So Far, Big Data Is Small Potatoes
blogs.scientificamerican.com
blogs.scientificamerican.com
How could software catch predatory lending practices written into policy and executed by bankers? Wall streeters were using their 'fastest computers' and 'biggest databases' to make and manage money. Even if they were using cloudera's CDH or some other big data framework, they will use it for the exact same reason. Maybe even find a way to maximize profits and exit before the next crash.
I'm probably nitpicking, but that's not a real 'big data' use case. Maybe when every computer passes the turing test and learns to spot corruption, they can throw that in a mapreduce job and solve the problems this guy is talking about.
From a commercial standpoint, people only care about making big data cheap and easy for enterprises to adopt. There are only a handful of really innovative, forward-thinking implementations of 'big data'.
There's more to predicting the future than crunching numbers, finding correlations, and making pretty graphs.
> Maybe when every computer passes the turing test and learns to spot corruption, they can throw that in a mapreduce job and solve the problems this guy is talking about.
I would be very surprised. History has shown that if we have intelligent agents (humans) with roughly equal access to data, we still have unpredictable results.