Profiles In Innovation

Rob Spectre: Net Patrol

Rob Spectre is a New York-based technologist building childsafe.ai, an artificial intelligence platform meant to help protect children from online predators.
Chris Mohney

What drew you to getting involved in protecting children in the digital space?
While I was working at my last software startup, I worked on a project to create a SMS shortcode for a big national human trafficking nonprofit for victims to text in and get connected with a service provider. I didn't know anything about the problem, but from what they said, it sounded like what these young people were going through was horrific. The ask was a Saturday afternoon of work, so I pitched in with what I could.

A couple weeks after they launched it, I heard back from them indicating that they just recovered a victim who had been sold for sex on the internet for five years. Their law enforcement partner arrested the pimp in the hotel next door, and the driver across the street, and the young woman was on a flight home. And that day—the day the nonprofit contacted me—the victim had just turned 18.

In that moment I learned two things: this problem was everywhere and growing every day, and a tiny amount of software engineering can have an enormous impact. After I learned about how bad child exploitation on the Internet was and how fast it was growing, it ceased being a professional choice—it became a moral imperative.

How would you describe your creative relationship with technology?
"Creative" is indeed the best way to describe it. Programming is the medium of expression that spoke to me ever since I was a kid. When I have an idea I want to get out in the world, my text editor is the instrument I pick up by default. It's why I really can't use a tablet or play video games that often—technology for me is not about consumption, but creation. It's probably also why I'm such a rubbish guitar player.

What is it like taking AI to such a dark place?
It is indeed dark with respect to the content, but what struck me when I first started working on this problem is how all this activity is out in the open. Kids are being bought and sold not on some hard-to-find corner of the dark web, but on the universally accessible, Googleable surface web, plain as the light of day.

Children are being groomed right now on the social media websites you and I use, advertised on the surface web marketplaces you and I know, and sold on the messaging apps you and I have on our phones. When folks first learn about this problem, they often ask, "Why are they doing all this out in the open?" If you'd ask those who are doing it, I imagine they'd say, "Why hide? No one is coming to get us."

What do you view as the greatest challenge over the next five years?
Everyone staying focused on what matters. We're living in the three decades that will determine whether humanity—and much of the rest of life—lasts long on this Earth. We've stumbled past the inflection point in our development as a civilization, where the problems humans create are solvable by the humans that follow.

Our institutions and technology can now cause irreparable harm, and we're saddled with thousands of brutally effective machines built to disrupt our attention. Some of these machines are technical. While reading this interview, your reader probably got two notifications on his or her phone, and five emails begging for his or her eyes. Some of these machines are political. 

Try to search the web for the latest climate change data without running into an oil-and-gas-funded landing page, or scroll through your social media feed without an unwitting relative sharing a divisive Russian blog post. And some of these machines, like the ones we encounter in our work at childsafe.ai, are economic. There are deeply rooted financial incentives to grab more eyeballs and sell more ads without regard to the unspeakable harm those same machines are causing our children.

Where we are headed there is no turning back, and it's never been harder to keep your eyes on the road.

How has the nature of your work changed over the course of your career?
The productivity of the individual developer is the biggest single change. I remember in my first gig 15 years ago, a "system upgrade" was three of my colleagues and I heaving big Sun servers into my hatchback and driving them to a data center north of Boston in the middle of the night. Now that same operation is pressing a "+" button in a web dashboard.

But that productivity extends not only to how far an idea can scale with a single developer, but the breadth of problem domains that developers can address. Computer vision, natural language processing, time series prediction, graph analysis—ten years ago you needed a PhD from a top-flight school to even attempt any of these. Now with the explosion of open source frameworks and cheap web APIs, any programmer can drop advanced machine learning in their creations.

To draw a analogy from manufacturing, it'd be like if a worker on an airplane production line ten years ago could be a rocket scientist today. There hasn't been anything like it in the history of vocation.

What would you like to achieve in your work that you haven’t yet attempted?
We're hiring for a fully remote workforce at childsafe.ai, and that will be an entirely new thing for me. I've served remote teams in a leadership role at mostly co-located companies before, but building an entire organization around a remote experience is something I'm very excited about.

Being able to hire engineering talent anywhere is such an unfair advantage—I can't wait to see what that looks like for every role critical to a software company's success.

What do you view as the risks or dangers associated with the intersection of human behavior and algorithmic analysis of that behavior? 
More broadly, the public concern about ethics in interconnected computing is rising every day. As we learn more about all these signals we unwittingly generate through our digital lives and how they are used without our informed consent, we all rightly ask the technology industry what it is they think they're doing.

As a technologist, I feel like we're in a Jurassic Park moment. We spent 30 years building the internet so preoccupied with what we could do, we rarely stopped to ask what we should do. 

