Explore, Experience, Express

Note: Feynman Technique can be taken as a macro workflow in Knowledge Engine, which can contain many smaller regular workflows.

According to Feynman Technique：
1. Pick a topic you want to understand and start studying it. Write down everything you know about the topic on a notebook page, and add to that page every time you learn something new about it.
2. Pretend to teach your topic to others, especially to youngesters who are not familar with the topic . Make sure you’re able to explain the topic in simple terms.
3. Go back to the books when you get stuck. The gaps in your knowledge should be obvious. Revisit problem areas until you can explain the topic fully.
4. Simplify and use analogies. Repeat the process while simplifying your language and connecting facts with analogies to help strengthen your understanding.

How can Knowledge Engine(KE) be used in this process? We understand that learning is to build up a model of a domain, and Feynman Technique is a technique for building such a model. KE, based on Significant Experiences (SEs) is sufficient in building up a mature model of a domain knowledge. Below we outline how KE can be used in the process of Feynman Technique:

1. pick a domain from KE (there will be domain tags in kE for people to choose from. TODO:), write down things you know about this domain in snippets, or search and find your relevant experiences from your past Significant Experiences (SEs). You can even try to build up some frames to represent your current total understanding of the domain, and in the process, you can probably identify a few key tags of this domain’s knowledge, which you can use to collect your SEs in your learning of this domain
2. You can use your frame to teach this domain to other people.
3. After realizing the gaps, read books, do some appropriate projects or do any activities that can help you fill the gaps. Revisit your frames and tags and keep udpating them to better reflect your current understanding of the model.
4. Keep enriching your frames and simplifying them at the same time. And repeat the process.

In the above process, you can see that KE can be used to support Feynman Technique. Furthermore, all the SEs are collected along the way to build up a rich model and speed up the thinking in this learning process. We can also expand Feynman Technique beyond teaching and books reading to include projects or other types of activities. Thus to discover your gaps and fill in those gaps, you don’t just rely on teaching and book reading, but also by other activites like working on projects, touring meuseums, travelling, watching movies, playing, or whatever you want to do in your real life.

Understanding of the model can be layer by layer. Sometimes it is difficult to obtain sufficient resources to build up the model deeply and completely, and you somehow still need to use it in some way in your work. So an understanding model at some level is justified at times. With KE, it allows such building up of models in flexible ways. So this is another aspect that we can expand on Feynman Technique to reach a more complete picture of learning.

If we understand Feynman Technique as centered around building a model, we know that learning activities can be very flexible. For example, you don’t have to read through a book from beginning to end. You can jump around or take a glimpse first to see if you can improve your model (of understanding) very quickly. (This is why life centers matter, because with life centers, you can approach it and interact with it in very flexible ways. Here we treat the model of understanding as a life center.) And everyone’s methods in how to build up the model quickly for oneself can be drastically different.

最近因为参加公司的代码节活动有更多的机会去熟悉和学习Paracraft编程。在这个过程中我惊奇的发现，像在线下导引课上课的那些小孩子一样，我会去做些看似没有意义的事情，比如不断的破坏方块，发现草地下面是什么样的方块，而再往下会是空的，即所谓的世界其实是悬空的，或者尝试钻进打的比较长的地洞后看自己如何能够走出去。

在周末的线下导引课时，很多孩子会有类似的行为。例如把角色的速度设置的很高或很低，然后看角色动起来什么样子。或者搭建很高很高的楼从上面跳下来等等。在大多数大人看来似乎是没有意义的瞎玩而不是有效的学习。我发现很多家长会很及时的进行阻止。我虽然一直相信学习首先是种探索活动，很多时候也可能会感觉孩子们的这样的活动比较无聊，不是在进行有效的有意义的学习，可能是沉迷于低级别的“玩”。

然而当我不知道是否有意识的在模仿这些孩子们的行为时，我发现这些其实看似没有意义的“瞎玩”中，却正是让我去全面的了解Paracraft这个软件。没有预先的规划，就是跟着直觉的好奇，想看看在Paracraft世界里随意捣鼓会是什么样子。像那些孩子把角色的速度调到很高的时候（如1000），角色跑起来世界的震动可能告诉了他们一些什么。像我把世界打穿了以后让我得以从Paracraft软件设计者的角度去思考他这款软件是如何设计的。这些维度很难预先规划，更很难被所谓的课程所覆盖。但在孩子似的“瞎玩”中却可以接触到这许多的维度。

