05/22 Nye digitale medier – muligheter og fremtidsutsikter
The title “New digital media - opportunities and future prospects” could have been used decades ago, but it has not become less relevant.
My perspective
I am especially interested in topics like Co-creation with machines, New teaching methods, and What will be essential knowledge in the future. I will draw on examples from history and from my own practice
The development of new uses for digital technology is for me closely linked to how digital interacts with physical environments. That is, situations where people create and control various objects using digital technology.
Through the use of technology, we change society. This provides education with significant opportunities, but also challenges: How do we prepare today's young people for a future where they will collaborate and co-create with machines?
I will try to give some examples, both by looking at the history and at my own teaching practice. I will also open up for a discussion about what will be essential knowledge in the future.
Exponential growth
We can often make projections and assumptions about what can be made technically possible, in the future. However, what becomes possible is not the same as the solutions that actually become successful.
At the heart of the technical development of digital technology is the phenomenon of exponential growth. The chessboard illustrates this phenomenon, which also has proven to be very central to the development of digital technologies.
According to the myth, the game of chess was invented by a servant of a Persian king. The king was very excited about the game and wanted to reward the servant. The servant is sais to have requested one grain of wheat for the first square of the chessboard, two for the next square, then four, and further a doubling for each square up to square 64.
What looks very moderate at the beginning of the chessboard will on square 64 end up several hundred times the annual production of wheat in the world. That with today's modern agriculture.
Moore's Law
The example from the chessboard shows exponential growth: something that doubles at given intervals. It is a phenomenon that we have a hard time imagining.
The same development also underlies Moore's law. Gordon Moore, one of the founders of the computer chip manufacturer Intel, came to the conclusion in the late 1960s that we get about twice as many transistors (read: computational capacity) every 20 months for the same price.
We will leave the chessboard, but first link it to a central event. In 1997, the computer Deep Blue chess won over chess champion Garry Kasparov. In 1997, Moore's law had been in effect for about 30 years. We were one third into the chessboard.
Today the effect of Moore's law has worked for 55 years, and we are halfway on the chessboard. In technological development, "the second half of the chessboard" was first used by Ray Kurzweil, referring to the point where exponential growth is really beginning to gain importance. From here, the increase is dramatic. It is at this point that the Persian king, according to myth, understands that he cannot possibly fulfill the servant's wish.
Robotization
Projections like this are quite common: "- One third of the workforce can be replaced by robots". This does not necessarily mean that a third of the workforce will be without work in the future, but it does tell something about changed skills requirements.
Everyone deserves a meaningful professional life, and what makes sense for the individual will of course vary. However, there are many tasks that most people will agree can be done by machines. Many will work closely with robots. Such technical solutions can be liberating, but they move and change requirements for skills and competence.
From my youth I have a strong memory of then my grandmother who told how liberating it was for her to get a washing machine. This happened once in the early 1950s. Some knowledge disappeared with the sink, but it is difficult to argue for returning to the previous state.
The report Robots @ School performed by Lego, among others, found that most children have no problems with the concept of having a robot as a friend. Most of us will probably get used to dealing with robots, not least if we can talk to them.
At Høgskulen på Vestlandet, we have varied experience with the use of simple robots in education. We can achieve a lot with affordable technical solutions. Costs also mean a lot if something is to get into the school. I will return to examples where students and pupils build and code robots themselves. First, I want to talk a little bit about the importance of speech interfaces.
In this example, we have worked with students, teachers and students in relation to using robots to code simple but meaningful dialogues using text to speech. In this way, we can give students a taste of how a chatbot works and the choices behind the dialogues that arise. We learn about this by deciding and coding what the robot should express and the conditions for this to happen.
The coding takes place here in the block programming language Scratch with its own blocks to code the Aisoy robot's text to speech.
Many express their resistance when I argue that we must learn how to have conversations with machines. Still, I do believe that understanding how we best express ourselves in order to get a machine to do what we want becomes necessary competence in most professions. Such conversations with machines will involve more of a verbal language and be very context-dependent. In the same way that we have developed a verbal and visual language adapted to the mobile phone, we will probably have linguistic variants that are adapted to communication with robots.
We should develop forms of communication that take care of humans and not just use machines on their terms. We can only do this by using technology and gaining our own experience.
Machine learning – "Nearest neighbor" algorithm
The next thing I want to look at is how machines learn. Here I will also look back to history. This takes us to a street in Soho, in London, and a pub named after the doctor John Snow. Outside of the pub is a water pump, a reminder of the cholera epidemics that ravaged Europe in the 17th and 19th centuries.
During an epidemic in 1854, John Snow registered all deaths from cholera with a black dot on a map. In this way, he was able to show that most of the deaths were related to a specific water pump. The reason I spend time on this story is that this method is an analogous example of how computers learn through the comparison of phenomena, with other phenomena already classified.
Such comparisons are the foundation of all systems that use automatic classification. We are all familiar with this, through various forms of customized content and advertising. It is one of the basic features of recommendation systems and content tailored to individual users. Given how dependent we are on digital media, a basic understanding of how such systems work is important from a societal perspective.
Such algorithms affect us all, even young children. Of course, we should not expect teachers in kindergartens and primary schools to teach about the technical aspects of recommendation systems. But we can work with such sorting algorithms in simple ways, and gradually link more complex knowledge to this.
In my case, this starts with a sack with woodbricks left from the workshop. A square and a rectangle. Both foursided, but so different that I choose to categorize them separately. The next objects are not completely square, and with a different color. I still choose to categorize this along with the square.
In this way, I categorize in much the same way that we can train machines in various ways through machine learning - people categorize, the machine learns and in many areas is superior to humans. Doctors who have to interpret different test results already have massive help from this type of technology.
