Khan Academy and Math Education

Two professors just released a parody video critiquing Khan Academy.  Its pretty boring and too nitpicky, so you can instead read the article about it in the Chronicle of Higher Education.  Their basic critique is that Khan Academy focuses too much on procedural thinking and not enough on actual thinking.  This is an issue, but traditional education suffers from a similar flaw. Often, schools just teach for whatever test it is that students must take, and ignore actual understanding of the material. To properly explore a topic, one would have to hire an very good teacher to teach a small group of students who are all at a similar level. This would be very expensive. A more realistic plan would be for students to use quality interactive software to learn the material and be able to discuss it with their peers, and when needed, be able to consult with an expert or advanced student. This balanced option could probably produce the best result for many students at the lowest cost.

The other complaint in the parody is the quality of Khan’s videos. It’s true that Khan’s videos are pretty simple, but he himself recognizes this. He stresses the main power of the site is the amount of data they have and their ability to add new features. For example, they added an intelligent quiz tool, and they will be adding additional interactivity in the future. This will improve the site, but there is also room for other players to contribute to online education. For example, if a site would make it easy for many people to create interactive educational content…

See also:
Math Education
The Future of Education 

Computer Science Curriculum

Teach Programming Instead of Computer Science

A basic problem with the Computer Science curriculum is that it does not teach students what they’re interested in or what they need to know professionally.  Most students would probably prefer practical programming experience but they get CS degrees since that is all that is being offered at their college. The schools don’t want to be a “trade school” so they instead require students to learn difficult and often unnecessary material. Besides making it more difficult for the students who do succeed, it ends up scaring off many people from a career in software development all together. Its time for more options to be offered and for some disruption in the education system:

Higher Education’s Online Revolution (WSJ)

Algorithms Writing Articles

An old post mentioned how algorithms are being used to help with diagnoses, but they’re also being used to write articles. Wired just published an article about Narrative Science, a company that creates programs which can generate articles. They mainly publish articles on finance and sports, since those are well-structured topics and very repetitive. You don’t normally have too many things happening in a baseball game or stock chart, so I could see why a computer could write such articles. They also have begun generating articles with other data:

Once Narrative Science had mastered the art of telling sports and finance stories, the company realized that it could produce much more than journalism. Indeed, anyone who needed to translate and explain large sets of data could benefit from its services. Requests poured in from people who were buried in spreadsheets and charts. It turned out that those people would pay to convert all that confusing information into a couple of readable paragraphs that hit the key points.

This is obviously quite impressive, though I wonder if all that information is necessarily better in paragraph form. Why can’t some of the data just be turned into some nice charts that people can view? Articles can sometimes “textualize” the content and make it take longer for the reader/viewer to get the necessary information. They can turn tables into paragraphs, but will they be able to soon write more general articles?

wired.com/gadgetlab/can-an-algorithm-write-a-better-news-story-than-a-human-reporter

Paul Graham’s Frightening Startup Ideas

Paul Graham discusses some “Frighteningly ambitious startup ideas” in a post on his blog (based on a speech he gave which is available here). He wants someone to replace 1)Google 2)email 3)universities 4)Hollywood 5)Steve Jobs 6)code optimization and 7)doctors.

For #1, he suggests a startups create a search for hackers, though I think it might be easier to focus first on certain high-revenue categories that Google doesn’t do so well at. Both #1 and #7 sound very difficult, I’m not sure if a startup will be able to solve them, but there are larger companies (like IBM) which are making progress in those areas.

For #2, I don’t know if email will be replaced anytime soon, but I assume more features will be added to it to help people manage its magnitude. Gmail and Hotmail already both have some features to help with this, and I assume such tools will improve.

#3 is an area I am especially interested in. I don’t know if universities are about to disappear, but I think the traditional lecture model will be upended.

PS
Another interesting link from today: Job Growth and Loss (LinkedIn). Looks pretty good for Tech stuff. Of course, some of the shrinking areas may be rebounding now, while some of the growing ones may be in a bubble (see the greenest one on the chart).

The Future of Education II – A New Curriculum

In a previous article, I discussed how interactive technology can be used to move beyond the lecture-system of education that has been dominant for so long. In this article I will discuss what material should actually be taught. I think the educational syllabus in many subjects should be changed to reflect the changing nature of knowledge. In addition, the educational curriculum should be changed to teach new skills that are relevant to the information age.