鸟山明创造了全球最受欢迎和商业收入最高的漫画《七龙珠》。鸟山明的成长历程中就有很强的这样“瞎玩”的元素，他对生活体验的丰富度，使得他可以创造出七龙珠这样充满生活趣味和幽默的漫画。关于鸟山明和七龙珠的创作，可以看这个知识包。蔡志忠也有类似的成长经历。

我想，这可能是未来的人类社会最迫切需要的能力吧，也是现有的人工智能研究最难实现的挑战。换句话说，具备这样素质的人是很难被人工智能取代的。话说回来，有150年历史的现代教育，其设计的目标本来就是培养大量的可以在流水线上和机器融为一体的熟练工人，也就是说培养“机器人”，现下的中国在这个方向上更是走到了极端。当人工智能能够实现更高的自动化的时候，这种教育制度下培养出来的人自然就会被真正的机器人取代掉。这就是所谓人工智能危机的本质。而这些孩子尚未丧失的看似“瞎玩”的探索式学习，正是抵御人工智能危机的最佳途径。

我一直说我很庆幸这一生选择了两件事，一个就是教育，让我得以更深入的了解我自己。在Paracraft课程里和这些孩子的接触，就让我对学习中的探索有了更深的认识。如果不是在讲授领域有丰富学习经验的老师，可能会很难认可学生们的探索性活动，会期望学生更多的按照自己的安排来学习。如果让我教小孩编程或者踢足球，从我个人的丰富学习经验上，我会很非常鼓励他们自己的探索性活动，比如自己拿着足球想怎么玩就怎么玩。记得我在小区的草坪上练练颠球的时候有时会有小区里的小孩想跟我一起踢，但我发现上了10岁的小孩比较普遍的已经丧失了探索的快乐，他们很希望我告诉他们应该怎样传球，怎样的动作才是正确的，然后很努力要做“正确”，程式化非常明显。我在游戏方面经验很少，当我自己有了Paracraft的学习经历后，我方能充分认识到孩子们的那种探索性活动。可见教师自己的丰富学习经验的重要性。

人的成长过程，有时是习得了一些东西，有时是失去了一些东西。习得的是好的，还是失去的是好的，需要我们的仔细观察和分析思考。和孩子们的接触，做教育工作，就是提供了这样的宝贵的机会。

我想新时代的教育老师的概念会发生转变，老师不是无所不知，而是有重要体验和探索能力。

我们常说教师应该传授知识，那么什么是知识？知识的一个属性就是有自己个人的理解，也即可以用自己的话来表达，其基础就是个人的重要体验。当我听到很多人不断重复一些人云亦云的话时，我知道他们其实不是在表达知识。当我听到他们偶尔表达出个人的重要体验时，可以感受到他们明显的兴奋感，这时才是在传递重要体验和基于重要体验的知识。所以说，在知识引擎里，我们认为知识的基本单元是重要体验，就好像生物体的基本单元是细胞一样。

学习能力里最重要的其实是探索能力，即发现新知识的能力。我们认为学习就是自学，其内含的意思就是每个人其实都是在发现知识的过程中，这个发现建立新知识的过程就是学习。而这种发现就是基于重要体验的，知识的构建也是基于重要体验的。教师就是能够比较好的引导学习者对某领域进行探索的人。

“看到小孩学习编程游戏和动画的各种行为，感觉彷佛是看到了生命之初的样子，观察他们就是在观察生命原始的状态，看生命如何生长。同时我心里升起的是尊重，对生命的尊重。如果不尊重孩子，就是不尊重自己。”

这是今天我在知识引擎里记录下来的一条重要体验，来自于今天上午在公司的Paracraft线下导引课。具体来说，当时看到小孩学习编程游戏和动画的各种行为，看到孩子在玩，在搭建东西，甚至有很好的设计能力，可以搭出很漂亮的建筑，或者看他们到处摸索软件的界面，到处在点（我们鼓励家长和孩子一起学习，但观察家长，家长们很少会像孩子们一样在软件上到处点。家长比较习惯跟着老师的节奏，或者需要专门的教材）。。。。。。看到这些，感觉彷佛是看到了生命之初的样子，观察他们就是在观察生命原始的状态，看生命如何生长，很多成人们已经失去的东西，看到生命的力量，心里突然升起对生命的尊敬。然后想到自己也是生命。如果不尊重孩子，就是不尊重自己。

我想这就是生命与生命之间的共鸣吧。我一直很庆幸自己选择了教育作为自己一生的事业，做教育给了自己非常难得的更深的了解自己的机会。当看到不同年龄阶段的朋友他们的学习状况，放佛是看到自己的昨天或者明天。尤其是很多早期的记忆，如果不是看到这些小孩子，可能会很难再去想起自己小时候经历的各种事情。但是当你看到那些小孩的行为的时候，很自然的很多记忆就涌上了脑海，能够看到自己在生命之初的样子。也许随着成长，经历和认识变得丰富许多，但作为人的最根本的一些东西却似乎是永恒的命题。所以做教育是非常幸运的，有机会这样充分全面的了解自己，了解人性，了解生命的各种可能。

我人生的另一个很庆幸的事情，是选择了软件编程。软件编程帮助我认识整个世界，也更加认识自己。这个话题，以后有机会再讲。

Learning can happen in real life. Learning can also happen in a specially designed learning environment.

