Future classroom environments

From: redpointai

Salman Khan, founder and CEO of Khan Academy, envisions a future where classrooms evolve dramatically, leveraging AI to enhance learning and teaching experiences [00:01:46]. He believes that schools could be among the first sectors to see mainstream adoption of AI for productivity [00:05:05].

Vision of Future Classrooms

Khan anticipates that in 20 years, classrooms will resemble what great classrooms already are today: environments where students are actively engaged, interacting, problem-solving, and working in groups, rather than passively listening to lectures [00:02:12].

A significant shift will involve immersive experiences:

• Students will be able to “immersively go into simulations” [00:00:20].
• They will be part of virtual worlds [00:00:22].
• It will be possible to “go back to ancient Rome” [00:00:23], akin to a “magic school bus ride” [00:00:25] [00:04:26].
• This will be facilitated by virtual reality (VR) and augmented reality (AR) becoming mainstream, combined with generative AI [00:04:15].

Khan also suggests that AI won’t necessarily require staring at a screen, but can become “ambient,” observing the classroom and understanding interactions [00:03:50].

Role of AI for Teachers

AI is expected to provide teachers with “superpowers,” augmenting their existing capabilities [00:03:11]:

• Time Savings: Teachers will gain time back from tasks like lesson planning, grading, and writing progress reports [00:03:18] [00:02:57]. Some school districts are already reporting saving teachers at least 5 hours per week [00:21:03].
• Insights and Management: AI can offer better insights into student progress and help manage classroom dynamics [00:03:24].
• Lesson Enhancement: Tools like Khan Academy’s partnership with Blakit allow KIGO to generate game-based questions (similar to Kahoot) in minutes, a task that previously took teachers half an hour or more [00:14:36].
• Student Accountability: AI can help teachers assign targeted tutoring sessions and hold students accountable for completing them, significantly driving student engagement [00:34:05].
• Recruitment and Retention: Schools are using the integration of AI as a tool for recruiting and retaining teachers [00:21:06].

“A teacher who has that kind of habit for me, it’s always been the case, who’s using Khan Academy to make assignments, hold the students accountable, look at the data, and then keep doing that, they they seem to get very very good results.” [00:15:05]

Role of AI for Students

Khan Academy launched KIGO, an AI-powered tutoring assistant and teaching assistant, which has been deployed to over 1.4 million students and teachers [00:00:37] [00:06:30]. Key aspects of student AI interaction include:

• Personalized Tutoring: Khan Academy’s long-standing goal is to replicate the personalization of a good tutor [00:05:36]. KIGO aims to approximate this through its Socratic method, focusing on good pedagogy and safety [00:06:40].
• Proactive AI: The next phase for Khan Academy is a “much more proactive AI” that greets students, reminds them of tasks, and offers help, moving beyond a passive “blank screen problem” where students don’t know how to prompt the AI [00:07:33] [00:08:19].
• Learning from Errors: Surprisingly, students have engaged with AI models by explaining their reasoning, leading the models to iterate and even self-correct mistakes, fostering a unique learning dynamic [00:11:10].
• Reduced Hallucinations and Errors: While models like GPT-4 initially had issues, Khan Academy has focused significant resources on improving AI accuracy, particularly in math, bringing the error rate down to about 2% when anchored to Khan Academy content [00:10:03] [00:12:46]. This is arguably better than many human tutors [00:13:45].
• Ethical Writing Coach: The “writing coach” tool allows teachers to assign writing with AI support, which acts as an ethical guide. It tracks the student’s process and can detect if content is simply copied and pasted from another AI, thus undermining various forms of cheating [00:15:39].
• Simulation and Exploration: Students can engage in AI simulations of literary or historical figures (e.g., Jay Gatsby, Harriet Tubman, George Washington) to ask questions and drive conversations [00:11:51] [00:15:25]. More curious students use it to explore complex ideas in subjects like advanced math [00:12:35].

Enhancing Engagement and Pedagogy

Engagement is highlighted as the primary challenge in education, more so than efficacy [00:10:09]. AI can significantly drive this by:

• Making Learning Fun: Creating AI-powered games, scavenger hunts, or escape rooms that require solving cognitively challenging problems within a broader engaging context [00:27:22].
• Personalized Pace: For curious and already engaged students, AI is a “dream come true” because they can ask any question and proceed at their own pace [00:31:55].
• Structured Learning: While some might advocate for simply giving students AI and letting them learn, Khan emphasizes the importance of structured content to guide the learning journey, especially for those who “don’t know what they don’t know” [00:36:23].

Overcoming Challenges and Future Capabilities

Building AI tools for classrooms requires significant work beyond just a “thin prompting layer” on top of models [00:21:41]. Challenges and desired capabilities include:

• Safety and Moderation: Implementing robust safety and moderation features, initially conservative but now more refined to avoid false positives [00:21:48].
• Math Accuracy: Continuous effort to improve math accuracy, particularly evaluation errors where the AI might incorrectly assess a student’s answer [00:22:07].
• User Interface: Creating intuitive and natural user interfaces [00:22:35].
• Memory: Enhanced memory for AI models to retain context about individual students over time [00:23:58].
• Advanced Voice: Integrating advanced voice capabilities into platforms for more natural human-AI interaction [00:24:26].
• Multimodal Capabilities: The ability for AI to “see” and understand a student’s handwritten work (e.g., on a tablet) and provide natural, conversational feedback [00:25:33].

Skills for the Future Workforce

Khan believes that fundamental skills like strong critical thinking, writing, reading, and math will remain crucial [00:42:59]. However, a key skill for the future workforce will be entrepreneurship [00:43:26]. This refers to the ability to take existing resources—including new AI tools—and combine them in novel ways to create new value [00:43:30]. People who can leverage AI to accelerate their work, even if it’s just getting 80% of the way there, will be highly valuable [00:44:03].

“I think the thing that’s really going to be I think it I think economists talk about entrepreneurship as a factor of production, right? This ability to put resources that already exist but put them in new permutations to create value that didn’t exist in the world before.” [00:43:21]

Khan also highlights the future of coding where AI can generate entire applications, moving towards a “vibe coding” approach where users prompt the AI to build software [00:46:33]. This represents a massive acceleration in productivity [00:18:28].