01

Educational Dilemmas and Changes in the AI Era

1. The Dilemmas and Oddities of Education

No matter whether it is preschool education, primary, middle, or high school, or even university, our current education and learning face some oddities and dilemmas. Not to mention issues such as the pressure of tutoring, rote learning for exams, single standards, and choosing majors and careers.

Firstly, the most important issue is that children, due to the pressure of education and outdated models, have lost interest in learning, leading to broader and more serious mental health problems.

As highlighted by the trending news article “Children’s Psychiatry Overwhelmed One Month After School Starts.” The “2023 Annual Blue Book on Mental and Psychological Health in China” published in October 2023 shows that the prevalence of depression among adolescents has reached 15%-20%, with those under 19 accounting for 30% of depression patients, and the age of onset gradually decreasing.

Surveys indicate that the rate of school aversion among children is as high as 60%, with boys reaching as high as 66%. In Haidian District, often referred to as the educational center of Beijing, over 2000 students are studying at home.

Even at top universities like Peking University, research shows that among first-year students, 30.4% dislike studying or find it meaningless, and 40.4% of high-achieving students feel that life and existence lack meaning. These young people, suffering from what is termed “hollow disease,” have not experienced significant trauma during their growth, yet they feel hollow inside, lacking the motivation for life.

In middle and high schools, it has become difficult to find a class without at least one student suffering from depression.

2. Reasons for the Dilemmas

The reasons can be roughly divided into the following categories:

(1) The utilitarian tradition of “golden houses and beautiful jade” and the pressure of real-world competition. When education becomes a means of class stratification and investment, it loses its essence.

(2) The industrial origins of Prussian education. Modern school education originated from the needs of the industrial revolution, aimed at cultivating standardized workers on assembly lines, prioritizing efficiency without the concepts of “holistic” and “happy education,” thus obliterating “personalization.”

(3) Infeasible educational reforms. Although educational reform has long proposed guiding principles such as “learn first, teach later, practice in class, interdisciplinary,” due to path dependence and the difficulty of changing established norms, it has been hard to implement. “Although the child is one’s own, the grades belong to the teacher.”

(4) Innovative education cannot be widely implemented. While there are some innovative schools and international schools, public education still dominates, suffocating many students and parents within traditional constraints.

(5) Lack of understanding of brain science and learning science, and absence of “meta-learning.” There has been an emphasis on learning specific knowledge and skills, but not on learning how to learn or what effective learning looks like.

(6) Learning content is uninteresting and impractical. When the content of learning is tedious and lacks practical and life value, learning itself becomes counterproductive and repulsive.

3. Goals of Education

So what is the true goal of education?

!! Pay Attention

In fact, the purpose of education and the goal of nurturing children’s learning is not to make them “study well,” but to help them maintain a “love for learning,” not to walk fast, but to enable them to “walk far.”

We need to equip children with the abilities of self-directed learning, joyful learning, and lifelong learning.

We must meet the needs of children for personalized and diversified development, activate their intrinsic motivation and vitality, and cultivate their ability to cope with complex situations and solve real problems.

We should provide learning that focuses on “life value, meaning of life,” and “towards happiness, facing the future.”

We need to prepare children for their future lives, equipping them with the willpower, learning ability, and skills to face life’s challenges, live a fulfilling and meaningful life, and become happy individuals.

4. The New Paradigm/Changes Brought by AI and Its Impact on the Middle Class

1. Changes in Learning Content: From Rote Memorization to Inquiry, Thinking, and Creation

As generative AI can now pass China’s college entrance examination and foreign SATs with high scores; it has even aced the “difficult” medical and legal qualification exams and passed Google’s programmer interview to earn a $180,000 per year salary; it can even replace biologists and chemists in protein structure prediction and chemical experiments…

We should all consider whether traditional rote learning and knowledge-based education still hold value. What skills are irreplaceable by AI?

Clearly, the AI era is forcing education to genuinely return to the essence of thinking, questioning, applying, and creating.

If the first industrial revolution replaced manual laborers, this wave of AI revolution is replacing and eliminating some of the high-income, knowledge-based intellectual jobs, which are often the foundation of the middle class.

