AI: Endangering Humanity or Widening Horizons?

What is Artificial Intelligence (AI)? Will Artificial Intelligence (AI) destroy humanity? Or will we use it to improve our lives? Read the full article below, that appeared in today’s (3rd Jan, 2018) The Assam Tribune newspaper. 

Artificial Intelligence or AI has become a heated topic of discussion lately. A number of intellectuals, scientists, policymakers and entrepreneurs have expressed either optimism or apprehension about the possible effects of AI on humanity. While countries like Japan and the United States have started extensively using AI for various industrial and managerial applications, its use in India has been fairly limited till now, but fast catching up. Accordingly, it can significantly affect lives in our country, both positively and negatively.

Artificial Intelligence, in simple language, is essentially about making machines think and act like humans. In other words, AI is about replicating human intelligence. Hence, the prefix ‘Artificial’. The term was coined by John McCarthy of MIT in 1956. If there is anything that separates us humans from the rest of the animal kingdom, it is our intelligence. This human intelligence includes the ability to gather information through our sense organs, analyse it and then draw conclusions and inferences from that analyses. All this helps us solve complex problems. It also enables us to learn from our mistakes.  The ordinary machines that we use in our day-to-day lives, be it the computer, the razor, the hammer and so on, are not intelligent. They cannot perform the analytical reasoning and complex problem solving mentioned above. This makes them ineffective beyond a point in solving different problems and human intelligence has to step in.

But with the advent of AI, all this is about to change. Machines have been programmed in such a way that they can learn (by acquiring knowledge and information), reason (using various algorithms and acquired data) and also self- correct. They do this by mimicking the way the brain performs this task. The brain uses systems of neural connections to build up memory and analyse information. Likewise, machines have been wired into systems called Deep Neural Networks that almost behave like the neural networks in the brain. This, therefore, opens up a wide possibility of applications, some of which I discuss below.

Ever wondered how YouTube appears to know what type of videos you like or Facebook recognises the persons in the photo you uploaded, even before tagging them? The internet giants (Google, Facebook, etc) and other web-based companies are frequently resorting to a branch of AI called Pattern Recognition. This involves the collection and use of a large amount of complex data (called Big Data) that is generated based on our online activities, and then using complex machine learning algorithms to arrive at possible patterns. It helps not just Google or Facebook to know what your interests are (and thereby, feed you with related ads) but also optimises your online activities by reducing search costs. AI is being increasingly used in aviation and space applications, where computers can not only figure out what is wrong with an aircraft or spacecraft but also fix the problems. Areas requiring heavy data computations, like weather forecasting, are frequently relying on AI-enabled machines. Automation of relatively mundane and low-skilled jobs is taking place at an accelerated rate, using robotics (which employ robots to perform these tasks), especially in countries like the US and Japan. AI has also entered the fields of banking, finance, law and even medical science- areas hitherto thought to be beyond the grasp of machines. Autonomous Drones and Machine Vision are being used in warfare and counter-insurgency operations to penetrate deep into enemy lines without any human involvement.

The ability of AI to mimic human thinking and to be able to perform almost every task that humans are able to has raised serious concerns about the future of mankind. On the one hand, there is a fear that machines might become too intelligent and overpower humans. Entrepreneur Elon Musk, whose company Tesla is making AI powered driverless cars, thinks so. He, therefore, calls for greater regulation on the use of such technology. Recently, two bots, Alice and Bob, developed by Facebook, had to be stopped as they had invented a new language and started communicating amongst themselves. This has lent credibility to Musk’s prediction, although it was found that the bots had not invented but actually shortened the original language to communicate better. On the other hand, there are people like Mark Zuckerberg, the founder of Facebook, who feel that such a doomsday scenario is very unlikely. He is against any regulation on the use of AI as that will scuttle innovation. AI, according to him, is absolutely important to solve some of the pressing issues facing mankind, including climate change, terrorism, hunger and so on.

However, the thing that is worrying policymakers and economists alike is the massive loss of jobs that we are likely to witness in the near future, thanks to AI. A recent report by the World Bank predicts that nearly 69% jobs in India and 77% in China will be threatened by automation. The number is likely to be even higher in labour surplus poorer developing countries like Ethiopia. Even manufacturing jobs in the US and other developed countries are at the risk of being overtaken by robots. What is more worrying is that even those jobs which require fairly high skills are at the risk of losing out. IBM’s Watson is able to detect diseases and perform treatments that leave even the best doctors astonished. Investors are using AI enabled languages to hedge funds in the stock markets, rather than relying on stock managers. Google’s Magenta is producing paintings that only the likes of Picasso can match. Automakers are relying on AI to design cars. What makes machines better at these highly skilled and creative tasks is that these machines do not have any biases that ordinary humans suffer from. So, they are far better at being creative, and also at using big chunks of data simultaneously.

Given that almost every job currently available is under threat, how do we respond to this evolving scenario? Clearly, AI cannot be banned. We do need AI to tackle many of the current challenges, like climate change, terrorism, etc. AI also enables us to do dangerous jobs using machines, like clearing mine-fields, producing explosives or cleaning sewers, thereby minimising the loss of human lives and that too, in a cost effective manner. Given that banning AI is ruled out, we must, first of all, regulate the use of AI (as suggested by Stephen Hawking and Elon Musk) so that it is not used for the wrong purposes. Second, AI will also open up a plethora of new opportunities as new sectors open up and new products become available. For example, the demand for engineers specialising in AI will shoot up in the near future, and so will be the demand for managers running companies that specialise in AI. In order to realise its demographic dividend, India must invest considerably more in STEM (Science, Technology, Engineering and Mathematics) education. This will produce the highly skilled workers who can drive the modern sectors of the future. Third, upskilling of existing workers has become a necessity. Otherwise, their current skill sets are liable to be made redundant by the rapid intrusion of AI technology. Fourth, there are certain areas where no machine can enter. Take, for instance, leadership qualities. The ability of a leader to motivate his/her colleagues or juniors is a skill which no machine can match. Hence, there is a need for people to work towards improving their leadership abilities, emotional quotient, and the ability to relate to others’ problems and suggest solutions. Teaching and fundamental research are other areas where humans have a comparative advantage, and India needs to increase investments in those.

Thus, Artificial Intelligence is here to stay. What we need to ensure is that they become complementary to humans and help improve our standards of living, rather than being substitutes for human jobs.

 

 

(The author is currently in the Department of Economics, Delhi School of Economics.)

Here’s the link to The Assam Tribune article: http://www.assamtribune.com/scripts/at.asp?id=jan0318/Page6