The 5 Pillars of Education

 Education  No Responses »
Jan 222012
 

It is my belief that the way education is delivered and consumed will experience radical changes in the next 5 years. If you dissect the elements of education you find 4 key elements all of which are dropping in price dramatically, with a fifth that is all about personal investment. I am not going to get into each one of these in any great detail. That will be done in future blogs. However, they are as follows:

  • Content (Pillar I) – There has been an explosion of free (or very inexpensive) content on the web. Major universities such as Stanford, MIT, Harvard, Yale and dozens of others are making their lectures and materials available at no charge. All you need is a way to watch is aPC, Mac, iOS, or Android device coupled with decent broadband connectivity available through any cable company. If you don’t own either, visit your local library … they are becoming data centers with PCs and free wi-fi access.
  • Collaboration (Pillar II) – While it is possible for you to learn on your own, it is a lot easier to learn through peer interactions (other students). Again, there are numerous free platforms to establish a public or private group to foster discussion, sharing, and collaboration. In addition, multipoint video conferencing has finally come to the masses. If you want to team up with others interested in learning something, it is up to you … there are no barriers other than your personal time.
  • Coaching (Pillar III) – There are several mentoring networks evolving on the web where learners and coaches can find each other. This is an area that is ripe for development. The internet is the ideal tool to pair groups of individuals with a hunger for knowledge with other individuals that are experts in a field and have a desire to mentor, coach, and lead. There are currently a number of freelance sites that lets you define a project (maybe learning) and ask for bids from freelancers or gurus in selected fields. Why not a similar network for learning. You will be able to pick from a list of providers based on their expertise and reputation as measured by previous customers.
  • Certification (Pillar IV) - MIT recently announced that they will make a number of their Open Courseware offerings available on a learning platform which will be tied to certification. The organization that does the validation will be separately branded from MIT. While they plan on charging for certification, it is expected to be very reasonably priced. When coupled with free content, a learning platform which will likely support collaboration and perhaps coaching, this move will offer education and validation of mastery to the masses.
  • Commitment (Pillar V) - It doesn’t take that much imagination to see how the barriers to education are falling rapidly. In fact, it is possible that, with the exception of device, bandwidth, and some certification costs … the only barrier is you. The 5th pillar is your commitment, attitude, drive, and ambition. Suppose access to knowledge and the ability to prove you have acquired it are essentially free. What personal strategies, attitudes, motivations, and behaviors are necessary to benefit from this new world? In my 15 years of running SetFocus, this factor is by far the most critical to overall success.
Future blogs will dive down into each of these pillars in relations to what is currently going on in the marketplace and  where education is going in the future. Sal Kahn, the founder of Kahn Academy, recently put together a video about his predictions for the future of education. I think he is on the right track. Check it out …
YouTube Preview Image

 Posted by at 6:06 pm

How Big is a Red Blood Cell?

 Biology, Nanotechnology  No Responses »
Jan 222012
 

Red Blood CellIt is amazing how small a nanometer is compared to some other very small things we know about. For example, the other day I got to thinking how does a nanometer compare to the size of a red blood cell. This interest came about because a saw a number of articles talking about how one day, we may be able to create nanorobots to interact and/or enhance our existing red blood cells. Here is what I found out:

  • 1 Red Blood cell is about 2-3 microns (micrometers … 1/1,000,000 th of a meter … 1×10^-6) in height
  • 1 Red Blood cell is about 6-7 microns in diameter

Carbon Nanotube

Since their are 1000 nanometers (nm) in a micron, the red blood cell would be roughly 2500 nm high and 6500 nm in diameter. Ok, this is still pretty geeky and may not be easy to see. How about putting into something that is more recognizable. Suppose that a nm is actually 1 foot (i.e. 12 inches), then the height and width of the red blood cell would be twice the height of the Empire State Building and about 20 New York City blocks in width. Now try to imagine the types of things you could construct with nanotubes which are about 1 nm in diameter and up to 132 ml in length. Using the 1 nm as 1 foot … this means we can construct tubes from a few floors to up to the equivalent of 88,000 Empire State Buildings high.

These nanotubes will be the ultimate building blocks of nanorobotics. Our kids and grandkids should understand this … oh wait a minute, they do … it’s called Tinker Toys!

 Posted by at 9:55 am

How Big is the Brain?

 Biology, Brain  No Responses »
Jan 192012
 

How big is the brain? In doing some research for my blog, here is what I found.

  • There are approximately 50,000,000 neurons per square centimeter (50X106 per cm3)
  • Each cm3 neurons have axons that reach out to create 1 trillion synapse (1×1012)
  • If each synapse can express 8 bits or 1 byte of information, than 1 cm3 contains 1 terabyte of data (1×1012)
  • Now, the human brain is roughly 1000 cm3 in size which means it can store about 1 petabyte of data (1×1015)
  • That is about 1/3 as much data as stored in the entire internet … in one average human brain.

So when will a computer the size of a brain match its capacity? By current rates of change it looks like we will be there between 2025 and 2030.