People are Not Hard Drives
The development of computers and the internet has changed the way people can access information and therefore changes the nature of the material people will need to learn. When knowledge was stored in books, perhaps it made sense to require people to memorize large amounts of information. However, since knowledge is now available at the touch of a smartphone (or even through speech), people can have easy access to the information they need without having to have spent years memorizing information. This information can obviously be much vaster and more up-to-date than anything they could have memorized. This does not mean that people will not need to learn anything and can just become an instant expert in any field by Googling any issue they have. (Try fixing your car’s transmission based on an eHow guide.)

An expert will need a solid understanding of the principles of his field and practice in applying them so that he can correctly draw on database of information to solve a specific issue. This is what students will need to learn instead of memorizing thousands of little details. Perhaps a student training to become a doctor no longer needs to memorize every inch of anatomy. The curriculum of many fields still remains focused on the same material as it did a few decades ago, but it may be time to revise it in light of modern technologies.

People are not Computers
Computers can do much more than just provide access to relevant information. Any problem that can be solved with clearly-defined steps can be programmed so that a computer can solve it. This fact also needs to be taken into account when designing a syllabus of study. Many areas of education involve students learning to mechanically implement set procedures and formulas to solve problems. This is particularly true in math-related areas such as mathematics itself, the sciences and parts of business and economics. These mechanical processes can all by definition be solved by a computer, so why pretend that these technologies do not exist? Human computers were once necessary, but they have since been supplanted.

Instead of focusing on being the computer, students should learn how to do the things computers cannot. They should learn how to use the computer tools and learn how to take real-life problems and convert them into a form that computers can solve. Computers are still quite poor at solving general real-life problems unaided by humans, and this is an important skill people will need. Even as computer programs improve, there is always an area where computers cannot solve problems, and people will be needed to work on them. Instead of teaching methods that are no longer needed, the focus of  education should be on the areas that are beyond the reach of computers.

This does not mean students should not learn any principles of a subject that a computer can solve. Students who are training to be an expert in a field should learn the basic principles of it so they can fully understand the material and be able to apply it in cases where a specific program does not exist. But it is not necessary for students to memorize a large number of mechanical methods for solving specific problems (without understanding them) when they will anyways just use computers to solve such problems. Plugging something into a formula or into a computer involve about the same amount of understanding, just one way is a million times faster. So perhaps it is not necessary to for high-school students to learn all those formulas and methods. Students should either understand a subject, know how to apply it, or learn what computers cannot solve, but there is no reason to treat them like mechanical computers themselves.

A New Curriculum
New technologies should cause more changes than just modifying the focus of specific subjects. The curriculum of subjects itself should be changed to meet the needs of the times.  Besides learning subject-specific computer skills, students need to learn general computer-skills that will allow them to use computers effectively in many different areas. There are many computer skills that many people do not know well, from the most basic (such as typing) to the more advanced (programming). Since these skills are often extremely useful, they should take precedence in an educational curriculum over less important subjects.

A Simple Example
Typing is obviously one of the most common and useful skills of modern times. People almost never write things by hand anymore, but instead type almost everything. Yet many elementary schools still focus on teaching cursive in 4th or 5th grade! While this was never very useful, nowadays it is completely pointless since no one writes in cursive. Many people never learn how to touch-type correctly and instead use the hunt-and-peck method to type throughout their life (I’m currently using a modified pecking method myself to type this myself). This is probably one of the most basic things that kids should learn when they are younger. There are also simple related topics people should learn, such as keyboard shortcuts and the purpose of certain keys, e.g. the “home” key. (I’ll teach that right now: It lets you go to the beginning of a line without having to press the arrow key 20 times. You’re welcome.) These skills are very basic, but learning them can probably save people hundreds of thousands of hours over the course of their lifetime.

Other Computer Skills
Typing is a simple example of the most basic skills people need, but there are many other areas people need to learn, such as greater proficiency in using operating systems and common computer programs. For example, many people do not know how to perform common tasks in Word or Excel (such as managing styles or creating graphs), despite their common practical use. There are also certain higher-level skills that are very important. For example, while there is a huge amount of information easily available on the internet, it is not always easy to quickly find the information a person needs.  There are skills and techniques people can learn to improve their ability to construct and filter searches, and to organize and index information they encounter. Google and other services cannot automatically find the best result for the exact thing a person is looking for, so people need to learn the skill of good searching and organizing.

Programming
Currently, high-schools force students to learn many difficult subjects in the math-logic realm, such as trigonometry and geometry, which they may never use. Yet there is a basic logic-based skill that would be of more interest and relevance to all: programming. Instead of learning so many details of math, students could learn programming, which can be used to do math and much more. While many people will not have any interest in programming full-time, most people will get some benefit form learning the basics of programming. This will enable them to do many tasks in the modern age, such as creating simple apps or websites, using macros in Excel, performing simple manipulations of text, or writing simple scripts for various scientific or business applications. Programming can be considered one of the basic skills that people should know in modern times.

