Jonas Christensen 2:52
Hind Benbya, welcome to Leaders of Analytics. It is fantastic to have you on the show today.
Hind Benbya 2:59
Thanks very much. Jonas. Thanks for having me.
Jonas Christensen 3:02
I am really looking forward to learning from you today because you are an expert in education but also in education, specifically of people in the data and data science space to This is Toby that's very dear to my heart. How do we train people to become the future leaders of our organisations. That is the topic for today. But before we get to that, could you tell us a little bit about yourself your career background and what you do?
Hind Benbya 3:30
Sure, my name is Hind Benbya I'm currently professor and Head of IS & business analytics at Deakin and I'm also the founder and director of the Centre for AI Artificial Intelligence and the Future of Business, which is a new multidisciplinary Centre at Deakin and I will tell you more about both roles shortly. So, in my role as head of IS and analytics, I lead the strategic directions and manage academic aspects of my area. This includes staffing, research, teaching and industry engagement. So it includes for instance, ensuring that our area as a leader in digital AI and analytics in Australia and the Asia Pacific recruits and retains the best staff, that our teaching is relevant and also connected to industry needs and best our research not only creates new knowledge that informs our teaching, but it also impacts industry and society and for my role of director of the Centre for AI and the Future of Business which really brings together a multidisciplinary group of researchers and industry experts with diverse backgrounds. So we have people from a law background, we have data scientists, we have people within AI by ground business as well as computer scientists. And the aim really is to be a catalyst for new ideas, insights and collaborations. In order to guide organisations and policymakers about how to leverage AI technologies more effectively, so the centre runs events such as our annual AI festival, it leads research projects and runs educational programmes. Prior to Deakin. I served as a researcher in digital business and department share of it and innovation in ambass monopoly in France, I have also been affiliated with a number of faculties in the US and UK such as UCLA, Anderson Business School, University of Oxford and Cambridge, among others. And before that, I worked as a consultant for many companies in the Silicon Valley in the US such as Hewlett Packard and Cisco, I worked with IBM in Italy. And I worked in St. microelectronics in Switzerland and Morocco. That's about me.
Jonas Christensen 5:51
So a very long and impressive background, I must say there. So how did you get into the world of analytics in the first place?
Hind Benbya 6:00
Yeah, well, sure, I initially started, they had a background in marketing and quantitative methods. So my very first master focused on using advanced quantitative techniques and statistics to understand customer behaviour. But at the time, technology was changing a lot of things beyond customer interactions. So this is how I wanted to be in a transformational field. And that is how and why I shifted into it with the second master information technology before undertaking a PhD. So this is how it's all started.
Jonas Christensen 6:37
Okay, so you're a quant by background as well. So, Hind I might shift to the topic of today, so we get straight into the detail. So you describe your role as the department head at Deakin of Information Systems and Business Analytics, that department and how you lead the strategic direction of the department, but also do academic aspects and so on. I'm really interested in what are students learning in the space of business analytics, right from short courses to degrees and PhD programmes. What are we teaching students today? What's important?
Hind Benbya 7:16
Sure. So first, we were the first in Australia to launch a Master of Business Analytics. And so we have a Master of Business Analytics, we have short courses on business analytics, we have the post graduate certificate on business analytics, as well as another one focused on artificial intelligence for business. So in terms of business analytics, our graduates cover pretty much the full spectrum out say, of analytics from descriptive analytics to design stations and predictive analytics, as well as prescriptive analytics to really learn advanced skills including predictive modelling, machine learning, data visualisations descriptive analytics, as well as statistics. But they also cover what I call the various applications or flavours of analytics in business, especially that this is an area that is becoming specialised, and pretty much embedded in every industry. So they cover for example, illustrations of people analytics, Decision Analytics, security analytics, among others. They also learn a number of tools and techniques from Tableau, Python SQL, among others. And what they do is that they also apply these learnings in capstone projects with diverse partners. So they work with different datasets on different problems with different industry partners to apply their knowledge. And the other thing that they do is that they have lots of them go into internships and research projects, with diverse partner organisations in industry, as well as not for profit organisations. So really, they get into working in real projects with specific problems faced by industry, that's pretty much cover what they cover in terms of analytical skills.