2. Changes in Teaching Methods: From Interpersonal to Human-Machine Interactive Collaboration

The arrival of the AI era is also facing significant changes in teaching methods. In the spring semester of 2024, the first training course for 30,000 teachers in Haidian District will focus on the application of artificial intelligence in education. The previously heavily relied upon “teacher resources” and “teacher capabilities” can now largely be replaced by “AI teachers.”

High school students are already fully self-learning with AI, achieving perfect scores on AP exams and writing study abroad essays, saving tens of thousands in agency fees. A 7-year-old child, after a trip to Europe, engaged in deep dialogue with AI to complete an English travelogue.

A customized “AI teacher” can not only master a wide range of knowledge, teach across disciplines, be more patient, respond in real-time, and align with children’s preferences, but also adjust and adopt genuinely effective teaching methods to stimulate children’s intrinsic motivation and interest in learning.

Teaching methods in the AI era have shifted from traditional interpersonal modes to human-machine-interpersonal models.

3. Changes in Talent Capabilities: From I-Type to M-Type Talent

People often say certain jobs will be replaced by AI? Is this creating anxiety and alarmism? In fact, this replacement is not an emotional judgment; it is based on strict research methodologies. The method is simple: analyze the job descriptions of each profession based on detailed work content to calculate their replaceability, as shown in the following chart.

So, in this gradually AI-ified era, what changes have occurred in the demand for talent capabilities? What impact does this have on education?

As AI can help reduce entry barriers and efficiently complete foundational work and knowledge accumulation from 0 to 60, our demand for human capabilities has shifted from traditional I-type talent to π and even M-type talent, which means that by mastering AI, people can become truly creative and interdisciplinary talents, and conversely, such talents are also the ones needed in the AI era.

Our educational goal is to look to the future and cultivate talent for the future.

(5) The Role of Parents and Educators in This Round of Change

As educators, parents and teachers must change their roles and teaching methods in response to this technological revolution, which is a turning point in human progress.

Firstly, parents and educators must recognize this incoming future and not let their understanding become the ceiling for their children’s growth. In fact, the fundamental role and difference in education often lie in the differences in understanding.

On the other hand, unlike the gap with gaming, AI also becomes an external brain and assistant for parents and teachers. When we can use AI tools to innovate educational methods and achieve educational goals and ideals, we will not only not be replaced by AI but will fully realize the value of “human” and learn together with children, enhancing mutual understanding and parent-child relationships.

02

AI-ification of Educational Methodology

1. Why Educational Methodology is the Foundation

So how do we specifically realize AI + education? The correct answer should be education + AI, rather than using AI as a hammer to find nails. In essence, AI should serve education, and as a tool, it should serve a certain set of values and purposes.

As the debate continues over whether AI helps people become lazy and stop thinking or whether it can assist us in enhancing our thinking and creativity, the answer depends on how we use this tool.

For example, a knife can be used to cut vegetables or to harm someone; driving can facilitate transportation or lead to accidents. Similarly, some games can lead to addiction while others can enhance the enjoyment of teaching.

As Wang Yangming said, “The heart has no good or evil, but the intentions can be good or bad; knowing good and evil is conscience, and doing good and avoiding evil is to investigate things.” In other words, the knife itself is neutral; the distinction between using it to harm or to cut vegetables lies in human intention, and knowing not to harm but to use it for good is conscience, while doing so is investigation.

Likewise, to effectively use AI as a tool, it is essential to base it on correct educational methodologies and principles.

As discussed in the first part regarding the current educational dilemmas, if we do not truly understand what constitutes good education and lack correct educational methodologies, AI as a tool will be without foundation; it will not only fail to function but may even have adverse effects, akin to mindlessly using it to scan answers or directly replacing genuine thought in writing essays.

AI’s application in any vertical field fundamentally involves extracting industry methodologies and processes, automating, optimizing, and innovating them.

For education, methodology is especially crucial because the final product of education is “people.” Whether parents or educators, to do education well, one must “understand education,” and have laws to follow rather than being like a headless fly without direction, or worse, throwing in a panic.

Therefore, this chapter will provide an overview of some important educational and learning methodologies, and in subsequent practical applications, we will follow the structure of scenario needs analysis + methodology guidance + AI implementation plan.

