亚里士多德逻辑自动化_接近亚里士多德

亚里士多德逻辑自动化

Calls for radical change in the education system are a dime a dozen.

呼吁对教育系统进行彻底改革的呼吁只是一角钱。

In my perspective, one idea stands out from the rest. In his 2000 essay entitled Aristotle (The Knowledge Web), American inventor William Daniel Hillis describes a computer program called Aristotle which would act as a sort of guide for those seeking to acquire new knowledge. In his words:

在我看来，一个想法比其他想法更突出。 美国发明家威廉·丹尼尔·希利斯(William Daniel Hillis )在2000年发表的题为“ 亚里斯多德(知识网络)”的文章中 ，描述了一种名为亚里斯多德的计算机程序，该程序将为那些寻求获取新知识的人提供指导。 用他的话说：

Aristotle would begin by asking you how much time you’re willing to devote to this project and the level of detail you want. Then Aristotle would show you a map of what you need to learn. Aristotle plans its lessons by finding chains of explanations that connect the concepts you need to learn to what you already know.

亚里斯多德首先要问您愿意花多少时间从事这个项目以及您想要的细节水平。 然后亚里斯多德会向您显示您需要学习的地图。 亚里斯多德计划课程的学习，方法是找到将您需要学习的概念与已经知道的概念联系起来的解释链。

This system would be curated and continuously improved by human experts and educators, but would ultimately allow for high quality, pertinent lessons to be distributed en masse to anyone who cares for to stream them. The machine wouldn’t have to do any actual teaching; it would simply show learners educational content created by human teachers which it thinks would best illustrate a given concept.

该系统将由人类专家和教育者来策划和不断改进，但最终将允许向所有希望流式传输的人大规模分发高质量的相关课程。 这台机器不需要做任何实际的教导。 它只会向学习者展示由人类教师创建的教育内容，认为这将最好地说明给定的概念。

Two decades ago, Hillis described Aristotle as “an idea whose time has come”. It seems like we now have the technological capacity to create such a program — so where is it?

二十年前，希利斯(Hillis)将亚里斯多德描述为“一个时代已经来临的想法”。 看来我们现在具备创建此类程序的技术能力-那么它在哪里？

亚里斯多德在哪里？ (Where is Aristotle?)

Hillis’ essay was originally published in 2000, when the technology required to create Aristotle was still far from realization. Its closest relative at the time was probably the internet as a whole, where it is possible to access a significant amount of instructional material on a wide array of topics.

Hillis的文章最初发表于2000年，当时创建亚里士多德所需的技术还远远没有实现。 当时，它最接近的亲戚可能是整个互联网，在这里可以访问有关大量主题的大量教学材料。

In the original essay, Hillis rejected the internet as a version of Aristotle on the grounds that it could not support guided learning and had no knowledge verification system. Nowadays, individual companies like Google and Facebook might have the data required to accurately recommend instructional material to its users. In a similar way to how the Youtube suggestions algorithm recommends videos users will want to watch, an Aristotle-like service could recommend pre-existing Youtube videos based on what will be most educational.

在原始文章中，Hillis拒绝将互联网作为亚里士多德的一种版本，理由是它无法支持引导式学习并且没有知识验证系统。 如今，像Google和Facebook这样的公司可能已经拥有了向其用户准确推荐教学材料所需的数据。 以类似于YouTube建议算法推荐用户想要观看的视频的方式，类似亚里斯多德的服务可以根据最具教育意义的内容推荐已有的YouTube视频。

However, the lack of fully guided learning completely undermines the claim that the internet itself is Aristotle. In order to learn something online, I typically need to have a mental understanding of the progression of topics I wish to follow. Then, I need to search for explanations of each concept individually. Even after this effort, it remains difficult to gauge my own understanding of a topic, because there is no one to test my understanding.

但是，缺乏完全指导性的学习，完全破坏了互联网本身就是亚里斯多德的说法。 为了在线学习一些东西，我通常需要对我希望遵循的主题的发展有一个心理上的了解。 然后，我需要分别搜索每个概念的解释。 即使经过这一努力，仍然很难衡量我对某个主题的理解，因为没有人可以测试我的理解。

On the other hand, there are certain apps and websites which meet some important criteria for Aristotle, of which Duolingo and Khan Academy are prominent examples. They contain knowledge webs and provide verified knowledge to their users within these structures.

另一方面，某些应用程序和网站满足了亚里斯多德的一些重要标准，其中以多林戈和可汗学院为代表。 它们包含知识网，并在这些结构中向其用户提供经过验证的知识。

Screenshot of the Duolingo dashboard with a progression of concepts: lessons unlock as previous ones are completed to a high-enough standard.

