INTERVIEWS by Stephen Ibaraki
Interview with Amir Banifatemi, Leon Strous, and Mike Hinchey - Globally Renowned Technologists Supporting the XPRIZE in Artificial Intelligence
This week, Stephen Ibaraki has an exclusive interview with Amir Banifatemi, Leon Strous and Mike Hinchey.
Amir Banifatemi joins XPRIZE with more than 25 years of experience in development and growth of emerging and transformative technologies. At XPRIZE he is the Prize Lead of the IBM Watson AI XPRIZE.
Prior to joining XPRIZE, Mr. Banifatemi began his career at the European Space Agency and then held executive positions at Airbus, AP-HP and the European Commission division for information society and media. He managed two venture capital funds and contributed to the formation of more than 10 startups with emphasis on Predictive Technologies, IoT, and Healthcare. Mr. Banifatemi is a guest lecturer and an adjunct MBA professor at UC Berkeley, Chapman University, Claremont McKenna College, UC Irvine and HEC Paris.
He holds a Master's degree in Electrical Engineering from the University of Technology of Compiègne, a Doctorate in System Design and Cognitive Sciences from the University Paris Descartes, as well as an MBA from HEC Paris.
Born in 1960 in the Netherlands Leon Strous is still residing there today. He is qualified as a registered EDP-Auditor (RE) in the Dutch Association of Registered EDP-Auditors (NOREA), and also as a Certified Information Systems Auditor (CISA) in the international Information Systems Audit and Control Association (ISACA).
He is a member of a number of professional societies and has been active in many different positions in the Dutch Computer Society (NGI) since 1988, including vice-chair in the Board for five years, and in the International Federation for Information Processing (IFIP) since 1994. In IFIP he has chaired Technical Committee 11 on security and protection in information processing from 2001-2007, he was Vice-President from 2007-2009 and in August 2009, he was elected as President for the 2010-2013 term with a re-election in 2012 for the 2013-2016 term.
Leon has co-authored and co-edited publications in the area of information security and chaired/organized several international security conferences.
His professional career started in metal and plastics processing and then went on to eight years with Philips Electronics in the Netherlands. His focus areas were administrative organisation, internal control and information security. Since 1993, Leon has been with De Nederlandsche Bank (DNB), which is the central or reserve bank of the Netherlands, as IT auditor in the internal audit department and as overseer in the oversight department of the cash and payments division, focusing on the security of payment systems. He also participated in committees of the European System of Central Banks (ESCB). Currently his main jobs are advancing business continuity and crisis management arrangements with the key players in the payments and securities clearing and settlement processes in the Netherlands, and liaising between the financial sector and the government concerning critical infrastructure protection programs.
Mike Hinchey is Director of Lero-the Irish Software Research Centre, a multi-location national research centre funded by Science Foundation Ireland and with a footprint in all of Ireland's universities. He is also Professor of Software Engineering at University of Limerick.
Hinchey holds a B.Sc. in Computer Science from University of Limerick, a M.Sc. in Computation (Mathematics) from University of Oxford, UK and PhD in Computer Science from University of Cambridge, UK. He is a Member of Academia Europaea, Fellow of the British Computer Society, Irish Computer Society, Institute of Mathematics and its Applications, Institute of Engineering Technology and Engineers Australia and Engineers Ireland.
Hinchey is the author/editor of more than 20 books (including two to appear in 2017), and over 200 papers on various aspects of Software Engineering and Computer Science. He holds 26 US Patents on various aspects of autonomous systems, code generation and computer hardware. In 2009 he was awarded NASA's Kerley Award as Innovator of the Year.
Prior to leading Lero, Hinchey was Director of the NASA Software Engineering Laboratory at NASA Goddard Space Flight Centre, Greenbelt, MD. Hinchey has been previously full professor or visiting professor in UK, Ireland, Sweden, Germany, USA, Japan and Australia.
He is President Elect of IFIP (International Federation for Information Processing), becoming President on 17 September 2016, Vice President of the Irish Computer Society and Vice Chair of IEEE UK and Ireland Section.
