Artificial Intelligence is a very dynamic field of research and development. Particularly during the last 2 - 3 years a great number of tools has been created and they have been so successfully applied that in fact every day new applications are created in many areas of human activity, among those, in the educational field.
Have you any experience to share?
I am just a student but i found Claude 3.5 and Claude 3.7 Reasoner to be excellent at Contextual Understanding especially in Sciences which got me the idea to develop a Custom version that acts as backend to create Visualizations of study Material
This is an example i made yesterday (In German)
Andreas Kurz Has you tried to compare this result, applying other IA, let's say, DeepSeek or Perplexity?
Perplexitys ability to create code has yet to be looked into by me further, i will update here once i did some tests. DeepSeek can create good looking Interfaces in React but it tends to create just a GUI with no interactivity but i am trying to get the perfect system prompt for the local model to create all necessary pages for the entire Visualization. Claude stays winning for me because of its excellent understanding of the context.
Attached are the three codes i got from the three (Perplexity, Claude, Deepseek) With the corresponding prompts (I had to prompt perplexity and DeepSeek to make one coherent code encompassing React, Js and CSS and render them in Claude while i only told claude to visualize because it automatically writes one file only)
- Perplexity prompt and Link: "Code a all encompassing (CSS, JS and React) Code that visualizes the Material at hand in an educational and interactive way. Only write one single code file";(Used the Pro Sonar model for this)(Video attached because it gave me 3 different files I had to first build in an IDE)
- Claude prompt and Link: "Visualize this educationally and Interactively"; https://claude.site/artifacts/dd1cb287-7edf-4c1d-b75c-74521b2b7ec3
- DeepSeek Prompt: "Code a all encompassing (CSS, JS and React) Code that visualizes the Material at hand in an educational and interactive way. Only write one single code file"; Video is attached since it didn’t make a react code but a HTML code even though I stated it should be react.)
The last attachement is the sample material uploaded.
Andreas Kurz Just a side comment: I have not asked about comparisons with Qwen, because I have just started trying it, but, up to now, I've get far better results than DeepSeek... Now in AI, It looks like we are getting some kind of snowball of vertiginous results...
Rey Segundo Guerrero-Proenza Qwen is amazing and one of the very few models to recognize Video Input.
Try out WanAI from the group behind Qwen: https://wan.video
China is really revolutionizing AI and proving that open sourcing knowledge will always win
I hope my Examples yesterday helped you understand why i prefer Claude for Educational Material
Professor Rey Segundo Guerrero-Proenza
I have been using several artificial intelligence options applied to research and other scientific and educational tasks. From chat bots on Telegram, Copilot, and older versions of Chat GPT. But I feel more comfortable working with Perplexity. Its use and recommendation by professionals, along with the verification of its operation with a much more scientific approach, have facilitated its use in various tasks over the past year.
Indeed, I am pleased to have found your works (
Article Buenas prácticas en los entornos virtuales de enseñanza-aprendizaje
yArticle Estructura de un EVEA adaptativo basado en las preferencias ...
) which I find very interesting and suitable to incorporate into my literature review for my Master's thesis in Educational Psychology. In this thesis, I am evaluating the proposal of a psychoeducational program in the form of a distance workshop for executive women on the theme of empowerment and gender.Book Programa socioemocional para habilidades sociales en la educ...
Maví Pérez García Very interesting your research; it is coping with a very frequent situation in the post Covid-19 scenarios.
On the other hand, I am curious: in which kind of task do you apply AI tools, let's say, Perplexity?
I have used Perplexity to filter and analyze the content of open-ended questions in national surveys, to prepare conference and workshop scripts, to find suggestions for workshop titles and refine dissemination and call messages, to generate analysis suggestions for survey and interview questions, and to plan the sequence of steps for organizational strategies related to motivation and job satisfaction
The rapid evolution of AI indeed mirrors a technological renaissance. In education, tools like adaptive learning platforms, automated grading systems, and AI-driven content generators are reshaping how knowledge is created, delivered, and assessed. While these advancements unlock unprecedented efficiency, they also challenge institutions to rethink pedagogical frameworks and prioritize ethical integration. The key lies in balancing innovation with critical oversight to ensure AI amplifies—rather than disrupts—the human-centric core of learning. 🚀📚 #AIinEducation
notable examples of AI-powered tools transforming education:
notable examples of AI-powered tools transforming education:
1. Adaptive Learning Platforms
- Khanmigo (Khan Academy): Acts as a 1:1 tutor, guiding students through math, coding, and critical thinking with real-time feedback. https://www.khanacademy.org/khan-labs
2. Language Learning
- Duolingo Max: Uses GPT-4 for personalized explanations and role-play simulations (e.g., practicing job interviews in a new language). https://blog.duolingo.com/duolingo-max
3. Content Creation
- Canva Magic Write: Generates lesson plans, quizzes, and creative prompts from simple text inputs. https://www.canva.com/magic-write/
4. Plagiarism & Integrity
- Turnitin’s AI Writing Detection: Flags AI-generated text in student submissions while providing feedback on originality. https://www.turnitin.com/solutions/ai-writing
5. Virtual Tutors
- Squirrel AI (China): Delivers hyper-personalized learning paths in STEM subjects using machine learning.
6. Automated Grading
- Gradescope: Streamlines grading for essays and coding assignments with AI-assisted rubric analysis. https://www.gradescope.com/
7. Research & Analysis
- Consensus or Elicit: AI research assistants that summarize academic papers and extract key insights. https://consensus.app/
8. Accessibility Tools
- Microsoft Immersive Reader: Uses NLP to support dyslexic learners with text decoding, translation, and readability adjustments.
9. Gamified Learning
- Quizlet’s Q-Chat: Creates interactive study modes (flashcards, quizzes) based on AI-curated content.
10. Virtual Labs
- Labster: Simulates science experiments (e.g., DNA sequencing) with AI-guided troubleshooting.
- Badges
- Science topic