Let’s get this square from the beginning. Talking about 2050 is a nonsense if you try to make social prediction. It is fun, but it is not rooted in science. There are so many factors at play that we do not control. However, it might be useful to make some hypotheses not for seeing in 30 years time how good, lucky, we have been, rather to get a feeling of possibilities and take actions to build one specific future out of the many that we can foresee.
Again, trying to build a future versus actually building it are two completely different things, but as you try to build something you are going to face hurdles, you discover alternative and this makes progress roll on with a minimum of steering.
Why choose 2050? Well it is far enough to provide an open slate and it is sufficiently close to base actions on what we have today. In this specific instance this exercise is happening within the IEEE FDC Symbiotic Autonomous Systems Initiative, an initiative that has a long-term horizon around 2050, a time when we can expect machines to have developed a sort of autonomous intelligence and self-evolution capabilities (it may happen sooner) and a time when humans may entertain a symbiotic relation with machines, seamlessly leveraging on machines capability that will be perceived as an extension of our self.
You can take a look at the recently published White Paper, even better you can become part of its growing membership to be in the loop, at the least, or –better- to contribute in the shaping of the initiative.
So, Education. It is a crucial area for IEEE: since its foundation IEEE has been focussing on promoting engineers’ education and now, as technology has become pervasive, extending its education scope to anybody who needs to come to reason with it, basically anybody.
Let’s start with a hypothetical symbiotic being, a seamless interplay of human and machine. Suppose this symbiotic organism needs to increase its knowledge to carry out a specific task or to be prepared to face a new situation. Who shall be “educated”? The human component or the machine component?
It is not a trivial question and it is actually being discussed today although in a slight different frame: shall a students learn to do the square root of a number or given the almost ubiquitous availability of electronic calculators just look at the answer on a screen? In the near future this question may become:
- should we learn a foreign language or learn to use a real time translator (meaning pressing a button…)?
- should we learn to write or a machine will do the writing for us? Learning to speak is enough.
Notice that Society is implicitly providing an infrastructure for outsourcing knowledge. Most of us wouldn’t know -I don’t- how to make bread! Do you know how to get grain seeds, how to sow, till, harvest them, make flour and so on. And I am just talking of plain bread. Most of the things that we rely upon every single day are beyond our capability of producing, and for most of them we don’t even have the “knowledge” required. Yet this is not a problem, our knowledge is about using something that someone else had produced and made available. We accept this implicitly because, by far, this works. Besides, there is no alternative. A single person would not have the possibility to possess all the knowledge that is now available and that is required to “run” our life.
We have come to accept this segmentation of knowledge and even our schooling system is geared towards a segmentation. You get the basic tools you need to learn, and then you apply them to learn some specific things. The tools available for learning have increased in the last decades and they keep increasing to the point that it is becoming impossible to learn all of them. Hundred years ago it was about learning to read and write and little else. Then you learnt the tools of the trade, the specific one in your profession.
Now young people have to learn how to use the Internet (only very few know how to… and we are not teaching them), have to learn applying specific tools to extract knowledge from a rapidly growing set of data. Soon they will have to learn how to use augmented reality and virtual reality, how to interact with collaborative robots, how to balance their knowledge with the one of artefacts. In the meantime the knowledge half-time (the time it takes for 50% of what you know to lose its value, become useless, superseded) is shrinking, it is now below 5 years in technology areas (as shown in the graphic, IT knowledge reaches its half life in less than 2 years!). More than ever in the past knowing how to ask the right question and “whom” to ask becomes crucial. (Read: Half life of facts- why everything we know as an expiration date. Samuel Arbesman).
Symbiotic autonomous systems should be able to tackle the issue of the exploding knowledge, the knowledge obsolescence and the rebalancing of knowledge by operating in three main areas (and very possible more):
- digital twins,
- designing training on the fly, and
- autonomously generate education content.
This is not going to happen thirty years from now, first steps can, and should, be taken now and this is the value of considering a longer term perspective.
I’ll deal with each of them in the following posts.