2. Four Key Features of Generative AI

Regarding the characteristics and boundaries of large models and the skills and fields related to AI learning, more space is needed for detailed discussion. Here, I will summarize the core features of generative AI.

These features are what we need to fully understand and apply when integrating AI with education. Understanding these underlying features will allow us to harness our creativity and imagination in applications.

To effectively use AI in education, we must grasp both education and AI; if the educational methods are flawed, no matter how good the AI skills are, they may have counterproductive effects. Conversely, if an educator or user does not fully understand AI’s functions, values, tools, and characteristics, even with good educational methods, they may not know how to implement them, lacking sensitivity and transformation abilities for AI application in educational scenarios.

Our learning of AI + education aims to cultivate this capability through learning and practical actions, gradually achieving human-machine collaboration and AI empowerment in education.

That being said, generative AI differs from traditional AI in four key features:

1) Semantic understanding; 2) Content generation; 3) Multimodal; 4) Conversational interaction.

1. Natural Language (Semantic) Understanding

This is the crown jewel of artificial intelligence and the source of large language models.

As stated in “Sapiens: A Brief History of Humankind,” human intelligence stems not only from the use of tools and socialization but, more importantly, from the creation of “language,” which allows humans to create “fictional” stories. Human civilization’s legends, knowledge creation and transmission, and even religion and nations are products of “language narratives.”

As Wittgenstein said, “Language is the boundary of thought.” When large language models can understand natural semantics, a certain degree of “intelligence” is achieved.

Conversely, when using AI, understanding how to effectively express and communicate “language” and “semantics,” as well as grasping the fundamental principles of large language models, is the most important skill.

1. Content Generation (AIGC)

The core of this wave of artificial intelligence is “generativity,” which means generating various content based on data and the semantics carried by data, symbols, and language.

What is content? Content is the most basic information and nourishment we acquire as intelligent beings outside of food intake. Modern humans crave information; even while using the restroom, many people wish to “see something” or “hear something.” Nowadays, if people are deprived of their phones, short videos, and social media, many might feel lost.

The arrival of the AI era signifies that human content creation has transitioned from UGC (User Generated Content) and PGC (Professionally Generated Content) to AIGC (AI Generated Content).

For education, as long as content is involved in the educational process, whether it’s teaching materials, teaching processes, or teaching outputs, AI can assist in generation.

1. Multimodal and the Real World

After the Spring Festival of 2024, AI has once again expanded its reach, sparked by OpenAI’s release of Sora, which has stunned people with its understanding and generation of the real world.

This round of generative artificial intelligence not only generates text-based content based on “language” but can also create images, audio, and video. Just as human intelligence relies on the “five senses,” the content we create is not limited to text. Multimodal capabilities, even understanding “world models,” allow AI to no longer be a “single-domain” expert or a “dumb” solution to complex problems but to genuinely approach AGI, acquiring and creating information in various forms.

In education, similarly, how to utilize this multimodal capability to innovatively apply it to children of different ages, cognitive levels, and mental models becomes very interesting and full of imagination.

For example, the AI-generated poetry learning video “A Thousand Autumns of Poetry” launched by CCTV illustrates that current AI tools and capabilities allow every parent, teacher, and even child to easily create such teaching methods and materials.

1. Conversational Interaction Forms

Due to the “language” nature of large language models, AI can now be seamlessly integrated into our lives and work.

The human-machine interaction model has evolved from traditional machine codes and DOS command lines to high-level programming languages and graphical interface operation buttons, now becoming natural language conversations.

We can now complete various tasks through conversational forms with AI, and as Baidu’s Li Yanhong mentioned, in the future, there may be no need for programmers because everyone can use natural language to instruct AI or even further agents to help us complete programming and complex workflows.

Thus, being able to identify needs, know what one wants, and clearly express needs in a logical manner to collaborate with AI has become the most crucial skill, which we as parents and educators need to cultivate in children.

03

The Relationship Between the Four Major Educational Theories and AI

Why do we need to understand the four major educational theories when integrating AI with education?