That moral question is why I started childsafe.ai. We have a hundred times the equivalent whiz-bang, next-level computer science that we used to get people to the moon right now trying to figure out whether or not you want to buy a new house or a new toaster or a new jacket. Orders of magnitude more code than was in the Apollo 11 module is built to get the statistically honed, computationally crafted message in front of you to buy something you don't need.

And yet, none of that radical technology is being used to keep our kids safe online. What if we used a tiny portion of that talent to keep our kids from being bought and sold online? That's what we want to find out at childsafe.ai.

How has technology allowed you to connect with people?
As a poor kid growing up in the middle of nowhere, technology was the way I was able to connect with other people who knew how to write software. Our computer lab in our tiny farm school only had ancient Apple IIes that were made before I was even born. The books that accompanied them were owner's manuals that would tell you how to turn them off and on, but not how to build things on them. When our town finally got internet access in the late 1990s, I was able to reach professional programmers that were doing this work all day, every day. The internet was a much smaller place back then, and I could waltz into some random chatroom and learn something I never would have been able to learn without it.

A kid growing up in 2018 does not have that experience. A recent study in the Journal of Adolescent Health estimated that one in five children between 12 and 16 are exposed to unwanted sexual imagery or video. One in nine are solicited sexually online.

It's like they are using a different internet than the one I grew up on. Fortunately, though, connectivity is two-way by definition. If there is a way for this to happen, there is a way for it to stop. Finding that way must start with us.

What do you view as your most personally meaningful accomplishment so far?
You know, I have a hard time with questions like this. Every win I've been fortunate enough to be a part of was the result of being a member of an excellent team doing their best work. Anything meaningful requires the concerted effort of many dedicated people, though we tend to celebrate whoever it is that is at the top of the org chart as the sole progenitor. Victories that are truly personal are about as valuable as a high score on a video game—real change needs a team.

What new ideas have you encountered when traveling?
One of the bits I love about visiting other countries is seeing how the workplace functions. I spent some time working in Russia, and every single person started and ended every single day with a firm handshake. I worked for a bit in Israel and loved the candid task conflict; everyone felt they could say their piece on a big decision, sometimes forcefully. From Stockholm to Shenzhen, colleagues gather over food and drink in an intentional way to reinforce their professional bonds. Getting to participate in those rituals and watching their effect on those teams is always a real treat.

What opportunity do you most regret not taking full advantage of?
Ooooh—this is a great one. Building software is opportunity with a universal adapter. If you have the good fortune of being in that business, you have a lot of choices on what to work on. With that abundance of choice, there are inevitable whiffs.

Professionally, I turned down being an early employee for the biggest video game in history, worked at two startups that burned over $60 million in capital that both completely flopped, and got woken up by a production outage every night for an entire year. If failure is the best teacher, I am the beneficiary of world-class instruction. Personally, I could have seen The Book of Mormon, Kinky Boots, and Hamilton in previews for less than $50 each. I elected to stay in each of those nights to write some code. I am the worst New Yorker.

What brings you the greatest satisfaction in your work?
When I first started working with law enforcement on human trafficking, I would naively feel accomplished when we made an arrest or got a conviction. After such an event, I started to notice I was usually alone in my excitement. Eventually, a lieutenant pulled me aside and said, "Rob, the arrest is just the beginning of our work."

The journey between becoming a victim and becoming a survivor in this problem domain is the longest and hardest you're ever going to see in the human condition. They have physical trauma, sexual trauma, emotional trauma, addiction, frequently while being the primary caretaker of one or more children of their own. By the time the door gets kicked in, unimaginable evil was already visited upon these young people. They aren't kids anymore—that innocence was stolen from them.

Now my greatest satisfaction comes in the moment when I see someone learn just how bad this problem is, and just how much they can do to stop it.

What organizations have you enjoyed working with the most?
When you work on what I work on for a living, volunteering with young people is an essential light in your life. I serve with hackNY, an organization building an inclusive and responsible innovation community in New York through undergraduate technology programs, and DoSomething.org, a global movement for empowering teens to make an impact in their communities by scaling up their own service projects. 

Every hour I spend with the kids in both orgs is like plugging my finger into a light socket—stunningly effective organizations empowering absolutely inspiring young people.

Who are some people you particularly admire?
This year I got to meet and work with Marian Hatcher from the Cook County Sheriff's office. She's got a cabinet full of accolades and achievements from countering human trafficking (including from the Obama administration), nearly every counter-HT officer in the country on speed dial, and the most effective survivor network in the country. She is a survivor of multiple sclerosis, cancer, and human trafficking herself. Several times a week, she gets calls involving a victim somewhere in the country needing shelter, and she picks up every time. Whenever I wonder if I could do more, I remind myself of what Marian is doing after what she's been through.

And, like a lot of people, I admire my ma. She raised two kids as a single parent in the 1980s and 1990s on $11k a year. I don't imagine I'll never do anything as hard as what she did.