEO, in part because of his focus on AI. Aadhaar and India Stack are expected to bring more people into the digital economy, which in turn will generate huge amounts of data, a fodder for AI.
There were three big takeaways from their conversation. One, we are nowhere close to artificial general intelligence, but we are going in the right direction. Two, we should indeed be concerned about AI taking away jobs. Three, there are ways to make AI more acceptable.
Artificial general intelligence is not around the corner, but we are climbing the right ladder
Binny Bansal, a co-founder of Flipkart, who joined Nilekani and Nadella halfway through, recalled that his IIT batchmates were not too keen on studying AI when he was a student there. But now AI seemed to have picked up. He wanted Nadella to give his views on what AI holds for the future. Nadella said, the most exciting development in the last five years took place in one specialised branch of deep neural networks, fundamentally giving us human perception level AI, whether it’s speech or image recognition. That’s just magical to see, he said. But, “I don’t think we can go and claim that artificial general intelligence is just around the corner. If anything I would say we are on the right ladder this time. I don’t think we are going back to AI winter as long as we don’t overhype it, and we are all grounded in where we are and we show real applications, solve real issues.”
“Ultimately, it’s human language understanding,” Nadella said. “That still doesn’t exist. We are not even close to it. We have a lot of parlour tricks but not anything that says we can actually have the ability to write like Rabindranath Tagore anytime soon. We will just have to keep taking steps on that ladder.”
Take a pause, and listen to Nick Bostrom, a philosopher at Oxford who has thought deeply about AI and its impact on human life, and the subject of a fascinating profile by The New Yorker to get a long-term perspective on AI.
Juxtaposing Bostrom’s views with Nadella’s comments, one might conclude that AI might be nowhere close to its immense potential, but it is on the right track. In fact, it has already started disrupting the way our society and economy works.
AI and the future of jobs
A few days before Nadella spoke about concerns regarding AI’s impact on jobs, Vishal Sikka, Infosys CEO, gave a presentation to investors and analysts at an event in Mumbai. In that, he told them that efforts worth the work of 2,650 full time employees were replaced by automation in the third quarter alone. The trend has just started. In manufacturing, it started even earlier and is much bigger. Foxconn, the Taiwanese manufacturer of iPhones, for example replaced 60,000 factory workers with robots.
Take a moment to savour that thought, as you watch this documentary, aptly named Humans Need Not Apply.
There are many such videos on YouTube, and it’s no wonder that there is a growing backlash against automation in the West. The big question is whether India should look at automation and AI with the same lens?
There is a way to make AI more acceptable
To view AI and automation only from a jobs perspective would be an incomplete view. And we will be blindsided into thinking that we will be able to wish it away. A more proactive approach would be to accept that it’s there, and see how best to make use of it.
Nilekani’s comments reflected that pragmatism. He said: “When you think of the challenge of taking the country with $2,000 per capita to $20,000 per capita, it is really a major challenge. You have to fix basic things like health, education and access to financial services. The classical way of doing that would have been to say let’s have more doctors, let’s have more (teachers and so on). That’s certainly not possible in the timeframe that we have. If you want to get everyone educated, and if you take 25 years to do it, then they will be adults and you can’t do it. The only way to square the circle is by using AI and cloud to deliver personalised health, education and finance to a billion people. That will drive the economy. For a country that has low per capita income, to use this as a strategic tool... is very important.”
Listen to the full conversation here.
That technology must help the poor is not a new insight. Mahatma Gandhi spoke about it even before independence. When Jawaharlal Nehru set up the elite Indian Institutes of Technology, he hoped something like that would happen. So, it’s easy to be sceptical when tech titans talk about the poor. But, one can assume that they mean it. They are aware of the growing backlash against hi-tech—in some ways underlined by the simmering tension between Donald Trump and the West Coast in the US. There is societal pressure. Political pressure is picking up. Technologically, it’s possible. It’s even possible to make some money out of it. So, why not?