Duolingo仪表板的屏幕截图，其中包含一系列概念：以前的课程以很高的标准完成后，课程会解锁。

However, both of these websites fail in terms of scale. Aristotle contains a substantive portion of human knowledge, with links drawn across disciplines. In Hillis’ words:

但是，这两个网站在规模上都失败了。 亚里斯多德(Aristotle)包含着人类知识的实质性部分，并跨越了各个学科。 用希利斯的话说：

With the knowledge web, humanity’s accumulated store of information will become more accessible, more manageable, and more useful.

有了知识网，人类积累的信息存储将变得更易于访问，更易于管理和更有用。

The scale of current educational endeavours pales in comparison with that envisioned by Hillis. Aristotle would likely require an open source, peer reviewed database of knowledge and tutorials painstakingly arranged in such a way that there is a logical flow from basic to more advanced subjects. To begin with, perhaps a single discipline, like history, could be documented. Still, such a Herculean effort might take decades to complete, with new instructional technologies and developments in the field itself potentially rendering entire portions of the project obsolete before completion.

与希利斯(Hillis)所设想的相比，当前的教育工作规模不大。 亚里斯多德可能需要经过开源，经过同行评审的知识数据库和精心设计的教程，以确保从基础学科到更高级学科的逻辑交流。 首先，也许可以记录一个单一的学科，例如历史。 尽管如此，如此艰巨的努力仍可能需要数十年才能完成，而该领域本身的新教学技术和发展可能会在完成之前使项目的整个部分过时。

And, as Hillis himself points out, there is the issue of compensation for the educators producing content. While a paywalled Aristotle would still be a significant improvement on the current (lack of) organization of human knowledge, it would undoubtedly be suboptimal.

而且，正如希利斯本人指出的那样，存在着教育内容产生者得到补偿的问题。 虽然有酬金的亚里斯多德仍将是对当前(缺乏)人类知识组织的重大改进，但这无疑不是最佳的。

知识网 (Knowledge Webs)

To some extent, personalizable pathways through a knowledge web are already embedded in our education system. Advanced high school classes typically have prerequisites which prepare students for the level of rigour and specialist knowledge contained within the course. This allows students to navigate course selection in a way which allows them to become progressively more knowledgeable about each subject they take.

在某种程度上，通过知识网的个性化途径已经嵌入我们的教育系统中。 高中高级课程通常具有先决条件，可以使学生为课程中所包含的严格和专业知识水平做好准备。 这使学生能够以某种方式导航课程选择，从而使他们对所修的每门科目逐渐了解。

The main functional difference between modern high schools and Aristotle is the level at which education is customizable. High school allows for personalization at the subject level (e.g. choosing between calculus and statistics), whereas Aristotle could be personalized more deeply (e.g. the arrangement of questions in a problem set). One might argue that the main difference between high school and Aristotle is between the primary decision-maker being the learner or an algorithm, but I don’t think this distinction is pertinent. In most high schools, subject choice is only a matter of personal preference on the surface. In actuality, educators and guidance counselors play a significant role in influencing (or mandating) selected courses, not unlike Aristotle’s informed decision making on instructional pathways.

现代中学和亚里士多德之间的主要功能差异是可定制的教育水平。 高中允许在学科级别进行个性化设置(例如，在微积分和统计学之间进行选择)，而亚里士多德则可以进行更深入的个性化设置(例如，问题在问题集中的排列)。 有人可能会争辩说，高中和亚里士多德之间的主要区别在于主要的决策者是学习者还是算法，但我认为这种区别是没有意义的。 在大多数高中，表面上的选择只是个人喜好问题。 实际上，教育家和指导顾问在影响(或强制)所选课程方面起着重要作用，这与亚里士多德关于教学路径的明智决策一样。

机器人老师 (Robot Teachers)

Teaching is often said to be a profession immune from automation — would the development of Aristotle render this statement false?

人们通常认为教学是不受自动化影响的职业-亚里斯多德的发展会否使这一说法错误？

While Aristotle could potentially reduce the necessary ratio of K-12 teachers to students, I argue that human educators will never go obsolete. For one thing, content delivery is only one function of the current education system. Maintaining discipline in the classroom, resolving conflicts between students, providing emotional support, etc. are all in a teacher’s job description, and I cannot envision any of this ever being automated.

虽然亚里斯多德有可能减少K-12教师与学生的必要比例，但我认为人类教育者永远不会过时。 一方面，内容交付只是当前教育系统的功能之一。 保持课堂纪律，解决学生之间的矛盾，提供情感支持等都在老师的工作描述中，我无法想象这是自动化的。

Hillis himself objects to Aristotle being used to educate children (at least initially) on the basis that one needs to have sufficient self-motivation to learn from a computer program. It is clear that teaching content and making students learn the content are two different tasks, of which only the former can be completed by a computer.

