Fast Forward Lab's Hilary Mason talked to TechRepublic about taking data and making useful products for clients.
It's an age-old battle: Cats versus dogs.
One day while working as link-shortening company Bitly's chief scientist, Hilary Mason and a coworker decided to run an 8-hour computational job to figure out whether more people shared pictures of cats or dogs on the Internet. And, why not?
"That just made me realize that the technology had become so cheap for this large scale analysis that we could play with it, and when you can play with technology, then really creative, awesome stuff can happen," Mason said -- the least of which is finding out that dog pics are shared most.
This realization helped Mason move toward where she is today, the CEO and founder of Fast Forward Labs in New York City.
Fast Forward Labs is relatively new, having been founded in June 2014. It's a machine intelligence research company. They work with clients to take stock of the data that client has, the data it should have, and what assumptions can be made about the data that could be validated by the data.
"From there, we drink a lot of coffee, come up with bad ideas and then try to filter them down to the good ones," she said.
For example, Fast Forward's first project involved natural language generation algorithms. A prototype product they produced took structured data about real estate and wrote advertisements. They even found insights about words like cozy, which means about 400 square feet smaller than the average, or big, which is at the average or slightly smaller, when describing real estate properties.
"This is the kind of thing that, even today, people are surprised that software can do," she said.
They put together a report, took it to a studio in Brooklyn that does letterpress, and published the findings.
Other projects include deep learning for image recognition, so software that can say there's a 99% chance that there's a puppy in a certain picture.
They also offer advising for clients, to help them figure out how something like that software could have an impact on their business.
"They're never ready to take that step-- there's always a lot of other work to get there, so our advising service really helps them get to that point and do so successfully. So, we basically become their nerd best friend," Mason said. They try to strike the balance of always looking at new things, as well as how they could work in the real world.
For Mason, computing and the real world first merged in Kindergarten.
"I went to a school with very little supervision that was lucky enough to have computers," she said. Those computers were Apple IIes. They booted to a BASIC prompt, and it took a floppy disc and some commands to get them to do anything.
That led to a conversation in the car with Mason's mother, where the kindergartner announced she'd like to either be an astronaut, computer programmer, or taxi driver when she grew up.
Obviously, Mason went for the programming option, and completed her undergraduate work at Grinnell College in Iowa.
Initially, she wasn't sure if she'd study English or computer programming. Either way, the school encouraged students to take classes across different disciplines. She's brought various aspects of that liberal arts education into her career.
"I think it teaches you how to think and reason and communicate, and then sends you out in the world to see what damage you're going to do," she said. "It's a good way to think about preparing people for a world in which things are changing all the time, and that's the world I live in now."
In a way, that idea plays into the structure of Fast Forward, which ends up being a place that not only makes a technical product, but explains it and gives context.
"We have a lot of mythology in our industry about how people who like math and like programming can't talk to human beings and I think it's complete bullshit," she said.
These days, Mason is splitting her time between traveling to meet with clients and getting the word out about Fast Forward, and making sure she takes a few days a week to do technical work. The former isn't exactly where she might have imagined herself, initially
"Telling the story of what the company does, figuring out how we best work with our clients, making sure it's a viable business; those are all challenges, and if you'd asked me five years ago if I would enjoy doing something like that, I would say you were crazy, but I do," Mason said.
In her own words...
How do you unplug?
"I read a lot. That's something I've done since I was a teenager. My parents still make fun of me for always having my nose in a book, but I think they're a bit hypocritical because my mom has the biggest paperback collection of anyone I know. So, that's something I do on the subway to relax in the evening. I also really love to travel. It's hard to say that right now when I've been traveling so much, but that feeling you get when you go to a place that you can't take for granted. I just spent Thanksgiving in Stockholm, Sweden. I went out there for a conference and stayed for the holiday. I tried foods I'd never tried before, got to learn a bit about the holidays there. There's something about travel that I find really opens your mind to more creative thinking. I also play video games, I play with technology--I've been building some hardware stuff for fun lately. If I ever lose that ability to play with tech, I will know that it's time to retire or take some time off."
What do you like to read?
"I am reading right now a mix of science fiction, which I have loved science fiction my entire life, and history of science books, with an occasional business book thrown in."
If you could try another profession, what would it be?
"Can I still be an astronaut? That's something I would totally do. The whole idea of pushing the frontiers of science and human exploration, and pushing your own limits as a human being. That's really appealing."
- Hilary Mason: Use data science and machine intelligence to build a better future (TechRepublic)
- Curt Savoie: Principal data scientist for the City of Boston. Data philosopher. Wikipedia spelunker. (TechRepublic)
- Eliot Van Buskirk. Data and music storyteller. Child opera singer. Ramen noodle obsessive. (TechRepublic)