Imagine being told that you are assigned to work on cutting-edge dossiers that match your personal interests and that you would be collaborating with a team of individuals that are smart, industrious, innovative, challenging and convivial. The result is an atmosphere (and tension) comparable to that faced by the team assigned to put a first satellite into space. But, in this case, the goal is to prepare Dzʿ and our graduates for a new world not yet fully understood, and as different from the old world as that which preceded the Industrial Revolution. The journey is as fulfilling technically as it is intellectually and philosophically.
But first, I should describe the actual job of the Dean of Academic Initiatives. The Dean of Academic Initiatives is a new position created in response to keeping Dzʿ proactive in areas flowing from advances in technology, particularly artificial intelligence (AI). The focus on AI comes in part from the Ministry of Education’s Plan d’action numérique which aims to kick-start Quebec’s education system in the use of new technology and prepare citizens for a society that is increasingly digital in nature. This means understanding and preparing for the role of AI, augmented reality (AR), virtual reality (VR), adapted pedagogy and online education.
At Dzʿ, the effort translates into developing partnerships with industry leaders, sponsoring and supporting research, establishing communities of practice, rethinking the design of classrooms and, among other things, investing in Makerspace. In short, it’s about providing leadership for the development of innovative high-technology initiatives across all sectors of the college in collaboration with the other managers and faculty and staff. Each of these challenges and initiatives would require a separate article to describe.
Education 4.0
The thrill of pioneering “education 4.0” at Dzʿ is matched only by the personal journey of discovery and insight that this new era in industry and education brings. Education 4.0 is the term used to describe the challenge institutions face in graduating students for “industry 4.0” where a workforce enhanced by the use of intelligent machines will focus on the individual and not the mass market as in previous eras. The need for adapting our pedagogical mission to meet the needs of education 4.0 is well established in numerous reports and publications, but the steps achieved so far in Canada are modest and usually subject to approval processes designed for slow incremental change – incapable of matching the rate of change in labour market needs.
What machines did for muscles, AI can do for parts of the mind.
What makes education 4.0 special is the fact that it is adaptable and focused on the individual. It has been described by analogy to the industrial revolution, where machines extended our reach beyond what could be done by muscles. What machines did for muscles, AI can do for parts of the mind. The result is an educational paradigm that hones in on personal needs and focuses on helping where difficulties arise. It is a “bootstrapping” process, whereby we develop and educate the AI unit that then helps us in our own education. A number of AI tutors are already deployed. The Georgia Institute of Technology was among the first educational institutions to use AI teaching assistants, and this pioneer project is being copied by at least one high school in the Montreal area, Collège Sainte-Anne. In the private sector, Korbit is a local company that is currently developing powerful AI tutors that will revolutionize pedagogy. Smart learning resources tailored to our pedagogical needs will mean fewer individuals are left behind.
At Dzʿ, only a few months into the challenge before us, there is much to talk about in terms of achievements. The AI focus of the Humanities and Public Life conference helped spark the minds of more than 1500 students who attended, while the last Coffee ’n Code workshops drew the interest of over 400 individuals who wanted to learn the basics of Python programming. We can also talk about the successful meeting of the first advisory group which included some of Montreal’s leading minds in AI and made us all blush when we weren’t terrified of our own success in gathering them.
The power of AI
But perhaps the personal journey is the more interesting course to follow for this issue of Academic Matters. Understanding the power of AI and its current impact on us as individuals is quite a process of discovery and it challenges misconceptions. It is not only about training neural net programmers. The challenge is to prepare our graduates for a whole new ecosystem of jobs, professions and career paths. Professions will evolve to maximize human qualities as machine learning takes care of simpler technical aspects. This shift in the scope of practice has been recognized by many different experts.
Most people think of AI in terms of robots, and images of Arnold Schwarzenegger come quickly to mind. But that is misleading. AI is not a thing of the future, it is currently inhabiting our world, it is powerful, it is largely invisible and it brings a vast improvement to the quality of our lives when used properly. You may not want to be surgically operated on by an AI-trained robot today, but even now robot surgeons are more adept in tricky operations, especially those which require a high degree of delicate maneuvering or extensive internal stitching.[1] Are you ready to be operated on by a robot? You may have already been!
AI systems abound, and they are pervasive in areas you may not expect, such as in supply chain management, insurance, marketing, health care, transportation, tourism and entertainment. They improve quality of life and I know that I enjoy dialoguing with Siri when I need something. The question turns to understanding the costs of such improvements. We need to equip our graduates to understand both sides of the equation.
Imagine sitting with your bank agent to secure a mortgage loan. With the AI tools available, chances are they know more about your spending habits, your likelihood to miss a payment, your likelihood to default and even your likelihood to accept a particular rate, than you do.[2] This has a huge impact on your ability to negotiate the deal you want. Implicit in this scenario is being pre-judged and held responsible for acts you have not taken. Students need to learn the ethics of AI. Happily, Dzʿ is already making great inroads in establishing itself as a major player in AI ethics.
