IGF 2021 – Day 0 – Event #2 IFIP 60th Anniversary: Future of Information Processing

The following are the outputs of the captioning taken during an IGF virtual intervention. Although it is largely accurate, in some cases it may be incomplete or inaccurate due to inaudible passages or transcription errors. It is posted as an aid, but should not be treated as an authoritative record.



>> DON PASSEY:  Hello, everyone.  Hope you can hear me okay.  Very good.  Well, we made it.


So far we made it.  Oh, I guess I shouldn't say anything as it's being recorded.

>> Against all technical issues we encountered.  Thank you to all the support team that made it possible for us.  To hold this session right now, um, we are at the final panel of the IF16 jubilant series.  IFIP is federation for information processing and they are this year contacting the 60 jubilee here.  My name is Elisabeth Schauermann.  I am together with my team at the informatic society that have been supporting the IFIP 60 jubilee activities and this is our great final panel on the topic of the sunset declaration which you will hear more about in just a minute.  But the declaration is on the topic of "Sustainable Education in a Digital Age of Rapidly Emerging Technologies", so something that really is at the forefront right now when we talk about how people interact with technology.

This panel is going to be moderated by professor Don Passey and supported by Johannes Magenheim.  And he's an honorary processor in India.  He's also the current chair of IFIPs technical committee on education.  Over to you, Don, and wish us all an insightful session.  Thank you.

>> DON PASSEY:  Thanks very much, Elisabeth for that kind introduction.  Hello and welcome, everyone, to this session.  We're going to talk about an initiative that was so the up a couple of years ago which was to look at the future of education with regard to developing technologies and sustainability.  I'm Don Passey and Elisabeth kindly introduced me.  I'm going to be moderating this session together with Johannes Magenheim.  I think I can only see three panelists at the moment, but hopefully 4 will appear in due course:  Mary Webb, who's the chair of the IFIP TC3 task force in the curriculum.  And Cathy Lewin who is working.  And also Bernard Cornu who is a past year of IFIP TC3.  So Johannes has good hands with regards to what we're going to cover in this panel.  I should also say that behind the scenes in terms of the work that's gone into what's going to be presented to you today, which is basically to do with the outcomes of the stage that we reached with the declaration, I would also like to mention three people who have had a significant role within this whole initiative.  Javier Osorio and Raymond Morel that was in a conference in Zanzibar in 2019.  And Christoph Rafe who provided organizational support and advice throughout the Zanzibar Declaration events.  I hope that I've introduced you to everyone at this point in time and what I'm going to do now is I'm going to hand over to Johannes who will give you a brief introduction and the stage that we have reached at this point.  So if we could move the slides on to the relevant part, please.  And one more.  Okay.  Over to you, Johannes.

>> JOHANNES MAGENHEIM:  Thanks, Don.  My name is Johannes Magenheim.  I was co‑moderator by the Zanzibar declaration process.  I tried to give you an overview of what's been running so far.  It was the IFIP TC3 conference on sustainable education of digital age of rapidly emerging technologies in Zanzibar in April 2019.  The conference topic is closely related to the United Nations STD 4.  Both in the conference itself and the subsequent, it is remitting and discussing intensively questions of sustainable development and the influence of current and future Information and Communication Technologies on various areas of society and demands and education.  We concluded that it is necessary to initiate a process in a series of webinars which these contexts further discuss and assist concerning the future education and policy and scientific strategies of the IFIP TC3.  The Zanzibar Declaration process focuses on the relationship between technological develop and a field of ICT.  It is influenced on different areas of society and are resorting to demands and education.  Particular how they have learning processes and extends to which they are basically technological principles can be made understandable for students.  For example, in science and education.  In a subsequent series of webinars, we have invited experts from IFIP together with practitioners, decision makers and researchers from the education sector and we asked them to discuss specific questions on this topic areas.