No matter whether it is Real Life Learning or artificial learning environment, as for learning, it is essentially about what activities to do and what experiences to generate and how to work with those experiences. Activities (including projects as in project-based learning) and experiences are the life centers in such learning environments (see Life Framework). Regarding activities, it is about interests and playing. Accumulating and reflecting on experiences generated during the playing is what learning is. (People should be able to learn while playing.) Teaching is to provide the map of experiences and immediate feedback to learner-generated experiences. If no teacher, experienced learners can provide part of that function.

Furthermore, being able to choose and navigate freely in those life centers according to one’s own feels are what constitute self-learning. Educators’ responsibility is to provide rich choices and to build up strong life centers that will enable rich interaction and thus free navigation. In other words, Educators’ responsibility is to build up a better playground.

Below we outline basic elements we think can be of a learning environment to make it a Selflearning Platform according to Selflearning Framework.

• Interest

Learning should be driven by intrinsic needs and goals, especially it should be driven by one’s own interests. As learning in real life usually follows this principle naturally, learning in a specially designed environment may not be the case. A Selflearning Platform should be based on learners’ intrinsic needs/goals instead of external ones and learning should be driven by learners’ own interests.

• First, play with it and know a lot of facts or phenomena

Learning in real life naturally starts with first knowing a lot of facts and phenomena of the subject as those are what form one’s interest in the subject. In a specially designed learning environment, learners should be exposed to a large number of facts and phenomena before starting systemic and theoretic learning on it, e.g. learners should first play with the subject sufficiently.

• Build up personal experiences, repeating the experiences of the master

Learning is to accumulate Significant Experiences in various depth and aspects within the space of the field and to reflect on them in order to achieve a coherent understanding of the whole. Masters of the field usually have gone through this process in their years of practice and learning. In a learning environment, the masters can provide their maps of experiences in the field, which learners can view and choose to build up their own relevant experiences. It is also possible for the masters to design the activities/projects for the learners to work on in order to gain relevant experiences. Masters can even provide feedback to specific experiences generated by the learner.

• Review and reflect on experiences, share and discuss

Significant Experiences gained in the activities need to be reviewed and reflected on periodically so a coherent understanding of the whole can be achieved. Social sharing of those experiences or knowledge built with them can be supported and discussions can be formed.

• Make a change to improve things or improve the environment

Being able to make changes to things or environments and improve them is a peacebuilding process in one’s mind. It gives one confidence and calmness. Deep inside everyone lives a craftsman or artist. Furthermore, good learning environments should allow learners to change the learning environment or the learning process itself (or every learner has one’s own learning process). It should also be possible for the learners to contribute the learning material for the subject.

• Quick and immediate feedback

Learning needs quick and immediate feedback. Good environments should be able to provide quick and rich feedback.

• Share, the social aspect

Although learning needs individual work and focus, social learning stimulates one’s interests and provides feedback.

• Rich activities

A good playground for learning should have rich activities for the learners to choose from so it is always easy for learners to find appropriate activities to work on to keep a continuously flat learning curve. Activities are not limited to projects. For example, student-led course-teaching can be a kind of activities. Judiciary committee meeting or school meeting of Sudbury Valley School is also a kind of activities.

Other than those artificial learning environments, real life is naturally a Selflearning Platform. Elements above can be digitized with software and internet to make Real Life Learning a strong Selflearning Platform.

We envision the future learning will be a mix of Real Life Learning and many artificial learning environments.

The environment can be online or having a physical locality, or mixed. It can be a complete and comprehensive environment such as Sudbury Model schools or Summerhill schools, or it can be an environment for learning one specific skill such as software programming or biology.

