Metacognition has the common definition of “thinking about thinking” (see The Best Posts On Metacognition). In other words, it is the self-awareness to know what our strengths and weaknesses are, and how and when to apply the former and compensate for the latter. Broadly explained, learners applying metacognitive strategies plan in advance for effective learning, monitor and make adjustments during the lesson/activity to maximize their learning, and reflect afterwards about which learning strategies worked and which did not for them.
AI has already infiltrated the workforce, so higher ed institutions have a responsibility to teach their students to use it responsibly and effectively.
Take control of your Book Creator dashboard with new tools to organize, sort, and search your libraries. Discover the most requested features of 2025—custom sorting, drag-and-drop, search bar, and personalized icons—designed to streamline your workflow and boost student creativity.
Juan Domingo Farnós (((La imagen representa un gráfico de conocimiento que visualiza cómo la integración de la educación disruptiva y la IA-AGI en la educación superior, dentro del contexto de la realidad aumentada (RA) y la realidad extensiva (RX), transforma el proceso de aprendizaje. En él se muestran conexiones clave como el aprendizaje personalizado, entornos…
Metacognition has the common definition of “thinking about thinking” (see The Best Posts On Metacognition). In other words, it is the self-awareness to know what our strengths and weaknesses are, and how and when to apply the former and compensate for the latter. Broadly explained, learners applying metacognitive strategies plan in advance for effective learning, monitor and make adjustments during the lesson/activity to maximize their learning, and reflect afterwards about which learning strategies worked and which did not for them.
Benchmarks like Massive Multitask Language Understanding (MMLU) have played a pivotal role in evaluating AI's knowledge and abilities across diverse domains. However, existing benchmarks predominantly focus on content knowledge, leaving a critical gap in assessing models' understanding of pedagogy - the method and practice of teaching.
This paper introduces The Pedagogy Benchmark, a novel dataset designed to evaluate large language models on their Cross-Domain Pedagogical Knowledge (CDPK) and Special Education Needs and Disability (SEND) pedagogical knowledge. These benchmarks are built on a carefully curated set of questions sourced from professional development exams for teachers, which cover a range of pedagogical subdomains such as teaching strategies and assessment methods. Here we outline the methodology and development of these benchmarks. We report results for 97 models, with accuracies spanning a range from 28% to 89% on the pedagogical knowledge questions. We consider the relationship between cost and accuracy and chart the progression of the Pareto value frontier over time.
We provide online leaderboards at this https URL which are updated with new models and allow interactive exploration and filtering based on various model properties, such as cost per token and open-vs-closed weights, as well as looking at performance in different subjects. LLMs and generative AI have tremendous potential to influence education and help to address the global learning crisis. Education-focused benchmarks are crucial to measure models' capacities to understand pedagogical concepts, respond appropriately to learners' needs, and support effective teaching practices across diverse contexts. They are needed for informing the responsible and evidence-based deployment of LLMs and LLM-based tools in educational settings, and for guiding both development and policy decisions.
Los investigadores probaron el conocimiento pedagógico de 97 modelos de IA: esto es lo que encontraron...
Gemini 2.5 Pro de Google y Claude Opus 4 de Anthropic ha ofrecido sistemáticamente un alto rendimiento (hasta un 89% de precisión) en los puntos de referencia de conocimientos pedagógicos.
Qwen -3 32B & Medium 3 de Mistral AI mostraron un gran rendimiento (70-82% de precisión) a un costo menor que los modelos "premium".
Los modelos de IA más pequeños y de menor costo, como GPT-3.5 Turbo de OpenAI y Llama -3.2 1B, tuvieron un desempeño constantemente malo (28-52% de precisión) en los mismos puntos de referencia pedagógicos.
Todos y cada uno de los modelos de IA, incluso los de mayor rendimiento, mostraron una menor precisión cuando se probaron en especialidades de diseño (preguntas de educación especial).
Algunos modelos, específicamente Gemini 2.5 Pro y Claude Opus 4, están "mejorando en pedagogía", pero todos los modelos aún tienen puntos ciegos significativos.
Confiar en modelos de IA más baratos y de menor rendimiento significa que ~una de cada dos recomendaciones pedagógicas será engañosa o incorrecta, mientras que los modelos superiores reducen este riesgo a ~una de cada diez.