The following objectives and principles guided and content organization of the webinars you will see on the slide explore future and challenges that arise from rapidly emerging technologies considered a range of topics covering important digital impact and challenges.  Covering many countries situations and local context and experiences and perspectives as possible and we say that the declaration will be a process that only will result in one documented that will be published.  The Zanzibar Declaration process should allow them to identify hotspots in the area of research, computer development and future conference topics.  So the webinar's panelists represent diversity.  We have in the webinars experts from different IFIP ITCs, different countries, continents, educational practitioners, and decision makers and gender perspectives.  Please next slide.

As a BASIS for finding topics, we use the matrix seen here on the one side on the left side, the various IFIP technologies and on the other, different social areas in which these technologies are applied with corresponding social effects and social impacts.  We then try to identify automatic clusters of IT and social impact areas that are, of course, not entirely free of overlap, but of great interest toward education.  This approach is illustrated here user webinar 1 as an example.  Please next slide.

You can seat rectangles which are highlighted and you will see that, for instance, webinar one focuses on big data analysis and machine learning related to different technological areas like big data, of course, but also machine learning and technology and recognition and so on.  On the other side, you will see different areas of social impact like cybersecurity, privacy, social surveillance and (?) and quality of information.  Of course, we did it in the same day for the other webinars.  I don't think we have time to go into details here, but you will see the other four have focuses and intersection between technology and social impact.  But always are focused on education issues.  So webinar 4 was on the impact of communication and economic and local (?) of society.  Webinar 3 was technical issues of (?) assistance and webinar 4 discussed power of AI methods and algorithms.  Always we well education and educational perspectives in mind.

Okay.  Topics of other webinars we have developed in same day according to same concept and I think I should now turn over to the floor to Don and me collaborators of the panel discussion.  Thank you.

>> DON PASSEY:  Okay.  Thank you very much, Johannes for that introduction to the process we have gone through.  So I now get to the fun part really where I get the chance to ask questions of the panel because what we found from the webinars were that there were certain questions which were arising really quite significant questions for the future.  And as the panelists fortunately have had the opportunity to be involved deeply in those discussions, what I'm going to do is to draw out questions and to pose these to the panelists and I'm going to pose them to individual panelists and then ask other panelists if they would like to respond.  And as Mary, one of our panelists has not yet arrived to the sections, what I'm going to do is go straight to Cathy, if I may and, Cathy, I'm going to ask you a key question if I can.  What we've been looking at is a range of emerging applications within these webinars such as data science, data analysis, data mining, Artificial Intelligence, autonomous systems, machine learning.  I could go on and on with the number of important emerging technologies that are arising.  I suppose a key question is, you know, of all of these emerging technologies, what potential is there for education and particularly for teachers?  We hear a lot about these technologies, but what is the potential for teachers?  How would you respond to that from your involvement in these events?

>> CATHY LEWIN:  Thank you, Don.  I'm going to talk and respond to this question from three different perspectives.  From the perspective of the opportunity, from the perspective of the teacher and from the perspective of the institution, the educational institution.  I think that ‑‑ I mean, it's a very broad range of that terminology that you mentioned there, Don.  This huge potential for the future and some of it is already in place in the classroom.  So from the students perspective, a big potential for personalization of learning through adaptivity, enabling the system to adapt to a student's knowledge, to pace their own learning, to access learning at any time.  And these systems have been around for a long time.  We have had (?) systems for decades, but they are now much more sophisticated.  So they can also easily identify strengths and weaknesses and really target the areas of learning that a student, an individual student needs.  And these systems can also curate the learning materials and provide a unique pathway through the learning that an individual student needs.  That's a big advantage of these kinds of systems.  In addition, there are systems that offer tutoring.  So I would be using natural language or, ah, they can also tailor this sort of activities through sensory recognition.  So facial recognition, for example.  A very common example of that is chatbox.  Enabling a student interact as if they were interacting with a human.  I have a colleague that's been working on a system that uses video data to try and detect levels of comprehension in students and then to pick up on whether a student hasn't really understood the learning they're engaging with and then to, you know, to a pathway accordingly or to remind them to provide them additional practice and so on.  Other ways they can help the students is in the form of guidance.  It might be through a chatbox, enabling a student to access information about time tabling, exams and so on or inquiries when they want to register for an institution if you're looking at higher education.  So access, easy access to information and responding to student questions about that.