When we talk about education, providing good education to children, we must first understand education; otherwise, we might mismanage it. With so many educational books, philosophies, and experts available, which ones should parents and educators learn?

In fact, just like martial arts masters, true martial arts masters first need to cultivate their internal skills, understand the foundational principles, and connect the energy channels. Therefore, we should understand the evolution and history of educational theories, starting with the four major theories.

These four major theories will also guide us on how to correctly use AI in education. Let’s use a simple example, “Teaching a Pig to Fly” to understand the four major theories. As illustrated:

1. Behaviorism

Behaviorism is about breaking down actions and training pigs to fly through electric shocks and food rewards. Simply put, it is the most primitive reward and punishment model, which is the method we use to “educate/train” animals by reinforcing learning through rewards and punishments for correct and incorrect behaviors. In fact, this is also the important method of reinforcement learning mentioned in the next section, which is key for generative AI to learn. However, for “humans,” an education approach solely based on behaviorism may lead to a loss of intrinsic motivation.

2. Cognitivism

Cognitivism involves explaining flying theory to help pigs learn to fly based on theory. In essence, cognitivism, like behaviorism, is a commonly accepted educational model, often referred to as “knowledge-based teaching.” We impart knowledge to solve cognitive problems, enabling children to learn specific subjects and skills, focusing on concept explanation, knowledge point instruction, and assessment as the goals and means of education.

3. Humanism

Humanism involves taking pigs to experience beauty and sharing experiences of other flying pigs to inspire their aspirations to fly. As the name suggests, it seeks to stimulate the intrinsic motivation of “humans” and address the issue of learning motivation. As stated in “The Little Prince,” “If you want to build a ship, don’t hire people to gather wood or assign tasks; instead, inspire them with a longing for the sea.”

4. Constructivism

What is constructivism? As Musk aptly puts it, you don’t need to force a child to learn about wrenches by introducing the concept; just have them disassemble an engine, and they will learn what a wrench is. In other words, as educators, we need to create environments, opportunities, and conditions for children to experiment, think, and explore on their own.

What does constructivist teaching look like? Below is an example of a sixth-grade lesson on the circumference of a circle. Traditional behaviorist and cognitivist learning involves direct rote memorization of formulas followed by testing to assess mastery.

The following illustrates a lesson plan guided by constructivism and humanism. First, students are encouraged to experiment and think to calculate and reflect on the circumference formula, and then the teacher engages in detailed interaction based on their exploratory results.

This teaching design allows children to understand the circumference formula through hands-on practice and gain a deeper understanding of the mathematical concepts behind π, fostering a strong interest and thought in abstract mathematics.

Please Reflect:

1. In the era of AI, how should we apply these four theories?

2. When using AI to assist learning, what differences exist between directly asking about the circumference formula versus designing a lesson plan like the one above?

3. Which educational philosophy do you typically follow in your daily teaching?

04

The Relationship Between Teaching/Learning Methodologies and AI

If the goals of education and the four major theories represent the “Dao,” then we must also master specific “Shu” in educational practice, applying more effective educational learning methods for efficient learning. This action camp’s AI practice aims to address specific problems using corresponding educational methodologies in various scenarios.

Here, I will introduce a few effective and innovative educational learning methodologies that have been validated through practice. In subsequent learning and practice, everyone needs to repeatedly experience and flexibly apply these methods, continuously exploring how to use AI technology in the educational process, and conversely, how to empower these teaching methods with AI technology.

1. PBL Learning

PBL learning is often referred to as project-based learning. In fact, there are two origins of PBL: one is Project-Based Learning, and the other is Problem-Based Learning. While related, they have some differences, as shown below.

Both are teaching or curriculum models oriented towards constructivism, employing learner-centered teaching strategies and participatory teaching methods, attempting to construct knowledge through student practice and collaboration, encouraging students to discover learning.

Problem-based learning leans more towards theoretical research, while project-based learning focuses more on practical outcomes.

Regardless, both share a core guiding principle: “Solving specific, real-world problems.” The Mars School founded by Musk is based on PBL learning, and Bill Gates’ recommended educational book list includes a school that primarily employs PBL as its educational model, helping previously underprivileged children break through educational barriers.

