As a new study from the University of Arizona about sex trafficking during the Super Bowl highlights, advances in data analysis are underpinning some powerful new ways of tackling very tough problems. Among all the stones hurled at the tech sector lately, this is an area in which it can take pride.



Some days we’re reminded of just how lucky we are. If you’re reading this, you’re probably pretty lucky. I’m pretty lucky. My daughter is pretty lucky. I toured a school this morning where we plan to send her next year and left thinking she’ll leave there ready to take on the world.

Then I got in the car and heard this story about sex trafficking on NPR’s “Here and Now” program and all I could think about is what a crappy place this world can be if you’re born into the wrong circumstances. What an absolute travesty it is that we treat fellow humans — especially women and children, maybe even family members — with such utter disrespect. And how I hope beyond hope that the techtopia we inhabit and that I spend my days writing about can actually make a difference in helping save some of these peoples’ lives before it’s too late.

NOTE: It’s never too late to join this movement by becoming an expert in computer programming. More details at

I’m cautiously optimistic that it can. That NPR story is about a research study out of the University of Arizona, which worked alongside a defense contractor to identify probable victims of sex trafficking during the Super Bowl and the underground networks through which they’re trafficked. The level of sophistication it took to carry out this research was pretty advanced (like, about-on-par-with-what-a-very-animated-guy-was-explaining-to-me-in-the-hallway-of-DEF-CON-last-year advanced), and that sort of sophistication doesn’t come cheap. (Businessweek wrote a fairly detailed story about the program about a week before the big game.)

The reality is that this level of sophistication comes only from governments spending billions of dollars on R&D, or from businesses doing it and spreading the wealth.

Google is a prime example of this. Its research in everything from distributed systems to big data to deep learning has led to significant changes in the way companies across the board practice information technology (take the existence of Hadoop, for example), and is now making it easier for law enforcement agencies and other organizations to fight and identify crime. As Google’s work in computer vision and text analysis gets better and begins trickling down, I can envision some remarkable advances in fighting crimes such as sex trafficking.

The NPR story referenced a project by the Minneapolis Police Department to photograph hotel bedding around the city so it could have a database against which to compare the bedding shown in images of child pornography or prostitution ads. A researcher noted the difficulty of trying to create algorithms that would identify ads that might be advertising minors for sale. Imagine an artificial intelligence system sufficiently intelligent to comb through images and ads itself and identify the common visual or linguistic features. That’s the kind of stuff that’s on the horizon.

Unfortunately, though, these aren’t the types of issues around which many companies can afford to build a business and maintain the ridiculous amount of research investment necessary to keep advancing the technology. Which is why when you come to a conference like Structure Data, which kicks off 13 days from now in New York, you’ll hear a lot of talk about the cutting edge of big data, machine learning, artificial intelligence and general analytics in the context of business. Business is what makes our world go ’round and keeps our technology advancing.

Still, I’m glad to say that among the earliest speakers we reached out to and confirmed for this year’s conference are Google Idea’s Justin Kosslyn, who’ll be speaking directly about Google’s efforts to fight human trafficking, and Palantir’s Ari Gesher. He won’t be speaking about this topic (at least not as a focus of his talk), but Palantir’s software was instrumental in the University of Arizona research mentioned above and the company is actively involved in Google’s anti-trafficking efforts, as well.


It’s easy to scapegoat companies like Google, Palantir and even Facebook for many things (driving anyone earning less than six figures completely of Silicon Valley and generally eroding our notions of privacy, for example) and often those criticisms are fair. They absolutely should not be dismissed. As Tim O’Reillypointed out in a blog post on Thursday morning, there are some serious benefits to the advances we’re making in data analysis and also some serious risks in letting it run unchecked.

But, honestly, on days like today I’m left thinking it’s not such a steep price to pay that the luckiest among us sacrifice a little privacy. If the technology will some day make it into the hands of people who can get the unluckiest among us out of some truly horrific situations, and maybe even help them get back on their feet afterward, who am I to complain about Google knowing my contacts, Facebook reading my inane posts or Disney knowing every step I take inside its parks. Everyone should be so lucky.

Photo credit:

Via Gigaom