The value of an ounce of gold, if all of its mass were converted to energy (E=mc^2) in the form of electricity (with 100% efficiency), and then sold at market rates (~5cents/kWh = $1/72e6J), would be:
1 oz. * (1kg/36oz.) * (9e16J/1kg) * ($1/72e6J) = $35 million
The value of a 170-lb. person would be:
170lb * (1kg/2.2lb.) * (9e16J/1kg) * ($1/72e6J) = ~$100 billion
Point: you get a lot of energy from a little mass.
Tuesday, July 29, 2008
Sunday, July 27, 2008
Math with the Stars
So I read in an old copy of Business Week that Teach for America, since being founded in
1990, has reached approximately 2.5 million children. These days, TFA recruits about 3000 teachers every year. Let's assume that the average TFA teacher teaches 50 students per year (elementary teachers would teach less; high school teachers would teach more). And they last for two years, so that yields:
3000 teachers/year * 50 students/teacher * 2 years = 300000 students impacted every year
Over 20 years, that would yield about 6 million students impacted. But we can figure in the early days, TFA was a lot smaller than it is today. So I think the calculation is about on par.
So that got me thinking: how else would it be possible to have such a large impact on education quailty in this country? And the answer, which has been brewing in my head for a while, is that high-quality education software could radically transform the way that young people learn in this country, and around the world.
I think the biggest opportunity lies with math. So let's consider a basic Algebra I course. the U.S. 8th grade enrollment is ~3.5 million. So that means that if you got 10% market share of all algebra students, you would be reaching 350k students per year. Already more than TFA reaches each year. And that's just one course. You could imagine doing lots of math courses (arithmetic, transistion math, geometry, statistics, calculus, etc.), reading comprehension courses at many different levels (read a passage and then answer questions about it--basically standardized test prep for reading), some history courses (American history, world history, comparative religions), social science courses (psych, econ), science courses across the board, computer science, and even foreign language (though Rosetta Stone is probably the gold standard in this market).
So the question is: how could you get 10% market share for a given subject material? I think there are a few key ingredients:
1) Must market (or at least target) kids and parents directly, rather than going through teachers and school districts, which are much slower to adopt novel teaching techniques.
2) The pedagogical approach must be research-based and proven to be effective, especially on standardized tests.
3) The product must have a low marginal cost.
So here's the vision: break each course down into lessons grouped into chapters. Each lesson is introduces by a video tutorial from two instructors: a subject-matter expert, and a "cool" host. The subject-matter expert should be someone who is 1) charismatic on camera and 2) articulate in explaining and teaching. The cool host is the one who is learning each lesson, asking questions, and working through problems. They would ideally be someone who has some initial B-list celebrity appeal, like Ashton Kucher, or Johnny Knoxville. Many of the lessons would be just the two of them together. But lots of lessons would involve some sample students (need around 4). And some lessons would be with a special guest, someone famous like Al Gore, Oprah Winfrey, Kobe Bryant, etc. The special guests could be stars from Hollywood, TV, politics, business, or sports. The idea is that kids would want to watch the lessons as a to see some of their favorite stars. It would also be great for marketing, because it would drum up a lot of media attention. If you make it something that kids want to see, then they will demand it before parents and teachers have even gotten on board with it.
After the video lesson, there would be a textual explanation of the material (which, frankly, a lot of kids might skip). All the text material would also be included in a textbook which could be purchased separately.
Then, the main event of the software is interactive problem-solving. At first, students would start with practice problems, where they could get hints, and explanations when they get a problem wrong. After getting comfortable with the practice problems, students could take a "qualifier" for each lesson. They would have to pass the qualifier to advance to the next lesson. Then, at the end of each chapter they would have to take a test to pass that chapter. The scores on the tests could be sorted into gold, silver, and bronze medals, with the requiremnt that a student earn at least a bronze before advancing to the next chapter. If they don't pass the first time, they could keep taking the test over and over again until they passed. If students are using the software with their school, the teacher could modify the course to rearrange the order of chapters or the requirements for advance.
So, to summarize, the key features that make this learning software an improvement over old teaching materials:
1) Students have an incentive to keep learning, as a way to see their favorite stars on the screen.
2) Students learn new material through a variety of channels--reading, listening, and watching, thus providing lots of ways to master the material.
3) Students practice skills until they master them, allowing for individualized pacing.
As for implementation, I think the key is to partner with a group with experience doing quality entertainment TV, like MTV. After all, education materials are competing with television (and video games, and movies...) for kids' time and attention, we have to use all the tricks that they use.
Initial production would be expensive, but doing a really good job on one course (say, algebra), and marketing it well (as part of a mission to ensure that every American understands math) would get a lot of kids to buy into it and that would generate a lot of revenue which could be used to support development of other courses. Let's say you charge $40 for the software (about what one hour of tutoring would cost), if you get 10,000 kids to buy it, that's $400K; get 100K kids to buy it, and you've got $4 million. And I think $1 million is a reasonable sum to pay for a low-budget MTV-quality videos. So I think it's very doable.
Back to TFA: the idea of taking smart, motivated college kids and having them work on the thing they know how to do best (do school) for the purpose of fixing one of the neediest problems in our country (bad schools) is not all that hard to come up with. Wendy Kopp just had the good sense to take that idea and put it into action. Math with the Stars isn't any brilliant breakthrough, but if it's a successful idea, it could take off and have a huge impact on education.