关于费曼学习法，似乎并没有“官方”的或者“权威”的表述。下面翻译了网上找到的比较具有代表性的定义，用我们自己的话，融入我们对学习的理解：

1. 选择一个想要去学习和理解的领域。首先在一个笔记本上写下你对这个领域已知的所有的东西，可以尝试你是否当下就能够达到对这个领域的一个理解模型（如果所知实在有限，不能立即达到一定深度的理解模型也没有关系）。并且每当你关于这个领域学到了新的东西的时候，就添加到那个笔记本上去；
2. 尝试去教别人这个领域，尤其是教低龄的小孩子。努力用简单的话语去阐释这个领域；
3. 返工。当你发现你不能够顺畅的阐释这个领域的时候，你应该能够意识到你知识结构里的缺漏，回到你笔记本上的那个模型去继续学习（通过书籍，互联网，做项目等等各种方式）；
4. 回顾和精简。回顾你对这个领域的整体的理解，做进一步的简化，并能够和更多的事实现象和体验连接起来，可以轻松的举例说明你的模型，可以很容易的识别新的现象和问题。不断重复上述过程，不断发现缺漏，弥补缺漏，做整体的回顾和整理，教他人，完善你的理解模型。

文末附费曼学习的英文原文。

我们可以理解费曼学习法是一种系统化思维的学习方法。系统化思维，就是对事物的观察和理解抱有整体观，不孤立片面的看事物，会去主动寻找不同事物间的联系。这篇文章里，费曼发现的问题，如物理研究生并没有在学物理，学希腊文的学生并没有在学希腊文，就是缺乏系统化思维，孤立片面的看待事物。而系统化思维恐怕是学习中的核心能力，是很多人所缺乏的。很多人缺少清晰表达事情的能力，或者不知道把自己会做的事情教给其他人，都是缺乏系统思维的能力。而善于表达的人一般系统化思维能力比较强，并且善于把自己学到的知识传递给他人。如果你对系统化思维还不太理解，可以尝试一个小练习，比如列举你现在所知道的所有的关于生物的知识，或者列举所有的你知道的关于宇宙的知识。

费曼不光指出了问题，还提供了锻炼系统思维的学习方法，即费曼学习法，其核心即建立某领域的理解模型。有了这个核心，学习就是一个非常灵活自由的事情，不需要一定是按部就班的进行，你可以一上来就尝试建模，试图对该领域建立一定深度的整体理解，发现有不足的地方，再有针对性的去学习。并且以建模为中心，你可以在很散的日常活动中碎片化的去学习，只要你很清楚你是在哪几个领域建模，你当前的领域模型是怎么样的，你可以不断的把在这些领域积累的新的知识（即我们说的重要体验）放到这个领域去，并通过经常的回顾去尝试再次建模达到新的整体的理解，这样可以做到“形散神不散”的从碎片到系统再从系统到碎片的不断迭代的学习过程。

很多人由于家庭的影响，从小养成的思维习惯会非常缺乏系统化思维，而在学校里的学习也并不能培养他们的系统化思维能力。即使是具备一定的系统化思维习惯和能力的人，也可以通过更清晰的认识系统化思维有效的提高自己的学习效率。而把这个建模的过程进行数字化，恐怕是人们在数字化学习时代真正需要的学习软件，帮助人更好的驾驭互联网时代的学习。

知识引擎可以用于实现费曼学习法。知识领域即是在某领域建立理解模型。老师通过知识领域展现自己的领域模型，学生通过知识领域展示自己所学，老师可以对学生的知识领域即领域模型进行评估。知识引擎实现从碎片到系统，再从系统到碎片的自由自主的学习。并且学习者从学到教可以平滑过渡，学生可以无缝接入对课程的建设，打破老师和学生之间的藩篱。并且通过领域建模，实现知识发现，大大的丰富新知识的创造和分享，每个人可以去“研究”自己感兴趣的领域，发掘那个领域的知识，有如美国当下不少大学里流行的”Design your own major”专业，让知识的发现和研究无所不在。

另外知识引擎可以提供多个大的和小的工作流推荐给学习者选用。比如整个费曼学习法可以考虑做成一个大的工作流，其中包含了很多小的工作流。这个想法还在构思中。用户是否容易接受，是否好用，是否有效等等都还需要探索一下。

Feynman Technique：

• Pick a topic you want to understand and start studying it. Write down everything you know about the topic on a notebook page, and add to that page every time you learn something new about it.
• Pretend to teach your topic to others, especially to youngesters who are not familar with the topic . Make sure you’re able to explain the topic in simple terms.
• Go back to the books when you get stuck. The gaps in your knowledge should be obvious. Revisit problem areas until you can explain the topic fully.
• Simplify and use analogies. Repeat the process while simplifying your language and connecting facts with analogies to help strengthen your understanding.