AI Education and Innovation Hub for Costa Rica

Costa Rica’s Digital Transformation Strategy expresses the nation’s commitment to position the country as a leader in advanced digital technology (https://www.micit.go.cr/digital-transformation-strategy-bicentennial-costa-rica). In order to accomplish this, the country needs a strong focus on STEM education, together with a path for the implementation of AI applications for education transformation.  Our project focuses on AI in education. This focus brings together technology to improve students’ success in STEM, professional development opportunities for teachers (and pre-service teachers) to help them effectively deploy technology in the classroom, and opportunities at the college level to work with some of the world’s leading researchers to develop effective AI technology to drive and enhance education in Costa Rica and the rest of Latin America. In this way, we create a virtuous cycle: Educational technology leads to a better educated workforce, which contributes to improvements in educational technology.

Costa Rica is one of the most literate countries in Latin America (https://www.bizlatinhub.com/expand-business-costa-rica/), has a strong public University system and has been increasing University graduates in science and engineering (https://nearshoreamericas.com/how-costa-rica-reinvented-itself-as-the-tech-epicenter-of-central-america/). The growth in Costa Rica’s technology sector has been defined by companies centered outside of Costa Rica. Other large international companies have just limited their investment in Costa Rica to call-center services  Given the dramatic changes to the nature of work and education that will accompany the widespread use of Artificial Intelligence, it is essential that Costa Ricans be prepared to lead in the AI-based economy with its own workforce and resources, and become an AI hub to the rest of Central American and Caribbean countries. Costa Rica has a strong entrepreneurial culture[2]  and ranked second in 2019 in innovation among Latin American and Caribbean countries (https://www.globalinnovationindex.org/media-infographics). A sustained partnership between the technology and educational community is required to get to the next level and drive dramatic growth in Costa Rica’s own technology industries. AI growing industry may be a more suitable driver of innovation and entrepreneurship for universities and educational actors in this region.

We have been working with academics from the main public universities in Costa Rica and the Ministry of Education of the country to implement an AI tutor that will improve middle and high school students and professor performance in mathematics and math teaching. Through a pilot program we will work in two ways: assisting school teachers with math learning, and improving teachers and pre-service teachers’ skills in math teaching and AI technologies. We will, then, develop further AI skills in higher education professors in order to encourage them to develop more AI technological solutions for other education fields. Through entrepreneurial solutions, Costa Rica will become an international hub for AI solutions to education and the regional center of research and development in AI. As part of our solution, we will organize meetings and workshops with teachers of all levels, Costa Rica’s Ministry of Education officials, and university faculty to share ideas about how to best support them in this work. We will also develop specific research programs and projects about AI applications together with these actors to enhance their research skills and knowledge on AI.

Our solution is to establish an international AI research and innovation hub in Costa Rica, based on Carnegie Learning experience with MATHia, an intelligent tutoring system for math. In the short term we will deploy Carnegie Learning mathematics curricula, including MATHia AI-driven software, to develop localized curriculum co-designs, and offer professional learning in middle/high school grades and university mathematics courses. We will also develop artificial intelligence (AI) capacity (especially for educational applications) at the university-level for undergraduate and graduate students, staff, and faculty, via a series of meetings, conferences and workshops that identify local opportunities, needs, and challenges, and development of at least two courses for both pre-service teachers and computer science (CS) university students. Our goal in the first year of this project is to establish K-12 pilots using the MATHia AI software for students at CONED (UNED’s national distance high school) and for students in the Guapiles region (located in the Northern-Caribbean part of Costa Rica).

We plan to support 50-100 teachers (including pre-service teachers at UCR and UNED) and 1500 students in this initial pilot. Carnegie Learning will lead 2 2-day face-to-face professional development sessions with the participating teachers. In the long term we will also develop enhanced STEM skills, educators with increased knowledge of blended learning and educational technology best practices, regional expansion of the approach, and a technology workforce driven by university graduates with increased capacity for using AI and data science to drive further AI education solutions and applications through entrepreneurial formats. We will also encourage sponsorship of students at Carnegie Mellon University learnlab summer school and the host of AI international conferences in Costa Rica.