Jonas Christensen 9:05
So it sounds broad and deep at the same time, and technical and non technical as well. So you get to learn coding and tools, but also how to apply that in a business setting. So I'm interested here, what are the critical must-learn skills for students wanting to shape the future of business with data analytics? And you might say, well, it's all we teach in the course. But if you sort of boil it down into some foundational skills, what would you say that they are?
Hind Benbya 9:35
Yeah, well, I think you mentioned the pure analytical skills that we have just covers the ones related to statistics, programming, the tools and the techniques they need, but they need also to have business acumen eyes most of the time. They work as translators and problem solvers use analytical skills in diverse business areas. So not only they need to understand the issues faced by different areas, they need to be curious to have problem solving skills, they need to be creative and be able to tell a story with the data. But I believe that more and more than I don't think this is something covered in a lot of programmes. They need to grasp complex thinking principles, which is, in a sense, a problem solving skill as well. But it's embraces probabilistic thinking, scenario analysis, to embrace the possibility of various choices and different outcomes. Mostly, it's never one choice or one decision, it's always different possibilities. So it's important to anticipate these differential scenarios. I believe this is one of the skills that's really, really important, in future leaders, and I believe this one is one of the most important.
Jonas Christensen 10:49
Aha, this is interesting, because one of the things that I think a lot of data scientists and analysts struggle with is, they might come up with a recommendation or prediction based on the data set, which is then taken sometimes as gospel by the business or as a statement of fact, but it's really often a statement of probability. Yes, getting the stakeholder to realise that can be a bit of a challenge. I face that challenge regularly. So I know it, what do you teach the students in this space specifically, in terms of them being able to handle those situations well?
Hind Benbya 11:26
Yeah, I think they work on differential scenarios and Decision trees and work on different contexts and baby out call that critical thinking, you know, like, not just telling one story, but what's basically the outcomes and can sequences, that particular situation. I think they learned that when they work with people from different backgrounds in different projects, because most of the time, it's not just working with the data sets, and and you know, suggesting various outcomes, but working on problems with different companies and trying to provide various recommendations. So I think they those skills are developed when they work with different companies facing different situations.
Jonas Christensen 12:07
Yeah. Okay. Another adjacent the topic is really about the ethics of AI. This is something that I suppose the early big users of aI have never been trained in. So the famous examples are, for instance, developers at Facebook, using some data there to do things that may be considered unethical, but they were not thinking about that at the time, because they were technical experts, but not experts in ethics, so they hadn't considered those things, but it's becoming increasingly important that we consider ethics, what are the topics you're teaching students there? And how are you suppose educating our future leaders and analysts to be using AI effectively and responsibly? Should I say data rather than AI?
Hind Benbya 12:51
Well, sure, I mean, you can do that in various ways that first analytics now are increasingly being integrated into curriculum teaching at all of those, and how they can be applied to different business problems. But I think what we did is that for graduates or future business leaders, what we do is that we offer a postgraduate certificate in AI for business. And this was designed specifically to provide graduates with both the knowledge and skills to implement AI, as well as to address the challenges related to responsible and trusted AI. So of course, they cover the strategic thinking around how do you design AI strategy and how to develop AI products? They cover machine learning techniques, and how do they build, evaluate and train a machine learning model. But you know, as just you have said, there are a lot of potential in AI. But there are also major concerns around the ethics of AI. And we have seen different instances which there have been cases of discrimination and bias because the datasets that were used there were pretty much biased. So what we do is that we embed principles of responsible AI in our courses, so managers are aware of how companies can build governance frameworks to protect themselves against AI failures. This includes issues around explainability there are a lot of framework around Explainable AI, accountability topics around data privacy and ethics, as well as they have at least a flavour of the various regulatory frameworks that are emerging from different countries. So given that this is an evolving topic, you know, regulations change every day in every day we hear about decisions from different countries. I think having a grasp of what's going on around the world, not just in specific countries is very, very important.