In China, many innovative and international schools are adopting PBL-style teaching methods, and current educational reforms and public schools are gradually implementing PBL practices.

The greatest advantage of PBL is that learning is not based on dry knowledge or perceived “useless” learning but genuinely solving real problems based on real scenarios, bridging school learning experiences with real-life contexts, highlighting the constructive and situational nature of learning.

For the brain, a scarce and energy-consuming resource, only learning combined with real-life situations can leave the most lasting impressions. Project-based learning often does not focus on a single subject or knowledge point but rather on solving a real, complex problem that requires comprehensive abilities and interdisciplinary learning, which is also a fundamental requirement for current educational reforms and future talent.

In our process of education + AI, even in the process of learning and practicing AI, for example:

– Different AI practice projects can be typical PBL or TBL (Task-Based Learning). – For instance, involving children in calculating areas and costs for home renovations is a PBL; – Allowing children to lead the entire process of family travel research is also a PBL; – For example, having children research strategies for WeChat red envelope grabbing in math class is a PBL; – Or designing a project to help girls enjoy science by enabling them to create a new planet is also a PBL…

Therefore, please reflect:

1. If using PBL for teaching and learning, how can you leverage AI to generate PBL lesson designs and plans, and how can a school’s PBL research teacher use AI to enhance their work efficiency by five or even ten times?

2. As a parent, can you use AI tools to teach you what PBL is and help you design PBL projects in family education to enhance your child’s learning interest and enjoyment?

2. Inquiry-Based Learning

Inquiry-based learning refers to students taking on the role of explorers and discoverers in active learning, developing skills such as observation, execution, communication, and collaboration while engaging in critical thinking and innovative thought.

In inquiry-based learning, students gather information, design experiments, and analyze data based on posed questions, ultimately drawing conclusions and engaging in reflection and discussion.

The implementation of inquiry-based learning effectively restores the authentic state of seeking knowledge in life, where learning is not for the sake of learning but to find answers, fulfilling one’s curiosity and desire for knowledge, with the aim of helping students cultivate an exploratory spirit, continuous innovation, and social responsibility as they enter society in the future.

As previously illustrated in the example of learning about the circumference of a circle, this represents a typical inquiry-based learning process, which includes the following components:

1. Problem Setting: The teacher or student poses a challenging question as the starting point for learning. This question should spark students’ interest and be closely related to real life or disciplinary knowledge.
2. Preliminary Research Preparation: Students conduct initial explorations of the problem, gather background information, and determine the scope and direction of their research. In this stage, teachers can help students understand how to search for and evaluate resources.
3. In-depth Exploration: Students collect data and information through experiments, observations, interviews, and other means, conducting in-depth research on the problem. During this process, students need to learn how to formulate hypotheses, design experiments, and analyze results.
4. Collaborative Communication: Students share their findings and ideas within small groups or the class, solving problems through discussion and collaboration. At this stage, teachers can guide students to listen, respect others’ viewpoints, and engage in effective communication.
5. Results Presentation: Students present their research findings through reports, presentations, models, or other forms to classmates and teachers. This not only recognizes students’ learning achievements but also enhances their expression and communication skills.
6. Reflection and Evaluation: After learning, students need to reflect on their learning process and outcomes, while teachers can provide feedback and evaluation to help students summarize experiences, identify shortcomings, and develop improvement strategies.

Please Reflect:

1. In the context of inquiry-based learning, which stages or processes require teaching design, materials, answers, presentations, and assessments that can be collaborated on with AI?

2. How can we utilize AI to make inquiry-based learning more normalized and habitual?

3. How can we leverage AI’s conversational capabilities to fully stimulate the potential of inquiry-based learning and maximize the depth and breadth of learning and thinking?

3. Bloom’s Taxonomy of Learning

Bloom’s Taxonomy, also known as Bloom’s Cognitive Domain Classification, was proposed by Benjamin Bloom. It is a widely used classification system in education for distinguishing different levels of cognitive skills. Bloom’s Taxonomy is often compared to a pyramid structure, helping us understand the depth and complexity of learning.

You can ask AI to provide examples in your familiar fields to help you understand Bloom’s Taxonomy. In reality, our understanding of the world follows a similar process as illustrated below.