AI’s “black box” and accountability
Listening to a speaker at an AI summit talk about the common aggravation with the “black box” driving artificial intelligence hit a chord with me. The “black box” refers to the layers of interacting units that sprout and reinforce connections with one another when learning in the same way that our neurons do. It’s an organic process, and the AI unit learns in rigorous training sessions through trial and error, like we do. Figuring out what happens when an AI unit learns to tell the difference between a fly and a ladybug is as complex as figuring out how a frog does it – ergo the black box analogy. But clearly the AI unit uses an internal logic. And the aggravation expressed in understanding the inside of the “black box” led to a number of personal reflections.
The logic used by the AI unit is consistent and coherent within its confines – the universe it knows about – but it is a foreign logic, not like ours. The question then becomes, is logic itself absolute and universal? If it were possible to expand our own universe would we discover that our rules of logic are a subset of some greater logic? Would the inhabitants of a 2D-flatland, for example, understand the logic of citizens in a 3D-world? Does AI have to limit itself to human logic? Whatever the answer, creating artificial intelligence gives us ample matter for reflection on biological intelligence, the universality of concepts and the speculation on alternatives.
The appeal of the “Turing test” leads to another personal reflection. Current wisdom holds that an AI unit can be recognized as “intelligent” and perhaps even sentient if it can convince another human to believe it, in a blind comparison test with another human (i.e. the Turing test). This is a goal envisaged by developments in AI. But if we turn the question around, does this not become a more interesting area of reflection? Is the Turing test the real arbiter that we make it out to be? Consider an advanced alien species applying their Turing test on us. Would we fail and therefore be considered less than an intelligent sentient species? How we deal with the world around us determines intelligence. Change that world and intelligence takes a different direction. Will we be able to recognize intelligence and self-awareness in an AI unit even if they do not match our own?
What these reflections underscore is that advances in AI usher in more than a technology that learns how to drive our cars autonomously. It touches on ethics, it touches on philosophy and it touches on the notion of right versus wrong.
I have omitted the dark side of AI on purpose. Most of us are aware of the power of AI in influencing election outcomes and in enabling an Orwellian regime. Preparing students to navigate safely in a world with increasingly intelligent agents about us is a formidable task, but one that we dare not postpone any longer. Siri is listening.
The potential of immersive virtual reality
AI is a big dossier, but not the only exciting one in Academic Initiatives. Consider teaching a physical education class in professional level Quidditch, or squash as played in Tron, or invoking a scene on a holodeck where your imagination is the only limit. Imagine conducting a physiotherapy session where tailored activities force you to use muscles or techniques that are needed for your rehabilitation but in a sport-like setting. Or discovering medieval chivalry in an arena attended by knights and nobility. This is the potential of an immersive virtual reality. Imagine what offering courses enabled by this technology could do for student learning (not to mention our reputation!) This is not the distant future, but the subject matter of a partnership we are currently developing with industry.
Imagine conducting a physiotherapy session where tailored activities force you to use muscles or techniques that are needed for your rehabilitation but in a sport-like setting.
Or imagine offering a course in a room where some of the students are visible and accessible, yet geographically distant. The advantages are huge. Students need not waste hours of commuting on days where they have only one class. Students who would not normally have access to post-secondary education would find themselves with new options. New technologies are making synchronous online education easier and more natural when addressing a class or organizing team-learning activities. A new type of classroom may soon emerge, one that resembles less the typical lecture hall, but by virtue of augmented reality, is more suited to and more focused on the individual. The people at SMART and Nureva are pushing that envelope and Dzʿ is listening.
In collaboration with the other English-language colleges, we are currently developing an initiative known as the eCampus. Based on the OntarioLearn model, this venture will allow partner colleges to offer joint programs of study and create opportunities to develop new programs and courses. If successful, this will bring learning opportunities to individuals in distant communities, or make training and upgrading skills available to the workforce after hours.
Biophilic classroom design
A final big project is based on an outcome of studies conducted in the United States suggesting that the design of classrooms can affect student performance and grades significantly. When learning spaces are designed to lower stress levels, student performance has been shown to increase by up to 11% in certain conditions. These biophilic stress-reducing considerations have prompted us to review our existing classrooms and study the impact of introducing natural elements into the classroom. A joint research project with the Université de Montréal is currently under way.
The variety of initiatives being pursued is inspiring and marks us as a progressive and mature institution. The new and emerging technologies empower us in our mission. With the impact of AI on our world and its potential as a learning tool we can agree that this is a defining moment in education. We are the pioneers. Like it or not, we are part of the fourth education revolution and there is an opportunity for each of us to get involved. What a thrill.
Endnotes
[2] A recent Frontline episode about AI revealed that in China, loans are granted by AI units (within 8 seconds) that look at 5000 data points gathered from a person’s phone, including how fast they typed the loan application and the degree to which their cell phone is charged. Apparently, people who keep their cell phones charged are less likely to be delinquent in paying a loan off while those who let their batteries run down are greater risk takers and are considered greater financial risks.