One more observation, that neuron that is the central switching unit is about 4 to 100 microns in diameter. Scientists are working on developing nanotechnology based devices to interact with our bodies. This means they will be between 1 and 100 nanometers in size. If we were to measure a neuron against 1 nanometer, the neuron would be between 4,000 and 100,000 times larger. To put it into a scale which is more understandable, if the nanometer was 1 ft., the neuron would be between slightly less that a mile to about 20 miles in diameter.

As we master nanotechnology and can construct nanoscale devices, could we augment the brain with advanced features like connectivity, extended databases, and faster processing? How about uploading and downloading data to and from external sources? Could we learn faster and more efficiently? All interesting and fascinating questions that are not 100 years into the future, but instead a few decades.

Here is one view of how a nanobot might interact with a neuron:

YouTube Preview Image

Fasten you seat belts, we are in for a bumpy ride.

 Posted by at 3:18 pm

The Power of IBM Watson

 IBM Watson  No Responses »
Jan 182012
 

IBM WatsonIn 1997 my company, SetFocus, began operations and prior to our first class had to outfit a classroom with about 20 PCs. Back then, the latest and greatest was a 166mhz machine for about $1,000.

In an effort to shed further light on the notion of exponential thinking, it occurred to me, ‘how does that compare to the IBM Watson computer’?

I also have this theory that all technology ends up at the checkout counter at the supper market … when might that occur?

Well, let’s take a look:

IBM Watson Stats:

  • Price: Approx. $3,000,000
  • Blades: 90
  • Processors/Blade: 4
  • Cores/Process : 8
  • Clock Speed/Core: 3.55ghz
  • Each blade is equipped with 256 GB of RAM or about 200 million pages worth of data
  • Watson can process 500 gigabytes per second, the equivalent of a million books
  • Total Cores: 2,880 (i.e. 90x4x8)
  • Avg Cost/Core: $1,042

So, here are a few observations:

  • The computing power of 1 core in IBM Watson = 20 PCs purchased in 1997
  • 1997 Cost of 20 PCs which equate to the power of a single core = $20,000

Now keep in mind, while that represents a 20:1 ratio … we are not really comparing apples to apples. My PCs where desktop systems while the IBM Watson platform is clearly server level quality. Even today, there is at least a 2:1 relationship between these types of platforms. So, on a apples to apples comparison, the ration is more likely 40:1 … quite a change in price performance over the last 15 years.

But, what about 15 years from now … 2027? What might IBM Watson look like? There are two ways to express this. First, how much computing power could you buy for $3,000,000 and second, what would the current capability cost.

From my earlier blog, we know that it is likely that computing power will increase by a factor of 8 each 5 year period. So, over 15 years we should see 8x8x8 or 512 times more power for the same price.

That would suggest the equivalent of 1.5 million cores at 3.55ghz vs. the current 2,880. In addition, IBM Watson 2027 could manage 100 billion documents. By way of comparison there are currently between 7 and 8 billion documents on the internet today. Even assuming a decent growth rate of documents, it appears that IBM Watson 2027 could access all of the documents stored on the web.

Conversely, what would todays IBM Watson cost. Well, using the same Moore’s Law assumptions, the price should be 1/512th … or $5,859. At that price, IBM Watson 2011 could be affordable to nearly any small business. Especially if it were paid off over 3 years … about $180/mo!

Clearly, my predictions may be off a bit … but probably not that much.

Oh, by the way, in 2042 you will be able to say, ‘let’s throw one of those IBM Watson 2011 units in the shopping cart’. No problem … that will be $11.44 please.

 Posted by at 2:20 pm

Looking Forward by Exponential Periods

 Exponential Thinking  No Responses »
Dec 062011
 

Been thinking a lot about how to create a vocabulary to communicate about the impact of exponential technology change on all of the products, services, relationships, economics, education, etc. that we deal with on a day to day basis.

If we assume Moore’s Law using a 20 month period as our doubling rate, there will be 3 doubles every 5 years or 2 ^ 3 which equals 8 as the multiplier for each Exponential Period (EP). For each EP, we need to understand Relative Processing Power (RPP), and $/RPP. In other words how powerful are processors getting and how does the cost of a processing unit decrease in cost.

Specifically if we use the term EP0 (i.e. EP=0) it means the following:

  • 5 Year Period:
    • EPBeg = (CurrentYr)+ 5 * (EP-1)= 2006
    • EPEnd = (CurrentYr)+ 5 * EP = 2011
  • RPP = 8 ^ EP = 1
  • $/RPP = $1000/RPP = $1000

So, if we think of the impact on something 10 years out, it can be expressed as EP2 (i.e. EP=2):

  • 5 Year Period:
    • EPBeg = (CurrentYr) + (5 * (EP-1))= 2016
    • EPEnd = (CurrentYr) + (5 * EP) = 2021
  • RPP = 8 ^ EP = 64
  • $/RPP = $1000/RPP = $15.65

This may sound a little convoluted, but stick with me. My plan is to socialize this notion so that when we speak about the future we think exponential change vs. linear.