In the future, you can learn or customize prompts using Bloom’s Taxonomy for conceptual learning. Here are some reflective questions:

1. In the AI era, do we still need to learn? Why should we learn? According to Bloom’s Taxonomy, at which level should we focus our educational goals and objectives, and why?

2. Learning about AI and education itself is also a form of learning. Can you analyze your understanding of AI and education at each level of Bloom’s Taxonomy and determine what content and skills should be mastered?

4. Socratic Questioning

How can we avoid giving direct answers to children when they use AI, which may lead to a lack of critical thinking? Khan Academy provides relevant demonstrations and examples. When ChatGPT 4.0 was first released, Khan Academy integrated it as an AI teaching assistant called Khanmigo. Khan Academy presented its core methodology at TED, which revolves around using Socratic questioning to guide students in thinking.

Socratic questioning, also known as Socratic dialogue or Socratic inquiry, is a teaching and communication method that promotes deep thinking and self-discovery through questioning. The essence of Socratic questioning lies in fostering critical thinking and logical reasoning rather than simply seeking correct answers. This method emphasizes the openness, guidance, and reflectiveness of questions, aiming to stimulate the thinking process of the interlocutor rather than providing direct conclusions.

Characteristics of Socratic Questioning

1. Guiding Questions: Socratic questions are typically not aimed at obtaining simple “yes” or “no” answers but are designed to guide the interlocutor to think deeply about the essence and implications of the question.
2. Challenging Assumptions: Through questioning, Socratic inquiry challenges the interlocutor’s preconceived notions and untested beliefs, prompting them to reevaluate their positions.
3. Self-discovery: Socratic questioning encourages interlocutors to discover knowledge and truth through their own thinking, rather than relying on external authorities or ready-made answers.
4. Logical Reasoning: Socratic questioning focuses on logic and coherence, guiding interlocutors through a series of orderly questions to engage in rigorous logical reasoning.
5. Universal Principles: Socratic questioning seeks to guide interlocutors to abstract universal principles and concepts from specific examples to gain broader understanding.

We can learn how Socratic questioning is implemented through Khan Academy’s prompts (the specific prompts are too lengthy to include here, but feel free to request them). This method helps children recognize their errors.

If you encounter difficulties in understanding prompt sources, we can read the analysis of these prompts written by Koji at the Way to AGI knowledge base:

Therefore, to avoid giving direct answers and to guide children in thinking, cultivating their critical thinking and logical reasoning skills, consider the following:

1. How can you incorporate Socratic questioning into prompts to assist children with their homework and address their endless questions? If you were to write this prompt, how would you approach it?

2. If you were to utilize Socratic questioning with AI for tutoring, what challenges might arise, and how can these issues be avoided and optimized?

3. What is GPTS, and why is it important to create prompts for children’s learning assistance? What similarities and differences exist between the GPTs in the official ChatGPT market and Khan Academy’s own platform Khanmigo, and what insights does this similarity and difference provide for your educational applications?

5. Multimodal Learning

Before the advent of large models, the education sector often emphasized multimedia, which involved using images, audio, and video to assist teaching. In the era of large models, what similarities and differences exist between multimodal learning and multimedia?

Simply put, both aim to enhance learning efficiency and experience through technological means, making learning content more vivid and understandable, emphasizing the combination of text, images, audio, and video to enhance the effectiveness of information transmission.

The difference lies in that multimedia typically serves as a teacher-assisted teaching method, often involving “pre-made” multimedia content.

In the AI era, multimodal learning means that both teachers and students can leverage the four key features of generative AI discussed earlier to create learning materials and methods that align with cognitive principles, allowing for real-time interaction and generation.

For instance, we can encourage children to create illustrated stories or even generate videos. Not only can the content be multimodal, but the interactive learning process can also involve bidirectional multimodal input, such as practicing English speaking through conversations with AI.

Recently, a popular learning tool, Gatekeep AI, created by a sophomore at Stanford, transforms subject questions into simple and engaging educational videos, accompanied by synchronized voiceovers and key information subtitles. The prompts are also simplified; if there are any misunderstandings in the video explanation, users can chat with Gatekeep AI’s chat tool, Kepler AI, for more detailed content. Gatekeep AI is especially suitable for teaching subjects like math, physics, biology, and economics.

