What should Education look like in the midst of Artificial Intelligence?

The video above is a recording of a keynote presentation I gave at the 2019 Education Innovation Summit. The topic was looking at the impact that Artificial Intelligence will have on the world and how the education system needs to change as a result.

I’ve also included an edited transcript below if you’d prefer to read it.

Hope you enjoy!


What should education look like in the midst of significant advancements in Artificial Intelligence? 

Good afternoon ladies and gentlemen, it is a great privilege to be with you today – thank you to Craig from BRIDGE, and to Virginia and the team for inviting me to share some of my thoughts. 

My name is Barry Morisse and today I stand in front of you as a researcher and ethicist in the field of artificial intelligence.  My day job is with a small company called En-novate which designs and facilitates global immersion trips for talented South African entrepreneurs and business people.  Our trips aim to expose the participants to what is happening in the major technological hubs.  We operate across 11 cities and in this vein, I’ve spent much of the last year travelling to some of the most influential startup hubs around the globe, trying to get a handle on where technology is going. 

Now my focus is on the field of artificial intelligence – a key technology that I believe is going to revolutionise our world – not just in a technical sense but also, and perhaps more importantly, socially, ethically and philosophically.  All of these are admittedly heady concepts.

So, I hope that you will forgive some of my idealism as I share my thoughts today.  I stand here on the outside looking in at the education system and as a result cannot possibly relate to the systemic challenges and constraints that all of you in the industry deal with on a day to day basis.  So I empathise with the fact that much of the optimism which I might share in this presentation may seem naïve and impractical.  That may be true.  However, the purpose of my talk is not to provide a magic bullet – but rather to start a conversation about artificial intelligence and how I believe the education system needs to adapt as a result.

Here we go.

Part 1: Why should you care?

Just the term ‘artificial intelligence’ has become a buzzword over the last few years and is a very nebulous term.  Although the field itself has been around since the 1950s, it’s only in the last decade where the rapid advancements have forced mainstream society to pay attention.  Unfortunately, as with all hyped technologies, the media frenzy often serves to muddy the waters and the term slowly becomes more and more vague – with plenty of misrepresentations getting in the way of clarity.

So if I try to bed down what artificial intelligence represents, my best attempt is that an AI system is a computer system that is able to perform tasks that traditionally required human intelligence at a level of competence that equals or exceeds that level of which humans are capable.

Let me illustrate with the example of self-driving cars.   

As I’m sure you know, there is a serious arms race happening at the moment between the likes of Google, Tesla and a myriad of other firms who are racing to create a fully autonomous car that can drive itself anywhere that we want it to go – without requiring a human driver at the wheel.  While to some that may feel like Sci-Fi, the truth of the matter is that the technology is very close to being able to do just that – and it just a matter of time until the software running in that car is able to drive much more safely and more effectively than any of us.  Thus, taking a task that we previously imagined would always require human intervention and rendering it obsolete. 

That is a clear example of artificial intelligence and it has serious implications for our society – which we’ll touch on a bit later. 

The point I want to make here though is that if we use that same logic – the simple pocket calculator could once have been considered artificial intelligence.   

Prior to computational systems being developed – the mere act of multiplication was considered inherent to human intelligence.  We then design a machine that is 100% accurate at a negligible cost and it renders human calculators obsolete.

The point is that once artificial intelligence is embedded in our day to day lives, we stop thinking about it as artificial and it simply becomes another tool that we use.  AI’s advancement is inevitable as we strive to make our lives better and due to its foundational nature at the heart of computer science – it will impact every industry out there.  Regardless of how little you know about it, or even how little you care about it – it is and will continue to impact your daily life on a micro-perspective and our society and workforce on a macro-perspective.  If we don’t prepare for it, if we don’t talk about its effects, if we simply choose to stick our heads in the sand – we will be on the wrong side of history.

Part 2: The Impact of Automation

And that is where the education system plays such a pivotal role.  We have to do everything within our power to prepare the generation of tomorrow for a world that doesn’t look anything like the world of today.

If we look back over the last few centuries, we’ve seen the same story play out again and again. Farming machinery augments and eventually replaces human labour. Henry Ford’s automobiles disrupt the horse-and-cart, rendering an entire transport system obsolete. The machine assembly line replaces its human version, emptying factories of almost all their unskilled workers. These, which were all huge innovations in their time, caused the same panic that we hear today. What will all those people do? How will they feed their families? Are their employers ultimately responsible for finding the solution or is this the modern form of Darwinism?

