The Need for A Higher-Level Goal When Implementing AI in Education
Updated: Nov 13, 2021
Navigating the Unchartered Waters of AI in Education
Artificial Intelligence (AI) is quickly emerging as a technology that is reshaping education, a trend that is only accelerating during the current pandemic. This is changing the landscape both in how we teach and what should be taught, but the widespread implementation of AI in education will have ethical and societal challenges that will be difficult to navigate.
Recently, we sat down with Tsai Sheng Ho (Jason), Chief Executive of Taipei Language Institute (TLI), to discuss how the advances in AI are affecting the education sector. TLI, founded in 1956 by Jason's father, is one of the world's leading education and media conglomerates, with Jason at the helm since 2012.
As a leader in the education sector, Jason understands the potential of AI technologies, but also the risks, better than most. To ensure that these technologies do more good than harm, Jason believes it is fundamental that Edutech leaders have a “value-based higher-level goal”, and we think that his approach and advice are valuable to everyone in the sector.
Education has different meanings in different cultures, argues Jason, but ultimately it can be understood as “enlightenment”. Education is about the wisdom to make judgement or choices within a complex environment, and the advances in education reflect that evolving wisdom of humanity. As we implement AI systems into our education models, it is crucial to ask ourselves what are the outcomes we are trying to achieve. From Jason’s perspective, the most important goal should be to help humans become more human.
“Education is about helping humans become more human”
With this goal in mind, Jason describes language as the key variable, and the main difference between us and machines. Language is at the centre of human creativity, the space between human intelligence and machine intelligence. For now, humans have the advantage when it comes to language, even though it is not impossible to envision a future where machines will be able to close this gap. Still, at present language is essentially an expression of human creativity, according to Jason. This gives us some insight into the values Jason has set for his company, and has important implications in the way he is implementing AI in education.
With a no-shoes policy in the office and meditation music playing in the background, Jason is clearly a CEO that will follow his ideals. His office is undeniably luxurious, with plush leather couches, pictures with famous personalities, and a secretary that serves expensive high mountain Taiwan tea, yet Jason chooses to work mostly in “the pen”, an open office where he sits shoulder-to-shoulder with his employees. This is one way he demonstrates his value-based leadership, and to him, it's this higher-level goal approach that is needed when implementing AI in education.
Educators must be clear on their learning outcomes in order to teach AI systems accurately, for implementing machine learning is like training children: learning is achieved through repetition, using the data that is given, and without clear objectives, the results can be very different from our expectations.
With the goal to make “humans more human through language”, Jason’s company is rethinking and reworking their entire curriculum and methodologies to incorporate AI as well as to prepare learners for the increasing presence of AI in all aspects of human activity. Jason believes language helps us become more human, which will be increasingly important as machines take over more and more tasks. Language is the fundamental way to create something, for when we learn a language, or when we use a new phrase we never had used before, we are essentially creating something we never had before. For Jason, language is quintessentially human, as it has played such an important role in our evolution. With this in mind, bringing language and creativity to people is the goal that he has set as the higher objective to pursue.
Jason’s methodology includes collecting feedback and revising teaching methods based on the analytics collected from students. His scientific approach uses inductive and deductive reasoning, combined with designing a hypothesis and using experimentation, which is a technique that works well with machine learning. He is now incorporating AI technologies to help with this process and customize lessons to tailor them precisely to each student’s learning proficiencies and abilities.
At TLI, the focus is on using AI for personalization, and to understand both learning and teaching behavior better. Data analytics is employed to understand how each student learns which makes it possible to personalize online practice and create specific work plans catered to individual needs. The company is also looking into employing psychometric markers to collect more data on how learners comprehend and retain what is being taught. The implementation of AI will make this personalization process effective and more lucrative, as more data is collected. At the same time, Jason recognizes the data protection and ethical issues that can arise with this, and stresses the importance of ensuring that all data collected is in line with the higher objective. Having a value-based objective in place makes it easier to make decisions about how data is used and for what purpose.
Similarly, Jason describes a system that will use facial recognition to analyze student responses and predict learning outcomes. With the help of these tools, we may eventually be able to understand how we learn as humans, and develop much more effective teaching methods using the power of AI. Yet, Jason is also very aware of the potential ethical concerns that arise with the use of facial recognition, but ultimately, he believes that we should not be afraid to use the technology just because of the potential concerns. He views AI as a “young species”, which with the right evolution path, can do more good than harm. The question is how to ensure an ethical implementation of these technologies.
Clearly, there is a need for accountability if Edutech companies like TLI are going to implement such technologies. Jason is willing to jump into this field because he is applying a value- based higher goal and willing to take responsibility. And it is this sense of responsibility regarding ethics that we think companies implementing AI need to have.
Jason’s confidence to embrace new technology comes from his objective. He asks what he values the most, and says that for him, it's about life and creativity, which he sets as his priority when implementing technological solutions.
With the objective of “making humans more human”, Jason wants to implement the latest technology in order to develop education technologies that can help anyone learn a language. It is this type of higher-level-goal-based thinking that companies must consider before investing in and implementing AI technologies in education. Unfortunately, not all CEO's have such commendable objectives.
Jason believes that despite the importance of maximizing profits in the capitalist society that we live in, we should all have a value-based objective independent of profit-related targets. He argues that this objective should be about making incremental positive changes before it is too late, and we both agreed that profits should never be a higher-level goal.
Education is the foundation of our societies. To navigate the challenges of implementing AI in education, Edutech leaders and technology developers need to have a higher-level goal that will always come first. For Jason, it is about “making humans more human”, by enabling people to be more creative through teaching them new languages.
We appreciate Jason's approach and and agree that there is a need for a higher-level goal when Implementing AI in education. This higher-level goal can serve as a way for Edutech leaders to make better decisions, akin to having a north star to guide you through the uncharted waters and ethical challenges of AI implementation.
Written by Jesse Arlen Smith President of Aiforgood Asia. Jesse is advocate for the ethical development of artificial intelligence and believes in driving innovation by empowering people, companies and societies to change for good.