Some may fear that adding these computer subjects to the curriculum may cause other subjects to be de-emphasized. In some cases, it may be possible to use more effective educational methods so students can learn both the old and the new subjects. However, people may need to choose which subjects have priority. Perhaps subjects should be empirically evaluated for what actual benefits they provide for the students. Studies have shown that most students do not actually improve their critical thinking skills over the course of the time in college. In a struggling economic climate and with extremely high prices for tuition, perhaps it may be time for the curriculum to focus on practical skills that students will actually use throughout their life.

The Future of Education

The development of the internet and related technologies has revolutionized many areas of the modern world, everything from research to shopping. However, the educational system has basically remained the same as always, with little improvement or innovation. Yet, things are starting to change. New technologies and developments in online education will help make learning less expensive and more effective and enjoyable.

Currently, most educational institutions follow a lecture-based learning model. In general, the teacher or professor recites the same lecture each time for every class he or she has on a specific subject. This raises a simple question: why can’t students just watch a video of the lecture?

The basic advantage or repeating lectures is to make the class more interactive so students can ask and answer questions while the material is being taught. However, this is often not so effective in practice. Different students understand different things, and one student’s question might just be an interruption for many other people. Therefore, a student might not ask the question, but then won’t be able to follow what comes afterwards.  Perhaps certain subjects could get a discussion-model to work for small, selected groups, but then the per-student cost for the class would be very high.

What actually happens in many cases is that students sit passively in class for most of a lecture, with little interaction and varied levels of attentiveness. Such a class does not really provide any advantage over a video of the same lecture, which students could watch instead. In fact, once the lecture becomes a video, it can be given once by the best presenter on the topic, and then shown to all future students throughout the world. This model has been particularly successful with Khan Academy, a website with thousands of free videos on various topics. By having the lecture on video, a student can watch at his own pace, pausing it, fast-forwarding or rewinding it when necessary. Many students preferred learning from these videos over any other approach.[1]

Khan Academy does not even show any lecturers in its videos, the focus is on the material itself. This might be a little too extreme, since a blackboard can get a bit boring at a certain point. However, educational videos can move beyond such simple graphics and display sophisticated diagrams and visualizations that can make the material clearer and more interesting. For example, the focus of a biology video can be on the actual organ or cell being studied, rather than on a lecturer or blackboard.

Videos themselves can be useful but they are not the ideal total system to learn new material. They offer little interactivity, and play at a constant rate, ignoring the student entirely[2]. Newer technologies allow the student to go through the material in a more active and engaged manner.

At a simple level, the student can always be required to be answer questions and complete challenges in order to move forward to the next topic. This way, he will always be involved in doing things, not just watching them. The system can quiz him and keeps track of what he knows, so he can be tested on it later. There can also be various gaming elements as students compete for high scores in their topic. Khan Academy has recently begun adding such features to their website, and has reported much success.

Besides just having quizzes, the educational content itself will be much more interactive than a simple lecture, text or video. The student will not just passively watch content, but will be able to manipulate and explore virtual representations and objects. Some topics can be learned through simulations and games where the student will be able to directly partake in the subject material itself. For example, when learning physics, students can play around with a physics simulator to explore how different laws work. They could then be required to complete a virtual activity that requires the correct physical knowledge, such as building a bridge or firing a rocket. This would make the subject more interesting and understandable than current methods.

Although computer-based education provides many benefits, there are also some apparent disadvantages that need to be dealt with. When learning a subject it helps to have someone to whom one can ask questions. Googling questions do not always turn up good results, and sometimes one needs specific help with a particular issue. A good online educational system would need to provide ways to get specific help. Some of the content and questions can expandable sections and links that provide more explanation when needed. General questions can be asked in online forums, where other students or experts can answer them. This model has worked well on sites like StackOverflow.com where strangers often give excellent answers to answer to other people’s questions for free. It can definitely work for a group learning a subject online, and there can be experts who review some of the answers or answer the most difficult ones.

If a student needs help with a specific issue, it may sometimes help to have an actual person he can ask the questions to. However, there is no need to have a teacher give a whole lecture just so he can answer students’ questions. The student can learn the subject online, and then if he runs into an issue that another student cannot answer, he can ask a designated expert on the topic.[3] When needed, these experts can also give specific feedback on a students’ work. Overall, this model of answering questions can be both cheaper and more effective than the traditional one.