To listen to the interview, click on this MP3 file link
PARTIAL EXTRACTS AND QUOTES FROM THE EXTENSIVE DISCUSSION:
Interview Time Index (MM:SS) and Topic
|:00:17:|| ||Thank you for sharing your deep experiences with our audience. Please introduce yourselves.
:00:20: Amir Banifatemi "....My name is Amir Banifatemi. I'm an entrepreneur investor and currently leading the global AI-XPRIZE, which is the first global artificial intelligence competition trying to showcase the most advanced approaches in artificial intelligence and also showcasing AI collaboration...."
:00:46: Leon Strous "....I'm Leon Strous. I am President of the International Federation for Information Processing (IFIP). I have a background in information security and business continuity and I work for the Federal Reserve Bank in the Netherlands...."
:01:03: Mike Hinchey "....I'm Mike Hinchey. I am a Professor of Software Engineering at the University of Limerick in Ireland and I lead Lero, which is the Irish Software Research Centre based in all of Ireland's universities. I am also president-elect of IFIP so I will be succeeding Leon as president in the next few weeks...."
|:01:21:|| ||We have XPRIZE in Artificial Intelligence and a tremendous program leading to TED 2020 where the top teams will compete to determine who the final winner is. Amir, as the General Director of the XPRIZE program for artificial intelligence, can you overview the XPRIZE for artificial intelligence?
:01:49: Amir: "....Not everyone knows about XPRIZE, but XPRIZE's interest has always been in transformative technologies that can have a global impact. XPRIZE is today the global leader in incentivized prize competitions (and many other prizes such as the Lunar prize or ocean cleaning or going to space), and is launching the first AI competition. The belief of XPRIZE is that AI (in collaboration with humans), is capable of solving some of the world's grand challenges and we need to democratize the expression of technologies and embed it into the field of AI to really enable a world of problem-solvers....XPRIZE has taken the leadership role to organize the global AI challenge where we offer every competitor and applicant an opportunity to come up with their own challenges and present their own milestones and commissions for success...."
|:03:52:|| ||Leon, can you talk about IFIP and why IFIP is a partner of XPRIZE?
:04:01: Leon "....The International Federation for Information Processing (IFIP) is a federation of Societies of ICT professionals and we aim at achieving a worldwide professional and socially responsible development and application of information and communication technology. We are a body of Societies and we were established in 1960 under the auspices of UNESCO after the first World Computer Congress in Paris. As the Mission Statement indicates, we work towards promoting technology for the advancement of mankind in general to see how we can apply ICT technology and applications for the benefit of all kinds of areas....In the area of other partners, the AI-XPRIZE organization is a wonderful opportunity for IFIP to expand our links to make sure that we can assist initiatives like this which is to the benefit of development...."
|:06:00:|| ||Mike, can you describe what you are doing to support the program?
:06:05: Mike "....As a federation of Societies, we have a global imprint in that we have member Societies from each continent, and that means we are in a position to help XPRIZE become a truly global competition and bring entries from around the world. More importantly we have over five hundred thousand individuals who are researchers and practitioners, parties in all sorts of areas of ICT, but in particular we have a technical committee (technical committee 12), which is devoted to artificial intelligence and has some very prominent researchers as members, but we also have a number of other TCs that will support this. So we can bring together expertise from around the world, top quality people who work in a lot of areas that are not solely AI (but related), and can contribute to making this a really great venture...."
|:08:42:|| ||Amir, Leon and Mike, can you talk about what are some of the grand challenges that could be solved or tackled with artificial intelligence?
:09:05: Amir: "....There is probably no shortage of problem-solving that AI can be involved with, but we expect competitors to surprise us with a large variety of applications. The main applications visible today are in fields such as healthcare and education....And also going to places where we can help with democracy, advocacy and providing a better platform for citizens to understand their choices....And all the improved applications such as transportation, robots or personal assistants and automated curriculum for education....We don't expect any teams to address any grand challenges directly of course. No one will finalize hunger or poverty but on the way to oversee those grand challenges, meaningful applications will be those that are wide-ranging and touching maximum people with their needs and really making progress very practically...."