For example, take any current technology; such as Siri (Apples new iPhone 4s app) and ask what it might look like by the end of EP1, EP2, EP3, etc. and what the impact of this evolving technology will have on economics, policy, education, health, and culture in general. It will probably be a little difficult to look past EP2 or EP3 since the amount of change is so great, but as we develop enough EP1 and EP2 views, it will form a foundation that will allow for more informed EP3 views.

The chart below shows the RPP and $/RPP for each EP; -5 thru +8. The EPEnd is shown on the X-Axis, with the Y-Axis showing RPP (in blue on the left) and $/RPP (in red on the right) on logarithmic scales.

How far into the future can you see?

Thinking Exponentially in terms of power and cost is key to understanding the future and its impact on everything we know today.

 Posted by at 12:33 pm

Thinking Exponentially

 Exponential Thinking  No Responses »
Nov 292011
 

If I accomplish one thing, I hope it is to get you to think exponentially. It’s not easy. Most of us use the past to predict the future.

For example, 5 years ago we did not have an iPhone. So, you can say in that period of time, we went from the coolest phone being a Motorola Razor (because it was thin) to the iPhone 4s that comes with Siri. That is a lot of change. Hundreds of thousands apps, a dual core processor, centralized storage of music, apps, pictures, etc. (iCloud) … the list goes on.

What does this tell us about the future?

Remember from my last post, each 5 years is expected to see change that is 8 times greater than the previous 5 years. Since the iPhone has gone through about 1 new version per year, it is not unreasonable to expect by 2016 we will be talking about the iPhone 9. What might the hardware include? Software? Connectivity? Apps?

While we can’t predict exactly what it will look and feel like in 5 years, but we do know that it will leverage not only the advances in processor speed, connectivity (4g today … 8g by 2016?), existing and evolving software base, but also allow us to solve many complex problems that are assisted by huge increases in computing power.

Here is the list of problems we can expect significant progress on in the next 5 years:

Speech Recognition – Check out Siri … not bad today. We can decipher words, sentences, and some context. More processing power (both local and in the cloud) will help us improve the contextual issues associated with language.

Speech Generation – Today we have some rudimentary speech generation. It has gotten better in recent years. Expect clear human sounding voice responses that are contextually based. Probably not good enough to fool you, but getting close.

Vision – Lots of information to process, extract patterns, determine context, and map to meaning. More variable and nuanced than speech. But over the next 5 years, movement of images via video conferencing will explode providing a rich base of information from which to extract meaning.

Artificial Intelligence – Many would say that things like Siri and Watson are not AI. In the strict sense I guess they are not. However, I would argue that they are clearly the beginning of widespread use of expert systems that will clear the way for real AI. Leveraging both hand held devices with cloud based computing and web content (including numerous databases of experiences) will supercharge this area.

Connectivity – We have already come a long way. It was long ago that even simple images took a long time to download and display. Now we can easily stream 1080p HD video over the 4g network. Going forward network speeds will progress to the point where every endpoint will be capable of 1080p HD video conferencing. Higher bandwidth means more collaboration of complex tasks.

Robotics – Take all of the above together leveraging each other and imagine the types of machines that will be possible. In particular, robots will become more dexterous, faster, and able to communicate with other robotics to perform collaborative tasks.

This is by no means a comprehensive list of problems that an 8 fold (in 5 years) improvement in computing power can solve. But it does represent the basic areas that can be leveraged to perform work that was once the exclusive provence of people.

800% increase in power and all that brings in 5 years. How about 6400% in 10. Think exponentially!

 Posted by at 12:35 pm

The Future Ain’t What it Used to Be!

 Exponential Thinking  No Responses »
Nov 232011
 

Have been meaning to start a blog about a subject that I find facsinating and can’t stop following. I have coined this the Technology Tsunami. Most people feel it coming but just can’t seem to fully articulate its source or full implications. We see it everywhere we look. In the toys we buy, the education our kids consume, the media around us, and most importantly the work that we do.

At the base of all of this is the exponential changes that are being ushered in by the advances in technology that allow us to pack twice as much power into a processor every 18 months or so … known as Moore’s Law. This doubling effect takes a while to build up. While the rate of change is fairly constant, the absolute change from period to period is pretty dramatic.

Using 2011 as a base year (1 unit of processing power representing a typical $1000 notebook computer) here is what follows assuming we double about 3 times every 5 years:

  • 2011 – 2016: 1 Unit grows to 8 … net change of 7
  • 2016 – 2021: 8 Units grows to 64 … net change of 56
  • 2021 – 2026: 64 Units grows to 512 … net change of 448
  • 2026 – 2031: 512 Units grows to 4096 … net change of 3584
  • 2031 – 2036: 4096 Units grows to 32,768 … net change of 28,672

Note that each 5 year period is experiencing changes in computing power 8 times greater than the previous 5 year period. Understanding the past is not sufficient to predict the future … you need to think exponentially.

So, in 25 years we will see our current processor power grow by a factor of 32,768. Holding the price of a processor fairly constant this would suggest that $1000 of computing power would cost about 3 cents.

What does it mean when so much power can be created at so little cost?

 Posted by at 12:36 pm
© 2011 Technology Tsunami Suffusion theme by Sayontan Sinha