Companies around the world are now starting to tap into the massive potential of AI and Deep Learning to decipher human speech, translate documents, recognize images, predict consumer behaviors, identify fraud, make lending decisions, help machines “see” and even drive autonomous vehicles. Doing this requires massive amount of relevant data, a strong algorithm and a concrete task. Let's see an example.
Mortgage lending is a huge data-intensive business. That’s obvious to anyone who has gone through the exhaustive documentation process. We are not far away from the massive adoption of having an AI powered app that relies exclusively on algorithms to make lending decision themselves based on millions of existing data on loans and mortgages.
Relying on credit scores and income alone to assess the eligibility of a person for mortgage qualification might not be the optimal outcome. For instance, people with the same credit score can be in fact less (or more) risky than each other when additional pieces of information are used in the lending process.
Deep Learning's uncanny ability lies in the fact it derives predictive power from data points that would normally seem irrelevant to a human loan officer. It's a sign of the limitations of our own minds at recognizing correlations hidden within massive streams of data. By training an app's algorithms on millions of data on their existing loans, a company can discover lots of smart eligibility criteria even if those criteria could not be explained in a simple way humans can understand.
This type of deriving intelligence that takes data from one specific domain and applies it to optimizing one specific outcome is what we generally refer to as “narrow AI” or Deep Learning. While narrow AI is impressive, it is still a far cry from “general AI” - the all purpose technology that can do everything a human can.
AI is increasingly disrupting different sectors and weaving deep into the fabrics of our daily lives - whether we realize it or not.