:10:53: Leon: "....Healthcare is becoming very costly and there's new health issues arising all the time. Fighting and preventing new diseases, making healthcare in general a bit more cost efficient, I see a big opportunity for AI solutions and methodologies to help in that area. Another one is in personal safety; you talk about traffic and everybody is keen on monitoring what's going to happen with driverless cars....There are the other XPRIZES of course, but AI will be a means, a methodology that can help in many areas...."
:12:57: Mike: "....If we are to think of many of the big challenges facing society, things like the environment, providing clean water, providing adequate healthcare, safety, etc. and some of the things that have been mentioned by Amir and Leon already, any of these are going to be advanced by artificial intelligence....We've come along with machine learning, intelligent machine vision and many areas like this, all of which produced great results, but now what we want to do is to step up the game and to get much better results and the grand challenge is a great way to do this. It's been very successful in areas such as medicine and drug discovery, to give people specific grand challenges to try and address. XPRIZE is doing this by allowing for grand challenges that will up the game of artificial intelligence and what it can actually achieve, perhaps by integrating multiple areas of artificial intelligence that at the moment are niche areas for researchers who need to come together to form a solution for something that's large enough to need that kind of investment in terms of people, intelligence and power...."
|:15:34:|| ||Let's examine the impact of AI. What do you think about AI and the bias in AI?
:16:15: Amir: "....We always try to capture intelligence as a universal skill and try to make machine intelligence like humans. This comparison between machines and humans is probably one way to look at how we can make machines more effective or intelligent. But how about we consider a different approach which is not comparing machines to humans and to try to make an absolute skill....machines better at certain things. We are probably not ready yet today to fully model the human thinking and creative power and to expect machines to act like that. We might get something surprising and phenomenal when we try to do that. So I think it's a timing issue, it's probably not the right time...."
:18:02: Leon: "....When we look at a machine reacting in a way that we don't like to the fullest, we are biased....That's where I support Amir and his view and it's also an issue of time, where we are probably not ready yet. I'm not saying that we should look at the machines like they are humans but it's an interesting, different feeling when a machine does something we don't like and when a person does something we don't like to a certain extent...."
:19:13: Mike: "....I have to agree with Amir and Leon that we have to think differently about intelligence. I worry about using the term intelligence sometimes, particularly when we are talking about machines or when we are talking about software, because we don't really know what intelligence is. We have ideas about how we think the human brain works and to a certain extent we are correct. Some studies have shown that some of the ideas we have about intelligence are wrong and we've made advances in cars because we realized we don't try and copy a horseless carriage, because it's not the same thing....So perhaps we have to move away from the idea of intelligence in the human sense and that intelligence will be something different. In terms of bias we've been looking at self-managing software and one of the things we are trying to do is find ways to express the range of decisions that the software might take, rather than programming in potential possible decisions it might make because that would have a bias...."
|:21:25:|| ||Let's extend this conversation to ethics. What are your views on AI and ethics?
:21:32: Amir: "....Following up on what you just said about gender bias on some machine learning programs and the question about ethics, it is interesting to observe who are the people programming those machines, who are the people designing software for those machines? Do you have gender representation? Do you have diversity representation? What is the cultural and social background? What is the cognitive process of the individuals programming the machines? So that could be an interesting observation to think about. On the ethics side, my opinion is that ethics will play a huge role in how the associate work machines can reason and give back and propose...."
:22:55: Leon: "....A well-known case or situation in the area of driverless cars is the car is in control and you are heading towards a collision with a pedestrian (and there's a decision to make), and there's no way to either avoid a crash with the pedestrian or to steer away from it and hurt the passenger in your car or the driver itself. Now the ethical issue is, who is going to program that? Because it's not the machine that is making the decision it's a human that tells the machine, okay in this case you have to do that. The machine can start learning from experiences but nevertheless it's still a human decision and that's what we have to probably include in all the projects and all the challenges and I think that's a big challenge for AI....."
:24:57: Mike: "....I think Leon's example is a very good one because in a very simple scenario you can get into a huge minefield of ethics. And as I mentioned before we do have a technical committee in IFIP which deals with ICT and societies so ethics would be a major part of this (and the role of ICT in the world), and how it affects the world and I would hope to see that many of the entries for this XPRIZE would bring in those issues and consider those issues as part of the challenge...."
|:25:16:|| ||What are your views on when AI reaches the point it can do some art, it can do literature or music? What does that mean in terms of literature and art and music and any of the creative elements that are there if AI is able to do it?