希利斯本人反对亚里士多德被用来(至少在最初阶段)对儿童进行教​​育，因为人们需要有足够的自我激励才能从计算机程序中学习。 显然，教学内容和使学生学习内容是两个不同的任务，其中只有前者可以由计算机来完成。

In an ideal world, I envision Aristotle working alongside human teachers. A program as powerful as Aristotle would be able to specify groups of students around the same ability level to whom content could be delivered in a group environment. Instead of a school of 500 children isolated on their Aristotle-enabled laptops, there could be smaller, subject-specific learning groups working towards educational outcomes commensurate with the knowledge and skills of each student.

在理想的世界中，我设想亚里士多德与人类老师一起工作。 一个像亚里斯多德一样强大的程序将能够指定具有相同能力水平的学生群体，在小组环境中可以向其交付内容。 可能会有较小的，针对特定学科的学习小组 ，而不是使用支持亚里斯多德功能的笔记本电脑来容纳500名儿童，而是根据每个学生的知识和技能来努力实现教育成果。

Learning groups could be comprised of students of different ages (but comparable ability level) and be switched around periodically to maintain parity in ability levels within each group. Collaborative work could be folded into Aristotle’s curriculum as human teachers see fit. This model would result in students meeting and collaborating with significantly more peers than is currently possible; socialization during lesson time would no longer be limited to the same class group of 20–30 throughout the entire academic year.

学习小组可以由不同年龄(但能力水平相当)的学生组成，并定期轮换以保持每个小组内能力水平的均等。 在人类教师认为合适的情况下，可以将协作工作纳入亚里士多德的课程。 这种模式将导致学生与同龄人的聚会和合作数量大大超过目前的水平； 在整个学年中，上课期间的社交活动将不再局限于20-30岁的同班同学。

Because Aristotle delivers content in a progression of understanding perfectly suited to each learner, there would be no students too far behind or far ahead to derive value from lessons. If a student is at a level distinct from that of any of their peers, that individual could simply form a learning group of one, with human teachers creating collaborative opportunities where possible. (This wouldn’t isolate the student entirely, as other subjects would still be taught in larger learning groups.)

由于亚里士多德以完全适合每个学习者的理解的方式提供内容，因此不会有任何学生落后于或遥遥领先而无法从课程中获得价值。 如果学生的水平与其他同龄人不同，则该人可以简单地组成一个学习小组，由人类老师在可能的情况下创造合作机会。 (这不会完全孤立学生，因为其他科目仍将在较大的学习小组中教授。)

As for the actual content covered in school, no student’s ability to learn would be capped at the level of understanding possessed by their teachers. Aristotle could deliver lessons about anything a student is interested in learning, from statistics to music theory. Ultimately, the severe inefficiency currently inherent in classrooms whereby students performing at a wide range of abilities receive the same level of instruction could be eliminated.

至于学校所涵盖的实际内容，任何学生的学习能力都不会受到老师所拥有的理解水平的限制。 从统计学到音乐理论，亚里士多德都可以提供有关学生感兴趣的任何课程的课程。 最终，可以消除教室中当前固有的严重低效率，从而使以各种能力进行表演的学生获得相同水平的指导。

I will admit that Hillis’ initial conception of Aristotle is far more grand than I describe it to be. A world with Aristotle might indeed be a world without schools at all. Such a program could make learning a decentralized, lifelong process, with Aristotle teaching skills as its users find a need or demonstrate an interest for them.

我将承认，希利斯对亚里斯多德的最初构想远比我所描述的要宏伟得多。 亚里斯多德的世界可能确实是一个根本没有学校的世界。 这样的程序可以使学习成为分散的，终生的过程，并在用户发现需求或表现出对他们的兴趣时利用亚里斯多德的教学技能。

While Hillis’ paper 14 years ago presumed that a digitized knowledge bank would be of foremost interest to motivated adults, perhaps those who dream of Aristotle should take on a more manageable challenge: creating a knowledge web of K-12 content and convincing schools to redesign their current educational models to make good use of this robot teacher.

希利斯在14年前的论文中假设，对于有动机的成年人来说，数字化知识库将是他们的头等大事，但也许那些梦想着亚里斯多德的人应该面对一个更易于管理的挑战：创建一个包含K-12内容的知识网，并说服学校重新设计他们目前的教育模式，以充分利用这位机器人老师。

翻译自: https://medium.com/age-of-awareness/approaching-aristotle-50716427f99b

亚里士多德逻辑自动化