One of the key roles that educators play is to train and prepare young people to join the global workforce.  You work tirelessly to equip your students with the skills they need to contribute successfully to the world.  In this vein, it is crucially important that we consider which types of jobs and tasks will be made obsolete by artificial intelligence in the near future.  The cold hard truth is that a vast array of routine tasks, computations and processes will become completely automated in the coming years and that the humans that used to perform these jobs will find themselves locked out of the market.

In 2018, a working paper by the OECD attempted to codify the risks to certain professions due to automation and came up with these results.  Now as you would imagine, the science behind these sorts of predictions is always up to serious debate – but what I find interesting is that while blue-collar jobs are most at risk, as is to be expected, there is a whole ecosystem of white-collar work that is also at serious risk.  What this illustrates is that the scope of change that is going to happen to our economy is wider than we’ve ever seen before.  And that’s why many have started to describe it as the fourth industrial revolution.   

It’s not doom and gloom though.  As with every revolution before this, the technology opens the doors for new industries, jobs and value to be created, improving human living conditions and, ultimately, pushing our species forward. Over a long-enough timeframe it should become clear that the advancements are incredibly beneficial, even though certain segments of the population may be left behind - essentially, because they couldn’t adapt fast enough.   Proactive action on our behalf can ensure that our students are not part of that group that is left behind.

 

Part 3: What Do We Change?

The hard part of all of this is figuring out how to prepare for an uncertain future.  How do we prepare for a world that we can’t quite imagine?  It’s been a trope for a while now that the education system as a whole hasn’t adapted quickly enough for a 21st century world and, for the most part, still functions as an industrial revolution machine – procuring the skills for the challenges of the past.

So I think it is uncontroversial to say that the system needs a serious update.  And although these sorts of systemic changes often give us stomach ache due to their massive scale and significant difficulty – I actually think it represents an incredible opportunity for Africa to break new ground and chart a new course for itself.  Education has and always will be the Archimedes lever that we can use to change the course of people’s lives and as such, remains the most important tool we have in creating the future that we want.

Let’s get a bit more tactical.

It’s clear to me that the education system needs to shift its focus away from skills where automation and AI will have a serious advantage.  It is naïve to think that we are going to be able to compete with machines on tasks that require computation, information recall, precision processing, etc.  We have to accept that those types of tasks are going to be offloaded to AI systems.  This is a scary concept of course, especially in a country like ours – where unemployment is so prevalent – but we have to be able to let these go in order to create space to craft new curriculums and new modes of teaching.

Again, I will emphasise that there is an opportunity here to forge a more fulfilling future for young people and by letting go of tasks that are easily automated, we stand to create whole new industries which calls more on our humanity.  But this requires bold action from decision makers in the space.

Ok, so assuming we’ve let those go – what skills do we aim to replace them with?

This is a very hard question.  Some people when tackling this question talk about skills that are ‘future-proof’ which I think is a terrible term.  As I mentioned earlier – the goalposts will continue to move and the AI of 2025 will be the pocket calculator of 2030.  And so, to predict that there is some sort of magic bullet that is immune to the march of technology is unbelievably naïve.  I don’t think any of us are in a position to suggest that, say, computer programming is a skill that will never be automated – for example.  The game of trying to pick careers decades in advance just doesn’t work.

And we’ve seen this in the millennial generation.  When my parents were being educated, they were prepared to work in a single career for their whole life.  That world no longer exists and the vast majority of people in my generation will work in 5 or 6 completely different career paths during their lifetimes – many of which we can’t even imagine as we sit here today.

I think that a more fruitful way of looking at it is a real attempt to prepare young people to be as flexible and adaptable as possible – instead of pigeon-holing them and preaching the power of specialisation.  What humans do better than machines is generalisation.  By that I mean – humans are incredibly good at large scale problem solving especially in cases where the problem has never been seen before.  Because we have the ability to draw parallels and patterns from seemingly disparate parts of our knowledge to connect dots and carve new solutions from those patterns.

These are the types of skills that I think we should be focusing on – and I want to call them meta-skills because they are one level of abstraction higher than specialised skills such as mathematics or language or programming.  Let me explain what I mean by taking you through four meta-skills that gives a human being the edge over the AI of the next decade.

The first one is the meta-skill of problem solving.  It’s an obvious one, a foundational skillset that all educators aim to nurture in their students, but the way we go about teaching problem solving needs to take account of what the world looks like today.  Can we teach our students to be resourceful by using the internet while also maintaining a sceptical eye on what they find?  Can we teach them to look at things from first principles and encourage them to challenge the status quo assumptions?  How can we champion the scientific method as a methodical and responsible way of understanding the world?  Problem solving should be a foundational focus.

