The 2022 school year was marked by the urgent need to reestablish routines, to reconnect, to unlearn poorly learned rituals, in addition to recovering protected work spaces and times for students, teachers, and parents.
In short, it was imperative to return to normality.
Some educational organizations knew and were able to work on this exercise of discernment and managed to "return to the future", that is, to recover the best of the past and, at the same time, assimilate innovations that are to come.
Others felt the irrepressible urge to “go back to the past” before the pandemic.
The return to school in 2023 once again confronts schools with this dilemma.
The pandemic is passing, but innovation continues, unstoppable.
A few weeks ago, OpenIA gave users free access to the third generation of its powerful new GTP-3 chat.
Millions of users tried to access it and caused a collapse that took weeks to repair, but not without the news going viral throughout the world.
The GTP-3 chat is part of a family of Artificial Intelligence applications - which include Siri, Alexa, Baidu, Prometheus or LAmDA - called Large Language Models (LLM).
LLMs are statistical tools that predict the next words in a sequence based on the context you've been given.
To do this, they use huge amounts of text data, sometimes on the scale of pentabytes, using hundreds of thousands of parameters that allow them to "learn" autonomously.
Technology giants have been competing for years to introduce Artificial Intelligence into their products.
Microsoft announced this week that it will embed a version of Prometheus in its Bing and Edge browsers, clarifying that its LLM is far more powerful than "basic GPT-3 Chat."
Google will fight back with Bard, its new AI chat that will be integrated not only with browsers but with other applications it owns.
The AI era is just beginning, and education should not be distracted.
Today's most powerful LLMs are capable of not only creating a poem respecting a certain style, drawing an illustration, writing an essay, performing calculations or answering concise questions, but also gradually improving the quality and relevance of the responses based on the "accumulated experience.
Inevitably, the user must know how to use the tool, giving the appropriate instructions and correcting the result so that it finally meets their expectations.
In this sense, the LLMs do not replace the human, but they greatly facilitate the development of certain tasks.
Their misuse, however, exposes them to risks, which include academic fraud, among others.
Among other urgent tasks, education has the task of training intelligent users.
The “slogan” sounds nice, but there is generally no consensus on what it means or how to solve this challenge.
Nor is it an easy or urgent task in the context of a country in which 4 out of 10 students have a poor level of reading comprehension, or in which connectivity and basic digital literacy are not ensured within schools or in the families.
The challenge is not only one of educational policy, but is also played out in each classroom.
Back to school represents a good opportunity for teachers to ask ourselves what it means for us and for our schools to train intelligent students.
It is an open, passionate debate, with a pluralistic vocation and an urgent need.
Dean of the School of Education of the Austral University
look also
Tests Learn: the education system has a fever
The risks of judicializing and politicizing education
/cloudfront-eu-central-1.images.arcpublishing.com/prisa/7EXQTDHWSVDVDPC6TDIQTSHWUQ.jpg)
