Jose Machicao

Jose Machicao

Location: lima, peru

About me

Complex Modeling Expert, AI Engineer. Graduated as MSc in Energy in the University of Wales, UK, and as MechEng in the Pontificia Universidad Católica del Perú (PUCP). Lecturer at the PUCP, Continental University of Florida and Universidad Continental, Perú. Researcher on Modeling with AI.

Interests

artificial intelligence, reinforcement learning, complex modeling

Skills

complex modeling, python, ai modeling

Activity

This made me realize that the idea of LLMs is very limited regarding what we think it can do or how it can help. The potential is much higher than what is presented

I would like to use it to avoid traditional education tools as quizes and may be replace them with student conversations with algorithms

The concept I learned is how to relate personal student experience, instructor experience and academic content in order to optimize cognitive level in the learning process, trying to guarantee not only content learning but experience about the content

Interesting. I am going to translate it to virtual environments

I teach them how to interact properly with LLMs to learn. I show them the wrong way to use LLMs and how that damages their learning. So they come to class telling me the stories about how they interacted with the LLMs and I comment specially regarding the weaknesses of LLMs and sometimes the strenghts.

The mapping of the student experience is one of the things this course teaches. We all know that students are complex and they eventually feel limitations to complete their learning, but to know what factors influence the building of those feelings is important.

If the framework is not clear it is very difficult to assign levels of compliance. But when the framework is clear about the objectives is easier to deliver a rubric. Also it is important to have a system to collect feedback from many evaluators, including the students themselves

I use rubrics usually with analytical scheme. I find very useful to link the personal experience with theory, and I demand that criteria from the student work. 90% of students usually understand it.

Theory is very good. But unfortunately all infrastructure is designed for traditional approaches. Code in Python seems to be a very good application of experience based learning.

I already apply Python to complement any type of learning. Also I use demos from apps in the web. Currently I am encouraging my students to use ChatGPT and encourage to tell me how the experience of cognitive chatting helped them to learn or what leaks they discover in the chat.

End of Content

End of Content