Tuesday, January 24, 2012
It has been said that, when you have enough data, you don't have to be too clever. Or, in other words, more data usually beats better algorithms. This is empirically true when it comes to programming machines. If you just know some good ways of finding information already stored somewhere, you can often save a lot of trouble on actually calculating it; if you have a sufficiently massive dataset, you don't have to worry about how to do good statistics with a representative sample, because you can just look at the entire population. But many people, myself included, implicitly apply this logic to the operation of their own brains, and, sadly, our brains just don't work like computers. Every day, I am immersed in internet news. While I'm reading it, I always feel like this is a very important thing to do, and how wonderful is it that I can be so well informed in this day and age. But in the long run, very little of it really matters. I never want to get to work on anything until I know everything that the internet knows about that particular topic; but in reality, there is a point, very near the beginning, past which more reading will not improve my performance in any meaningful way. A human needs to be clever, to know how to use the information he has more than to gain all the information there is. One of the greatest challenges for education now may be not getting access to accurate information or figuring out how to study well, but rather filtering out what we actually need to know and successfully ignoring everything else.