The proposed approach is innovative because it creates a transformative, virtuous cycle linking deployment, in Costa Rica, of high-quality, artificial intelligence (AI)-driven instructional materials for mathematics, co-design of localized curricula, high quality professional learning for teaching professionals in middle and high schools (as well as university settings) with the incubation of university capacity for AI and data science research and development to drive practically important outcomes. Data from Costa Rican learners (and eventually, the broader region) will be used in courses created by the project to develop educational technology, AI, data science, and application development skill sets for both pre-service education students and computer science students. University learners and faculty will be empowered to use this capacity to support curriculum implementations, providing relevant, actionable analytics and progress monitoring to teachers and teacher-trainers, who also receive enhanced training in blended learning as a part of this approach. University students, researchers, and faculty will transform the Costa Rican technology R&D landscape, contributing to a high quality AI workforce while also contributing to enhanced STEM outcomes for more junior learners who will then rise to the university level as a part of the pipeline fed by this virtuous cycle. Pre-service teachers will go out into the region and promulgate blended learning best practices while computer science students with practical training in AI and data science with real-world datasets will drive new entrepreneurial ventures in educational apps or in any of a variety of other fields of human endeavor.

Project activities center on deploying Carnegie Learning mathematics curricula, including MATHia AI software, localized curriculum co-design, and professional learning in middle/high school and university mathematics courses. Deployments occur with the development of artificial intelligence (AI) capacity at the university-level for undergraduate and graduate students, staff, and faculty, via a series of meetings/conferences that identify local opportunities, needs, and challenges, and development of at least two courses for both pre-service teachers and computer science (CS) university students. 

Project outputs include localized MATHia content, mathematics course completions, teachers and trainers versed in the approach of blended learning, “big” datasets of MATHia usage for analysis in courses to be developed by the project, AI in education course materials for pre-service teachers and CS students, all by locally-generated content, datasets, and contexts.

Short term outcomes include mathematics for students using localized curricula relevant to Costa Ricans, enhanced teacher and trainer pedagogical content knowledge, and enhanced knowledge of educational technology, AI, and the confluence of these areas for university stakeholders.

Long term outcomes manifest as a pipeline of K-12 students with enhanced STEM skills, educators with increased knowledge of blended learning and educational technology best practices, regional expansion of the approach, and a technology workforce driven by university graduates with increased capacity for using AI and data science to real-world outcomes.

Costa Rica is perfectly suited because:

• Public universities CONARE    
• Strong focus on education nationally
• Universities are autonomous 
• UNED runs a national distance high school (CONED) 
• Public universities sponsor a network of scientific high schools 

Currently in Costa Rica, we are preparing the project to be implemented, so there are zero people currently been served by this solution. In one year we are expecting to have at least 50 to 100 teachers involved and 1500 students. The solution fully implemented will serve at least 30% of the teacher force of the country including high school and public universities and 200,000 students around the country.   

In the long term we will develop enhanced STEM skills, educators with increased knowledge of blended learning and educational technology best practices, regional expansion of the approach to Central American and Caribbean countries, and a technology workforce driven by university graduates with increased capacity for using AI and data science to drive further AI education solutions and applications through entrepreneurial formats. We will also encourage sponsorship of students from Latin American at Carnegie Mellon University learnlab summer school and the host of AI international conferences in Costa Rica and other Latin American countries.

1. Currently, the Costa Rican educational system is implementing a curricular transformation focused on developing skills within the teaching-learning process. This process involves training for teaching teams that the Ministry of Public Education must execute. This may affect the availability of time and space to train teachers by another initiative.

 2. From a technical point of view, there is no culture of self or mutual learning in the teaching personnel within the Ministry of Public Education of Costa Rica.

3. The current crisis generated by the COVID-19 pandemic will possibly limit resources for public education and teacher training processes in new technologies or in the education process supported by artificial intelligence.

By establishing a solid network of partners. Pedagogical mediation for the development of skills requires the design of more dynamic, diverse and innovative educational environments, as well as didactic activities, focused on the integral development of people; this opens up possibilities for training and learning processes through learning at distance and online and the use of Artificial Intelligence. The project seeks to train teachers to apply these techniques, which can be used virtually by students and to monitor their performance more closely.

The project must consider strategies to enhance Costa Ricans increasing needs of online and smart education, besides the production of academic material that will facilitate teachers' work. The project will also build capacity among Costa Rican teachers to train each other on the use of AI and other technology.

The initiative must contemplate the search for economic resources so that global effects such as those we are experiencing with COVID-19 do not limit the development of distance learning activities.

This solution builds on a collaboration that began ten years ago. Two members of our team, Maynor Jimenez and Ismael Morales, attended the AI in Education conference in 2010 and subsequently attended Carnegie Mellon’s LearnLab summer school (https://learnlab.org/index.php/simon-initiative-summer-school/). At this time, Carnegie Learning was collaborating with Carnegie Mellon University to evaluate the effectiveness of intelligent tutoring system software use in public schools in several Latin American countries (including Chile, Mexico, Ecuador, El Salvador, and Brazil in addition to Costa Rica). These pilots were very successful and continued in Chile until last year (supported by Ignacio Casas of Pontificia Universidad Católica de Chile; see https://link.springer.com/chapter/10.1007%2F978-3-319-19773-9_32 and https://dl-acm-org.ezproxyberklee.flo.org/doi/pdf/10.1145/2207676.2208597).