Jonas Christensen 14:47
Yeah, absolutely. And the legislation is catching up shortly around the world in this space, but they unfortunately we have had the sort of, I think in a in a previous episode, I called it a digital oil spill where data has gone out or been used in the wrong way and created issues for lots of people that have been adversely impacted. This is a big topic that we could talk about for hours. But I actually want to divert a bit because otherwise we will, we will end up talking about it for hours. We've learned now about how you educate students and the sort of typical topics that you cover. I'm interested in your research arm of the department. What areas of focus do you have when it comes to research in the space?
Hind Benbya 15:34
Yeah, well, interestingly, our research areas are pretty much aligned with our teaching areas. So it's pretty much the same areas. So we focus broadly on three topics or theme which has the key themes and expertise of our area. One of them is around AI and the future of work. The second is around digital innovation transformation. And the third is business analytics for enhanced decision making. So in terms of analytics, for example, I can cite recent research projects by colleagues. So one of my colleagues ,Tomas, has been working in this area around creative analytics, also how analytics supports creative decision making. And the objective is really to understand how organisations can make the creative decision processes more evidence base, while retaining the best features of artistic intuition and human creativity. So it's an integration of analytics, as well as intuition and the creativity context. Another project is around sustainability analytics, and how analytics can support sustainable energy use and emissions reduction of electric vehicles. Other colleagues are working on social media data analytics in tourism marketing. So mostly here, the focus is on novel approaches to discover and capture comprehensive information about tourist behaviour based on data that is posted by tourists on social media platform in terms of artificial intelligence. The focus is on developing responsible AI practices, frameworks around ethics of algorithm, and mostly how AI can be leveraged across business and for social good in various industries from recruiting to innovation to decision making. And finally, the one around digital, I can cite research in cybersecurity conducted by colleagues. So for instance, data breaches can severely damage as you know, a firm's reputation and its customers' confidence. So the research conducted here suggests ways to address those issues before even they happen. And finally, I can cite also a couple of examples we're doing with practitioners, one of them is developing digital twins, and addressing the challenges faced by managers to scale their digital twins projects. And as you know, digital twins is first and foremost a data problem, of course, you accurate simulations and physics based models, as well as you know, analytics and AI. But it starts with data. So the project offer guidelines on designing is dealing with the challenges companies face to develop maturity with digital twins. And another example is around design in digital humans. So here, it's really designing digital humans that could could be accepted, you know, in different companies that that have that human side. And so, as you might see that our research is not only managerially relevant, by also talk about some of the challenges faced by business and society in general.
Jonas Christensen 18:59
So quite a broad remit. I think your first few use cases, they were very managerial or business focused, but I can see a lot of them, towards the end, they're also perhaps more and more technical in nature, and also, the societal impact shines through. Given that I think my next question is really relevant in so how do you see data analytics and AI playing a role in business in the next 10 years, and how must today's business leaders prepare for that?
Hind Benbya 19:31
So I think here it's it's important to say that while there have been some progress in AI over the last few years, so for instance, we start to see AI and particular machine learning algorithms used across diverse products in various industries. Everyday you see an example in finance and manufacturing, healthcare and recruitment. We have also moved From separate tasks towards models and applications that not only combine language speech, but also computer vision. And we see more and more these AI systems that are not only capable of conversing with humans, but they rely on realistic computer generated avatars that can generate convincing texts. They converse with humans, but they also can account for various human emotions, you know, they reognise when somebody's frustrated or angry. Now, I think despite all of this progress, we're very very, very far. And I got it very, very, very, very far from the AI that is promoted in the science fiction realms, which I believe, in some sense creates both and realistic hype and expectations around AI, as well as it creates, I believe, a lot of scepticism from diverse stakeholders. So people have in mind the AI they see in science fiction movies, they don't understand that AI can be both used in very mundane activities, as well as could be used for much more advanced activities and for innovation and other things. So I think it's very important first to educate the workforce about both the potential and opportunities of this emerging technology, as well as the challenges and risks they may pose for business. So you know, the jobless future narrative due to automation did not and will not materialise at the scale and scope that was previously anticipated. But what we're really seeing is a change in the nature of work and how his work has been conducted. And so there is more the need for reskilling of employees and the necessity to work with smart machines. Because they're here, I mean, there is not gonna go away, there is going to be more and more use of AI in the future. And I think it's important to recognise, to educate leaders that this is not kind some of a race against machine. And we will never have this human or superhuman AI. So it is what it is. It's a complementary tool with its own strengths and weaknesses, that leaders need to learn how to harness strategically. So I think that's very important educating about AI, what it can deliver, and how to address the challenges and risks it might pose for organisations.
