Throughout our lives we are confronted with experiences that have no meaning on the surface, but when presented with time to learn, improved understanding of the experiences themselves and repetition, our brains are empowered. It is this level of knowledge and expertise that makes seasoned trial attorneys so successful, longtime surgeons great at what they do, and why we learned early on that sticking your finger in the campfire is a horrible idea.
When it comes to predicting catastrophes and saving lives, machine learning – specifically deep learning – holds the keys to something great that, until very recently, couldn’t be unlocked. But the times are changing.
Whether we’re looking at signs of a potential avalanche, a massive flood or a terrorist attack, technology has evolved to the point where we have the power to train computers using deep learning to sift through massive amounts of data then intelligently extract key nuggets of actionable intelligence. Subject-matter experts train while supercomputers do the heavy lifting, but where does the real time data come from?
A recent phenomenon that’s bountiful from an intelligence perspective is the fact that many people post on social media, namely Twitter, to report everything from floods to shootings to suspicious individuals. And funny enough, at times tweets go out before calls to 9-1-1 and at least 10-15 minutes before news outlets report on incidents.
Even more frequently, people tweet about things that have no context/meaning to them, but to a trained eye elude to something much greater. In this case that trained eye is a computer.
Imagine having the power to analyze millions of tweets to crowdsource pre-incident indicators then act on them before damages occur or people are injured/killed. Here are some examples:
1) Within a 2-mile radius of a large dam, 3 people tweeted about water overflowing and posted images of flash floods.
2) Five independent Twitter users posted about an individual in a grey sweater leaving a backpack in Time Square.
3) There are strong winds and a small fire breaks out a few blocks from a company’s corporate headquarters, and images are being posted on Instagram.
The moral of the story is that deep learning has some futuristic, magical qualities that may seem far off today, but are actually being implemented by companies like Soteria Intelligence and have the power to change the world of public safety as we know it.