The educational revolution of artificial intelligence

Among the many documents cited by the Handbook of Artificial Intelligence in Education are the Guidelines and Recommendations for Use in Schools – which many teachers take advantage of Easter to read quietly, as its authors realize that the development of artificial intelligence is very rapid, so it may become Chart 10 methodologies and activities for assessment and learning The use of generative artificial intelligence published by the Open University of Catalonia (UOC) is particularly useful, as concrete actions such as the introduction of oral examinations into the assessment process are proposed. Examples include “introduction videos that include metacognitive processes on the part of students, as well as interviews to evaluate the content and validate student authorship – as the text indicates -. In addition, tests that favor personalization and enrich the assessment process.”

Communicate with students

In a similar vein, a UOC document included in the department handbook suggests increasing feedback throughout the training process. The idea is to “create several points of contact between teachers and students (both individuals and groups) before the activity is delivered, so that the formative nature of continuous assessment is reinforced and the content of the work submitted is linked to the learning process undertaken,” detailed in the diagram. Thus, the text also aims to include the use of ChatGPT in the planning of activities: to obtain an answer to a specific question and for the student to evaluate the contribution of the machine and identify its weaknesses; To highlight the most obvious aspects, or to complete the resulting text, all in a dialectical way. In other words, adapt your teaching to avoid ChatGPT. Ask them something a machine can't answer.

These are proposals that combine the competency-based approach of the new curriculum with forms of learning that value the PISA tests (whose poor results, announced in December, created a division that is still simmering) and that, despite the fact that On paper it is one that should be applied to the entire educational network, and all indications are that this has not yet happened, at least in a general way.

One of the first things that many of the teachers consulted by this newspaper brought to the table after reading this advice is that to be able to make this adaptation and constant monitoring of students' work, what they need is time. “Personal work, which we all call for, requires a reduction in proportions. With the current number of students in each class [la ràtio mitjana a l’ESO és de 28,5 alumnes per aula] “It is very difficult to implement these proposals,” in fact, a recent survey prepared by CCOO indicated that 62% of Catalan teachers attribute poor PISA results to high grades.

Although the weight of the guidance document falls on the opportunities of artificial intelligence, it devotes – as the school claims – a section to risks and challenges, which it summarizes in seven: privacy, surveillance, lack of autonomy, biases, overconfidence, digital divide, and data protection; Risks that are part of Catalonia's AI strategy.

On the issue of bias, this new guidance – developed by the department's Task Force on AI in Education, in collaboration with specialists from institutes, schools and organizations such as CSIC's AI Research Institute – points out that there are many examples in AI, such as requesting an application Generating images of the image of an educated person and always returning to white men. In addition, AI-generated documents may contain distorted or “overly stereotypical” images and opinions, which may be discriminatory or affect the recipient’s perceptions, the experts continued.

Working with data

Besides teaching, the document also indicates that the educational system could benefit in the near future from artificial intelligence as a support to personalize learning[puttingthestudentinthecenterwhichisthesloganofrecentyearsortodevelopeducationalpoliciesbasedongeneratingagenerationofeducation“knowledgefromworkingwithdata”Thusitcitesaspectssuchaspredictinginterventionsdetectingtheprofilesofteachersandcentresandimprovingadministrativeprocesses[posarl'alumnealcentreelmantradelsúltimsanyso”perestablirpolítiqueseducativesapartirdelageneraciódeconeixementsdeltreballambdades”Aixícitaaspectescomlaprevisiód'intervencionsladetecciódeperfilsdeprofessoratidecentresilamilloradelsprocessosadministratius[وضعالطالبفيالمركز،وهوشعارالسنواتالأخيرة،أو”لوضعسياساتتعليميةتعتمدعلىتوليدجيلمنالتعليم”المعرفةمنالعملمعالبيانات”وبالتالي،فإنهيستشهدبجوانبمثلالتنبؤبالتدخلات،والكشفعنملفاتتعريفالمعلمينوالمراكز،وتحسينالعملياتالإدارية[posarl’alumnealcentreelmantradelsúltimsanyso”perestablirpolítiqueseducativesapartirdelageneraciódeconeixementsdeltreballambdades”Aixícitaaspectescomlaprevisiód’intervencionsladetecciódeperfilsdeprofessoratidecentresilamilloradelsprocessosadministratius

In parallel with the publication of the guide, the administration, realizing how amazingly quickly this world is progressing, has collected this material, which expands on the web page on AI in education, and created four training courses for teachers in which it is noted that at the moment “AI is not going Beyond ChatGPT.