Jonas Christensen 22:29
Yes, so really learn how to use it in your business, don't be afraid of it. And in your research, you describe the potential of AI in terms of four dimensions or business capabilities. They are automation, engagement, insight, or decision making and innovation. Could you tell us about what these opportunities are specifically and the challenges that AI brings across these four dimensions?
Hind Benbya 22:55
Sure. So what I'm gonna share is maybe a few examples here and happy to share with the readers the full articles, because there is there is much, much much there to discuss. And I think we can spend the whole day talking about them, and I will not be able to cover them during the whole day. So let me give a few examples. The first one is that AI is not limited to automation, it has the potential to enhance engagement with customers and employees in other ways to support decision making, as well as to create novel outcomes. And we see that more and more in diverse industries. So let me start perhaps with an example, let's take the automation capability, which really revolve around the use of technologies to support what we call structured or semi structured tasks. So these tasks often are repetitive. They are labour intensive. Some of them might include physical or manual work. But they also more and more include cognitive task thinking decision making, and they rely on diverse technologies. So you've got physical robots, you've got robotic process automation, and more and more the use of machine learning and natural language processing to automate various functions. So this is all good. I mean, we can use AI for automation in various tasks. But we also need to pay attention that it changes the interaction of humans with machines. And we don't know a lot yet about how workers interact, for instance, with physical robots, how they are going to alter their routines in order to accommodate robots movements in the workspace, or given the increase in reliance of automation tools. We don't know how much these technologies might render organisations mindless because they rely too much on these technologies, and they are becoming capable of outperforming humans in certain tasks. So we don't know yet. A lot of about how AI can be incorporated in strategic decision making or how organisations incorporate AI agents as members of board, for instance. So these are all at say, let's say challenges or problems that need to be developed in the future, we don't know a lot about all of these emerging questions. Now, let's talk about the second capability, the one related to engagement. So organisations design more and more and rely on conversational technologies, which needs to be somehow human-like, have human-like attributes to be accepted. So they need to have personality, they need to have a form to ensure that the customer experience is enjoyable and effective, but touches on the ethical aspects. We mentioned that before, assuming that people trust these technologies, would that make them more of remember, whenever when following the tools, suggestion? And how would accountability be managed? So I think this AI technology is more and more are also not only used to understand what individuals and groups say, their language, but they also how they feel their emotions and their response to human emotions. Now, that also raises ethical questions about whether machines should even display emotions that they don't actually have. So it raises a lot of questions and new challenges from an ethical perspective, from an accountability perspective. So for instance, if we take the decision making more and more decisions are automated with machine learning, and that's, you know, who is responsible for the implications of the actions that are either automated or based on insights that come from AI. And the regulation here differs. So for instance, if you take a just a few weeks ago, in the UK, they mentioned that they're going to be reliant on, you know, automotive vehicles. And that's the decision in 2025, in that the decision that is taken is that humans are not going to be accountable human drivers in case of accidents, but it's going to be the companies that are relying on electric vehicles. So making decisions around accountability is important, not only from a trust perspective, but also it reduces uncertainty, it reduces any questions people might use, or might have, again, in terms of decision making about this technology. So I think these are important decisions and challenges that companies need to come up with concrete frameworks on how to deal with these challenges.
