Once a year or so, I do a Google search for “teens some_crime arrest” where some_crime is murder or assault or car jacking, etc. My goal is to try and gain some insights into how machine learning might be used to prevent crime.

A few weeks ago, my recent experiment yielded results that were similar to previous experiments. My first impression is that there is far more teenage criminal behavior than I’d expect. Perhaps this is because in the area where I live, there is essentially zero serious teenage crime — teens putting detergent into a water fountain at an apartment complex to make bubbles is about as serious as it gets.

I live in an area dominated by large tech companies such as Amazon, Google, Facebook, and others, and so the demographics are high education, strong work ethic, cultural diversity (about equal proportions of White and Asian families in my neighborhood), stable two-parent households, and relatively high incomes.

Google search for “teens murder arrest”.

My second thought is that, unfortunately, I didn’t see any obvious way that machine learning can be used to prevent crime. In fact, it may be the case that machine learning isn’t well-suited for preventing crime. But that doesn’t mean it’s a waste of time to think about the idea.

Google search for “teens car jacking arrest”.

So, my experiment didn’t lead to any concrete results but I’ll keep thinking about how machine learning can be used for good purposes, such as reducing Burglary Looting Murders. Ultimately, machine learning — and all science for that matter — should enable people to have better lives.