couple-thousand-dollar jobs for hotels,” Oh says. He remembers “coming back at night, implementing some functionality, and coding away. I was poor—it was very much an extension of grad school—staying up all night with fellow co-workers, getting stuff done. It was about doing what you really loved.”
Oh says he would love to start another company if the opportunity arises. “Certain things are bubbling up. I want to see if some of these ideas stick or not,” he says.
In talking about these possibilities, a couple of themes emerged. First, Oh learned at EveryScape that advanced technology does not necessarily solve problems. “The progression of our innovation, from grad school to now, is that the technology got simpler and more useful, and less 3-D,” he says. “I still believe 3-D is not ready for prime time. If you’re trying to scale up your user base, or crowdsource, it becomes important for things to be so simple to use so people will adopt it. That’s the direction I want to go.”
What he’s really saying is that computing technology will only take you so far in a business. When it comes to solving real-world problems—say, recognizing objects or text in photos with very high certainty (more on this below)—“you still need a human in the loop,” he says. So technology-wise, look for Oh to pursue a combination of automation and crowdsourcing of some kind.
The second big theme he’s exploring is the “mapping of online to offline.” From Google to Facebook to Groupon, there has been a progression of using social and location-based technologies to bridge the gap between the digital world and the real world. Of course, this theme is as old as the Internet itself—but now it’s happening on a very local scale, with very fast impact on stores and businesses.
To make all this more concrete, Oh says, think about what Google is doing. For instance, he says, Google’s Street View will soon use optical character recognition to read store signs in its images, so as to be able to link directly to information about pictured businesses (such as a Starbucks or McDonald’s on a street corner): how to get there, where to park, hours of operation, and so forth. The value to the store might be to improve its foot traffic by 1 percent, say.
So it sounds like Oh is interested in working on a new venture that combines a social or human layer with automated data (visual or otherwise) about local businesses. “It’s more about platforms and content than technology,” Oh says. “That’s the whole direction of where the Internet is going today.”
It’s also where his old company is going, so Oh has to mind his non-compete clause. We’ll see how that affects what he does next (and when). “I have to be very careful there,” he says.