Jonas Christensen 27:46
Yeah, there's some pretty big gnarly things in that. So just the insurance implication of who's responsible for for certain actions, you mentioned, autonomous vehicles as an example. We've all got car insurance, or hopefully, people have car insurance, or various types of personal liability and whatever else is there damage or whatever it's called, but who's responsible when you're not driving the car? Is the company or person who issued the algorithm? And so Tesla or what have you, or is it the fault of somewhere else? And how do you decide that? And how do you ensure that is very complex, really. So there's quite a lot in that. And I'm interested in how we bring today's business leaders and executives at speeds when it comes to this stuff, because it's not just about understanding or appreciating that the data has value. Everyone's heard the phrases of "data is the new oil", and so on. But it's the ability to understand how to apply it, but also, especially the risk management around it, because it is such a different type of risk management. And where I come from, I compare where we are with analytics, AI, machine learning, whatever you want to call it, to where we were, maybe, say 30 years ago, when it came to IT. So back then, people were getting personal computers sort of en mass. And they were becoming more standard on office desks and at home, internet banking was becoming more popular. And you had to be computer literate, all of a sudden, to be working in the business. I remember my mom going on this courses to learn how to tap on a computer rather than typewriter and so on. And back then many of the senior leaders, executives, they knew that it was important, but they didn't necessarily have any hands on experience. Even this the first CIOs of that time, they might not have actually known much about what computers were and how to deal with them and how to use them. And we're a little bit similar today in that situation where the boards and executives for big companies, they don't have hands on experience with building and utilising analytics and so on. So yeah, to frame up the question again, how do we bring today's business leaders up to speed when it comes to this?
Hind Benbya 30:16
Yeah, I think, well, first, there are a lot of programmes out there to support leaders understand both the advantages of data driven decision making and how AI can be used in organisations. I don't think, however, that there is a single path to becoming a data driven company or an AI based company. So there is different paths in here. So some firms might focus on building the right data team, while others invest in technology or build analytics into their digital projects. And it all depends on whether we're talking about a company that is a small or a larger company. However, I believe that like AI, it's important to start small. So you focus on first internal initiatives focused on employees, rather than focusing on customer facing and external facing capabilities, you build credibility and show the benefits because there are a lot of sceptics, especially when you talk about AI, if you have people that are sceptical about AI, you cannot just come up with a grand project, you really need to start small and use a portfolio approach in which you focused on short term initiatives, while gradually learning and building the skills needed for longer term initiatives. So I think also, it's important to use an appropriate structure. And here again, there isn't one best structure, there are various models, a centralised model, Centre of Excellence, dispersed. It really depends on the needs of the organisation. So some amount of centralization is necessary, because it helps bring efficiency. But it's all depends on the analytics maturity level in your organisation, I think it's important also to I mean, a lot of people talk about creating a data culture, which is one of the important areas. So if you really want to create a data culture, I really liked the example of Google as a company here, because they really trained the leaders in the organisation, and they started small from one area, and then they embedded it throughout the organisation. And people were provided training in order to understand the importance of that. So gradually, you know, they were able to build this data culture, because it wasn't focused on a functional area, it was spread, they build that capability throughout and over time, while really starting very small, initially working on a few initiatives. And I think also, one thing that is often important is that to embed people from different business functions, so that they come up with what the area of focus they want to work on. So really just involvements of people from different functions, I think is key also to creating what we call a data culture.
Jonas Christensen 33:21
Hi, there, dear listener, I just want to quickly let you know that I have recently published a book with six other authors, called demystifying AI for the enterprise, a playbook for digital transformation. If you'd like to learn more about the book, then head over to www dot leaders of analytics.com/ai. Now back to the show. Yeah, so you've probably seen companies do this? Well, and you've seen companies with great ambitions, not get there not execute on those ambitions. What are the typical sort of pitfalls and roadblocks that you see people stumbling over? How do you avoid those?