And finally, the big area that's very beneficial is the provision of feedback and assessment.  So being able to provide instant responses to students and response to the activities that they're undertaking.  And that can offer them insight into their progress overall.  It can then be used to adapt the system to their areas of weakness where they need further practice.  And there are other specific uses such as correcting pronunciation when you're learning a foreign language, for exam.

So that's the sort of main areas in relation to a student.  And then if I move on to a teacher, so one of the benefits for teachers is the automation of tasks.  So that relates to the assessment and feedback, grading of pieces of work and systems have become more sophisticated and able to even make assessments of, you know, open written answers.  So essay marking and so on.  So that's one area.  And obviously that leads to a big benefit in terms of reduction in workload and time saving.

But it also provides the teacher with insights about students process.  So with learning analytics and visualization and dash boards they can see which students are struggling with certain concepts and so on.  So the visualization and the learning analytics really plays a key part there.

Another aspect in which these kinds of systems are being used it relates to supervision.  So remote assessment which obviously with the receipt pandemic has become more important.  So enabling people to undertake or students to undertake assessments remotely, but be monitored, there are ethical issues in relation to this.  You can look at facial recognition, the systems can also detect keyboard strokes and so on.  Obviously there's been with plagiarism and cheating, but they're developing all the time and behavior management is another issue which is more pertinent for school institutions.  So there are various systems which enable teachers to record behavioral incidents and then they get an overall picture again through learning analytics and dash boards to see which students are causing more problems than others.  There are tools offer seating plans, for example, for schools and enabling, you know, that amount of data to be gathered very easily and the patents to be identified very easily and then for teachers to be able to make those decisions about, you know, where to Seattle the children or which children need more attention and so on.

And then just more general kind of support activities.  So the systems because they are gathering so much data about students can be used to support predictions, to analyze that data overall and to as I said before, offer insight for teachers to make decisions.  Also these systems can generate content or identify content that's out there.  So another time saving aspect putting that information together from across the internet into the learning system that's being used.

So finally in a university or school management perspective, again it's about identifying lots of patents in the status, gathering data from the students at the institution, looking at their performance, looking at their attendance, looking at their wellbeing and all of these aspects can be support.  So for example, any universities these systems are used to help predict and prevent student dropout.  Again as I referred to you before, they can be used to support admissions, time tabling, attendance, monitoring of homework and so on.  So that's ‑‑ I probably haven't covered everything, but they were the things that sprung to mind when I was thinking about the potential benefits of these systems in educational institutions.

>> DON PASSEY:  Thanks, Cathy.  I think that's a really useful set of potential that are coming out already from these range of applications.  This in a way produces a very positive view, I think.  But it also ‑‑ I think it also presents a very sort of developmental view that in essence we need to be aware of how things are continuing to emerge and how we're going to tackle that in the future.  And I'm sure that we're going to come back to that question within this discussion in a range of ways.  But thank you for that start set of ideas which I think was wonderful.  That's a very wide ranging set.  I don't know.  Margaret or Bernard, would you like to add anything further?  Add any comments or add any other possibilities for potential?  I know Cathy has given us a wonderful list, there but whether you have any more that you can offer in response to that question?  Margaret, please.

>> MARGARET LEAHY:  Thanks, Cathy.  It was really nice to have it put telling like that.  I have this system for this.  I have this system for that.  But in listening through the webinars and listening to what you said, the one thing that keeps ‑‑ and thinking about it, the one thing that keeps coming into my head is how these systems will complement the teachers expertise.  They're not to replace the teacher, but they actually will enable us to do things in sharper and better ways.  I was thinking of a special needs teacher.  I was a special needs teacher in a former life.  As such a teacher, it takes a very long time to develop your expert ease.  Dyslexia was the area I was working in.  While it's a common term to work to both to actually learn what it is, how to deal with very specific and complex needs can take a very long time.  So if back then, if I had the sophistication of the systems in place now, you would have help with diagnosis, you would have guidance as to where to from here with the particular child to enable to you track the progression, identify gaps and to develop in a spiral approach possibly in a more efficient and quicker way because as a young teacher, you were always struggling to find out where to what to do and develop such a bank of expertise would be very useful.  That's my (?).

