Google unveiled the generative artificial intelligence models LearnLM designed for educational purposes during the Google I/O 2024 developers conference.
Google stated that it developed the LearnLM models through collaboration between the research department at DeepMind and Google’s research institution “Google Research,” built on the well-known Gemini models with the aim of providing innovative and personalized learning experiences.
LearnLM features include the ability to educate students through dialogues on diverse topics, and Google incorporates them into its applications such as YouTube, Gemini, search apps, and Classroom.
In a speech during the conference, James Manika, Head of Research, Technology, and Society at “Google Research,” explained that LearnLM models aim to transform learning experiences into exciting personal experiences.
Google collaborates with teachers through a pilot program in Google Classroom to enhance and simplify lesson preparation, and provide educational materials specifically designed for students’ needs, using LearnLM models.
In addition, LearnLM models offer the “Circle to Search” service on Android devices to help students solve basic math and physics problems, allowing them to ask questions and get answers while watching educational videos on YouTube.
LearnLM models will enable users to create customized chatbots within Gemini applications that offer study guidance and training activities tailored to each learner’s preferences. For example, users can create customized chatbots with educational personalities.
Google also plans to establish partnerships with educational institutions such as Columbia Teachers College, Arizona State University, New York Tech University, and Khan Academy, aiming to expand the scope of LearnLM in products different from its own.
However, LearnLM models currently face some challenges, such as difficulty in handling students’ correct answers, and adjusting to the hallucinations and misleading information they may provide.
