What’s Wrong with Big Data?

We know that the whole world led by tech sector is surfing the Big Data wave. Billions of bytes of data is being collected, cleaned up, indexed and mined by the time I will finish this sentence. According to Wikipedia, as of 2012, roughly 2.5 exabytes (2.5×1018 bytes) of data is getting collecting every day. The promises of mining that data range from cloud intelligence health care to driverless automobiles.

What could possibly be wrong with Big Data anyway?

There is nothing inherently wrong. But there are two problems I can point out.

  1. Big Data is too big. It is incomprehensible. Big Data remains infertile unless you can redact Big Data to Small Data, which can be summarized, visualized and put to use. Right there is the metaphorical big elephant of our Big Data living room. How do we handle that complexity? Google, Facebook and other tech giants are already scratching their corporate heads to unscramble this puzzle. Doable, but expensive and often could be erratic.
  2. secondly Big Data convinces us that we all can become mini-statisticians. “Show me the data” is the new mantra. We often mistake correlation to causation. With Big Data, this tendency can get even more accentuated.

There are plenty of avenues for making use of “small data” available to us. One of them is our on personal data. What do we eat, how much we exercise, what are our biometric readings etc. Data collection is going to be easier than ever with the emergence of wearable technology. If we could collect our own personal data and dive a little deeper, we can derive intelligence about ourselves and it can have a positive impact to our lives. On a personal level, it is more worthier a quest than crunching mega-billion abstract data points.

A big shout out to all those Life Logging ninjas! And to the good folks at Quantified Self.

Baby Carrots

A popular item that found its way to our kitchen was baby carrots. It was pretty, cute and crunchy. It took a while, before we figured out that baby carrots are not really “baby” carrots, rather baby-cut carrots. Also, it was disappointing to know that these fresh snacks were less nutritional than the regular carrots. Most of the antioxidents, it seems are found under the vegetable skins and most of that is lost,  since the skin is peeled off to create the appearance of deliciousness.

It seems it was invented by Mike Yorosek, a california farmer mainly to reduce waste from his farm produce. Not-so-good-looking carrots, that otherwise would get rejected in the super market, are picked up, washed and scubed off the imperfections to market as a value added product. Story goes, that the carrot consumption almost doubled, while wastage has nearly eliminated. It was a marketing brilliance!

Humans are evolved to love pets, toys and even things that remind us of babies. Yes, even baby carrots. Those who had a genetic make up to hate babies were already pruned off of our geneology map. I am not sure, if our ancient circuits get fired by hearing “baby” of the baby carrots, but they are insanely popular these days.

It may be a slick marketing strategy, but nutritionally, we figure it is a less than ideal for us. Now-a-days, we leave the temptation at the supermarket and happily settle for ordinary looking carrots.


Hunting for New Users

Yesterday Twitter announced that their user growth was flattening up. Needless to say, their share prices dropped. Now Twitter CEO says that his number one priority is to grow the user base. I think it is a great idea.

It is a great idea not only for Twitter, but also for Facebook, Google and other corporations, which provide services to connect people. If not today, in near future, their user growth is going to stall.

It seems there are not many ways you can make more money from this business model.

  1. Upsell your services to existing users (advertisers). In order to do this, dive deeper into peoples’ data and share it with advertisers.
  2. Create new users, who have enough buying power that attracts advertising revenue, while safe-guarding existing user base.

To upsell their services, they need to mine more personal data, which can also create more backlash in the form of user attrition and regulatory scrutities. So the trick is to mine more data without appearing to be endangering the perceived privacy, while providing more value per click to the advertisers. This, I think, is a game of diminishing returns.

Creating yet another billion users with enough buying power to attract advertisers is even harder. To achieve this, these corporations have to increase their market share. For Google and Facebook, that would mean creating a whole new market and then lay claim its majority stake, because they already enjoy near monopoly on the existing market. A company can grow as much as its industry, not an inch more.

Google and Facebook already realized this and have been working on this. Beyond philanthropical goals, projects like Google Loon, is ultimately aimed to connect the “unconnected” and thus to create new markets. Which is exactly what Facebook is trying to do in China.

If Twitter wants to grow their user base substantially, they might want to sail abroad too. Not an easy thing to do, especially for Twitter, which doesn’t provide many glitters in their services, but is defenitely worth the effort.

College or no College?

In a recent New York Times article, conservative columnist David Brooks argues that giving college education to hard working folks is going to increase their income to $58000 per year. Some studies have suggested that the income disparity between college grads and high school grads is only going up. As of 2013, the income gap may be as high as $17,500 in the US. No doubt that it makes perfect sense to send your children for college. The answer to the “college or no college” question, of course, is a no-brainer, higher education brings home higher pay. Or it seems.