>> DON PASSEY:  Thanks, Margaret.  It highlights this great need for us to evolve ways of sharing what we're doing and how we can do that because I'm sure that you know the examples that both you and Cathy have identified would have potential for many, many teachers and institutions and therefore, it is a question of how ‑‑ one of the questions is how we handle that in the future, I think, given the diversity of possibilities that continue to emerge.  But again, it's ‑‑ it's ‑‑ it's encouraging to hear of that sort of range of potential.  So thanks for that.  I don't know whether you would like to add anything to that, Bernard, at this point.

>> BERNARD CORNU:  Not anything important.  But my feeling was that behind all these terms, behind all this concepts and data mining and so on, each of them has a strong consensual content and these are difficult actually.  Difficult concepts.  So how can we ask teachers and learners to immediately incorporate them, include them in their daily learning, daily teaching.  It's a big question.  So the problem is that I don't have the answer to these questions, of course.  I think one of the main ‑‑ the increasing difficulty of technology in education is that not only technology is getting more and more complex, but the concepts of informatics, the concepts behind the technology also more and more complex and we need that people understand or at least part of them.  This is my question.

>> DON PASSEY:  Okay.  Thanks very much, Bernard.  I think that's a very useful question to bring up and it's certainly very key to this whole concept of sustainability, I think, for the future and how we manage that in terms of sustaining not open where we are, but where we're going to go.  I'm sure that we will be coming back to those sorts of questions again within this discussion.  But thank you for bringing that up.  I think that's very, very important.

So now, Margaret, I'm going to come to you and ask you a question, if I may.  It has to do with the grabbing interest that there is in Artificial Intelligence and it's potential applications.  I think that we hear a lot about Artificial Intelligence at the moment.  Some people would regard that as hype and other people would regard it as real opportunity and real potential and the way of the future.  So, you know, from your perspective, from what you have gathered in terms of insights from these webinars, to what extent do you think we might be developing something here or might be thinking about developing something, which is creating a reliance more on sort of outcomes and predictions from the digital systems rather are we reducing a concern for critical thinking in these respects and I'm thinking here not just of teachers, I'm thinking of students, I'm thinking of managers, I'm thinking of administrators.  What do you think in regards to that?  Are we moving to our idea of critical thinking away from the human to the artificial system?

>> MARGARET LEAHY:  Okay.  I would say not, Don.  In thinking about TI think I raised the eschewed answer in a slightly different way, but it will answer your question.  Before I start, a colleague and I are developing a basework in Ireland.  One of the emerging trends and something that we need to tackle in our next strategy, we have put artificial intelligence and big data.  It is something we are thinking in the short while.  But to answer the question, there is a danger of creating a reliance and outcomes and predictions.  I think it is becoming increasingly important to develop an increased awareness and understanding of AI.  That was something that came through in all of the webinars and I think particularly educators as gate keepers of formal education can exercise a critical judgement in deciding how best to make use of predictive learning analytics and education.  And really that really is utmost important when it comes to school context.  A majority of the people there are teachers or management may not have control over the data that's actually generated.  So what is concerning is that on occasion, schools and teachers or parents and students themselves can be provided data without properly understanding the ramifications of giving over such data.  So to prevent the creation of a reliance or a lack of criticality, I think it's imperative that the education sector engages in the debate as Bernard said as to how data and AI and can and is to be used.  In doing so, we need to think ‑‑ excuse me ‑‑ about the need to address concerns related to data use and protection, privacy and ethics.  In other words, who collects, controls, selects, interprets and uses the data?  So I started to think at the various levels within the system.  So yes the providers and policymakers they need to develop an overall vision and strategies and how to use technologies with regard to data and we know that organizations such as those of the OCD are beginning to look at this currently.  But most importantly, there is a need to develop teachers understanding of the potential and challenges around the use of data and AI for teaching and learning assessment.  Teachers in particular need to understand the big ideas in AI, that is.  So they can let their students understand the key ideas underpinning AI so that their students in turn can make decisions and raise issues in relation to data and the use of AI.  So the key question that should be to the front and foremost of all educators minds needs to be do we really understand what is being measured and more importantly why.  Okay?  And then any application of AI needs careful consideration as to how, where and when human interpretation is needed.  And one of the contributors in the webinars, I used a very, very nice sentence to describe that.  He talked about the importance of explainability for the recipients and social accountability for machine learning and AI systems.  There are implications of how the AI technology industry develops their solution.  How these companies develop AI applications.  Now, we can say that's not a new question.  That question has always been there that the development of the new software has always included underlying ethical issues.  (?) phrased it very nicely.  He says AI technology needs to be informed by pedagogue.  It is insuring that teachers and students are empowered rather than marginalized by technology.  So for me, what that implies is that educators should be part of the teams that develop and test AI tools to insure they are appropriately used in education context and that they're underpinned by appropriate learning principles and designs.  So it's necessary to understand what the data is saying, interpretations of the context is vital.  I think that can be done with a variety of competencies across a range of professionals.  So I hope in some way I have answered your question through that.