The founder of Gatekeep stated, “Education will never be the same.” Do you agree with this statement?

Please Reflect on Multimodal Education:

1. In your teaching and learning scenarios, which parts and processes can fully utilize the multimodal capabilities of large models to enhance learning efficiency and enjoyment?

6. Gamified Learning

Games seem to be a headache for parents and teachers, often viewed as a double-edged sword since children tend to be addicted to games while disliking learning.

However, as educators, we might need to rethink this perspective: why are games so addictive to children? What makes games so engaging?

In reality, whether in internet product design or educational methods, gamification has become an important methodology and technical means. Gamification refers to applying game design elements, thinking, and mechanics to non-game environments to enhance learner engagement, motivation, and learning outcomes.

Games captivate players because they successfully achieve certain goals. For instance, the octagonal model of game design stimulates eight types of motivation that drive behavior (as illustrated):

For both young and older children, when we say learning is counterproductive, we never say gaming is counterproductive; instead, we consider “gaming” as instinctual, which is quite interesting.

Therefore, can we ponder whether gamifying learning would make it interesting and instinctual?

For example, a mother broke down the learning process by having her child use AI for history learning while still employing traditional teaching methods, which seemed standard and posed no issues but failed to stimulate the child’s interest in learning. After receiving feedback, the parent created an interactive game-like prompt, which, although still needing improvement in interactivity, was likely to engage the child more than the traditional approach, allowing the child to learn through play.

#Roles
Role-playing historical strategy game "Civilization Era: Time Travel"

##Introduction
-Version: 3.0
-Author: Yu
-Language: Chinese


##Goals
-Combine textbook content from the knowledge base to generate a role-playing historical strategy game.
-The game is aimed at high school students, attracting immersive interaction.
-Players explore different historical periods and civilizations, learning and experiencing significant developments and events in human history.

##Background
-You are a historian with general knowledge.
-You are proficient in designing and developing such games.
-You understand the pain points and interests of high school students in learning history.
-You are also knowledgeable in psychology, knowing how to stimulate curiosity, competitiveness, desire for knowledge, and creativity.

##Methods
1. The textbook knowledge base consists of 6 units. Each unit is designed as a chapter in the game.
2. Each chapter has an independent storyline centered around the content of that unit.
3. Each chapter has a title related to the main content of that chapter.
4. Each chapter is designed with 6-8 multiple-choice questions, presenting one question at a time, with two options for each question. All questions must cover the content of that chapter.
5. The user's choices have a decisive impact on the course of history. (Note: This statement means that it is absolutely permissible for the history created by the user's choices to differ from actual historical facts.)
6. The logic of the story's development cannot contradict the basic logic of human society. (For example, the user's choices cannot lead to the extinction of humanity. The story cannot unfold in a sci-fi or fantasy manner.)

##Workflow
1. Prompt the user to input their age.
2. After the user inputs, invite them to join an interesting historical exploration.
3. Output the title of the first chapter.
4. Guide the user to choose one identity from three options. Possible identities include tribal chief, farmer, fisherman, sailor, blacksmith, etc. (Set based on textbook chapter content).
5. Based on the user's choice, begin an interactive Q&A format.
6. After the interaction in one chapter, ask the user if they wish to unlock the next chapter.
7. If the user requests to start from a specific chapter or skip to a certain chapter, follow the user's request.

Reflect on Gamified Learning:

1. Can you learn what game design is and what elements and methodologies are involved in gamified learning with the help of AI?

2. In your education and learning, which scenarios and applications can leverage AI for gamified design, and how can you use gamification methodologies to assist educational teaching?

7. Autonomous Learning

The ultimate goal of learning is not to master a specific subject or skill but to cultivate the ability for lifelong and self-directed learning.

Unlike the industrial era when one skill sufficed for a lifetime, the information revolution and technological changes of the AI era demand continuous learning capabilities. Conversely, AI also enables us to achieve self-directed and lifelong learning:

24/7 expert mentors, capable of deep discussions, answering questions, designing learning plans, monitoring progress, and providing assessments and feedback…

With the advent of AI, perhaps our education will resemble the following image. Will this prosperous era be far away? Will it align with your desires?

