The Slovenia Times

Some developments in the artificial intelligence fields are raising quite provocative questions



Q AI is high on the agenda in most sectors, how our privacy is affected by the development and application of AI?

A I think significantly. One of the challenges with these technologies is that we spend a lot of time reflecting on their capabilities, but we do not pay enough attention to the social and culture consequences that they have. For instance, the news that we get and the information we receive, the political debates we have; these are all shaped by the online platforme that we use and the algorithms behind that drive them. As my brother, Jamie Susskind, puts it in his book, Future Politics - software engineers are now becoming social engineers as well.

Q From your perspective medical issue, a legal issue, or tax issue could be replaced by the AI. Why?

A First, you need to stop thinking in terms of jobs - lawyers, doctors, teachers - and instead think in terms of all the different tasks that make up a job. When you do this, it becomes clear that, in any job, many of the tasks that are done are routine and can be automated even today. The second is that many of the non-routine tasks that remain in the job - the sort of activities that require creativity, judgement, empathy -- can now be done by machines as well, using lots of processing power, data storage capabilities and advances in algorithm design.

Q So the AI just assists to the jobs and cannot replace them entirely? 

A Last year McKinsey looked at hundreds of occupations and found that while only 5% can be fully automated, about 60% of those occupations were made up of tasks of which 30% can be automated. The fundamental question is will there be enough demand for those residual tasks that remain. In the medium term, through the 2020s, I think the answer is yes; beyond that, I am less sure.

Q Is the AI as smart as much algorithms we put into it - so just powerful as we empowered it or it could be something more of it?

A I distinguish between two waves of AI. In the first wave, AI systems were based on copying human beings and their reasoning process - it involved sitting down, getting a human being to explain how they performed a task, and then trying to capture that explanation in a set of instructions for a machine to follow. What is interesting by the second wave of AI is that these systems no longer try and copy human reasoning. Again, they are using lots of processing power to run algorithms through large bodies of data, hunting for rules to follow that may or may not resemble the ones that human beings happen to follow.

Q Which projects nowadays excites you the most in the fields of AI?

A There are two fields: one a field known as affective computing, where researchers are trying to design systems and machines that can both detect and respond to human emotions. They are now systems, it is said, that can distinguish between a smile of genuine joy and one of social conformity, or between a face showing real pain and fake pain, better than a human being. People tend to assume that tasks which require empathy or personal interaction from human beings must be out of reach of automation - but these developments raise interesting and provocative questions about that assumption. The second field that interests me is known as computation creativity, where researchers are trying to design systems and machines that can perform tasks that might require creativity from human beings; compose poems, write stories, generate works of art, and so on. And again, people tend to imagine that machines could never perform a task that requires creativity from a human being. But once again, the developments in this field challenge that.

Q Based on said, what are the tasks that in your view are the last that will be automated?

A Imagine a world in which machines can do everything better than human beings. In that  world, I imagine, there will still be tasks that we value not only because of the outcome they achieve, but because of how they are carried out as well - in particular, that they are carried out by human beings. For instance, someone who walks into the Sistine Chapel and looks at the painting on the ceiling is likely to say both "gosh, that painting is beautiful" but also "is it not amazing that a human being painted that!". Machines could almost certainly paint a replica of that ceiling, but we would not feel the same sense of awe and excitement that we feel when we know a very talented human being painted it. I think there will be some residual tasks like this, that we could automate, but we will choose not to, because we value the very fact that human beings will have done them.


Dr. Daniel Susskind, Fellow in Economics at Balliol College, University of Oxford,
Prof. Paul Claudel, IEDC-Bled School of Management,
Prof. Danica Purg, Dean & President of the IEDC-Bled School of Management.


The future is very uncertain.

Because of that leadership will become more important, helping us to navigate that uncertainty and find a path through it.

Q How does Artificial Intelligence and affect leadership?

A The future is very uncertain. Because of that leadership will become more important, helping us to navigate that uncertainty and find a path through it. I think there are two fundamental challenges for leaders. The first, again, is not to think about their future in terms of jobs but in terms of tasks, to not think about their overall role as a "leader" will disappear - that will not happen - but instead to ask, "what particular tasks do I do in my role as a leader that might change in the future". The second is to reflect on how the activities that those you lead will change as well; these will change dramatically too.

Q How does look like the geo landscape of the AI: China is beating the US in AI - Europe experts are challenging both? And there is also the startup industry of Israel that is targeting today's hottest tech sector...

A It is interesting that, in the past 24 months or so, almost every developed country has published an AI strategy of some sort. I think in many cases, though, they are "talking the talk", but not yet "walking the walk" in terms of providing serious support to the AI sector. At the moment, the AI sector does seem to be dominated a race between two countries - China and the US.


More from Nekategorizirano