Learning to realize how smart we are

By Rose Luckin | China Daily Global | Updated: 2019-08-01 08:01

With rise of AI, education must broaden its horizons so that the non-replicable richness of human intelligence can be cultivated

A few weeks ago, there was the headline in New Scientist that declared "AI achieves its best ever mark on a set of English exam questions". This made me wonder why students need to pass exams, now that AI can pass them too? This question crystallizes a bigger challenge: Now that we have created all this smart technology, what does it mean for us to be intelligent?

The fourth industrial revolution is powered by amazing developments in AI, from cars and lorries that can drive themselves, to diagnostic technologies that can identify cancerous tumors faster and more accurately than we can. We have successfully created machines that can behave in very smart ways. Admittedly they are limited to the one activity they are good at: a self-driving car can't diagnose cancer and a medical diagnosis system can't drive a car. But AI, is nevertheless, able to behave in ways that we recognize as intelligent.

Much of the publicity about the fourth industrial revolution focuses on predicting which jobs will be lost and how many people's livelihoods will be at risk. Transportation and storage are areas of the economy where much will change and many jobs are predicted to go. Education on the other hand is predicted to be the least likely area of employment to suffer the perils of automation. One could therefore believe that education will not change, but this would be a mistake. Education will and must change dramatically. It will change, because more AI will be used in classrooms and lecture halls. It must change, because what and how we teach must evolve to keep up with AI. Put simply, it has to ensure that students are properly prepared for the AI augmented world.

We need to stop trying to predict exactly how many jobs in which parts of the economy will be lost. Instead, we need to focus on understanding what it really means to be intelligent. We need to pay attention to developing the rich variety of our human intelligence that AI simply cannot achieve.

At the nub of the issue concerning intelligence and education, we find learning. It is the ability of machines to learn that has brought about the commercial scaling of AI, and it is the ability of humans to learn that enables them to develop sophisticated intelligence. The rise of AI should make us think about the extent to which education is limited to the sort of learning that generally happens at school and that we measure in relatively narrow ways, such as by means of exams. The thinking behind these ways of measuring progress, is the thinking that has spawned the sort of knowledge focused AI that was reported in the New Scientist article I mentioned at the start of this article. We now need to cast the net of progress measurement in education much wider so that we can truly recognize the richness of our human intelligence. Recognizing the rich and unique nature of human intelligence is also extremely important if we are to prepare people for the anticipated changes to their lives, largely due to AI.

We need to think about intelligence as an interwoven network of seven highly interconnected elements. We certainly still need knowledge and understanding about the world, we might call this academic intelligence. This is what helps us to solve complex problems, provided we do not limit our knowledge to subject silos and we learn about the ways in which physics is related to biology and geography is connected to history, and so on. We also need social intelligence to help us to communicate and learn from and with other people. In addition, we need a set of meta intelligences that we can think of as intelligences that help us to better understand ourselves.

There are four elements to these meta intelligences: first, there is our understanding about what knowledge is, what it means to know something, what good evidence is and how to make decisions based on this evidence - we can call this meta-knowing intelligence. Second, there is our understanding of our own thinking and of what we do and don't know, which can be described by the term meta-cognitive intelligence. Third, there is our understanding of our emotions and our motivations, which includes our ability to recognize our emotions and the emotions of other people. We refer to this with the term meta-subjective intelligence and it is particularly connected to our social intelligence through which we can develop an awareness and understanding about other people and our interactions with them. Fourth, there is our understanding of the way in which our physical bodies interact with our environment, which is described as our meta-contextual intelligence. All the four meta intelligences include both our understanding and our ability to control the way we behave in the world. These meta intelligences are not automatable with current AI technologies, and may never be automatable.

These six elements of our intelligence should not be thought of as individual intelligences, but rather an interconnected network of different elements. Finally, if we develop these different elements of our intelligence effectively, then we can develop the most important and seventh element of our human intelligence: our accurately perceived self-efficacy, which we can think of as the sum of all the other elements of our intelligence. This seventh element means that we know what we know, we know our ability to succeed in any specific situation. We are good at predicting which tasks we can be successful at, both alone and with others, and we know how to learn to succeed with new tasks. This is not something that AI can achieve.

The advent of AI at scale means that we need to rethink the way we think and talk about intelligence, especially our precious human intelligence. As a result, we need to redesign our education systems so that they value and nurture all the parts of our human intelligence much of which we can't automate with AI.

The author is professor of Learner Centered Design at UCL Knowledge Lab and director of EDUCATE. The author contributed this article to China Watch, a think tank powered by China Daily. The views do not necessarily reflect those of China Daily.