Generative AI is often demonized as something that makes people stupid. But new tools from AI companies are doing just the opposite, Satyen K. Bordoloi has discovered.
The 1990s were an exciting time to be a teenager. New research and learning advances have taught us children how to learn better. One of the key discoveries was the use of mind maps. Mind maps were used to create a visual map of a topic to better understand and remember it. The idea was that learning would be better and faster by tapping into the visual part of the mind. Thirty years later, generative AI companies seem to be doing more than that and taking the idea to a whole new level.
So far, we’ve been taught another lesson: when we think about the impact of artificial intelligence on education, dystopian images come to mind. Scenes of students cheating in class, passively receiving undigested information, and their critical thinking skills atrophying as machines do the intellectual heavy lifting. This narrative of technological dependence and cognitive decline is a common dystopia that everyone flaunts.
but, Google has launched a new initiative called “.Learn your way“” that casts doubt on this concept.. Their recent research paper, “Towards AI-enhanced textbooks,” offers a dramatically different vision of education. Generative AI does not replace thinking. Enhance learning by creating educational experiences that are customized to the needs of individual students, rather than the collective needs of today’s education.
Their idea is simple, but not revolutionary. Because what they are asking is what would happen if textbooks could adapt to individual learners, rather than forcing all learners to adapt to a static textbook. Their team has developed a system that transforms traditional educational materials into: A dynamic, multimodal learning experience Tailored to each student’s grade level, interests, and preferred learning method. This is not just an incremental improvement in educational technology, but also a fundamental rethinking of the education system’s most important and enduring tool: the textbook.
break the uniform mold
Some traditional textbooks have what Google researchers say are “fundamental limitations.” That is, textbooks are a uniform medium. In a classroom of 30 students with different backgrounds, interests, and learning styles, you can’t expect one static textbook to serve everyone equally. Adaptation of teaching materials While addressing individual needs has always been a high ideal, human educators naturally face practical constraints that make true personalization at scale nearly impossible.
Learn Your Way addresses this challenge by implementing the following: 2-step AI generation process. First personalize content based on input, then transform it into multiple representations. The system begins by asking all learners two simple questions. It’s grade level and personal interests in categories like sports, music, games, etc. These attributes serve as the basis for the modular development of the system.
Personalization engine
of the system Grade level adaptation It uses algorithms to adjust the complexity of the text while preserving factual accuracy and maintaining the conceptual integrity of the problem. This “re-leveling” process Flesch-Kincaid grade level calculation formulaThis means that students encounter the content to be learned in the complexity of their own language, and this is done without reducing the conceptual content.
Equally innovative is interest-based personalization, which rewrites conceptual examples based on students’ expressed interests. For example, Newton’s third law of motion can be explained to sports fans through the example of cricket, or to students who love art through artistic techniques. This means that an individual’s existing knowledge serves as the basis for subsequent learning and performance, as stated in the research paper.
Advantages of multiple representations
The first step is to personalize the core text, but then Learn Your Way generates multiple ways to present the material. This includes narrated slides that cover core content in a concise and engaging way, audio-graphic lesson vials that simulate conversations between teacher and student avatars, interactive mind maps with hierarchical visual representations that allow students to explore relationships between different concepts, immersive text that is personalized text enriched with embedded questions, and personalized examples such as using cricket to explain Newton’s laws and other interactive elements.
The reasons for this multimodal approach are the same as for mind maps mentioned at the beginning of this article. Presenting the same concept in different ways stimulates different parts of the mind and enhances learning. As learners move between different formats, concepts are reinforced through different routes.
The broader education AI landscape
This isn’t the only thing Google is doing. Some are working hard to leverage artificial intelligence to enhance human learning. Take Khan Academy’s Khanmigo for example. This is a tutoring assistant that uses Socratic questions to guide students through problem solving, rather than providing direct answers. The aim is to develop students’ problem-solving skills and critical thinking through interactive dialogue.
Duolingo uses AI to create adaptive language lessons that adjust difficulty based on your performance. It also accelerates language acquisition by generating contextual examples and providing instant personalized feedback. Similarly, Quizlet has an AI system that generates personalized study plans, offers a variety of assessment formats, and identifies knowledge gaps to transform passive memorization into adaptive learning.
Overturning the “insensitivity” narrative
There is a big fear that AI is desensitizing us. That would make us all intellectually lazy. That’s true to some extent. But tools like this show that this doesn’t have to be the case. AI will act as our intellectual wings, not our intellectual shackles.
Think about it. In the 1990s, I was forced to memorize text written in black and white on paper. Let me tell you, this was not only undesirable, but extremely tedious and tedious. By gamifying learning with AI, color, images, and video, we first make learning fun. Second, by personalizing education to each learner’s abilities, we ensure that no child or learner is left behind. The randomized controlled trials cited in this study also demonstrate this “learning improvement” with AI-enhanced approaches, providing empirical evidence that well-designed AI educational tools can improve learning outcomes, rather than detract from them.
Therefore, we are in a future where learning technology is as exciting as entertainment was decades ago. In the future, generic curricula may be abandoned in favor of fully individualized curricula that adapt not only to a child’s knowledge but also to their learning style, learning pace, and even emotions. Rather than being the nuisance it is today, AI could become the collaborator and assistant that teachers desperately need, providing detailed insights into student understanding and recommending targeted interventions.
And above all, advanced AI can help students see connections between different subjects and connect mathematics and historical events or politics and artistic expression, thereby sparking creativity in ways not seen in previous generations. Even better, learning never stops. With AI, we can truly do what we should be doing anyway: becoming lifelong learners.
Therefore, the promise and potential of AI lies not in replacing human teachers or automating learning, but in addressing the scalability of personalized education. Learn Your Way is therefore not artificial intelligence, but augmented intelligence with systems designed to augment and enhance our cognitive abilities rather than replace them.
In this respect, the future of AI may not be about humans becoming more machine-like. Rather, it may be that machines are finally able to respond to the diverse and beautiful complexity of the human mind.