Although these efforts were educationally successful, the lack of local infrastructure prevented sustained implementation. In November 2019, Maynor Jimenez hosted el V Encuentro Regional de Enseñanza de la Matemática (V-EREM) in Guapiles and invited Carnegie Learning’s Steven Ritter and Amy Lewis to deliver workshops to teachers focusing on the use of educational technology and intelligent tutoring systems. In addition, Dr. Ritter was able to give lectures on educational technology at UCR-San Jose, UCR-Guapiles and TEC. Meetings were also arranged with the rector of UNED and the Ministry of Education.

These conversations result in the realization that a sustained approach to supporting educational technology in Costa Rica will require the cooperation of the business, research and education communities, leading to our current proposal.

Maynor Barrientos Amador: Coordinator of UNED’s Research Program on Fundamentals of Distance Education and of the International Research Network on Distance, Online and Open Education.

Ismael Morales Garay: Professor of Mathematics at the State Distance University  and the University of Costa Rica.

Maynor Jiménez Castro: Professor of Mathematics at the University of Costa Rica (UCR) and the Public Educational Ministery of Costa Rica.

Rodrigo Arias Camacho, President of the Costa Rica’s University of Distance Education.

Steve Ritter: Founder and Chief Scientist, Carnegie Learning

Stephen Fancsali: Director of Advanced Analytics, Carnegie Learning

Amy Lewis: Senior Director of Content Design, Carnegie Learning

Maynor Barrientos, 17 years of experience in distance education, a researcher in various topics related to distance education, and organizational networks. More than 25 years of experience in higher education.

Ismael Morales Garay, 21 years of teaching of math experience in the State Distance University (UNED) and 8 years at the University of Costa Rica (UCR). Researcher in AI educative technologies in the last 15 years. 

Maynor Jiménez Castro, 14 years of teaching of math experience in the University of Costa Rica (UCR) and researcher in AI educative technologies in the last 20 years.

Steve Ritter - Co-Founder of Carnegie Learning; 25 years developing, evaluating and researching intelligent tutoring systems for mathematics.

Stephen Fancsali - 10 years of experience in learning analytics and educational data science research.

Amy Jones Lewis - 7 years of teaching high school mathematics, 10 years providing teacher professional development, 5 years developing instructional resources for mathematics.

Carnegie Learning, Carnegie Mellon University (Human-Computer Interaction Institute, host of LearnLab Summer School), UNED, CONED, School of Mathematics of UCR, Ministry of Education Costa Rica (Guapiles)

Costa Rican National Council of Rectors (CONARE), Costa Rica’s Ministry of Science and Technology (MICIT)

Key customers: K-12 schools, universities (pre-service teachers, AI researchers and faculty). 

Funding for each element of the solution comes from different places. K-12 curriculum costs are competitive with standard textbook models and can be funded from existing educational curriculum funding. UNED operates one of the largest educational printers in Latin America and can act as a reseller and distributor for Carnegie Learning texts and other educational materials. At the University level, we expect government and industries to support graduate-level scholarships for particularly talented students. CONARE, the consortium for public universities in Costa Rica may also provide undergraduate support. In addition, the members of our team and our colleagues at Carnegie Mellon University have been very successful in securing research funding that can support graduate students and faculty research in artificial intelligence.

Carnegie Learning has been in business for over 20 years. We expect that our network acts as a reseller for Carnegie Learning in Latin America. Costa Rica's University of Distance Education has existed for 43 years and the University of Costa Rica over 75. This past November, we met with the Ministry of Education, and had preliminary discussions about financial possibilities.

UNED already has a publishing organization for educational texts and CONARE (Costa Rican National Council of Rectors) have annual financing calls for research and extension projects. 

In the short term, we believe that this initial seed funding will attract additional support from other foundations, but we see this as being self-sustaining in the long run, based on revenue from text and software.

To get budget sustainability for faster impact - enough to kick-start the project; to improve our ability to present our solution to foundations, government officials, and key players; and enhance quality through feedback from the TPrize prestigious teams.

Other partners in the regions, productive sectors, local governments by using public universities’ networks

Gates Foundation

Schmidt Futures (AI for Good)

Imaginable Futures

Fundación Omar Dengo