The second meta-skill is that of creative thinking.  How do you create something truly original?  How do we create the spaces for our students to take risks and try things that may seem unconventional or foolhardy?  On the whole we tend to unconsciously punish creative thinking and champion conformity – simply because it’s easier to manage large groups of students when we have serious standardisation, rules, procedures, etc.  That tendency needs to change – because the unique individuality and creative energy that each student brings with them is the key for them to find a place to contribute in a world of machines. 

The third one is learning how to learn.  In a life where you change careers a number of times, the ability to keep learning and keep retraining yourself in order to stay relevant is crucial.  Their adaptability in a changing world is dependent on how effectively they can take in new information, assess its relevance, accuracy and impact on their current situation and then forge new mental pathways as a result.  We are going to have to constantly be re-inventing ourselves throughout our career.  So this skill of learning how to learn is one where I think we should be allocating a lot more of our time. 

And the last one is what I’ve called scaling our humanity.  The best way I can describe this is with a story.  I took part in a programme in Berlin last year focused on the future of work.  The incubation hub that was running in invited 15 young people from around the world to brainstorm solutions for the threats that automation was posing to the administrative workforce of BMW – the car company.  At that stage BMW had over 600 people in their global head office whose jobs were for all intents and purposes – obsolete.  So they gathered a group of thinkers from across the world to try and find unique solutions.

I was greatly encouraged by the discussions that took place there.  Initially, I expected that we would discuss the usual solutions that get bandied about – upskilling, learning to code, transitioning into different roles etc. But what was really refreshing was the thoughtfulness that echoed in those discussions about how we can scale our ‘humanity’.

The nature of the technological shift means that tasks that require the human touch will become ever more valuable. Human creativity, relationship-based service, or a range of other tasks that require empathy, judgment, intuition or an intrinsic understanding of human behaviour all become increasingly rare and valuable.  With this perspective, instead of surrendering to the idea that the rise of AI is an inescapable path towards human obsolescence, AI may in fact liberate us from the types of tasks that we didn’t really enjoy doing anyway and free up time and resources to scale the human-centric tasks to a level we have never seen before.  For example, what if a company’s human resources department wasn’t bogged down by the administrative tasks of dealing with payroll, leave forms, medical aid and the like, but could rather focus all their energy on proactive care for their employees – looking after their physical and mental health. Thinking about AI in this way allows a radical paradigm shift about what a human resources department can be and allows those employees (who went into HR because they wanted to interact with people) to do more of the tasks that they find fulfilling.  That’s where the opportunity lies.

These are the kinds of meta-skills are what we should be talking about, rather than the content of the next science textbook.  The internet has democratised access to information and its now available instantly at a volume that we can’t quite fathom.  That’s not the bottleneck.  The bottleneck is how effectively we can curate that information and use it to solve novel problems, scaling our uniquely human qualities.

 

Part 4: What is our responsibility? 

The future will be inextricably linked with AI systems and as we develop them we must make some very difficult decisions as to how to deal with the externalities.

Some thinkers in the field have coined the term ‘human-centered AI’ which is of course terribly cheesy – but I think has some great tenets sitting behind it.  The basic ideas is that computer science needs a fresh dose of humanity.  This comes in three pillars:

1.   The development of AI should be guided by a concern for its impact on human society.

2.   AI should augment human skills, not replace them.

3.   AI must incorporate more of the versatility, nuance and depth of human intellect.

Now I don’t bring these up here to suggest that educators are the ones who need to steer the development of AI.  Instead I bring it up to show that educators can play a major role in preparing the human side of the equation to thrive in a world where AI systems are ubiquitous.  Rather than replacing us, AI can make us better at what we do.  The technology is full of potential for the human race – if we prepare ourselves accordingly.

If we are going to bury our heads in the sand and continue on the path that we are on, we are going to produce a generation of people who are thoroughly ill-equipped for a 22nd century world and who will be made obsolete due to advancing technology. 

However, if we can equip our people with the meta-skills needed to thrive alongside the machines, we can empower them to take charge of their own destiny and use the AI tools to solve problems that we could never have dreamed of solving – pushing the human race forward.

I don’t have all the answers – but I know what the right direction is.  We have the responsibility to start sounding the alarm and to start turning the ship now. 

Thank you.