So shouldn’t we campaign for mass college education as well? Wouldn’t it be better for the common good to help another million or two to earn college degrees? On the surface it sounds like a logical thing to do but it may work out for worse. Here is why:

  • It is well known that the US has pretty darn deep student loan dept. On an average, a college grad accumulates as high as $26,000 on student loans. Those who make to the higher paying jobs may or may not make enough money to justify the huge debt that otherwise they would carry around for a long time.
  • Flooding with college grads would bring the average wages down. As a consequence, unemployment among college educated would go up. Good news for employers, who are constantly looking for cheap labor. Nothing exciting there for job seekers.
  • Unemployed from college grads will inevitably compete for low income jobs. Even in low paid job sector competition increases resulting in lesser pay. You can see how quickly unhappiness spirals up.
  • Creating a lot of college graduates will not make their lives easier. But creating a lot of jobs will. Before mass producing college grads, we need to create the demand first.

It is like selling stocks in a bullish market. If you own a hundred shares of a stock that’s doing very well on the stock market, yeah, you can make quick bucks by selling them on an uptick. Say instead, you are fortunate to own a million shares of the same stock. If you try to sell them all at once, that will force the prices go down and you lose.

You would still want to send your child to college. She / he will have better chances of a successful career. But as a public policy, mass college education, will create more debts, unemployment and unhappiness.

What we really need is to create a job market in which job seekers can find well paying jobs with a high school level education. These are the folks who work hard, spend all and keep the economy thrive. Unlike the conservative rhetoric, this will eventually help the economy grow.

A New Kind of Education

The past 200 years of school and college education have been the backbone of our worker centric economy. Society and individuals invest money and resources to produce analytical minds of various kinds, which employers could pick and choose for their labor force. All these years, we have been focusing on mostly analytical skills with little or no importance on emotional and creative skills for our curriculum for schooling.

We get our smartness from 80 billions of neurons and 100 trillion interconnections. We gain leverage by connecting these brains with 7 billions of others forming a massive network of wet machines. The problem now is that the machines and silicon brains are catching up. They too have billions of transistors and billions of interconnections. It seems as though Moore’s law is not going out of fashion anytime soon. On any analytical scale, computers have grown to beat us down. Then, why do we care to compete with them – that is an impossible task anyway. I think the time is ripe that we switch gears and focus on complementing machines instead of competing with them.

Instead, we should start investing in a new kind of education, where we focus mainly on the skills that computers can’t excel. At least for now. Like emotional skills, non-linear and spatial thinking and creativity.

Do you still want your kids go to medical school , engineering school or law school? Think again.

A Request to Pocket App

Pocket is a gorgeous app, available in almost all mobile platforms and the web. It helps you to save online reading for later offline reading. You can tag the articles and have a little organization around it. With simple and elegant controls, it provides clutter free reading pleasure. It also can remember where you left off individual articles. If you haven’t used it, please check it out. Trust me, you would love it.

Pocket Hits

Once every few days, Pocket automatically curates a set of articles, which match my reading interests and sends as an email newsletter. They call it “Pocket Hits”. For me this news letter content is almost always right on target. I love to read right out of my inbox. It’s as if the algorithm can read my mind. Under each article, there is a link to save it to your Pocket account. (see picture.) Usually the newsletter contains 10 article links, for which you have to click 10 times to save them all. But it is worth the effort. Since the articles are saved for offline reading, it is perfect for subway reading, when you are completely disconnected from the web.

But …..

Wouldn’t it make sense to add an extra link like “Save all to Pocket” ? What do you say, Pocket team?


Two Tools for Teaching Children to Code

Why should we teach our children to code?

  • Because it gives them the ability to see what’s going on under the hood.
  • Because it sharpens their logical reasoning.
  • Because it empowers them to create instead of consume.
  • Because they may just like it and someday they might want to make a living out of coding.
  • Because it makes them “cool” among their peers.
  • Because it helps them to work better with computer professionals even if they chose a different line of specialization.
  • Because it is pure fun. Coding gives them instant gratification. Working code can show them magic unfolding in front of their eyes.

I am sure I wasn’t persuasive enough but please give it a try. There are tons of resources out there online to help you choose the right toolkit. It will be a true gift to your children.

To begin with here are a couple of links to checkout. Scratch from MIT is really easy and fun for young kids of age 5-8. App Inventor (created by Google and later donated to MIT) lets you create real Android apps using some of the event driven programming concepts. It is appropriate for ages 9-15. A child who can do Legos can do coding as well with either of these tools. Both are great tools for teaching.