Hind Benbya 34:04
Yeah, I think it's related to the earlier question, I believe, it's a mistake to say, today, we're going to be a data driven culture, and it's gonna be everybody in the company needs to start from there on, we're gonna make huge investments in technology and money. And everybody's gonna do that. I think that's the wrong way to go with it's pretty much like in AI, where you cannot just, you know, transform your organisation towards an AI organisation. Why do you have not proven that it works in a small scale, so it really takes time to develop small initiatives and to prove that they are working and to gradually involve different people and train the leaders so that there is support also from a management perspective and to create that the right culture and structure so it's right really a variety of ingredients out there. I think in terms of of pitfalls or roadblocks, I think realistic expectations around AI and scepticism are one of them. And so really, to be able to change people's perspective you need specially from management, you need to demonstrate that it works on a small scale and shows benefits. And the other thing I believe, is that expecting AI to predict everything, like predicting human behaviour, I think that's pretty much unrealistic. You always need to integrate data driven with human judgement, and not expect to be able to predict everything and anything. So I think these are some of the pitfalls I have seen along the way with some companies.
Jonas Christensen 35:48
Yeah, okay, that rings very true to me. And I'm sure a lot of listeners can relate to those points. Before we finish up, I have a couple of questions left. But I'm interested in your unsolicited advice. If you were to give, suppose your top three pieces of advice for any business leader out there wanting to really get started on turning their organisation into a data driven organisation, what would they be?
Hind Benbya 36:16
Well, I think the first one would be to identify where data is used and where drives value, and it creates, you know, a difference. So I think that would be one of the areas where to start from, I think building the right data team is also important. And the third one, I say, do not leave that centralised or within just small areas, but I think creating a data culture is also important. Everyone in the organisation needs to understand its value, the so training people providing them opportunities to work on projects in their particular areas. I believe all these three are very important.
Jonas Christensen 37:03
Yeah, look, I couldn't agree more that one of the things that I talk a lot about to senior leaders in my organisation is we need to create the ability to produce analytical output at an advanced level, but also in a systematic way. And we also need to turn the organisation into an organisation that can consume that output in a similarly advanced and systematic way. So we need production and consumption. And that needs to really come together. It is a team sport. It's not just technology. It's different. Yes. So we're sort of at the end of our questions, I have a couple of questions left. The first one is one I always ask of any guests on these analytics, which is Who would you like to see as the next guest on leaders analytics, and why?
Hind Benbya 37:58
So maybe rather than citing people, I think we don't hear a lot about how analytics and AI are used in some specific industries. So the fashion industry, you know, they're using it to create novel outcomes. Skincare, sustainability, but I think that would be the focus of our AI Festival this year. It's our AI Conference, which is planned at the end of November. So we're gonna see a lot of those topics covered. Now to suggest to you a couple of guests, I would say maybe people analytics to understand engagement, well being before just you know, like optimising existing resources. So maybe David Green for people analytics and Sol Rashidi for AI and analytics in cosmetics and skincare, and all of the industries that we don't hear a lot about.
Unknown Speaker 38:58
I love it. Thank you Hind, this is really good. I haven't had any of those topics covered on levers or analytics, because you arrived, it is not something that we hear about often. So I will definitely be reaching out to those two. My last question is Where can people find out more about you to connect with you and get a hold of your content?
Hind Benbya 39:17
Sure, when please connect via LinkedIn or Twitter. I share some of the content there, but I'm happy to share longer versions via email. So yeah, connect via LinkedIn and Twitter.
Jonas Christensen 39:28
Great, and I'll make sure to also put a link to the research that we discussed earlier in the show notes so listeners can definitely find that there and explore the wonderful world of the research that comes out of Deakin University in Australia, a University of which I'm actually graduate myself, so I can I can vouch for that university. I had a great time there.
Hind Benbya 39:53
You're a great alumni Jonas!
Jonas Christensen 39:55
Yeah, that's right. So, Hind, thank you so much for your time today. We've learned a lot from you, all of us, myself and listeners, and we wish you all the best with your career and the research that you and the team do at Deakin University. And lastly, thank you for helping us educate the future of business leadership to be more analytically driven, and to really understand and appreciate the wonderful world of AI.
Unknown Speaker 40:28
Thank you very much Jonas, for this lovely conversation and the yes let's keep in touch and let's the conversation keep the conversation ongoing. It's been a pleasure having this conversation with you. Thanks again.