I will be honest, when I first began using ChatGPT, I viewed it as just another productivity hack. A smarter Google, perhaps a useful writing assistant at best. Ask a question, receive a neat little…
That’s a complete reversal of how it used to be. Colleges used to be measured by how many books they had in the library. Access to courses was restricted. If knowledge was power, controlling …
Hey, friends! If you’re an educator, you really need to be using NotebookLM. This game-changing tool from Google helps you instantly turn your lesson materials into clear summaries, custom study guides, engaging podcasts for your students to listen to, and thought-provoking discussion questions. In this video, we walk you through how to use it right away in your role to save time and engage your students. Enjoy! xx Molly & Lauren
⌚️TIMESTAMPS
0:00 – Intro to NotebookLM 0:37 – Summarizing Content for Students 4:08 – Creating Study Guides & Review Questions 6:31 – Generating Audio Overviews 8:12 – Sparking Student Inquiry & Discussion
🧐WHO ARE WE? If we’ve not yet met, hi! 👋🏼 We’re Molly and Lauren, the faces behind Designed for Learning. Our goal is to make professional development fun, relevant, and most importantly, useful in the classroom! On this channel, we share strategies, tools, and practices that are classroom-tested by educators to help you create powerful learning experiences for ALL of your students.
⬇⬇⬇
___
// 🍿 WATCH NEXT:
➡️ How to use AI to Reduce your Workload as a Teacher → •AI as your Teaching Assistant
➡️ How You Can Use Canva to Create Teaching Materials → • Steal Top Tips from a Teacher
➡️How to Start Incorporating AI into the Classroom → • Check out How to Teach Students about AI
2 | No matter where you are in your AI journey we want to support you! Selecting the right tools for school-wide tools, AI literacy, meaningfully using AI in the classroom, and safely and responsibly using AI with students…we can help you with all of it! → https://dforlearning.com/ai-resource-library#pd-options
3 | As a Canva for Education training partner, we provide custom professional development services tailored for educators to transform your classroom. → https://dforlearning.com/canva-partner
AI Agents can make the workplace more productive when fine-tuned for different personality types, but human co-workers pay a price in lost socialization, ongoing research by MIT information technology specialists has found.
MIT researchers led by Sinan Aral, a professor at the MIT Sloan School of Management and Harang Ju, a Sloan postdoc associate, found that humans using AI raised their productivity by 60%—partly because those workers sent 23% fewer social messages.
Investigadores del MIT Sloan School, liderados por Sinan Aral y Harang Ju, realizaron un estudio con más de 2,000 personas para medir el impacto de agentes de IA en tareas creativas. Utilizando la plataforma Pairit, el experimento consistió en que equipos humanos, con y sin el apoyo de agentes de IA autónomos (capaces de razonar, planificar y recordar), desarrollaran campañas publicitarias completas.
Los resultados fueron contundentes: los equipos que trabajaron con agentes de IA fueron 25 % más rápidos y un 60 % más productivos. Además, produjeron textos de mayor calidad, aunque las imágenes generadas por humanos seguían siendo visualmente más atractivas.
En este informe especial, analizamos una advertencia que está generando un gran revuelo en el mundo laboral y tecnológico: Un reconocido investigador de inteligencia artificial asegura que, para 2027, la mayoría de los trabajos de oficina podrían desaparecer debido al avance imparable de la IA.
Exploramos los datos y tendencias que sustentan esta predicción, las reacciones del mundo empresarial y académico, y las posibles consecuencias económicas y sociales de una transformación tan radical. ¿Estamos realmente preparados para un futuro sin empleo administrativo? Acompáñanos en este análisis profundo que desvela los riesgos, las oportunidades y los desafíos urgentes que plantea la automatización avanzada. ¡No te pierdas este reportaje crucial sobre el porvenir del trabajo en la era de la inteligencia artificial!
El trabajo de OFICINA tiene sus días contados, debido al avance de la IA. El trabajo de oficina incluye las siguientes tareas: Notarial, Negocios, Ventas, Desarrollo softwares, Programación, Diseño Gráfico, Análisis financiero, Contabilidad, Finanzas, Consultoría de RRHH, Mercadeo y Gerencia de procesos administrativos...
No está muy claro el panorama ni el nivel de impacto. Pero sí hay unas claves de desarrollo que están marcando el paso hacia un futuro altamente automatizado...
¿Cómo se preparan nuestras universidades para adaptarse a los grandes cambios que se avecinan?
As AI capabilities advance at crazy speeds, the level of entry for setting up a business is disappearing. Here's how to help your kids take advantage of this.
Matthew Kay provides actionable insights on how to structure discussions, promote critical thinking, and create a safe classroom to encourage thoughtful debate.
To get content containing either thought or leadership enter:
To get content containing both thought and leadership enter:
To get content containing the expression thought leadership enter:
You can enter several keywords and you can refine them whenever you want. Our suggestion engine uses more signals but entering a few keywords here will rapidly give you great content to curate.
Your new post is loading...
![[PDF] Benchmarking the Pedagogical Knowledge of Large Language Models | Education 2.0 & 3.0 | Scoop.it](https://img.scoop.it/PRCkb8TEx-QjLUR_Id4Lyjl72eJkfbmt4t8yenImKBVvK0kTmF0xjctABnaLJIm9)