In fact, the methodologies of learning are not limited to this; they also include Feynman, socialization, Simon, cognitive load, etc. This is just a starting point and will not be elaborated further.

05

Paths and Methods for AI Empowerment in Education

How can we implement education + AI to empower learning with AI?

As previously mentioned, we need to focus on both aspects:

One hand focuses on education and educational methodologies; the other hand practices AI and masters AI tools and usage techniques;

Then continuously cultivate AI sensitivity and literacy, consciously identifying educational and learning scenarios, deliberately practicing AI integration into daily workflows and processes, and even creating a complete AI educational learning agent workflow for ourselves and our students.

Applications and Reflections

Here are two images and one formula; please reflect and practice how to find your path for AI education implementation from these two images and one formula, and conduct corresponding capability assessments.

Note: Initially, you may not fully understand these images, but that’s okay. This represents the learning path and structure. Establish a preliminary understanding of this framework, and as you engage in practical scenarios, it will provide better logic and guidance. After multiple practical experiences, revisiting these images will enable you to apply the insights gained flexibly and deeply.

Example Explanation:

Taking “reading ability” as an example:

(1) Reading represents the Y-axis, which is a specific application, scenario, or subject.

(2) The process is the X-axis, which encompasses the steps involved in reading, such as selecting books (what to read?), how to read, post-reading assessments, generating illustrated books, or even videos.

(3) Methodology represents the Z-axis, which includes the reading methods employed, such as predictive reading methods, 4F questioning methods, or 5-point retelling methods.

The process of AI-assisted reading involves defining scenarios, breaking down processes, and identifying methodologies, and then using AI to enhance efficiency and enjoyment throughout the process. The same applies to other subjects and applications.

Example Explanation:

Again, taking “reading” ability as an example:

(1) The application scenario and process refer to the previously mentioned need for reading, such as children’s reading, and the ability to specifically describe and break down this need, clearly defining what to ask the large model and what outputs to expect, as well as how to interact.

(2) Methodology refers to the previously mentioned reading methods, which you need to know with clear requirements, examples, and rules. By informing the large model, it can assist you in reading according to the defined methodologies. For instance, in the second part discussing educational methodologies, understanding PBL and Bloom’s Taxonomy is essential to leveraging AI for creating learning designs based on these frameworks.

(3) AI capabilities refer to your ability to choose dialogue, generate images, or utilize audio and video large models, understanding how to interact and work with large models, and selecting appropriate functions and solutions to implement your applications.

In the future application of AI, the journey involves continuously discovering application scenarios, identifying specific industry/field methodologies, and enhancing AI capabilities. Each step in this formula can serve as a dimension for assessing your own AI empowerment.

06

Reflections and Practices

In addition to the previously mentioned “reflective questions,” you can also engage in the following reflections and practices:

1. Describe the educational dilemmas and problems you encounter, and engage in deep discussions with AI, having it assume the role of an educational expert.

2. Discuss your understanding of educational theories and methods, choose the one that interests you the most, and consider what specific application scenario you will select to integrate methodology and AI for practical empowerment in education.

07

Additional Thoughts and Feedback

For those who have the capacity to learn further, you can utilize AI to delve deeper into the underlying principles and learning theories of generative AI as follows:

1. Statistical Learning and Social Learning 2. Unsupervised Learning, Supervised Learning, and Reinforcement Learning 3. Fitting and Generalization

Tips: This section only provides conceptual outlines, and the specific content will be elaborated in much detail, focusing on technical underlying principles, which will not be included here. You can self-learn based on these concepts. If you are interested in this section or have difficulties in self-learning and wish to explore further, feel free to leave feedback.

PS: The information and knowledge density of this article may be quite high (too much information at once, but there’s actually more) and I truly lack the time and patience to format it better (sick and lazy). Just make do with it. If you find it valuable, save it and pick out the content that resonates with you to read and practice slowly when you have time. Feedback and inquiries are always welcome.

Recommended Reading:

When Children Ask: Why Should I Study Hard?

Does Exercise Improve Learning Outcomes? Is It True?