1990, has reached approximately 2.5 million children. These days, TFA recruits about 3000 teachers every year. Let's assume that the average TFA teacher teaches 50 students per year (elementary teachers would teach less; high school teachers would teach more). And they last for two years, so that yields:
3000 teachers/year * 50 students/teacher * 2 years = 300000 students impacted every year
Over 20 years, that would yield about 6 million students impacted. But we can figure in the early days, TFA was a lot smaller than it is today. So I think the calculation is about on par.
So that got me thinking: how else would it be possible to have such a large impact on education quailty in this country? And the answer, which has been brewing in my head for a while, is that high-quality education software could radically transform the way that young people learn in this country, and around the world.
I think the biggest opportunity lies with math. So let's consider a basic Algebra I course. the U.S. 8th grade enrollment is ~3.5 million. So that means that if you got 10% market share of all algebra students, you would be reaching 350k students per year. Already more than TFA reaches each year. And that's just one course. You could imagine doing lots of math courses (arithmetic, transistion math, geometry, statistics, calculus, etc.), reading comprehension courses at many different levels (read a passage and then answer questions about it--basically standardized test prep for reading), some history courses (American history, world history, comparative religions), social science courses (psych, econ), science courses across the board, computer science, and even foreign language (though Rosetta Stone is probably the gold standard in this market).
So the question is: how could you get 10% market share for a given subject material? I think there are a few key ingredients:
1) Must market (or at least target) kids and parents directly, rather than going through teachers and school districts, which are much slower to adopt novel teaching techniques.
2) The pedagogical approach must be research-based and proven to be effective, especially on standardized tests.
3) The product must have a low marginal cost.
So here's the vision: break each course down into lessons grouped into chapters. Each lesson is introduces by a video tutorial from two instructors: a subject-matter expert, and a "cool" host. The subject-matter expert should be someone who is 1) charismatic on camera and 2) articulate in explaining and teaching. The cool host is the one who is learning each lesson, asking questions, and working through problems. They would ideally be someone who has some initial B-list celebrity appeal, like Ashton Kucher, or Johnny Knoxville. Many of the lessons would be just the two of them together. But lots of lessons would involve some sample students (need around 4). And some lessons would be with a special guest, someone famous like Al Gore, Oprah Winfrey, Kobe Bryant, etc. The special guests could be stars from Hollywood, TV, politics, business, or sports. The idea is that kids would want to watch the lessons as a to see some of their favorite stars. It would also be great for marketing, because it would drum up a lot of media attention. If you make it something that kids want to see, then they will demand it before parents and teachers have even gotten on board with it.
After the video lesson, there would be a textual explanation of the material (which, frankly, a lot of kids might skip). All the text material would also be included in a textbook which could be purchased separately.
Then, the main event of the software is interactive problem-solving. At first, students would start with practice problems, where they could get hints, and explanations when they get a problem wrong. After getting comfortable with the practice problems, students could take a "qualifier" for each lesson. They would have to pass the qualifier to advance to the next lesson. Then, at the end of each chapter they would have to take a test to pass that chapter. The scores on the tests could be sorted into gold, silver, and bronze medals, with the requiremnt that a student earn at least a bronze before advancing to the next chapter. If they don't pass the first time, they could keep taking the test over and over again until they passed. If students are using the software with their school, the teacher could modify the course to rearrange the order of chapters or the requirements for advance.
So, to summarize, the key features that make this learning software an improvement over old teaching materials:
1) Students have an incentive to keep learning, as a way to see their favorite stars on the screen.
2) Students learn new material through a variety of channels--reading, listening, and watching, thus providing lots of ways to master the material.
3) Students practice skills until they master them, allowing for individualized pacing.
As for implementation, I think the key is to partner with a group with experience doing quality entertainment TV, like MTV. After all, education materials are competing with television (and video games, and movies...) for kids' time and attention, we have to use all the tricks that they use.
Initial production would be expensive, but doing a really good job on one course (say, algebra), and marketing it well (as part of a mission to ensure that every American understands math) would get a lot of kids to buy into it and that would generate a lot of revenue which could be used to support development of other courses. Let's say you charge $40 for the software (about what one hour of tutoring would cost), if you get 10,000 kids to buy it, that's $400K; get 100K kids to buy it, and you've got $4 million. And I think $1 million is a reasonable sum to pay for a low-budget MTV-quality videos. So I think it's very doable.
Back to TFA: the idea of taking smart, motivated college kids and having them work on the thing they know how to do best (do school) for the purpose of fixing one of the neediest problems in our country (bad schools) is not all that hard to come up with. Wendy Kopp just had the good sense to take that idea and put it into action. Math with the Stars isn't any brilliant breakthrough, but if it's a successful idea, it could take off and have a huge impact on education.
Sunday, July 13, 2008
Language acquisition, the primary goal of education
What are the key goals of education? I would argue that they are to teach languages. The acquisition of language is what separates humans from other animals. Language is inextricably linked to the way that we think. Thus, by learning languages we expand the capacity of our minds. I see 5 critical areas of language that should be cultivated in one's education:
- Native: How to write and how to speak in one's native tongue. This includes both persuasive as well as narrative froms of writing/speaking.
- Foreign: By learning a language other than our native language, we gain a better understanding for words, and are able to access cultures in the world more dissimilar from our own.
- Mathematics: The language of mathematics is critical for problem-solving and clear thinking.
- Programming: Computer languages teach how to break a complicated problem into smaller steps.
- Music: The language of music opens up a world of expression.
Subscribe to:
Posts (Atom)