:26:47: Amir: ".....Whether we go to phases where machines can produce extremely intelligently characterized or creative content, I think should not be the debate. It's just the matter of time. The real debate in my opinion is how would we leverage this? How can we enable humans to be more creative and more thinking and more architect and less manufacturers of content....I think the real opportunity is for us to think about other humans as in control and more creators and architects again...."
:28:31: Leon: "....If something is created by a computer when do you say it's a genius? I mean because humans will decide whether it's attractive or not to buy it, or whether it is worthwhile producing it....If you look at AI in the creative environment, in music, literature and other areas, if it's contributing to the pleasure of people, to the well-being in that respect by enjoying literature, by enjoying paintings, why not? The difficulty will be how much market will there be left for the humans who have to be creative?...."
:29:39: Mike: "....I'm not sure we can say what genius is, but we certainly know that the great writers, great painters and great composers all had some sort of extra something to their thinking and they've been able to produce something that is loved sometimes over centuries....But AI gives us the option to perhaps improve these recommendations. Also if we like a particular piece of music, we can envision situations where AI based machines would create music similar to that we might like, so that we are not just getting new recommendations of music that already exist but we are getting music that's written to suit us or books or stories being written to suit the style that we like that wouldn't have existed otherwise. In many ways it's going to open up the availability of many more things. But then as Leon points out, how is that going to stifle creativity amongst humans and what will the implications be for humans? I guess that brings that back to ethics and society...."
|:32:26:|| ||There is increasing automation through AI, which is going to appear in the different tools and vehicles and devices we have and it's going to be applied to warfare. What are your views on that?
:32:40: Leon: "....It is more or less the case for any technology which is being developed....I think the challenge for us as ICT professionals is to make sure we advocate and do our best to promote responsible behavior of the professionals. First, when we are developing technology and applications and second, it may be assisting in finding ways to reduce risks and to limit application areas if that's at all possible. But I don't see a way to completely prevent it and that's where we have definitely an important role. (By the way, it's one of the missions of IFIP as well to promote professionalism of people involved in ICT in many ways)...."
:34:51: Mike: "....Late last year IFIP issued a statement regarding the use of such technology in warfare and in aggression....If there is a great opportunity for misuse of the technology we are talking about there is also great opportunity to enable peace, and we want to make sure we have defense mechanisms that are based on this technology and that we have properly trained and aware professionals who understand the implications of what they do and how it can be misused...."
:35:39: Amir: "....At the end of the day we are humans and we have to make sure we go back to the ethics and design principles that we integrate into learning and programming and AI in general the basic principles that have impacted them and also that takes care of humans...."
|:36:52:|| ||If you have a big breakthrough in five years what does it mean to us as humans and government and business and all the different parts of who we are if there is this 'thing' that can do everything and really will surpass us in capabilities?
:40:00: Leon: "....That's one of the main questions and it actually relates to a lot of things that we've said today in terms of ethics and professionalism and in terms of human involvement. The thing that we might ask ourselves is should we develop or create everything that is possible, or should we be more responsible as we are looking into applications and technology that will benefit us and will not potentially harm us....I think that's a big question and I know people believe that things can't be stopped but I believe that if we have a good sensible discussion on it we can at least limit some of the effects that are not desirable...."
:41:28: Mike: "....My big concern about something that's learning by crawling through the web is that a significant proportion of what it finds is going to be wrong; will it know the difference and how will it make a decision between what's right and what's wrong when it finds contradictory information - will it just use the preponderance of evidence or will it have some way of inferring that it is the wrong information...."
:42:03: Amir: "....I'd like to bring a complementary perspective to this and the perspective would be in line with evolution and general lack of evolution. What if machines would be more efficient than us and would have more performance than us, would that be a bad world or should we be able to harness and to find ways to live in a world where you have machines with general intelligence...."
|:44:12:|| ||There is this movie called Her. You have this sentient who is guiding this gentleman and it took on a life of its own. What are the implications of artificial intelligence towards this concept of death or the concept of love? What is the impact of artificial intelligence of these ideas that we think of as purely human?