In kindergarten and primary school, we can use the categorized bricks to build "algorithm animals". The students then discover that there are many intersecting categories. The mathematically justified ones are not always the same as those that, for example, can be linked to different parts of the "animal".
From here we can start discussing with the students, about how machines learn and what significance it can have in different areas of society. An example is vehicles with self-driving functions, which take us further to the phenomenon of autonomy.
Autonomy
Autonomy is about self-government. The Faroe Islands, which belong to Denmark, have a large degree of internal self-government. It is thus an example of a form of political autonomy.
In the case of technical systems, these are described as autonomous when they are able to make decisions and act on their own. Machine learning and artificial intelligence are an important prerequisite for autonomous systems.
We will have many debates about what tasks machines can and should do. One mission of education is, among other things, to facilitate that we as a society have good discussions about the use of technology. This is to the highest degree an interdisciplinary and general educational project about machines' opportunities to act on their own, and how we should set boundaries for the machines.
I want to show examples from a type of project that we have carried out a number of times with students, teachers and pupils. Here, the physical design is a central part of the projects. Students go through a design process where they build a simple robot.
First a student has modeled a prototype in clay and then transformed the three-dimensional figure into a two-dimensional sketch. Then we see a group of teachers who build their own robots. A preparation for a project that these teachers later carry out with their pupils.
In this way, the pupils have concrete experiences with the connections between form and function: ranging from the adaptation of the parts, the assembly of the various parts and how these are connected to the electronics so that the robots can be controlled by a BBC Micro:bit, a small computer that can control servos and interpret information from various sensors.
A key feature of these projects is to give the robots a visual appearance, while at the same time this appearance is reflected in a form of behavior, defined through the code the students decide.
We look at some simple, concrete functions that all students can understand and relate to More complex connections arise by putting many simple elements together: An example of a block code where students in 6th grade built and programmed their own robots. In this example, the students have not written all the code themselves, but we spend time with the students explaining connections between functions and program logic. Then students can adapt this logic to their purposes.
Code properties
Pupils in 6. grade discuss and code the properties of the automatons The conversations about properties and how this should be expressed become very interesting. Not only in relation to the robot's functions, but also as a beginning understanding of how to create and control systems that act on their own. The pupils in the picture discuss specifically how to translate the properties they envision in the form of code. How does a shark move in this case?
I also show another project using the same technology. This project is aimed at creating prototypes of machines that can collect plastic waste. Plastic in the sea is a significant problem on the west coast of Norway, and close to the students' own experiences. Here we see an example of a sketch where a student has ideas for a solution. We discuss this and translate the sketch into a concrete machine. Of course, we are only talking about prototypes on a small scale, but a concrete starting point for further discussions.
In such projects, one will be more focused on building and coding for specific functions. At the same time, the project fits well into a larger interdisciplinary framework, which among other things touches on environmental challenges and issues of sustainability.
Further technological development
To sort of conclude, we must research to look a little further ahead. This will of course be more like speculation.
We will have to adapt to the idea that everything will be possible to "fake", because machines are good at finding patterns and exchanging elements. Much of what we see, hear and read can be made by a machine, and what we observe is not necessarily rooted in a physical reality.
"Fake" is a term we have already come to know. Unfortunately, we need to learn more about it.
Machine translation, even in real time, will be common. Google Lens and similar services already provide us with real-time translation of visual verbal text. Direct translation of oral speech is not yet perfect, but it is constantly approaching.
Robots for multiple purposes will be affordable and able to solve more and more tasks. This applies to a large extent to software robots, but also physical robots. It becomes important for all education, not just vocational subjects.
How do we relate to a seemingly intelligent robot? Should they have any rights as members of society?
A preliminary conclusion
One of the characteristics of digitization is how algorithms and other technical solutions are applied in very different fields. A simple drone is controlled by technical solutions that share many features with precise control of rockets, whether these are used for peaceful purposes or as weapons.
For teaching, there is an enormous interdisciplinary potential here. At the same time, it is also this potential that makes the consequences of technology development unpredictable. An engineer who creates a solution for a toy may end up contributing to the development of modern weapons.
In the future, teaching must place even more emphasis on skills where humans are more capable than machines. A model for aesthetic learning processes can provide some key words for thinking about this: Empathy is something that is special to people. Maybe also imagination and creativity.
We should also keep in mind how the technological development features we have looked at can make it possible to expand the learners' opportunities to learn together with machines.
In twenty years, computers can be more than hundred millions of times faster than when Deep Blue won over Gary Kasparov in 1997. Such an upcoming, radical increase in how efficient computers will challenge the subjects we teach today.
As a teacher of arts and crafts, I wonder what will happen to manual work and manual skills, but also skills in foreign languages, arithmetic skills etc.
Perhaps the most important will be human empathy, the ability to see needs, and perhaps different forms of code will be our most important second language. Everything has to start small, as with the students discussing how desired properties can be translated into simple code.
In an attempt to look a little into the future, I have also tried to look back. There is a lot to learn from history. In the 1970s, Alan Kay was a key figure in the development of the personal computer. The illustration shows his ideas about childrens' use of computers, at a time when computers filled an entire office.
Kay also shows us a way if we are to try to say something about what is to come in the future: the best way is to take an active part in and shape development. Through teaching, we have great opportunities here, but also a responsibility that can be perceived as overwhelming.
Digitale medier og materialitet
Dersom du har fulgt meg så langt, og leser norsk, vil du trolig ha glede av boken jeg har skrevet om hvordan digitale medier møter ulike fysiske omgivelser. Ta en kikk på Digitale medier og materialitet