AI, Machine Learning and challenges ahead of us.

For me, the coffee break is the most interesting and thought-provoking part of my workday. I work with a group of actuarial analyst, mathematicians, data analyst and computer programmers. We sit around and discuss about some of the latest industry trends, such as: AI, Cognitive Intelligence, Machine Learning and how it is going to shape our work as well as life in distant future.

Some of my colleagues think we're on the verge of something that could cause a profound change in the way we do business or software development, and that is machine intelligence. 

The Paradigm Shift

Artificial intelligence used to be about putting commands in a box. You would have human programmers that would painstakingly write the code considering various cases. You build up these expert systems, and they were kind of useful for some purposes. The issue was lack of scalability of the system. Basically, you got out only what you put in. But since then, a paradigm shift has taken place in the field of artificial intelligence. And that is Machine Learning.

“Machine learning is a type of AI that provides computers with the ability to learn without being explicitly programmed. Machine learning focuses on the development of computer programs that can teach themselves to grow and change when exposed to new data”. The result is A.I. that is not limited to one domain; the same system can learn to translate between any pairs of languages. The vision pioneered by Marvin Minsky (Perceptrons) with advanced Neural Network ,which is Deep learning indeed, seems more obvious today.

Machine Learning & Deep Learning in practice

The best example will be the one which we use every day: Google. Google showed that it is possible to find information by using a computer algorithm, which is based on machine learning. Since that time, there have been many commercial successes of machine learning. Companies like Amazon and Netflix use machine learning to do product recommendation that you might like to buy, movies that you might like to watch. Have you ever wondered how companies like LinkedIn and Facebook sometimes will tell you about who your friends might be and you have no idea how it did it, and this is because it's using the power of machine learning. These are Deep Learning algorithms that have learned how to do this from data rather than being programmed by hand.

The future ahead: Machine is going to take over Mankind?

It is indeed true that next decade will see a massive job reduction in all sort of routine & Algorithmic work. With the advancement in Big Data, Deep Learning & Robotics, more and more of our jobs will be replaced by robots and algorithms. Machine will be able to do those jobs much better and faster. We have seen this in last Industrial Revolution when human labor was replaced with steam powered machines and assembly line manufacturing. Thus, there will be social disruption, and the better computer get at intellectual activities, the world will experience changes like never before.

Armed with Deep learning and AI the machines will be able to read, listen and even THINK, as a part of their self-learning process. Now, a super intelligence with such technological maturity would be extremely powerful, and at least in some scenarios, “it would be able to get what it wants”. We would then have a future that would be shaped by the preferences of this A.I. Question is - what are those preferences?   Do we a Neo and Trinity then, to break The Matrix?

Scary – Isn’t it?

The issue of job loss needs to be addressed by re-skilling or up-skilling the employees across industries, because there will be lot of opportunities in data and machine learning fields, in coming days.

I do not see AI or Machines as our contenders, rather I believe in human-computer symbiosis – the idea pioneered by J.C.R.Licklider. His vision was to enable man and machine to cooperate in making decisions, controlling complex situations without the inflexible dependence on predetermined programs. As a human being we are good at creative and non-linear thinking but terrible at scale, computation and volume. Licklider never felt that men would be replaced by computer-based beings. As he wrote in that article: "Men will set the goals, formulate the hypotheses, determine the criteria, and perform the evaluations. Computing machines will do the routinizable work that must be done to prepare the way for insights and decisions in technical and scientific thinking".This approach, focusing on effective use of information technology in augmenting human intelligence, is sometimes called Intelligence amplification (IA).

Of course, perceivably things won't go wrong in these particular ways. But the important point here is: if we create a really powerful optimization process for objective x, we better make sure that our definition of x incorporates everything that we care about. So that even when the super intelligent AI escapes our control, it is still safe because, fundamentally it shares human values.

Courtesy: Wikipedia, Youtube