Summary
Highlights
Cristóbal Cobo introduces the concept of anthropomorphism in technology, where human traits are attributed to AI, drawing parallels to Frankenstein. He highlights how AI is challenging our traditional definition of intelligence by performing tasks previously thought exclusive to humans, such as recognizing voices, texts, and faces. He emphasizes that technology is developing so fast that today's information may be obsolete in a week.
Cobo explains Artificial Intelligence as an umbrella concept, a multi-purpose technology that, like writing or electricity, becomes invisible once integrated. He then delves into Generative AI, which can produce sophisticated texts, videos, and images, mimicking human creativity. He raises concerns about the biases present in the data used to train these AI models and the challenge of distinguishing human-made content from AI-generated content.
The speaker challenges the notion that technology makes people obsolete, instead suggesting that those who learn adapt and avoid obsolescence. He warns against the over-reliance on AI for fundamental skills like reading and writing, which are crucial for structuring thought. He introduces the concept of 'AI fluency,' advocating for understanding how AI systems work, basic functionality, and critically evaluating the information they produce.
Cobo discusses the negative consequences of generative AI, including over-reliance by students, manipulation, biases, lack of transparency (the 'black box' problem), and privacy concerns. He also highlights environmental impact and potential job displacement. He illustrates how AI can subtly impact human cognition, such as the potential decrease in reading and writing skills if AI takes over these tasks, citing concerns from experts about the shift towards oral communication with less rich vocabulary.
Cobo elaborates on the implications of AI for education, noting that adolescents use AI more than adults and often prefer AI tutors for their accessibility and immediacy. He raises concerns about academic integrity, excessive trust in AI, and limited human interaction. He presents examples of AI tools for administrative tasks (like lesson planning with Magic School) and pedagogical uses (like AI tutors that ask questions or allow interaction with historical figures).
Cobo compares human cognition to GPT's capabilities, stating that while AI memorizes vast amounts of data, humans excel at connecting experiences. He advocates for fostering learning experiences over automatic responses, contrasting it with 'junk food' for the brain. He uses Bloom's taxonomy to visualize how AI can contribute to different levels of cognitive engagement, from low-level plagiarism to high-level cognitive contribution combined with AI.
He suggests new approaches to education, such as shifting from traditional written exams to oral exams, real-time assessments, or project-based assignments like podcasts. The focus should be on the process and journey of learning, not just the final product. He recommends that educators integrate AI transparently, guiding students on how to use specific technologies and documenting their learning trajectory.
Cobo stresses the need for educators to have time to experiment with AI, discuss its integration with colleagues, and distinguish between administrative and pedagogical uses. He emphasizes prioritizing human cognition, critical thinking, curiosity, and establishing guidelines for AI use. He shares advice from a Harvard professor: embrace AI to understand its benefits and risks, foster trust with students, and maintain a critical attitude.
He presents five practical AI tools: Poe (customizable chatbot for analyzing multiple texts), Gamma (AI for generating presentations), ChatPDF (for interacting with PDF documents), Chat to (a browser extension for summarizing and querying videos), and Turnitin (for grading multiple-choice exams). He encourages an open dialogue about these tools in the classroom and collaborative exploration among students and teachers.
Cobo outlines five challenges: developing AI policies, moving beyond instrumental use to foster critical thinking, providing resources for teachers to explore AI, understanding the dilemmas of AI use, and building community observatories for AI tools. He concludes by reminding the audience not to humanize AI, as machines commit errors, imitate but do not think, reflect opinions but do not hold them, and can cause misinformation. He ends with Frank Zappa's quote: 'The mind is like a parachute, it only works when it's open.'
Génesis Emmanuel Sierra asks about the challenges AI poses for schools without access to this technology, and how it can hinder skill development. Cobo acknowledges the concern about the digital divide, noting that AI can both amplify and reduce such gaps. He suggests that gradual change and open discussions within the teaching community are crucial for navigating this transition.
Maestro Saúl asks for recommendations to motivate teachers who are reluctant to embrace digital skills and AI. Cobo emphasizes that reluctance is a common reaction to new technology. He suggests that change happens best within the community of colleagues, through shared experiences and learning from those who are more knowledgeable. He concludes by reassuring that gradual adoption and community support are key to overcoming such hesitations.