Two tear changes and struggle in higher education

Discussion ● Helge Hoevik

We must change to maintain it. Otherwise, the market will do it for us, writes Helge Hoevik.

Two of the main ones
Societal changes are tearing and struggling in higher education, globally and in Europe. The first concerns new language machines. The second is that this is now getting a European touch.

Of course we won’t come Apart from the rapid spread of generative artificial intelligence, artificial intelligence. These language machines are built on top of a new, unified global medium. It was established using client-server technology from the 1980s onwards. But things only accelerated when Amazon opened its online bookstore to all kinds of traffic and interaction across the World Wide Web since 2006.

Generative artificial intelligence It uses Language Large Models (LLM – Large Language Models) trained on countless fragments of text and images collected, – yes, specifically from the WWW.

training
The machines produce electronic communications. A (small) part of the method mimics the neural patterns of the brain. Once you have picked up a large dictionary, everything can be copied to your computer or mobile phone. “Skrablebøtta” GPT4ALL is freely available and can be downloaded and installed without special experience.

With increasing familiarity
For GPT, language machines can be ordered to produce scripts that turn on their heads in familiar forms of teaching.

As an example we can
Take this (prompt) to developing student assignments in data analysis in social studies:

You Sample generator for datasets that reflect the social, economic and psychological conditions in a country. Prepare many different examples of these data sets that are useful for teaching social work students about data analysis.

Here we take With two of the twenty (or one hundred) answers:

Socioeconomic data set: the income level, education level, occupation and housing conditions of different population groups in a country. It gives a comprehensive picture of the social differences and inequalities that can affect the state of people’s lives.

Psychological dataset: level of anxiety, depressive symptoms, self-esteem, and coping strategies among different demographic groups. It can help identify risk factors and the need for psychosocial support.

in front of that
The teacher gives a one-way lecture on the subject, and–preferably with slightly stubborn attempts to get questions from the audience–she can let the students choose a subject area from among them. On their own or in groups, they work with the teacher’s digital assistant to explore the topic, building data matrices for testing in statistics software such as (the free) Rstudio. From there, they move on to finding real data from services such as Mikrodata from Statistics Norway or Survey Bank, which OsloMet handles in cooperation with Sikt.

Then the weight is changed
From reproducing what has been read and lectured to reasoning and testing – with subject teachers trained and encouraged – the art of conversation about understanding and procedure.

Language machines are included
Now lifelong partnerships with search engines where Microsoft, after countless tries, makes Google worthy of ranking. Research becomes more like a dialogue to find out things. Here is an example in English:

How many university students are there in the world today?

Bing responds by pointing to the sources used:

According to a paper published by UNESCO, together with the International Institute for Educational Planning and the Global Education Monitoring Report, the number of students in higher education globally more than doubled between 2000 and 2014 to 207 million. (1). Today there are about 220 million students in higher education in the world, up from 100 million in 2000.2. By 2040, there will be approximately 600 million students enrolled in universities around the world (2)

Likes
health services and housing, education costs have exceeded their ceiling. The price of increasingly powerful educational technology is stagnant or declining.

Such a deficiency
The balance is not related to the reading board in the school, but to the number of teachers and the entire university institution. Academics in the public university system should know their visiting hours. We must change to maintain it. Otherwise, the market will do it for us.

the big other
Change is the development of a multipolar digital world. It is Europe in particular that matters to us here, because we may be talking to great allies, but the rising tensions are something you have to be aware of and keep an eye on. She drew attention when it emerged that the Americans had tapped the phone of former German Chancellor Angela Merkel. But it was the first big policy with GDPR rules on privacy.

United States Congress
He put in place a set of regulations here so that the security service could monitor anyone using an American digital service. It is mitigated by administrative orders (previous orders). As a result, META/Facebook has now been fined NOK 14 billion for breaching privacy in Europe. The law, which must be renewed in the fall of 2023, has remained in force so far.

in June 2023 The European Parliament will vote on other legislation that, in the same way as the General Data Protection Regulation (GDPR), will cut the wings of all AI companies, especially American ones. This opens the way for the development of the modern European computer industry in general, and various artificial intelligence solutions in particular. Maybe that’s part of the purpose.

very European and
In Norwegian academia, this likely has the consequence that we have to build from weakly secured systems (eg Canvas from Instructure). But there will also be a great need to develop sound digital knowledge systems. This will be seen in the development programs and interest in creating a European – and Norwegian – digital specialist literature which forms the basis of digital teaching assistants.