Posted on 14/Sep/2016
FacebookTwitterGoogle+WhatsAppEvernotePocketKindle ItBufferLinkedIn

The Digital Post spoke with Paul Chong, Director of Watson group at IBM, on the future of the popular supercomputer combining AI and sophisticated analytical software.

The Digital Post: What is the story behind IBM Watson?

Paul Chong: It all came about by chance in 2006, when a couple of guys were sitting in a pub watching a quiz show and they came up with the idea of taking on the champions of this show. Then, they started to look at the challenges.

At that point, we had a lot of experience and background in analyzing natural language programming and machine learning, so this opportunity came at the right time for us as well.

Eventually, we took it to the quiz show in 2011 and we were successful. By then we have had one million dollars and had already started to think about where we could apply the technology into different industries in a very transformational way, taking what is a lot of unstructured data and giving it some sense.


TDP: Apart from the health care sector, what are the other industries where IBM Watson is working on at the moment?

PC: We are present in 17 different industries right now, including financial services, where we’ve been typically involved with retail and utility companies.

Now we are trying to create a platform services for Watson AI machine, allowing a larger audience made up of users, developers, and startup entrepreneurs to use it for business purposes. The idea is to decompose the technology, provide it as a set of APIs, very easy to use, so that everyone can use it. What we’re trying to do is creating a really intuitive platform.


TDP: Where do you see Watson in five or even 10 years’ time?

PC: I think we’re going to see an evolution of those services, particularly in terms of numbers and quality of the services provided. For example, one of the big challenges with AI is the amount of time of supervised learning that you need to do. Supervised learning means that there is an intervention by humans to train the data and the models.

What you’ll start to see is a great improvement, namely less data will have to feed the system to teach it, and there will be less intervention from human beings. I also think services will become more intuitive to use so that businesses can take advantage of them by understanding the type of outcomes they want.


TDP: AI technologies, and computing technologies in general, are raising concerns about the impact they may have on employment.

PC: You have to look back in history in order to know what’s going to come in the future. Take the industrial revolutions for example. On various levels, we’ve always seen a situation where, as human beings, we’re very adaptable, and we start to find out what our new roles are going to be, how we are going to exist, and what work will look like in the future. We have always witnessed that humans eventually find another level to operate on.

We are now going through a cognitive age, which will require a great deal of adaptability. Governments will soon start thinking about it for the next generation. What type of training and education we are going to imagine for them? We should avoid training them to be working on machinery or doing certain mandate roles, such as accountancy, which, among others, will be automated within the next twenty years.

We have to start questioning ourselves and taking steps now. Regulators, governments and educators have to start thinking about what types of jobs are going to exist in the future, while companies and countries should focus on their competitive technology.

(Visited 41 times, 1 visits today)