>> DON PASSEY:  Thanks very much, Margaret.  Yes, indeed.  You know, I've always been concerned about these concepts of predictions because, you know, predictions can become self‑fulfilling processees in one sense.  We have to be very, very careful with regard to that.  And if what we see is coming out of Artificial Intelligence systems is a prediction rather than something which is actually asking us a question or asking us to think about this as a possibility, you know, I think that's a very in a way concerning way to be going.  I don't know.  Cathy or Bernard, would you like to come in fairly briefly on this?  Cathy, please.

>> CATHY LEWIN:  I want to repeat something that Margaret said earlier which is how important it is to recognize the systems as being something that supports the teacher.  So you don't want to rely on the predictions.  There needs to be some sense checking that goes on.  And these systems are only as good as the data that they have collected.  So the data has to be unbiased.  The data has to be complete.  You know, there are as Margaret said, there is a need for people to take responsibility for that and to insure those checks are in place with all of these kinds of systems.  But going back to my first sentence about the teacher, you know, using these systems to support what's going on, I don't think any of us are suggesting that these systems should replace teachers at all and there are lots of elements of teaching and learning that these systems are not yet able to provide, you know, more social interaction is the key one for example.  That is such a key part of learning when learning students collaborate and interact with each other and with their type.  So yes.  These systems have potential but we do need to be aware of the limitations of them.

>> DON PASSEY:  Thanks, Cathy.  I think those are very, very useful points.  It reminds me of this concept of we do have to be careful not only about the data that go in and the data that go out, but we also have to be careful about how we think about the data that come out and whether in fact it is informing us or not.  And often, we can't know that unless we understand what is at the background of this.  So rubbish in and rubbish out.  We have to be aware of, I think, in this instance.

I'm aware that we're moving on.  What I want to make sure is that participants in this event have an opportunity to also ask questions.  So I'm now going to go straight to you, Bernard.  I'm going to come and ask a question of you directly if I may.  This is very much to do with this idea of the whole concept of the Zanzibar Declaration of sustainable education of rapidly emerging technologies.  This is addressing very many different topics and on concepts.  There are concerns here with regard to technology, software resources, knowledge areas, learning, teaching, pedagogue, teacher training, teacher competencies, et cetera.  And there are clearly necessary changes in developments that have to consider all of those sorts of areas and in a sense, all at the same time or how they can be integrated or worked across time.  So I suppose the question I'd like to ask you is, you know, from your insights and the background that you have, is it possible to deal with so many of these changes that need to be taken into account because they're involved in different ways.  There are different rhythms and paces that are going on here.  There's a different pace with regard to technology and with regard to teacher training, for example.  How is it all of this possible do you think?  That would be my question of you, please.