One apparently minor issue that arises when with computer-based education is the opportunity for distraction. Before computers, people only had their pens to distract them while studying. (In fact, for some students, scribbling is one of the main skills learned in school.) Yet now, thousands of distractions lay just a click away at any moment.  While this may help alleviate the boredom of a student stuck in class, it raises a fundamental issue: How will people learn how to scribble? Also, it raises another important issue: How can one focus in the age of Gmail, Facebook, YouTube and non-stop news? In one way, moving education from the dry lecture to a more interactive model may help it compete better with modern distractions, but they will still remain a major problem.

Any technology that puts people at a computer to learn will need to take these issues into account. On the superficial level, time-management programs, such as RescueTime and LeechBlock, can be placed on computers to help prevent students from wasting time. To help encourage people to actually do work, there can be general tasks they must complete, so they can’t just do everything right before the tests. This will help encourage deeper learning of material than the current test-focused models. On the more fundamental level, people need to develop proper focus and work habits, but this is something which becomes easier when the material is more interesting.

Currently, education is extremely expensive. Over the last few decades, costs in education have risen far higher than the pace of inflation, but without improving their results. For example, at Yeshiva University, tuition rose from $16,000 to $33,000 between 1999 and 2010.[4] Obviously, if people switch to a completely online education system, they will be able to save a significant amount on tuition. But there can also be significant savings if schools switch to a “blended model” where students learn most of the material on a computer but still meet in person for some study groups and classes. University costs will be able to be dramatically cut, while the education may become more effective and enjoyable.

The move to online education is already happening, and it is already working.  A 2009 meta-study by the US department of education found that “…on average, students in online learning conditions performed modestly better than those receiving face-to-face instruction.”[5] Online education is not only far less expensive than traditional methods, it also has better results.  These results will only improve as more sophisticated interactive content and methods are developed.

 


[1] See Wired’s article on Khan Academy, at http://www.wired.com/magazine/2011/07/ff_khan/all/1

[2] It is true he can pause the video, but one doesn’t have the same control of “processing” the content when it’s a video, and it is hard to refer back to a previous part.

[3] This would be similar to the “shoel u’meshiv” of a traditional Beit Midrash. One expert will be able to serve a large number of students for each subject.

Computers Begin Taking Over

Soon computers will be diagnosing patients’ problems and answering their legal questions…

Watson, the “Jeopardy!”-playing computer system, is getting a job… WellPoint said it plans to use Watson’s data-crunching to help suggest treatment options and diagnoses to doctors. It is part of a far broader push in the health industry to incorporate computerized guidance into care, as doctors and hospitals adopt electronic medical records and other digital tools that can record, track and check their work.

 

http://online.wsj.com/article/SB10001424053111903532804576564600781798420.html

 

 

Another day, another 72+ compromised companies, NGO’s & GO’s

I am convinced that every company in every conceivable industry with significant size and valuable intellectual property and trade secrets has been compromised (or will be shortly), with the great majority of the victims rarely discovering the intrusion or its impact. In fact, I divide the entire set of Fortune Global 2000 firms into two categories: those that know they’ve been compromised and those that don’t yet know

http://blogs.mcafee.com/mcafee-labs/revealed-operation-shady-rat

Every short while, there's more news of large number of companies and organizations being hacked, normally with hints that it came from China. In this case:

The interest in the information held at the Asian and Western national Olympic Committees, as well as the International Olympic Committee (IOC) and the World Anti-Doping Agency in the lead-up and immediate follow-up to the 2008 Olympics was particularly intriguing and potentially pointed a finger at a state actor behind the intrusions, because there is likely no commercial benefit to be earned from such hacks. The presence of political non-profits, such as the a private western organization focused on promotion of democracy around the globe or U.S. national security think tank is also quite illuminating. Hacking the United Nations or the ASEAN (Association of Southeast Asian Nations) Secretariat is also not likely a motivation of a group interested only in economic gains. 

They used spear-phishing to get into everyone's computers. I think its time to start re-thinking the very structure of the internet. 

To Err is Human…

Problem: In the last few years, many congressmen have gotten in various affairs and scandals and been forced to resign. It seems its too difficult for these powerful people to avoid getting into trouble. Yet people still want to vote for the platform the representatives ran on. 

Solution: Instead of having human representatives, people should be able to vote for algorithmic representatives. Computers can’t sin, so there won’t be any scandals. The computers will have set principles that they will vote by, and will be able to follow their campaign pledge better than any human. Humans are also too slow, and can never pass a law in time for it to work (e.g the Stimulus). An algorithm will just kick in some predetermined measures at a certain time, which will be much more effective ;)