:44:38: Leon: "....If you look at human emotions the question is can you translate them into a machine? I'm pretty convinced that technically it's possible by machine learning and by programming. But then you have to look at it from the other side. The machine can be taught, but it's also a human issue and that's back to human nature and human psychology....People love and often see objects in a way they react emotionally so I'm pretty convinced people can love a machine or hate it. But in terms of human emotions it's only a matter of time before the distinction will become a lot vaguer than it is now. At the moment it’s pretty clear for many cases but in time when things improve, for example like you mentioned about robots falling down and people getting concerned about whether they will be hurt. Who would have thought that 20 years ago...."
:48:57: Amir "....I was just reflecting on the fact that humans can help or form friendships in robots or machines or programs, but it also tells us about our humanity or about our ability to project emotions and develop affection....I think many people will take this opportunity as a chance to work in this field and develop relationships between man and machine; not try to create a contrast between them but complementary and focus more on humanity...."
:50:00: Mike: "....Whether this is home video or the internet I like to think we use artificial intelligence for many more important things than that but certainly the human factor comes into it. Many people name their cars. Many people think of their cars as 'she' and maybe it's natural to anthropomorphize things...."
|:50:30:|| ||We had this financial crisis a few years ago with the banks, etc. and we have troubled economies and the impact it has on the global economic health. Concurrently to all of this you have this rise of AI and machine learning being used in hedge funds to make economic predictions and they are getting better and better. What is that going to do to the economic system worldwide if you have AI better than humans in terms of making predictions, forecasting events and even being able to give you a higher rate of return? For example in the stock market. Does that remove the foundation of the economic system in the world today?
:51:30: Amir: "....It will definitely happen because any system that humans have designed can be improved with machines, so there is no reason to think that financial systems, projections or any form of mathematical simulation cannot be done by machines. The opportunities again, how can we harness this and put the humans in the creative role and AI in the controlling role of using that assistant....."
:52:23: Leon: "....What I'm concerned about is that it is an issue of trust and letting go. I agree with you that machines can probably do a lot better than humans in predicting in faster calculations and so on, but it means that humans who are in charge of decisions in companies or even for your personal enterprises will have to let go of the control...If we reach that point I'm sure we have a new economic system which is run by machines who hopefully do the right things, but people will no longer be able to understand what's happening...."
:53:27: Mike: "....We like to think that the machine is an enabler and an assistant, rather than actually taking over. There is a danger that it could, but again it's got to be programmed by somebody so it's going to have some sort of bias perhaps which maybe will want to come through...."
:55:32: Leon: "....I think we have be careful that our audience is not getting the wrong impression because we are raising concerns and we are talking about difficulties and big questions which are there of course, but let's not forget that AI can also bring us major benefits and that's actually where we started. How can AI help us solve the major world problems in terms of food, safety, environment, health, etc., and I think that is definitely a combined effort in saying how can we use it to benefit and at the same time what are some of the questions we need to look at? But let's be careful that people don't think we are against it or we see too many problems...."
|:57:00:|| ||In closing, I'm asking each of you to share some wisdom (perhaps compacted into something that could be tweeted out), about the future of AI.
:57:15: Amir: "....The opportunity of this prize is really to encourage interdisciplinary thinking and collaboration across the globe to a path to define intelligence and lead us into our future...."
:59:08: Leon: "....In my view AI will become a great aid to solve difficult problems in many areas. That's definitely where initiatives like AI XPRIZE to challenge project teams, researchers, people working on this to come up with wonderful, bright ideas. We fully support that and the interdisciplinary aspect of it needs to be included in a way to address the issues that we have raised during our talk...."
:59:52: Mike: "....AI has produced some amazing results that are being used in a number of niche areas to great success. The AI XPRIZE gives us the opportunity to raise the game and to have a greater influence on AI in a whole range of areas where it may have only touched in the past, to produce great results and benefits for all of society...."
|:01:00:24:|| ||We thank each of you for sharing your insights, deep thoughts and wisdom with our audience. I know your schedules are really tight and we appreciate you taking the time to dialog and share about the opportunities with regards to artificial intelligence.