>> BERNARD CORNU:  Thank you, Don.  And I think it is a question for many people for a long time.  I have worked with computers and ICT education for most ‑‑ more than 40 years now and I have been impressed by the time it takes to make education, to make teaching evolved, to make learning evolved, why technology evolves so quickly and asks for immediacy and rapidity.  I had a feeling of a paradox or contra diction between sustainable and rapidly emerge technologies.  Sustainable refers to a pros is which is table ‑‑ process which is stable, which lasts for long while technologies are rapidly emerging and rapidly changing.  So this leads us to the question of time, the question of pace in the changes.  And I had this question for a long time in my head, at least.  In the end of the (?), when all education systems tried to integrate Information and Communication Technologies into education, three issues were addressed.  Equipment, the main question at the time was in most countries to provide schools with computers.  Resources, education software were developed very quickly at that final and teacher training.  How to prepare and train all teachers to immigrate ICT in their daily practice.  Many countries have questions separately one after the other.  While some other countries, the most successful in deed were able to addressed them simultaneously in parallel.  The three issues did not develop at the same pace and this is why integration of ICT education was so slow, so difficult.  But nowadays, because the evolution of technologies and of digital concept, the question is even much more difficult.  There is a kind of permanent conflict between the times that the region technology imposes and the time education needs or even technology and time and education.  The time of machine and the time of the human, short term, long‑term.  Digital technologies imply and need immediacy why technology needs time, long time.  The pace of informatics itself is not homogenous technology.  New devices appear, networks lead to emergency instant reaction, no time for reflection.  Artificial Intelligence, private, even the concept of knowledge and competencies, et cetera, are more slow.  In education, the time of a learner is also different from the time of the teacher.  There is also a gap between learner and teacher.  Nothing new here, but this needs particular strategies for teachers and being aware of that.  And, of course, different learners also have different rhythms in acquiring knowledge and competencies.  So the time of pedagogue, the time of teaching must adapt.  Sustainable in the rapidly emerging technologies needs many changes.  According to different times, the role of a teacher has to evolve in change.  This needs strategies and time.  The learner also has to change.  For instance, he has to get acquainted to digital learning, to distance learning, to collaborative learning and this is also a particular time.  Even in site schools, we have the same question.  The time of the classroom is not the same as the time of virtual environments.  (?) of 50 to 60 minutes virtual environments function according to more flexible times and pace.  One can learn when he or she wants or how he or she wants.  So changes in education, changes in learning need also the time of a social adoption.  Society has to adopt this before it is really ‑‑ it can be united properly.  It needs acceptance from learners, acceptance from teachers, acceptance from society.  They are driven the changes in technology mainly by economic factors and ICT companies, but society has also to take the time of integrating all these.  And education also evolves and there are many other parameters.  Ethics, digital devise, differences in the local level and global level.  This is also consequences on time and pace.  All the different items listed in the columns and the lines of the Zanzibar metrics have their own rhythm, their own pace.  In the real world, the system usually evolves or functions at the pace of the slowest component, not the quickest.  So we need to be aware of the different times and to deal with them in order to muster the e merging technologies and to insure sustainable education.  I think that we have to enrich the Zanzibar metrics with statements and hints about time in all the different cells of the metrics and then to be strategies according to these times.  Otherwise, technology will run faster and pedagogue will not be able to follow.  Don't let technology impose its space.  So we know we can answer your question.  Yes it is certainly possible to deal with all changes, but one must take into account the different rhythms, the different paces, the different times, sustainable education is the main aim.  Sustainable education must give the right pace to the system to teaching, to learning and one has to adapt technology to the pace of education and learning.  Thank you very your attention.

>> DON PASSEY:  Thank you, Bernard.  Thank you very that very, very insightful response.  In a way, it says to me that, you know, in some respects, we can't rely upon previous models.  We have to rethink where we are and where we're moving to in terms of these models because we have to understand the detail to a much greater extent.  Perhaps then we've realized before that in a sense, we've had an idea of a model that's been based upon this non‑rapid emergence in some respects whereas now we have to try and find a way to look at this which is going to take in so many different factors into accounts.  So I think ‑‑ I think that's very, very helpful.

Margaret, Cathy, I don't know whether you would like to comment briefly before I hand over the next part of this session to Johannes.  Would you like to add anything on this particular issue?  I think it's a very important one.  Cathy?

>> CATHY LEWIN:  Yes.  This just to share some of my experience really in relation to development and it's not always ‑‑ it's not always moving forward.  So in my experience, I've been involved in research on the use of technology in schools for, ah, many, many years now.  And I've been to schools for different projects and I've returned to schools later in type, several years later.  The first time I went to that school, it was things were clearly developing, technology was embedded across the educational system.  You go ‑‑ you can go a few years later and there will be a change of leadership, change of priorities, change of funding and it's completely difference from my perspective looking at the use of technology, it's taken a backward step.  It's not just about how things progress over time, but also being aware of the fact that you can, you know ‑‑ it can regress backwards as well as move forwards.

>> DON PASSEY:  Okay.  Thanks, Cathy.  That's very helpful.  Margaret, you have a comment?

>> MARGARET LEAHY:  Yeah.  One quick observation and I agree with you, Cathy.  As Bernard and Don were talking, the word that was coming into my mind was urgency.  I think with the use of this at the moment, you think of all algorithms controlling various things that we don't fully understand or often there will be an opaque decision making underpinning some of those AI processes.  I think there is an urgency to begin to understand ‑‑ we us a lot of times, but I call it AI broadly.  There is a need and urgency for education to begin to tackle these big ideas in order that we can prepare our young people to succeed and thrive in the future.

>> DON PASSEY:  Thanks, Margaret.  That's very helpful.  And I'm aware that Mary is now being able to join us, which is great.  So welcome, Mary, as the fourth panelist.  What I'm going to do now as you just joined us and I hope that Johannes and participants will be fine with this.  What I'm going to do, Mary, I'm going to ask you a key question which follows on from what's been talked about here, if I may.  And I think this follows on very, very much from what Bernard and Cathy and Margaret have been talking about with regard to rhythms and paces and how things are being coped w it comes back to this whole idea that we have a whole range of developing technologies that are coming on whether they be data science, data analysis, Artificial Intelligence, big data, autonomous systems, et cetera.  You know, and as Bernard was saying, there are different rhythms that are encountered with regard to these and different stakeholders are involved in different ways with regard to these.  From an educational point of view, how do you think education is coping at the moment with this range of emerging terms and concepts that are being brought forward?  What do you feel is important here in terms of education in this respect for the current and the future?  Can I pose that question to you, which I think comes very much from the discussion that we've been having?

>> MARY WEBB:  Sorry, everybody.  Sorry.  I got stuck into the system.  So I hope that didn't throw things too much.  Yes.  So this is a question that we've been puzzling about for quite some time and TC3 and one of the panelists particularly in the first webinar that was focused on.  I think everyone is agreeing that these terms are really, really important for education.  They're important for everyone.  And certainly in education, we need to make sense of them.  We need to make sense of the massive opportunities that data science and Artificial Intelligence will provide.  And I think one of the key ideas that most people were discussing was that we need these things to empower teachers, but also to empower learners.  And there are various ways that can happen.  Clearly there are areas in place where students are having to propose new ways that they might learn.  A lot of the systems are still in development.  I wouldn't say we got very widespread use of AI certainly in the UK and some of the other European countries and particularly in schools at the moment, but there are lots of systems being developed.  There's a massive development in China in this respect.  So they're here already and we need to make use of these opportunities and be aware of the limitations and potential challenges that they provide.  And I just was catching the end of what Margaret was saying there about the fact that some of these things are opaque.  They're not transparent to the users.  And this I think is one the biggest issues that we've been thinking about.  How do we insure that people understand what is going in the background to the extent that they're able to make good use of these for learners, for their learning and teachers understand what is going on but also that the learners begin to understand what's happening, when they're being advised by systems that are making some kind of sense of how they're learning and what they need to learn.  So as well as using the systems for students to learn, we think it's really, really important that students start to understand the data science and the machine learning and the way in which it works probably by using simulations if there are massive developments taking place in new opportunities for students to understand the different types of machine learning by training systems and Google's got one of these.  I think that's one of the things we've all suggesting needs to happen quite quickly.  We need to develop our curricular so that students begin to get an understanding of how these systems can support their learning, but more generally what the systems are doing which is going to be really important for society in the future.  And there is one other thing I will add and that is the actual definitions are problematic because we have different definitions in different fields of endeavor.  If some people only really are discussing what AI does but others need to know a bit about how it works and certainly an education.  We think we need to know about something about how it works and clearly the methods that machine learning is using are very complex statistical ones.  It's not going to be possible for everyone to understand those statistics, but they can begin to understand the implications of those and the kinds of ways in which they work and the kinds of issues they throw out.  So I'll stop there and I'm sorry I missed the first.  I hope I wasn't repeating what people already said.

>> DON PASSEY:  No, no.  That's very helpful.  Thanks, Mary, for those insights and the comments to that question.  I think that is sort of highlighting, you know, both the potential that we see and the challenges that we see ahead for us and certainly we have been able to start to discuss some of those key points within this session.  Johannes, I'm wondering are there ‑‑ would you like to pick up in case there are any questions from the participants involved?  Would you like to pick up at this point?

>> JOHANNES MAGENHEIM:  Yeah.  Unfortunately, we don't have any questions from the participants.  If you would like, ask a question by itself reflecting the discussions we had.  I think we have the opportunity and education to introduce concepts of machine learning and data analysis and different objects like sciences and biology and so on to examin questions of ecology and environmental damage, for instance.  The question is if students and teachers are using those concepts and the different objects, it is really important first to look to disciplinary concepts how they can work together on different projects and get things together not only from capital science, but from other subject areas and the other question is to what extent is it necessary to reveal the black box of AI (?) and of machine learning so that students will be able to understand how these concepts work and what will be the place in the curricular to deal with these questions and data science or can it be done within all subjects or central issue for computing science education.  That's my question.

>> DON PASSEY:  Okay.  I think you probably raised quite a number of questions there, Johannes.  But thanks very much for those.  What I'm going to do is I'm aware we have just a few minutes left and I'm going to invite Mary to respond to your questions first especially because Mary has been working on the task force with TC3 in terms of commuting curriculum.  If you don't mind, Mary, I ask that you respond to that first literally within a couple minutes.

>> MARY WEBB:  Okay.  Yes.  I think the black box issue is really, really important and the black box issue is the fact that explanations from AI systems, systems using machine learning, the kind of deep learning that is being developed very rapidly these days are very difficult to achieve.  So very often, the explanations are built on top of the system.  They're not actually able to go to explain exactly what's going on, if you like, in the background.  And sometimes they're not totally accurate.  So one of the issues for education is do we say that we will only allow AI systems that can provide proper explanations?  Do we allow these black boxes in education?  Because there are other way that machine learning can work.  More sophisticated applications that are being developed depend on these black boxes.  And I think people are saying we must make certain the explanations can be completely transparent.

So yes.  I think we do need to be very aware of that and we probably need to actually set some limits as to what people will do with these machine learning systems to make sure that the explanations are coming through in a way that's not just accessible to teachers, but to learners also.  And that's going to require some visualization techniques, some ways of presenting the data.  I'll stop there.  I probably hit my time.

>> DON PASSEY:  Thanks, Mary.  Margaret or Cathy, again, in response to that question, would you like to make a comment of literally 1 minute of comment, if you'd like.  Or Bernard.

>> BERNARD CORNU:  Yes one sentence.  There are more and more black boxes in education and technology.  If you want to use these boxes, people must be confident in these black boxes.  To be confident, in order to be confident, you must train them to reflect to have a minimum understanding, to have a certain knowledge, not of all the content, but just the minimum for being confident.  Social networks are not enough for that.

>> DON PASSEY:  Thank you, Bernard.  I think that's a very, very nice comment to be finishing on here.  I suppose the other comment they would like to offer at the end here is it does seem to me that we are becoming more and more reline the upon networks and networking and I wonder to what extent we actually have developed that for the benefit of all at this stage would also be a comment that I would add.  So at this point, can I thank you all for participating.  Can I thank the organizers of the Zanzibar Declaration events and panelists and moderator and the IFIP 60 organizers who have been behind the scenes on this.  You have there are much.  And to IGF for their support and particularly to thank all participants who have been involved.  So thank you thank you very much for a very useful and enjoyable session.  Thank you.